Many excellent texts on the investigation and management of individuals with HIV infection now exist^[1] and are well worth consulting. This section is intended as a synopsis from the developing world perspective, but some relatively rare conditions have been included because prompt recognition can be helpful in management in some cases and at least may save a protracted misdirected work-up in others.

HIV testing

Medical reasons to test:

As a preliminary to consideration for ARV treatment.
Knowing that a person is HIV positive will allow for more goal-directed/streamlined management of a specific medical problem.
As the ‘source’ patient – needlestick injury. In this case the potential benefit is to a third party (the person sustaining the injury)

Times when testing is NOT appropriate:

A confused patient where the test result will not alter medical management.
A dying patient.
Where the patient is unconvinced of the benefits of testing.
When the patient is not ready. There are some situations where bringing up HIV issues frightens the patient into running away from treatment of a curable condition (e.g. TB pericarditis). Be sensitive when starting to ask about the subject, and if you think that this sort of situation may arise, consider postponing testing until the patient has developed more confidence in you.

Remember that:

Testing without informed consent (i.e. adequate pre-test counselling) is still regarded as inappropriate – there is an unrevoked National Policy document on this^[2]. There is a move towards ‘opt out’ testing but this is still not standard practice and raises a number of ethical issues, as well as having mixed benefit in various studies^[3].

Agreeing to be tested by an individual doctor/sister, is not necessarily agreeing that the information be made available to all caregivers. Discuss with the patient.

Counselling checklist

You should at the very least have read both of the references^[4],^[5] cited below before embarking on the procedure. These references were chosen because they are readily available and cover most of the issues in adequate detail.

Procedure for HIV testing

Have a clear reason for doing it in this patient at this time.
The patient should be adequately counselled by someone with experience (preferably yourself) gained either from completing a course or adequate reading.
Obtain signed informed consent.
Draw the blood yourself and label the specimen.
Complete the request form yourself, checking that name on tube is same as that on the form (and that it is the correct folder for that patient!)
Depending on the logistics of your particular situation (i.e. how long it takes to get a result) organise a follow-up appointment with the patient, or in the case of an in-patient, state when the result is expected to be available.
Sit down and explain the result carefully once it is available.

Common mistakes:

Writing ‘probably HIV positive’ in the folder, and then sending the patient elsewhere for counselling and testing.

Half counselling. If you raise these issues, you must be prepared to have at least 10 minutes to spare; broaching the subject and then leaving it dangling causes distress and often contributes to a subsequent unwillingness to be tested.

Unclear explanation to the patient of the value of testing at this time – most people who don’t understand the reason for testing will refuse.

Failing to get the patient to commit to a time and place to return for results.

Telling the patient that you are concerned about HIV and then taking blood for something else. The patient may assume this was his/her HIV test and interpret you subsequent behaviour as indicating positivity. It is always advisable to explain why you are taking blood from any patient for any reason.

Initial evaluation of the HIV positive individual

Assessment:

Evaluate the clinical stage of illness.

Ensure that counselling has covered the relevant issues.

Explain what can be offered by the health system, and the importance of presenting early with new illnesses.

Symptoms of TB or other illness?

Examination:

Record the weight and temperature.

Oral thrush (Don’t forget to look under dentures.)

Significant lymphadenopathy (criterion: could you do a needle aspiration of that node?)

Baseline investigations:

Urea, creatinine and electrolytes.

Syphilis and Hepatitis B serology.

FBC and CD4 count

PAP smear.

CXR if the patient has respiratory symptoms or there is a history of significant weight loss (conventionally taken as 1.5-3 kg.)

TB PCR on sputum if cough for > 2 weeks, even if the CXR seems normal.

Nutrition in HIV positive persons

Weigh patients at every opportunity –this is a quick, cheap, and highly reproducible measure of health and arguably no clinician looking after AIDS patients should be party to weightless clinic notes. Weight loss should never be assumed to be due to HIV until careful thought has been given to the possibility of co-existing infection, particularly tuberculosis.

Poor appetite:

Think about medications, particularly highly prevalent and usually unnecessary additions such as iron supplements. (The usual anaemia in AIDS is the anaemia of chronic disorder where ferritin levels, reflecting iron stores, are high due to iron trapping.) Stop anything not essential and see if it makes a difference – you can always re-start it.

Constipation. Ask about this and fix it – inadequate fluid intake may contribute.

Hepatitis. If there is any evidence of upper abdominal discomfort, consider checking liver enzymes in anorexic individuals, particularly those on medication.

Depression. Although everyone knows the two are linked, it is puzzling how infrequently any attempt is made at addressing this in individuals with a disinclination to eat.

Access only to single large ‘convenience’ meals – e.g. when a caregiver returns from work. Smaller snacks that are more regular may be better tolerated.

Difficulty eating

Thrush should be looked for and treated.

Consider xerostomia (dry mouth) due to dehydration or salivary gland disease.

Suggest softening foods with water, milk or gravy.

Advice about using softer foods such as porridge is usually unnecessary – it’s nearly always a staple already.

Acute necrotising ulcerative gingivostomatitis can be very troublesome, and differentiating bacterial from herpetic causes is not always easy. Patients often are treated with a succession of amoxicillin, metronidazole, and acyclovir over the course of some weeks, usually without dramatic benefit. Antiseptic mouthwashes may be of value.

Pain on swallowing due to herpes may respond to acyclovir – this is no longer a particularly expensive intervention.

Don’t forget about the possibility of a pseudobulbar palsy in an individual with previous strokes – a nasogastric tube may be appropriate.

Anabolic medication

The best anabolic medication is prompt and efficacious treatment of underlying illnesses, and then anti-retroviral therapy if appropriate.

Small studies on various medications such as anabolic steroids and thalidomide waft in and out of the literature. In general, they show statistically positive results of modest if any clinical value, e.g. mean weight gain of 1 kg. They all have side effects and are mostly expensive.

HIV and adenopathy.

Symmetrical adenopathy is common in early disease, but is not often seen in hospital practice. Adenopathy is particularly concerning if:

It is asymmetrical
It is associated with pain or constitutional symptoms
The patient clearly describes progressively increasing node size.
The nodes are large. This is subjective, but symmetrically palpable pea-sized nodes in an emaciated patient are less likely to be important that a 2-3 cm diameter single supraclavicular node.

Causes of enlarged nodes include:

Extra-pulmonary tuberculosis (commonest)
Lymphoma
Cryptococcosis
Kaposi’s sarcoma

Perspective – HIV adenopathy.

Locally, there is a very high prevalence of tuberculosis, and patients with adenopathy usually present because they are sick. In this context, the finding of adenopathy should be taken very seriously. The size of nodes in HIV-associated adenopathy is usually less than 3 cm. In a study of HIV positive individuals^[6] tuberculous lymphadenitis was most often cervical (99%), and was symmetrical in nearly one third (29%). Axillary nodes were also commonly enlarged (82%), but palpable epitrochlear nodes were only found in 36%.

Perspective – epitrochlear nodes in HIV.

A Zimbabwean study^[7] on epitrochlear adenopathy (not necessarily due to tuberculosis) found axillary and submandibular adenopathy in 75%, and epitrochlear nodes in 84%. This was a prospective study of 259 patients admitted to an internal medicine ward over a 3 month period, and was performed in 1994, when ‘early’ HIV was relatively common.

Only axillary, submandibular, and epitrochlear nodes were scored, and so comparisons of clinical utility have to be confined to these nodes.

Epitrochlear nodes were only slightly (9%) more commonly felt than nodes in the other two sites. Big epitrochlear nodes (>1 cm) were only found in 36%.

It is unclear if more than one observer scored the nodes, and although it was stated that the doctor was blinded to clinical data, it is difficult to mask emaciation and skin rashes.

Post-test probability of HIV in presence of epitrochlear/submandibular adenopathy

Hence feeling for epitrochlear nodes, as opposed to other nodes, only really does anything dramatic for post-feel probability if you are unsure of the possibility of HIV beforehand. If there are already other suggestive features the presence of epitrochlear nodes doesn’t help much.

Initial management of adenopathy.

Unless there is clear evidence of pulmonary tuberculosis from which to get positive sputum, the first line of investigation should be a needle aspiration biopsy sent for either AFBs or TB PCR or both. (for TB PCR draw up some TB transport medium (Middlebrook, 1.5 ml bottle) with the used FNAB needle and syringe, and re-inject it into the transport bottle. This specimen is adequate for both PCR and TB culture.) The aspiration of ‘pus’ from a chronically enlarged node that isn’t draining a septic focus is highly suggestive of TB.

If still negative, then consider an excisional lymph node biopsy done by someone sufficiently experienced to biopsy the node and not subcutaneous tissue. The reason for this is that significant adenopathy that is tuberculosis negative on serial aspirations becomes more likely to be due to HIV-associated lymphoma or Kaposi’s, and this is often difficult to diagnose without the whole node.

When not to do a needle aspiration node biopsy.

There are a number of situations where doing a needle aspiration node biopsy is either unlikely to be helpful, or is potentially dangerous.

A node that is draining spontaneously – just send a pus swab for AFBs

A patient who is restless, confused or violent – the chances of a needle stick injury are increased, so rather sedate the patient very adequately first or even wait a day or two for things to settle down – the aspiration biopsy is rarely urgent.

A very ‘deep’ cervical node. Sometimes such nodes can be very rewarding, but if you lack experience, rather leave them for someone else.

A patient with a bleeding problem (usually due to thrombocytopenia) does not pose an absolute contra-indication, but care is needed, and thought should be given as to how to apply pressure if bleeding does occur (i.e. don’t go for a poorly accessible node!)

Management of nodes where FNAB is unhelpful

Try repeating the FNAB once. If you are an inexperienced operator, it might be worth asking someone else to do the second one. If this is still negative, the probability of active tuberculosis is less, but further management will depend on the patient’s clinical condition. In a well patient with a preserved absolute lymphocyte count or CD4 count, it might be reasonable to observe for a few weeks. In an ill patient consideration should be given to doing an excision biopsy to exclude lymphoma. Rarely, other surprises are found in such excised nodes (e.g. cryptococcosis, Bartonella infection, nocardiosis).

HIV and C-reactive protein (CRP)

HIV-associated hypergammaglobulinaemia is associated with a raised ESR in HIV infected persons; if there is a desire to check for evidence of an acute phase response due to another infection, a CRP is better than an ESR – both have a sensitivity of more than 90%, but the specificity of the CRP is also over 90% (for the presence of an acute infection), whereas that of the ESR is less than 20%^[8].

The CRP is elevated in many other infections besides straightforward bacterial infections – for instance, it is greater than 80g/l in at least 20% of patients with pneumocystis infection^[9]. HIV disease progression has also been associated with rising CRP levels even in the absence of evidence of associated bacterial infections^[10].

There is generally so much overlap between levels in bacterial and viral infections that setting clinically useful threshold levels is problematic^[11],^[12],^[13].

A CRP < 8 mg/l in an HIV positive patient suspected of pulmonary TB at a clinic is 95% (95% CI 74 to 99%) sensitive in ruling out the diagnosis^[14]. The same does not apply to intracranial bacterial infections (abscesses and meningitis) where a normal CRP does not exclude infection.

Knowing the CRP from having measured it recently previously, and subsequently finding it to be elevated, can be helpful in situations where there is uncertainty about new bacterial or other opportunistic infections. In situations of clinically uncertainty, checking it and then finding it to be rising over time can also be valuable.

Cardiology – HIV

The very strong associations between HIV and various cardiac pathologies are not that well publicised, and thus merit review^[15].

HIV and arrhythmias.

QT prolongation has been associated with both co-trimoxazole and pyrimethamine. Co-trimoxazole can cause significant hyperkalaemia in high dose. In addition, the hypokalaemia associated with chronic diarrhoea is eminently treatable if recognised.

HIV and pulmonary hypertension/right heart failure.

Primary HIV associated pulmonary hypertension occurs, but also consider the possibility of pulmonary thromboembolic disease. Prevalence data is still unclear, but some reviews have suggested it may be present in between 0.5%^[16] and 5% of asymptomatic patients^[17]. Prognosis of patients with established HIV associated pulmonary hypertension is guarded, with an early review of cohort studies providing a median survival time of six months from diagnosis^[18]. It is still unclear if ARV therapy makes any impact on survival, but it is appropriate to consider its use. Warfarin therapy is also unlikely to contribute much, although it may have a role in patients where there remains uncertainty about background thromboembolic disease.

HIV and cardiomyopathies.

Both an HIV associated myocarditis and cardiomyopathy have been described. Again, prevalence is unclear, but it was as high as 5-10% in some series. The aetiology is often unclear, apart from the well-known but rare viral causes. HIV itself may be involved. Don’t forget the possibility of beriberi – if in any doubt give some thiamine anyway. There is also a tenuous association with amphotericin use.

HIV and pericardial effusions.

Pericardial disease has been reported to occur in between 20 and 40% of HIV positive individuals, depending on the selection criteria of the study. Locally, tuberculosis is likely to be the commonest cause, but it should be remembered that a wide variety of other organisms have been associated with the condition (staph, strep, salmonella, chlamydia, nocardia, cryptococcus, toxoplasmosis, CMV, herpes). Kaposi’s and lymphoma are also associated.

If in doubt about an alternative infectious origin organise a pericardial aspirate, unless there is other evidence to suggest Kaposi’s, as in this group pericardiocentesis may be dangerous (bleeding risk). Although steroids are often given for tuberculous pericarditis, their benefits and harms in the HIV positive individual are as unclear as in HIV negative persons.

HIV and ischaemic heart disease.

This is generally associated with the dyslipidaemia related to the long term use of the protease inhibitors, and in areas where ischaemic heart disease is rarely seen, consideration should be given to this entity in patients chronically on PIs presenting with the usual symptom complex.

HIV and rheumatic heart disease.

Because rheumatic heart disease is relative common in resource constrained countries, there is a cohort of patients who would ordinarily be candidates for valve surgery but who are HIV positive. There is an understandable reluctance on the part of cardiac surgeons to operate on HIV positive individuals, in view of the guarded medium term prognosis. When it is remembered that the mortality rate in patients with replaced heart valves is still quite considerable anyway, a case can be made for operation on all HIV positive individuals with normal nutrition and preserved or restored lymphocyte function.

HIV and blood pressure.

The metabolic syndrome associated with protease inhibitor use may have hypertension as a component. Of more concern in resource-constrained environments, however, is the frequent finding of patients with hypotension.

Cachexia is commonly associated with a low blood pressure, and the absolute value is less important than assessment of tissue perfusion – if the patient is fully alert with warm peripheries and good urine output, then the blood pressure, more or less regardless of the absolute value, is adequate. More harm is probably done by over enthusiastic attempts to ‘normalise’ a low value with saline. In hypoalbuminaemic individuals with normal hydration, such infusions often result in pulmonary oedema.

Dermatology – HIV

Overview

Many HIV positive patients present with skin complaints, or have incidental skin disorders noted on examination. The majority of these conditions are relatively benign, but there are a few clinically important ones that need to be recognised.

Itchy rashes.

An itchy skin is common. Treat for scabies with topical benzyl benzoate. If it doesn’t settle with this, it may well be itchy bump disease, more formally known as papular dermatitis of HIV. It may respond to 0.1% betamethasone topically, but don’t put this on the face – use 1% hydrocortisone. Anti-histamines are often tried – their value is unclear.

Norwegian scabies is readily recognisable once you have seen it. It is otherwise known as crusted scabies, and is fairly infectious. There is extensive crusting and itching, often largely sparing the areas which can’t be reached by scratching fingers. It may involve the head. Give 5% sulphur ointment topically for a few days, then benzyl benzoate application for 24 hours. Wash off and repeat the benzyl benzoate after three days. It usually responds quite nicely.

Drug hypersensitivity reactions

These are common. The usual culprits are TB medications, co-trimoxazole, phenytoin, carbamazepine, or, nowadays, some ARVs (especially the NNRTIs). There is clearly an inflammatory process going on in the skin, and the swelling and inflammation of the face can be quite striking. LFTs may be elevated, and there is often an eosinophilia.

Management usually involves stopping all drugs and waiting for things to settle down. If there is a genuine need to re-introduce therapy, this is sometimes possible under very careful clinical and laboratory monitoring (LFTs and eosinophil count). For mild reactions to NNRTIs and co-trimoxazole (no systemic component) it may be possible to continue treatment, but monitor carefully and stop immediately if there are any signs of progression (e.g. fever, RUQ discomfort, worsening rash, leukocytosis).

Vesicular eruptions

Those that look like molluscum contagiosum may well be that condition, but cutaneous cryptococcosis is common too – aspirate some vesicle material and send for an India Ink stain.

Herpes zoster in HIV

Herpes zoster is common in HIV-AIDS (yearly incidence 2.5%^{^[19]}). In the elderly, post-zoster pain syndromes are common and may be debilitating, but in younger HIV positive patients the prevalence of neuropathic pain secondary to shingles is less common (about 4-6% at 6 months.^{^[20]}) There is some evidence that the incidence of postherpetic neuralgia is similar in HIV negative and positive individuals^[21].

In patients presenting with uncomplicated uni-dermatomal zoster the duration of rash is shorter if anti-virals such as acyclovir or valacyclovir are given within 72 hours of onset of the rash. The odds ratio for reduction in pain at 6 months is about 0.54^[22] which will lead to an approximately 2.5% absolute reduction in risk of pain at six months in young individuals. The NNT is thus 40.

Management consists of paracetamol 1 g 6 hourly, and amitriptyline 25 mg nocte, up to 100 mg nocte. (Adequate pain relief in 47-67%, NNT 1.6^[23].) The strength of evidence for carbamazepine is weaker. Calamine lotion is a cheap sympathy item without demonstrable efficacy, and patients are often quick to point this out. Don’t give huge doses of narcotic analgesics unless they are clearly necessary (failure to respond to simpler measures).

Don’t succumb to the need to treat ‘secondary infection’ with antibiotics. This condition often becomes quite pustular but antibiotics are rarely necessary because the appearance you see may be due to the viral infection itself. If you do want to give something, povidone iodine topically is adequate, with silver sulphadiazine an alternative in those with iodine allergies.

For disseminated zoster (several non-contiguous dermatomes) it is acceptable practice to use acyclovir (or equivalent) either 800 mg 5 x/day or intravenously if the patient is very ill. The presence of zoster pneumonia is also regarded as an indication to start specific therapy intravenously, although again the evidence of efficacy is scanty.

Histoplasmosis

Other fungal infections such as histoplasmosis (and its newly described cousin Emmonsia) can present with a variety of skin manifestations ranging from a ‘chronic’ impetigo look-alike to variations more similar to cryptococcosis. Diagnosis of the non-cutaneous illness is more difficult – non-specific symptoms of fever, weight loss and malaise predominate, and the organisms crops up on blood culture, marrow culture, or occasionally other cultures. Treatment is with amphotericin followed by itraconazole, but because of dose interactions with efavirenz, it may be appropriate to change the ARV regimen and give lopinavir-ritonavir in place of efavirenz while on itraconazole. Emergomyces africanus (formerly thought to be an Emmonsia spp) infection presents very similarly and is treated the same way.

Kaposi’s sarcoma

The purple papules are readily recognised. Do not confuse with the common slate-grey buccal pigmentation seen in normal individuals. For management, see the section on oncology in HIV.

Kaposi’s mimics

A patient who develops what appears to be rapid onset Kaposi’s should be considered to have bacillary angiomatosis until disproven. The two conditions are readily differentiated histologically even though similar macroscopically. In the absence of biopsy (often because of patient reluctance) it is probably reasonable to give a trial of therapy with erythromycin in standard doses. Bacillary angiomatosis will melt away after a couple of weeks; Kaposi’s will not.

Patients with deep bruising are occasionally mistakenly considered to have Kaposi’s. While coagulation disorders may be the underlying problem, a further rarely considered differential is scurvy, which is easy to cure – look for perifollicular haemorrhages and a history of poor vitamin C intake.

Lesions due to the almost universally present post-inflammatory hyperpigmentation (benign, and present in HIV negative individuals too) lack the characteristic purplish colouration of Kaposi’s lesions and are usually only a problem for caregivers who have yet to see Kaposi’s. The lesions are not raised.

Gastrointestinal disorders in HIV

Abdominal pain in HIV

The HIV positive individual who presents with sub-acute or chronic abdominal pain is often a diagnostic dilemma, and it may be of some help to have a perspective on some of the conditions that can occur^[24],^[25]. The normal differential of an acute abdomen is applicable in HIV positive individuals, and the prognosis for recovery with surgery is usually similar to that of HIV negative patients (e.g. appendicitis, perforations). In the less acute situation, consider the following:

Salmonella typhi with its complications.
Abdominal tuberculosis (often but not always associated with ascites and retroperitoneal adenopathy. The finding of splenic hypodensities on ultrasound is strongly suggestive of this condition.)
CMV infection is often associated with vasculitis and episodic severe pain.
Herbal enema usage.
‘Pus anywhere’ – subphrenic abscess, PID-associated pelvic collections, psoas abscess.
Intra-abdominal bleeding secondary to thrombocytopenia.
Hepatitis B or C infection
Sclerosing cholangitis
Neuropathic pain – zoster without the rash yet (it appears the next day) or so called zoster sans herpe – no rash but dermatomal pain distribution.
Gastritis related to medication.
Pyelonephritis and cystitis (check the urine!)

In terms of frequency in the developing world, prevalence may vary with geography: the South African study quoted above found that 25% of 44 cases were due to disseminated tuberculosis, and about another third (14% each) was due to CMV or cryptosporidiosis.

CMV infection

This is ideally diagnosed on biopsy, or in the case of suspected CNS disease, on PCR of the CSF. Conventional serology is unhelpful, and neither pp65 antigenaemia nor serum PCR adequately reflect organ involvement. (In one study^[26], pp65 had a sensitivity of 91%, but a specificity of only 73%, falling further to 65% if CD4 < 100.) In biopsy proven GIT disease or CNS polyradiculitis with positive CSF PCR it may be possible to access ganciclovir which is given in a dose of 5mg/kg 12 hourly for 14 days. Valganciclovir is potentially also available (discuss) – maintenance dose in patients with good response 900 mg daily until CD4 > 100 on ART.

Mouth problems associated with HIV

Mouth problems usually present with pain or difficulty swallowing or both.

Candida infection is the most common cause and apart from the usual adherent whitish plaques (and beery/doughy odour) may present with ulceration, generalised oral erythema, and circum-oral dermatitis or angular cheilitis. Odynophagia (pain on swallowing) in the presence of oral candidiasis is presumptively due to the same infection progressing down the oesophagus, but requires treatment with fluconazole 200 mg daily for 14 days.

Major aphthous ulcers can be very disabling. Their aetiology is uncertain. There may be associated perineal disease. Treatment options include topical local anaesthetics, adequate analgesics and in severe instances, after careful thought about alternative diagnoses, consideration of the use of prednisone. Biopsy may be appropriate prior to this^[27].

Herpes ulcers – the clinical appearance is typical. The condition is usually self-limiting, but acyclovir 400 mg 8 hourly or valacyclovir 500 mg 12 hourly is a reasonable option if things don’t seem to be settling promptly. Adequate analgesia is important.

Kaposi’s sarcoma causes purplish, ‘bruised’ lesions of hard palate or tonsillar area, and is cosmetically unsightly but only in need of urgent treatment if it causes airway or swallowing difficulties.

Oral hairy leukoplakia, which is a cosmetically unsightly, but otherwise apparently harmless, causes white striated lesion on the sides of the tongue. Do not confuse with thrush. It resolves with ART.

Oral candidiasis

Candida develops resistance to anti-fungal agents^[28] and routine prophylactic therapy is not advised. Treatment should follow a stepped approach, starting with the cheapest agents (which in fact cure the majority of cases) and only progressing to the newer drugs if symptom relief is unsatisfactory in three or four days. Although it is gratifying to remove every vestige of plaque, this is probably not necessary if adequate attention is also given to conventional oral hygiene, nutrition and treatment of intercurrent infections.

Nystatin suspension 1-2 ml (1 ml = 100 000 units) 4 x/d swilled around mouth before swallowing. Remember it acts locally! Nystatin often compares unfavourably to other agents (e.g. clotrimazole) but this may relate to poor adherence.
If nystatin is unavailable, gentian violet is worth a try^[29] (1.5 ml of 0.5% aqueous solution twice daily – 43% response rate).
Fluconazole 200 mg daily for 14 days in refractory cases. Ketoconazole is not a good alternative. It is an enzyme inhibitor that reduces cortisol synthesis and has been described as precipitating adrenal crisis in patients with reduced reserve^[30]. Its absorption is also erratic in hypochlorhydric patients.
Very occasionally severe discomfort makes any swallowing impossible and amphotericin 0.5 mg/kg/day (e.g. 30 mg daily in 200 ml 5% dextrose) IV for 7 days may help^[31]. An alternative is intravenous fluconazole. Pain on swallowing associated with oral thrush may be due to oesophageal candida; however if there is no visible thrush or no response to systemic anti-fungal therapy (fluconazole for 3 weeks), then there may be aphthous ulceration of the oesophagus, which may respond to prednisone 10 mg daily for 5 days (aetiology unclear).

Perspective – oral candidiasis

The relative response rates of the various agents in uncomplicated oropharyngeal disease are unclear. Information is seldom available in trial reports on whether entering patients were previously exposed to the control medication. Secondly, the drive to evaluate fluconazole against everything else means that modern trials comparing the older agents are sparse.

A review^[32] which suggested that a cure rate of over 80% could be expected with fluconazole also made the point that topical agents should be tried first, especially in the light of increasing evidence that Candida species are starting to develop resistance to azoles^[33].

The use of oral amphotericin in patients resistant to fluconazole is less than dramatic – response rate of 43% at 28 days have been reported^[34], using relatively modest criteria of response (no mouth pain and less than 5 plaques).

The dose of nystatin is contentious. Recommendations vary from 1 ml to 5 ml (100 000 to 500 000 U) 4x per day. There seems little evidence that bigger is better, but the corollary is also unsubstantiated. Pastilles are probably the way to go, rather than embarking on dose minimisation studies, but while these are still expensive it is probably reasonable to aim for 1 or 2 ml doses carefully swirled in the mouth for a full minute. There is some evidence for a ‘post-antibiotic’ effect.

Oesophageal candidiasis.

This is usually treated on a history of chest discomfort on swallowing associated with oral thrush, but it should be remembered that other conditions such as aphthous ulceration of the oesophagus also present with odynophagia. If symptoms fail to settle on anti-fungal therapy, oesophagoscopy should be considered. Treat with fluconazole 200 mg daily orally for 14 days, as oesophageal candidiasis rarely settles with nystatin or amphotericin lozenges. In patients unable to swallow, intravenous fluconazole is available, but amphotericin 0.5 mg/kg/d (e.g. 20 mg) daily for 5-7 days also works. In practice, it is quite rare for patients to be totally incapable of ingesting fluconazole.

Perspective – oesophageal candidiasis

Time to response with fluconazole is poorly reported. One small study of only 18 patients found a complete symptomatic response in 39% at 5 days, 89% at 7days, and all (100%) after two weeks^[35]:

Parotid disease in HIV.

Swelling of the parotid glands and submandibular salivary glands in HIV positive individuals are sometimes symptomatic, and causation and management may be puzzling. Causes:

Infection – bacterial parotitis related to poor oral hygiene and dry mouth in late AIDs is usually associated with some degree of pain and inflammation as well as the swelling. Treat with broad spectrum antibiotics – e.g. amoxicillin and metronidazole.

Lymphoid infiltrates in a Sjogren’s-like syndrome as part of DILS (diffuse infiltrative lymphocytosis syndrome) – bilateral parotid enlargement with dry eyes and dry mouth (sicca syndrome). There is a tenuous association with polymyositis, stronger association with lymphocytic interstitial pneumonitis^[36].

Cystic lesions (often part of DILS) – if cystic on ultrasound, they are usually benign lymphoepithelial lesions and if very troublesome they may respond to low-dose radiotherapy. Keep away from surgeons.

Peri-anal pathology in HIV

Perianal pathology in HIV can be so painful that it impacts on quality of life. Consider candidiasis herpes and other STIs, but non-specific painful ulceration is common. Consider therapy for herpes infection (e.g. acyclovir or valacyclovir) or, if lesions look particularly infected looking lesions, give a trial of antibiotics including anaerobic cover (e.g. amoxicillin and metronidazole). Pus swabs are only rarely helpful. In refractory cases or if there is evidence of fistula formation, a formal biopsy may reveal treatable disease such as tuberculosis. Topical therapies for pain vary but usually include some form of local anaesthetic for application prior to defecation. These can also cause a contact dermatitis which may confuse the presentation.

Diarrhoea in HIV

The malnutrition associated with protracted diarrhoea leads to weight loss and accelerated and sometimes irreversible reduction in immune function. The problem may reflect a combination of malabsorption, lactose intolerance from well-meaning attempts at dietary supplementation, herbal medication use, and infection with a variety of pathogens which vary geographically. It is nearly always treatable, but in later stages of the disease tends to recur with increasing frequency.

A key problem in acute diarrhoea is differentiating between self-limiting disease only requiring supportive care, and a more serious septicaemic illness with diarrhoea as one of its manifestations. Blood pressure alone may not be a useful sign as many HIV patients are marasmic with long-standing ‘normal ‘low blood pressure, and the hypotension may not be due to dehydration.

Key warning signs:

Tachycardia which does not start to settle with fluid replacement
Altered sensorium
Very high or very low temperature
Slow clinical response to initial rehydration

If in doubt, rather do a blood culture and treat as septicaemic with ceftriaxone 1g IVI daily.

Take a careful history including about diet and medication.

Look for thrush and asymmetrical adenopathy (suggestive of underlying TB, although this is seldom a cause of severe diarrhoea in its own right).

Management of acute gastroenteritis

Record the initial pulse and BP, and check that the patient is improving by checking again later. Check for a postural change in BP or pulse.
Intravenous fluid replacement is usually necessary if there is associated vomiting. Choice of an initial infusion rate is empiric and depends on the patient’s weight, blood pressure, and cardiovascular and renal reserve. A starting response might be to prescribe one litre of saline every six hours. The fluid used is debated – saline or dextrose-saline is probably adequate.
Check the potassium and start to replace early. In the absence of a blood result, and where there is good urine output in a very ill patient suspected of having marked fluid loss, it may be appropriate to consider empiric low- dose potassium replacement. (e.g. 20 mmol of KCl added to each litre of saline infused over 4 to 6 hours.
Antibiotics are controversial. Very sick patients, or those who appear to be deteriorating in spite of fluid replacement, should be considered for intravenous ceftriaxone 1 g daily In the somewhat less ill with suspected colitis (bloody diarrhoea with pain in a colonic distribution) possibly due to Salmonella, ciprofloxacin 500 mg 2x/d for 7 days may be appropriate.
Give thiamine 100 mg IM daily if the patient is malnourished and you intending giving IV dextrose.

If there is a longish history (more than 2 weeks), and the patient is not too ill, send stool for MC&S concentrating particularly on microscopy. If no treatable pathogen is identified, as is often the case, repeat the MC&S and give a quinolone (ofloxacin or ciprofloxacin) for 7 days. Giardia is uncommon but is occasionally picked up, and is a justification for continuing to add metronidazole even in areas where amoebiasis is now relatively rare. There is little strong idence for any of the various empiric regimens, but one option is metronidazole 1 g/d for 3 days.

Consider gastrointestinal tuberculosis, and discuss the possibility of empirical therapy. Stool for AFBs has an unacceptably high false positive rate, so results should be treated with extreme caution, if the test is done at all.

If the above doesn’t work, then consider starting an antiperistaltic (e.g. loperamide); never start it in anyone who is febrile or acutely ill.

The finding of a treatable disease seldom rewards very detailed GIT work-up with colonoscopy/sigmoidoscopy and histology, although various pathogens may be discovered. One exception is CMV colitis, which is potentially treatable. Attention should be given to education about home nursing in the later stages of the illness and liaison with the home-based care unit should occur.

Perspective – treating the treatable.

Some organisms are modestly sensitive to specific therapy:

In Africa, non-typhi salmonella septicaemia is a common cause of morbidity, with reported frequencies ranging from 20 to 50% of all septicaemia (not only enteritis) admissions^[37].

Hepatitis and hepatic pain in HIV

Consider drugs, biliary tract disease or viral hepatitis. Do an ultrasound, liver enzymes, albumin and INR. Drug induced disease is a serious concern and in any situation where there has been a significant enzyme rise, serious consideration should be given to stopping therapy while the problem is sorted out.

Non-hepatitic RUQ pain in HIV

HIV associated cholangiopathy and acalculous cholecystitis are important considerations. An ultrasound showing biliary tract abnormalities supports the former, as does a longer and more indolent history. Presentations that are more acute should raise suspicions of cholecystitis even in the absence of gallbladder wall thickening on ultrasound. Most of the causes for this are difficult to treat except Salmonella and Campylobacter, and possibly toxoplasmosis.

HIV cholangiopathy mimics idiopathic sclerosing cholangitis, except that there is a long list of associated organisms. It is unclear whether these are causative or colonists, although a few case-reports exist of resolution after treatment of specific infections. ERCP is the best way of making the diagnosis (liver biopsy is usually unhelpful) and if it demonstrates a single ampullary stricture, sphincterotomy may help. If there are multiple strictures, it won’t. The disease tends to occur late in HIV infection and treatment is generally symptomatic.

HIV and perineal disease

Painful ulcers are often due to herpes simplex, and may respond to a course of acyclovir in a dose of 400 mg three times per day for a week.

Other possibilities are various other STDs, and an empiric course of doxycycline is popular although not often particularly effective.

Sitz baths and topical anaesthetic cream may provide some symptomatic relief.

Approach to suspected abdominal tuberculosis in HIV.

Patients present with varying combinations of fever, abdominal discomfort and weight loss with night sweats. Abdominal examination may reveal ascites and hepatosplenomegaly, or may be rather non-specific, with nothing more than mild generalised tenderness. Ultrasound is helpful – the finding of retroperitoneal adenopathy is suggestive, especially if breakdown in a node is visible. The finding of hypodensities in the spleen is highly suggestive.

Perspective – alternative diagnoses in patients with RUQ pain.

CMV is a possibility, but another less well known illness that is seen occasionally is infection with Bartonella henselae. This organism causes bacillary angiomatosis (purple-red skin nodules which can be confused with Kaposi’s) and may present with a clinically very similar abdominal symptom complex. The hallmark on histology is the demonstration of cystic blood-filled spaces, so-called peliosis hepatis, and the organism can sometimes be demonstrated with special staining. (Warthin-Starry.) Serology is also available. Treatment is 4-8 weeks of erythromycin or doxycycline, although the use of a rifampicin-isoniazid combination has also been described.

Haematological – HIV.

Thrombocytopenia associated with HIV.

The two most likely possibilities are ITP and TTP.

ITP

Relatively common as an incidental finding, occasionally a presenting illness. (It presents like ITP.)
In 95% it is due to peripheral consumption and in 5% to inadequate production, i.e. marrow failure^[38].
In the absence of evidence to suggest another marrow problem (e.g. tuberculosis) there is little point in doing a bone marrow biopsy.
If the patient is asymptomatic from a platelet perspective, and the count is over about 40 000, then no therapy is necessary.
If there is a bleeding problem or the count is dangerously low (e.g. less than 10-15 000) then give prednisone 60 mg/d; when count comes up to within a safer range, wean therapy.
If the count is >20 000 and the patient is well, watch to see direction rather than treating.
Remember that both tuberculosis and rifampicin are causes of thrombocytopenia.
HIV associated thrombocytopenia is an indication to start ARVs.

TTP (thrombotic thrombocytopenic purpura)

Thrombocytopenia with a raised LDH and evidence of haemolysis on the peripheral smear raises the possibility of TTP. See under the general haematology section for more detail. ( Haematology )

Chronic anaemias and pancytopenias associated with HIV

The most common anaemia is the anaemia of chronic disorder, a normochromic, normocytic anaemia with an Hb usually of more than 7g%. Other possibilities are:

Megaloblastic anaemias, usually nutritional (folate, B12) MCV is usually >100, but in mixed anaemias may be normal (the RDW will be increased). Both stavudine and zidovudine can cause a raised MCV.
Drug toxicity (may be a pancytopenia – zidovudine, and to a much lesser extent lamivudine, some TB meds and cotrimoxazole and amphotericin may cause anaemia)
Haemolysis
Decreased marrow production: may be an HIV associated pancytopenia of marrow replacement in tuberculosis or other opportunistic infections.
Leukoerythroblastic anaemias (anaemia with left shift – nucleated red cells, with high with left shifted white cell series)
Iron deficiency is rare and iron therapy is not indicated unless deficiency is confirmed. Ferrous sulphate is quite an important gastric irritant, and in individuals with borderline nutrition, it may genuinely do harm.
Parvovirus B19 associated red cell aplasia is likely if the PCR for this virus is positive (antibody testing is less reliable.) Treatment is centred on starting ARVs promptly. There are case reports and tiny case series suggesting that intravenous immunoglobulin (e.g. 0.5 g/kg weekly for 4 weeks) may be of value^[39], although relapse is likely unless the ARVs settle it down first.

It is rarely necessary to transfuse patients with mild chronic anaemia, and the sudden fluid load can be dangerous. The controversy around the potential for transfusion associated immunomodulation (TRIM) to have a detrimental effect^[40] (increased mortality, and increased susceptibility to infection have been mooted) remains unresolved due to the observational nature of most of the studies, the paucity of data in HIV itself, and the lack of information about whether the risk associated with older transfusion practices still apply using buffy coat depleted blood.

Approach: ask for a peripheral smear looking for megaloblastic or leukoerythroblastic features. Take a drug history, and stop any potentially offending agents, and then see if the FBC responds. Send off an LDH and haptoglobin if haemolysis is suspected (the smear often gives a clue.) If the answer is not clear from the previous steps, then consider organising a bone marrow biopsy. (The approach is similar for pancytopenias and for combined anaemias and thrombocytopenias.)

Haemolytic anaemias.

Consider this in a patient who seems to be dropping Hb particularly rapidly, and has a markedly elevated LDH and modestly elevated unconjugated bilirubin. The reticulocyte count will be elevated. The serum haptoglobin is reduced and urinary haemosiderin is often present – this latter is a relatively easily available test, although laboratories are often not that keen to do it. The Coombs test may be positive (Ask the blood bank to do it.) Many medications have been associated with it, but in HIV positive individuals consider the autoimmune haemolytic anaemias, which may have a mixture of warm and cold antibodies, and also microangiopathic haemolytic anaemia – lots of fragments and microspherocytes, associated with thrombotic thrombocytopenic purpura (TTP) and the haemolytic uraemic syndrome (HUS). In the autoimmune variety, it is reasonable to give a trial of prednisone in a dose of 1-2mg/kg as this may switch off the haemolysis quite quickly.

Hyperviscosity syndromes and vasculitides in HIV

Patients presenting with evidence of a vasculitis either have a rash suggestive of a leukocytoclastic vasculitis, which is quite often associated with medication, or they present with evidence of peripheral ischaemia – e.g. gangrene of fingers or toes. Do CRP, PTT, INR and total protein. It is also possible to measure viscosity directly. Management is usually directed at prompt initiation of ARV therapy.

Metabolic syndromes in HIV.

Diabetes and HIV.

When both diseases are common, the combination will occur relatively frequently. Thin patients rarely cope with oral agents so switch to insulin early, or even better, start with it. When due to therapy with protease inhibitors, metformin may be quite effective.

HIV and possible Addison’s disease.

Thin hyponatraemic hypotensive individuals with AIDS are common. Finding full-blown Addison’s is however relatively rare. Measured random cortisol levels tend in fact to be high, prompting consideration of relative cortisol deficiency in situations of receptor-level resistance to the actions of cortisol and its derivatives. It is, however, important always to consider hypocortisolism in HIV positive individuals, as it is relatively easy to treat once diagnosed. See the endocrinology section on Addison’s.

HIV and severe cachexia.

Severe cachexia should always prompt a search for either tuberculosis or gastroenteritis. Some individuals are simply starved due to inability to access nutrition (no money, no form of disability assistance) and can respond quite dramatically to a little food. Beware of sudden metabolic changes – e.g. intravenous fluids, or even high calorie diet without thiamine. Go slow with normal, tolerated foods and don’t try to fix everything overnight.

Neurology – HIV.

HIV associated dementia.

Gradual onset forgetfulness and mental slowing associated with motor apraxia (clumsiness) are features suggestive of HIV associated dementia (AIDS dementia complex). This is worth identifying early as it is an indication for antiretroviral therapy which may arrest or even reverse the process^[41].

Slow onset changes, over months.
Cognition difficulties (concentration problems, forgetfulness). The usual mini-mental tests screen for the predominantly cortical lesions of Alzheimer’s disease and are not ideal, so alternative screening tools which also evaluate subcortical function have been proposed.
Minor motor apraxias (reduced speed/fluidity when tapping the thumb and index finger together).
(There is a variant (vacuolar myelopathy) that presents with spastic gait before the mental changes, and is associated with absent ankle jerks and brisk knee jerks – do a serum B12 in all such cases, and if in doubt treat as subacute combined degeneration while awaiting the B12 result.)
The CSF may show a few lymphocytes and a slightly elevated protein.

Perspective – screening for HIV associated dementia

The adapted table illustrates a system which has had some limited cross-cultural validation^[42] and is reported to have a sensitivity of 80% and specificity 57%, LR+ 1.86, LR- 0.35.

Therapy of HIV associated dementia is directed towards securing anti-retrovirals, but these are probably inappropriate for those individuals who have progressed to a stage where there is major loss of intellectual capacity although there are anecdotes of response even at this stage.

CNS penetration of the various ARVs varies – stavudine, zidovudine and indinavir seem adequate, but lamivudine penetrates less well. These theoretical considerations have not yet been shown to be clinically important.

Chronic headache in HIV positive individuals

Important treatable causes for headache in patients with HIV infection:

chronic meningitides – tuberculosis and cryptococcal meningitis, syphilis.

space-occupying lesions – tuberculomata and toxoplasmosis.

It is important to remember that many such patients present with relatively insidious features of mild headache and malaise, rather than the severe headache and prostration that tends to accompany pyogenic meningitis in the non-immunocompromised individual. For instance, only 50% of patients with toxoplasmosis are febrile.

The absence of neck stiffness does not exclude meningitis.

Investigation consists of excluding a space-occupying lesion clinically (conscious, no focal signs and no papilloedema) and then doing a lumbar puncture.

If you can’t exclude raised intra-cranial pressure clinically, then consider a CT scan.

The CT may not be able to differentiate between toxoplasmosis and tuberculosis, but toxoplasma serology and the CD4 count may be helpful. (Toxoplasmosis becomes more likely the lower the CD4 count, especially if less than 100, and tuberculosis is more likely with higher counts. On CT, toxoplasmosis classically shows multiple lesions, but in one third of cases the lesion can be solitary.

The CSF may have completely normal cell count and chemistry in the presence of active cryptococcosis.

Management:

Cryptococcal meningitis – see separate entry.

Tuberculous meningitis – standard 4 drug therapy (INH, rifampicin, PZA and ethambutol) in conventional doses for 3 months, then INH/rifampicin for a further 6 months. Don’t forget to send CSF for TB culture.

Toxoplasmosis – see separate entry.

Cryptococcal meningitis in HIV

This is a common illness in HIV infected patients and warrants prompt early management in order to improve outcomes. The core intervention is initial treatment with Amphotericin and flucytosine, followed by maintenance fluconazole, but then getting the patient onto a stable effective ARV regimen. In the absence of the latter, mortality at one year will still be about 40%.

.Cryptococcal meningitis -background

The diagnosis is based on a high clinical suspicion in any HIV positive patient with headache or behavioural changes. Stiff neck is not common, and its absence certainly does not exclude the diagnosis. If in doubt, do the lumbar puncture, where mild changes are common – slightly elevated protein, a few lymphocytes, sometimes even a normal glucose.

However diagnosis is often now based on positive reflex cryptococcal antigen testing (CrAg) in the laboratory in the face of a new HIV diagnosis with low CD4 count (<200 cells/microL); these patients are often asymptomatic, but should still be considered for lumbar puncture as finding CSF involvement means treatment should be more intensive than using fluconazole, which is the management for an isolated positive serum antigen in the absence of symptoms or CSF disease.

An isolated 6th nerve palsy as a feature suggestive of raised intracranial pressure may occur. Neck stiffness is rare (only found in 10% to 20%.) CSF usually shows a lymphocytic predominance although occasionally there can be very few or no cells. Sometimes the CSF protein is also normal, and in some series, there may be no CSF changes at all in as many as 10% of patients.¹ The lateral flow CrAg has high sensitivity and specificity (99%) and is the diagnostic test of choice. Occasional false negatives have been reported² With some of these ascribed to the ‘postzone’ phonomenon (excess antigen) rather than the more familiar ‘prozone’ (excess antibody) causing the false negative.

Low positive CrAg titres can persist weeks to months after treatment. The CRP is too insensitive to have any role in excluding the diagnosis.

It is not uncommon for patients with cryptococcal meningitis to have microbiologically proven concurrent tuberculosis, but remember that the chest X-ray appearance of pulmonary cryptococcosis may be mistaken for tuberculosis.

Disseminated cutaneous cryptococcosis looks like disseminated molluscum contagiosum, but the diagnosis can be made simply by aspiration of vesicular fluid and examination with India ink.

Managing cryptococcal antigenaemia is worthwhile. Patients treated with fluconazole had a 5% lower absolute mortality (18% versus 13%) in the REMSTART trial.³

Treatment of positive CrAg if normal CSF

Induction

Fluconazole 1200 mg daily for two weeks

Consolidation

Fluconazole 800 mg daily for 8 weeks

Maintenance

Fluconazole 200 mg daily for a year.

Treating cryptococcal meningitis

Induction therapy

Preferred: In the first week give amphotericin 1 mg/kg/day plus flucytosine 100mg/kg/d (in 4 divided doses of 25 mg/kg/dose) also for 7 days. In the second week give fluconazole 1200mg per day.

Consolidation therapy

Fluconazole 800 mg day for 8 weeks.

Maintenance therapy

Fluconazole 200 mg per day for 12 months, and stop after that if viral laod suppressed and CD4 > 200 cell/microL

This regimen gives best outcomes⁴. If flucytosine is not available, then two weeks of amphotericin and fluconazole 1200mg per day is the default next option. If amphotericin is unavailable or cannot be used, then fluconazole 1200mg and flucytosine for 2 weeks is a third regimen to consider. Fluconazole monotherapy is clearly suboptimal.

If already on an ARV regimen including tenofovir, swap this agent to either zidovudine (normal Hb) or abacavir while on amphotericin.

Patients with refractory headaches or visual changes in the early part of therapy usually respond to daily serial lumbar punctures with CSF drainage till pressure normalizes – this is an essential component of treatment in such patients.

CSF pressure management: Cryptococcal meningitis – CSF pressure

.Cryptococcal disease – digging deeper

HIV and the slightly abnormal CSF that is India Ink negative

Most lumbar punctures are done because of symptoms, usually of headache, although some may be done in the work-up of individuals with unexplained pyrexia or CT-demonstrated granulomatous CNS disease. The clinical context will help to determine the response to equivocal CNS findings.

A mildly abnormal CSF, with a few lymphocytes and a slightly raised protein, may be due to HIV infection itself. The exact cut-offs for ‘few’ and ‘slightly’ are poorly defined.

India Ink examination is operator dependent, and particularly at night when technicians may cross-cover to areas with which they are less familiar, false negatives can be obtained with both this and the cryptococcal serology. If in doubt, ask nicely for a review in the morning or after the weekend. If still in doubt, a repeat LP may be worthwhile.

Tuberculous meningitis usually presents at an earlier stage of HIV infection, so there is usually a brisker lymphocyte response in the CSF. If in doubt, send the fluid for TB culture. If you need to make a decision about TB treatment you may not be able to wait for a culture result, in which case the decision will be made in the context of other evidence (e.g. mediastinal adenopathy) and the degree of illness of the patient.

Don’t forget to do the VDRL or TPHA.

Very rarely, CNS lymphoma may be suspected with such CSF results. It is reasonable to send the fluid for cytology, although interpretation in such circumstances is often difficult. CT may be of help.

In a sick patient where you don’t seem to be making any progress, it is perfectly reasonable to repeat the LP after a few days.

HIV and seizure disorders

The differential of fits in HIV positive individuals includes all conditions seen in HIV negative individuals, but also includes those conditions found commonly in HIV, such as the chronic lymphocytic meningitides, toxoplasmosis, and lymphoma. Very sick febrile patients, particularly with metabolic imbalances such as hyponatraemia or hypoglycaemia may present with fits, which in some instances may be focal.

In general, management of new onset fits in HIV positive individuals is the same as in HIV negative persons, except that the probability of finding remediable focal pathology on CT is higher. Therapy with phenytoin or carbamazepine is still a reasonable choice once treatable causes have been eliminated. However if therapy with ART is being planned in the near future, the agent of choice is either valproate or lamotrigine. The enzyme induction of the other agents may make the serum levels of r some of the ART agents subtherapeutic.

Suspected meningitis in HIV positive patients with focal signs: can the LP wait?

In patients who would otherwise be considered for LP, the presence of focal neurological signs should delay the process until a CT scan can be performed to assess safety. In the absence of prompt access to a CT scanner, however, it is still sometimes necessary to make a value judgement about whether or not to do an LP or to treat empirically. Prior to the HIV era, treating empirically for bacterial meningitis was a reasonable strategy in this situation, but currently the differential is a lot broader, with particular concerns about excessive delays in initiating treatment for cryptococcal meningitis.

Under such circumstances, consider the following:

Is the neurological deficit ‘fresh’? An asymmetrical face where the weakness came on abruptly a year ago is probably an unresolved Bell’s and is thus not a contra-indication to lumbar puncture; drowsiness and a hemiplegia present for a few hours clearly is.

Will the LP result affect management acutely? A well individual with headache for months and a deficit for days can probably have LP deferred until after a CT.

Even after considering the above, there will still be a small number of very ill patients in whom an LP should definitely not be done before CT, but in whom waiting for the CT without treatment is expected to have serious adverse consequences. In such circumstances treating empirically with a third generation cephalosporin, anti-tuberculous therapy, and dexamethasone may be appropriate. You should be prepared for considerable diagnostic difficulty if the patient does survive.

HIV-associated focal brain lesions

Most developed world management strategies for focal brain lesions centre on the use of a therapeutic trial of treatment for toxoplasmosis in the setting of focal lesions identified in the brain parenchyma on CT scan. In less affluent areas, a number of issues arise:

Is the prevalence of toxoplasmosis similar?

Are there other lesions that are more common?

Is a ‘therapeutic trial’ with co-trimoxazole monotherapy always adequate?

Has the developed world model been adequately validated in local settings?

The answer to all these questions is probably no.

A modified version of the ‘treat as toxoplasmosis first’ approach is to assume it is toxoplasmosis unless there are features to suggest otherwise:

The presence of associated basal meningeal contrast enhancement is suggestive of tuberculosis, but is only present in 72%^[50]. (This small short duration local study with diagnosis confirmed only by clinical response suggested that 53% of patients had tuberculosis alone, 16% had tuberculosis plus another disease, and 19% had neurocysticercosis. Only 6% were thought to have toxoplasmosis.)

TB elsewhere suggests TB in the head as the cause of the lesion. (But remember that severely immunosuppressed patients can have more than one illness at once.)

The ‘classical’ tuberculous lesion is cortically situated, has a thick irregular wall with rim enhancement and a hypodense centre.

The single (TB) versus multiple (toxoplasmosis) criterion is not always reliable.

Toxoplasma serology showing a high IgG titre is suggestive of that disorder. (The IgM is usually negative as this is unlikely to be a new infection in that individual.)

Consider the other differentials – primary CNS lymphoma, progressive multifocal leukoencephalopathy, neurocysticercosis, cryptococcosis (cryptococcoma), and syphilis (gumma). Infarcts can also be a bit confusing at times, particularly when there is increased contrast uptake at the edges (luxury perfusion).

Perspective – referral bias

Studies of focal brain lesions in HIV positive persons done by neurosurgical units may under-represent patients with tuberculosis, perhaps because patients with lumps in their heads and tuberculosis elsewhere tend to be treated by their initial care-givers as having only TB, and are thus not referred on for investigation. The referred pool might be considered to contain only those patients where there isn’t another diagnosis immediately available, and there is a hope that the neurosurgeon might be persuaded to do a biopsy.

HIV and stroke

In the absence of a clear history of abrupt onset of lateralising neurology it is very difficult to tell the difference between a true cerebrovascular accident and many other focal HIV-associated CNS pathologies (e.g. tuberculoma, toxoplasmosis granuloma, progressive multifocal leukoencephalopathy). Many of the latter are treatable, and so it is advisable to organise a CT scan for all such patients. If the history of abrupt onset is clear, then it is more likely to be vascular in origin, but a CT is often still necessary as the cause may still be an infectious vasculitis, requiring lumbar puncture for its diagnosis. In the presence of focal signs, LP should be preceded by CT scan to check that the procedure can be performed safely.

(In a stroke unit in a referral hospital in Cape Town^[51], the proportion of the aetiologies was roughly one fifth each: cardioembolic 14%, CNS infection 28%, coagulopathy 19%, HIV vasculopathy 20%, and unknown 19%. The prevalence of infection in less selected series is likely to be higher, because patients with obvious infection elsewhere are less likely to be referred. HIV vasculopathy manifests either as aneurysms or stenoses. Extracranial disease may occur with higher CD4 counts and is postulated to involve a leukocytoclastic vasculitis of vasa vasorum, whereas intracranial disease is more focussed on the intima, perhaps a cytokine mediated or direct HIV effect, and occurs in individuals with generally lower CD4 counts. )

HIV and cranial neuropathies

The commonest cranial neuropathy is an isolated Bell’s palsy. Occasionally it is bilateral, which can be a bit tricky, but watch the eyes while asking the patient to attempt to close them – if the eyes roll upwards (Bell’s phenomenon) then bilateral lower motor neurone pathology is likely.

Prognosis and management of isolated LMN VIIth lesions is similar to that in HIV negative individuals – early acyclovir and corticosteroids are suggested in HIV negative individuals on the assumption that the illness is due to a herpes 8 virus. Evidence of major clinical benefit from this strategy in HIV positive individuals is scanty.

Other cranial neuropathies occur and can be puzzling until thought of as single entities rather than features of major intracranial pathology.

HIV and paraplegias and pseudo-paraplegias

Much of this is covered in the following section on HIV and peripheral neuropathy – patients who can’t walk and have a peripheral neuropathic process are often unfortunately labelled as having a ‘permanent’ cord lesion even in the face of reduced reflexes and tone in the legs – this is arguably a ‘pseudo-paraplegia’ and more careful examination usually gives an indication of the true state of affairs. Zidovudine myopathy should also be considered in patients on this drug – a creatine kinase is helpful.

True cord lesions are not that common, other than transverse myelitis associated with syphilis, toxoplasmosis or CMV and tuberculosis. A myelopathy as part of the initial acute HIV infection has been described, and may reverse with time.

HIV and peripheral neuropathy

This is unfortunately an area where the taxonomy^[52] is still way ahead of the treatment; however it is still worthwhile trying to make a thoughtful diagnosis.

AIDS-associated neuropathy

This is a symmetrical distal sensory polyneuropathy (DSPN). Prevalence is influenced by patient selection and how hard one looks. In a study from Greece incorporating electrophysiological testing, it was found in 36% of patients, but was only clinically apparent in 13%^[53]. It presents with slow onset bilateral neuropathic foot pain. The earliest objective signs are reduced or absent ankle jerks and reduced vibration sense. It presents relatively late in the course of HIV infection, and it is important to look for it in patients being considered for anti-retroviral therapy, as documentation of its presence prior to commencing treatment simplifies the differentiation from drug-associated neuropathy. It can also appear to worsen on starting ARVs, or the combination of TB treatment and ARVs.

Antiretroviral neurotoxicity (‘nucleoside neuropathy’)

Stavudine and didanosine are the main culprits, and clinically it looks very like AIDS-associated neuropathy except that pain is very prominent and progression may be more rapid. It usually improves within a few months of withdrawing the drug, but the decision is always about how severe the neuropathy needs to be before a regimen is changed. In general, a progressing neuropathy of sufficient severity to interfere with the patient’s life warrants swapping to another agent.

Perspective – distinguishing DSPN and nucleoside neuropathy

A small study^[54] (30 patients) found that a lactate of >2.2 mmol/l had a sensitivity and specificity of 90% (LR+ 9, LR- 0.11) for identifying patients with nucleoside neuropathy:

With such a small sample size, the confidence interval for a specificity of 0.9 was 0.6 to 0.99, so the result should be viewed with some circumspection, but it may be worth considering in puzzling patients.

CMV polyradiculopathy

This is relatively easy to recognise. It presents with rapid onset back discomfort and leg and back pain, often associated with pain in the perineum and inner thighs. There may be sphincter involvement, and it often progresses to a flaccid paraparesis. CSF examination shows polymorph predominance. There are case reports of good, although slow (3 months) response to a combination of ART and ganciclovir^[55]. There is little evidence on the use of ART alone. In very late HIV, CMV may also present with a mononeuritis multiplex type picture.

Nutritional neuropathies

See under the sections on the individual vitamins, but consider Vitamin B12, niacin (pellagra) and pyridoxine (especially if the patient is on INH). Neuropathic beriberi (thiamine deficiency) also manifests as a sensory neuropathy, although the motor weakness can be quite marked in some patients. It is usually difficult to dissect out which vitamin is the main culprit, and it is often more than one. The main concern is that all of these conditions are potentially completely reversible if recognised and adequately treated, but the homeopathic doses of vitamins in many of the multivitamin tablets available are considerably less than conventionally recommended replacement doses, although they may be adequate for prophylaxis.

Give Vitamin B12 1 mg IM and thiamine 100 mg IM for a few days, and only stop weekly B12 once you get back the pre-treatment B12 result showing that the patient never had the deficiency in the first place. B12 neuropathy can progress very rapidly in deficient patients given adequate calories but no B12, so err on the side of caution.

Thiamine can be given as 100 mg orally daily. Niacin is often difficult to access, but is particularly important if there is evidence of dermatitis or dementia – an alternative is to use regular Vitamin B Complex 4 tabs 3x per day. Pyridoxine is usually easy to access, and 100 mg/d may induce a very gratifying response within a month or so. (High dose pyridoxine – e.g. 2g/d – is reported to cause a peripheral neuropathy…)

Vasculitic neuropathy

This progresses in a stepwise fashion, and pain usually precedes the motor or sensory deficit. Check the nail fold capillaries in patients presenting with mononeuritis multiplex-like lesions. (Mononeuritis multiplex is just involvement of single nerves one after the other in different parts of the body.) Also look for evidence of renal involvement (haematuria, raised creatinine). If this is a serious diagnostic consideration, a sural nerve biopsy may show evidence of vasculitis^[56]. The condition responds relatively well to prednisone 1 mg/kg.

Perspective – nailfold capillaries.

The capillaries at the base of the nail sometimes give a clue to the presence of a vasculitis^[57]. They normally form a palisade of equal loops, but in the presence of a vasculitic process there may be missing loops or tufting or even haemorrhages. To see them, look where the proximal part of the nail emerges from under the skin – this is the nailfold. Use an ophthalmoscope turned up to highest magnification. Some clinicians believe that a drop of lubricating jelly placed on the nailfold improves visibility, but it often seems to make no difference.

Herpes zoster neuropathy

The rash usually gives this away, but in the absence of rash the characteristic dermatomal distribution of pain is very suggestive.

Acute inflammatory demyelinating polyneuropathy (Guillain-Barre)

Behaves similarly to GB in the non-HIV positive individual, and is variably reported as occurring relatively early in the HIV process. Unlike the situation in the HIV negative individual, a mild CSF pleocytosis (lymphocytes) is usually seen and does not preclude the diagnosis. As in HIV negative individuals, corticosteroids are probably of no benefit^[58], and the benefits of intravenous immunoglobulin are expected to be similar to that in HIV negative individuals (statistically significant but clinically relatively modest improvement at considerable cost).Acute GB occurring as part of the immune reconstitution syndrome is now relatively well described – it occurs early after starting therapy and should not be confused with a medication induced painful sensory neuropathy. (In other words, don’t be tempted to stop the antiretrovirals!)

Acute motor-sensory axonal neuropathy presents like an acute GB but with more obvious sensory changes.

Chronic inflammatory demyelinating polyneuropathy

This occurs relatively early in HIV infection (CD4 counts often greater than 300). The onset may be more gradual, with loss of all or most deep tendon reflexes; if there is marked distal painful sensory changes you should think of an alternative diagnosis. Both a few cells and a raised protein are often seen. It is conventionally described as being responsive to both immunoglobulins and corticosteroids. The latter is of considerable interest if true, but unfortunately the controlled trial evidence of benefit can best be described as modest^[59]. In patients without any sensory changes at all, and particularly if the motor changes are asymmetrical, a label of multifocal motor neuropathy may be more appropriate, in which situation corticosteroids may possibly lead to deterioration^[60]. All else being equal it is probably reasonable to try prednisone 1 mg/kg/day in patients with symmetrical motor changes and at least minor sensory changes. (Intravenous immunoglobulins do provide benefit, but this is usually temporary and often requires repeat dosing every few months^[61], which ends up being very pricey.)

Miscellaneous causes of distal weakness in HIV.

Syphilitic myeloradiculitis

Don’t forget to do a VDRL on both blood and CSF in any patient presenting with neurological signs – cord ‘vasculitis’ with radicular sensory symptoms should make you particularly suspicious of this entity. Response to therapy is sometimes very gratifying and remarkably complete.

Tuberculous myeloradiculitis

Patients on treatment for tuberculosis elsewhere can present with relatively acute onset weakness of the legs with features suggestive of a cord lesion but with sometimes quite marked sensory changes in a radicular distribution – e.g. severe pain round the waist or going down the back of the legs. Just like cerebral tuberculomata ‘flaring’ on TB treatment as an inflammatory response starts to occur, these spinal lesions may do the same^[62]. Diagnosis is with myelography or contrast CT/MRI of the cord. Treatment is to continue the antituberculous therapy and add prednisone 1 mg/kg orally.

Racemose spinal neurocysticercosis

In areas of high cysticercosis prevalence the two conditions may co-exist, and in the setting of reduced cellular immunity, there may be a higher incidence of the racemose variety. Spinal versions of this entity can present as a cauda equina syndrome^[63].

Critical illness neuropathy

In patients who have survived severe sepsis and remain bedridden for some time, consider this somewhat ill-defined illness^[64].

Progressive multifocal leukoencephalopathy

This condition can have a diverse presentation, and is due to infection with a parvovirus called JC virus (not to be confused with the Jakob-Creutzfeldt slow virus infection which has the same initials but causes a different condition).

It presents with a constellation of physical findings, but hemiparesis, speech problems and cognition problems are common. There may be visual defects, incoordination and sensory deficits. Parkinsonian rigidity and cerebellar disorders are also rarely described. The mental changes range from mild reductions in attention span or memory, to personality changes and even overt dementia. The CD4 count is commonly (but not always) low – e.g. <100.

CT scan shows characteristic white matter changes – scalloped hypodense areas in the parieto-occipital region are very suggestive. There is no mass effect. The diagnosis can be confirmed by doing PCR for JC virus on a CSF sample. Sensitivity is about 90% in patients not on ART, but may be lower (60%) in those already on treatment^[65]. In HIV positive patients with very low CD4 counts, sensitivity approaches 100%.^[66]

Case-reports and small case series are starting to appear suggesting that ART may modify the disease progression, but there are also reports of IRIS-associated flares on treatment.

HIV associated acute disseminated encephalomyelitis (ADEM)

This is a rare monophasic inflammatory demyelinating disease^[67] which also occurs in HIV negative individuals. It is characterised by abrupt presentation, behavioural symptoms, altered consciousness, and new and fluctuating symptoms during the initial illness, but unlike multiple sclerosis, it is a once-off (hence ‘monophasic’) condition^[68]. Optic neuritis may occur, and acute transverse myelitis is well described as part of the presentation. There is usually a CSF pleocytosis (without oligoclonal bands) and neuroimaging shows white matter lesions which are usually multiple. Its clinical relevance is that if recognised (and distinguished from PML, there may be some response to prednisone 1 mg/kg.)

This disorder should not be confused with anti-NMDA receptor encephalitis^[69], a rare autoimmune disease in children and young adults, presenting with psychiatric symptoms and often orofacial dyskinesias.)

HIV seroconversion disorder.

This is often puzzling at presentation. Most individuals who contract HIV are either asymptomatic or develop a relatively minor flu-like illness. Those patients who are sick enough to come to hospital are often quite ill, usually with features of an acute encephalitic illness – fever, altered sensorium or behaviour, and little to suggest an alternative cause for the fever. There may be quite marked tender adenopathy, and even a pharyngeal exudate, raising suspicions of infectious mononucleosis. The condition usually resolves within a few weeks.

Importantly, the HIV ELISA may well be negative, and remain so for several weeks. Be careful of labelling such individuals as HIV negative. If in doubt, a PCR may be appropriate.

Toxoplasmosis.

The risk of developing toxoplasmosis is a function of the level of immunity – the yearly risk in one study was 33% for patients with a CD4 count below 50 compared with 8% for those with a count above 150^[70]. Locally, the disease appears less common than this, but this may be because patients succumb to other infections first.

Features: Patients present with fever (40-50%), headaches (50%) and confusion (35-50%). Focal neurological signs will be present in 70-90%, perhaps reflecting selection bias, with the most common (50%) being a hemiparesis. Other possibilities are fits, cranial neuropathies, cerebellar and extra-pyramidal signs, and speech problems.

Serology: IgM antibodies are rarely detected using standard ELISA methods, and assays that are more sensitive may give false positives. IgG serology is usually positive (>95%) but the titre is highly variable. Reports on the use of PCR on CSF show vast ranges in sensitivity, so it is not particularly useful. Alternative diagnoses should be considered if the IgG is negative or if the patient has been on co-trimoxazole prophylaxis.

Radiology: The classic radiological description is of multiple bilateral ring enhancing lesions on CT, often involving the basal ganglia or occurring at the junction of the cortex and medulla. Up to 40% of patients with CNS lymphoma have radiologically indistinguishable findings. Although pulmonary involvement in patients presenting with neurological disease is quite rare, when it does occur, the chest X-ray findings often look like pneumocystis, although effusions and adenopathy may occur.

Treatment: Conventionally, therapy is with pyrimethamine 50 mg daily and sulphadiazine 2g 3x/day. Sulphadiazine is difficult to source, and bone marrow suppression secondary to pyrimethamine necessitates the use of (very expensive) folinic acid (10-50 mg/day). An alternative is co-trimoxazole in a dose of 4 tabs 12 hourly^[71]. Therapy should continue for a full 4 weeks even if the patient is feeling well, and thereafter continue co-trimoxazole 2 tabs 12 hourly for 3 months. Subsequent to that, secondary prophylaxis, also with co-trimoxazole, should continue at a dose of 2 tabs daily for life.

In patients convincingly allergic to co-trimoxazole, evidence of efficacy of other regimens is weak. As a general principle, therapy should be with at least two agents.

A common recommendation is for clindamycin 600 mg 8 hourly for four weeks in combination with pyrimethamine (100 mg on the first day, then 25 mg per day for four weeks). Other agents which have been used include dapsone 100 mg/day, pyrimethamine-sulphadoxine, doxycycline and azithromycin. Oddly enough, a combination which is relatively cheap and easily available (clindamycin and dapsone) has not been formally tested.

Perspective – prognosis in toxoplasmosis

The study by Torre referred to above was stated to be a pilot study only – patients either received co-trimoxazole or a combination of pyrimethamine (50 mg/d) and sulphadiazine (60 mg/kg/day.) Diagnosis was on ‘clinical’ and CT grounds only – neither serology nor biopsy was used in setting inclusion criteria. There were 37 patients in the pyrimethamine-sulphadiazine group and 40 in the cotrimoxazole group, and dexamethasone was given to 89 and 88% respectively. At 30 days, there was a complete clinical response in 66% of the former (complete CT response in 39%) and in the cotrimoxazole group the 30 day complete clinical response was 62% (CT response also 62%.)

The study is unfortunately underpowered to allow one to conclude that the two therapies are equivalent, and is confounded by the use of steroids that may have had differential effects on the CSF penetration of the two regimens (reinforced by the differences between clinical and radiological responses).

Psychiatry – HIV

HIV and sedation

AIDS patients often present with either confusion or psychosis, and in such states are a risk to both themselves and others. Conventional sedation should be undertaken, bearing in mind the usual principle that it is easier and safer to keep a patient mildly drowsy by regular dosing than to give intermittent therapy only once symptoms recur.

Special caveats in AIDS patients are that they are often thin and malnourished with a low serum albumin, so doses may have to be adjusted downwards. As a corollary, an ineffectively sedated patient wandering around the ward and biting both patients and staff will not endear you to anyone.

Check the blood glucose and correct it if it is low. Consider septicaemia.

Give thiamine 100 mg IMI

Sedate with a single agent e.g. a butyrophenone such as haloperidol in a dose of 5 mg IMI if still of average weight, but half this if malnourished. If it hasn’t worked in 15-30 minutes, repeat the same dose.

HIV and acute confusional states

Consider:

Acute infections – either meningitis or septicaemic processes. Patients are usually but not always febrile. Another possibility is HIV seroconversion disorder, which sometimes presents with an encephalitic illness; however the serology in these patients is initially negative.
Metabolic disorders – uraemia and hyponatraemia, and very occasionally hypoglycaemia.
Nutritional disorders – don’t forget about pellagra and Wernicke’s. If in doubt, give both nicotinamide and thiamine.
Drug reactions – some unexpected medications are listed as causing confusion – e.g. the quinolones such as ofloxacin or ciprofloxacin.
Conversion disorders – soon after being told that HIV positive, sometimes even with very good counselling…
Rarely, AIDS dementia complex, although on history there is usually something to suggest the process was actually more chronic.
Co-existence of genuine psychopathology in a person who just happens to be HIV positive.
Substance abuse.

Depression in HIV positive individuals

Apart from greater difficulty in recognising this, diagnosis and management is the same as in HIV negative individuals. A number of the somatic complaints (e.g. headache) may signify something more serious, such as cryptococcal meningitis.

Treatment should be with a tricyclic antidepressant such as amitriptyline, starting with 25 mg nocte and increasing slowly to about 100 mg per day if tolerated (sedation and dry mouth are common). Be careful of starting SSRI without checking for drug interactions with ARVs.

Renal – HIV

Acute renal failure in HIV

The causes of acute renal failure in HIV positive individuals are broadly like those in HIV negative persons; the relative frequencies are somewhat different. Causes in one study^[72] were sepsis (52%), nephrotoxins (23%) and miscellaneous (25%). In that study early mortality was 60% and mortality at 3 months was 71%, probably reflecting the severity of infections and immunosuppression. Consider^[73]:

Acute tubular necrosis – hypovolaemia and hypotension associated with sepsis and diarrhoea.

Interstitial nephritis – medications such as rifampicin, co-trimoxazole, quinolones and the penicillins and cephalosporins, anticonvulsants such as phenytoin and carbamazepine, diuretics, and cimetidine. Also may be due to a variety of infections, including syphilis, tuberculosis, and toxoplasmosis. There can be a more indolent Fanconi syndrome picture (NRTIs such as didanosine, lamivudine and stavudine) Tenofovir also induces acute renal failure with proximal tubular dysfunction.

Acute glomerulonephritis, including rapidly progressive GN – same as HIV negative individuals.

Vascular nephropathy syndromes – thrombotic microangiopathies. Haemolytic-uraemic syndrome can be associated with quite brisk haemolysis.

Chronic renal failure in HIV

Causes include focal glomerulosclerosis, mesangial hyperplasia, minimal change GN, and a few rare cases of proliferative GN. Focal glomerulosclerosis is now another disease that can be added to the differential diagnosis of nephrotic syndrome with large kidneys on ultrasound. (The other two are amyloidosis and diabetes)

Nephrotic syndrome in HIV

The most common causes for nephrotic syndrome in HIV positive individuals are focal segmental glomerulosclerosis, and the more recently described HIV-associated immune complex disease, for both of which there is no clearly effective treatment.

There is some evidence that ART may slow or arrest kidney disease progression, and ACE inhibition is probably appropriate to slow the rate of proteinuria progression. In the absence of ART, it is unclear whether prednisone 1 mg/kg contributes anything much, although a couple of small cohort studies have reported benefit.

Perspective – renal histology in HIV nephropathies

Most studies are retrospective biopsy cohorts suffering from selection bias (who gets referred to the nephrologists, and who do the nephrologists decide to biopsy?) A local study of this nature^[74] revealed HIVAN (HIV associated nephropathy, a form of FSGS) accounted for 27% of a series of 99 biopsies, and the next most common histology was HIVICK (HIV immune complex kidney disease) at 21%. Other histologies found in HIV negative individuals were also encountered (membranous, post-infectious, IgA nephropathy.) There is little information on therapy for HIVICK, which is histologically similar to some pathologies found in SLE, but a small study^[75] from well into the ART era found that 10 out of 14 patients had progressed to end stage renal disease within one year.

Malignancies in HIV.

Kaposi’s sarcoma.

Therapy is usually not indicated for single or grouped skin lesions. Single lesions in life-threatening sites (e.g. causing upper airways obstruction) may respond to radiotherapy. Radiotherapy to a limb may sometimes lead to severe lymphoedema.

The main indication for medical therapy is for patients on ART with visceral involvement (lung, GIT) or severe skin disease. ARVs should be started before chemotherapy, which on its own is unlikely to prolong life. Therapy should ideally be given by, or in consultation with, an established oncology unit.

Vincristine 1.5 – 2 mg (1 mg/m²) SC or slowly IV fortnightly for 2 doses.
Bleomycin 10 U/m² (max 15 units) slowly IV fortnightly for 2 doses.

If there is a good response, a case can be made for continuing therapy. If there is no response after a month, do not give more treatment with these agents and concentrate on other forms of palliative care.

Perspective – efficacy of therapy for Kaposi’s

In the absence of good quality anti-retroviral therapy, treatment of Kaposi’s is rarely very successful in the medium to long term. Most regimens are based on information from short term studies which report an ‘initial response rate’ varying rather disconcertingly from 20% to 100% with a range of different agents. An anthracycline is sometimes added to the regimen but the evidence of clinically important additive benefit is unconvincing.

Lymphoma in HIV

Lymph nodes that are AFB negative on aspiration should be re-aspirated. If still negative, then a formal excision biopsy is usually appropriate unless there is a clear aetiology such as disseminated Kaposi’s. Excision means the whole node – removing a random sliver may make the pathologist’s job impossible. Lymphoma in HIV positive individuals may present in somewhat atypical ways as well – e.g. rectal or biliary involvement, or even as a primary effusion lymphoma where there is disease involving serous cavity linings (pleura, pericardium or peritoneum) without any adenopathy.

Prognosis

HIV positive individuals not on ART experience increased toxicity (relative to HIV negative individuals) when treated aggressively for lymphoma. There is some evidence that reduced dose chemotherapy regimens may achieve similar outcomes to more aggressive regimens, but even so, median survival is 8-9 months^[76].

It is still not entirely clear what the influence of ART on therapeutic response is likely to be in the long term, but if ART is going to be offered and the patient has lymphoma, it is worth discussing with an oncology unit.

Respiratory – HIV.

Pneumocystis pneumonia.

(The Pneumocystis carinii versus jiroveci debate continues.)

Studies from the 1980s found that patients with CD4 counts <200 had a yearly risk of developing pneumocystis of 18%^[77], but this is in an area where pneumocystis was the AIDS-defining infection in 60% of patients and affected 80% at some stage^[78].

Consider the diagnosis in a young, relatively healthy person with a sub-acute (days to weeks) pneumonia. Remember that the history is not all that reliable and it may present as an apparently catastrophically acute event. Pleuritic chest pain is unusual (<30%) although central discomfort (‘can’t take a full breath’) isn’t.

Early hypoxaemia is very suggestive (cyanosis or decreased oxygen saturation on exertion – pulse oximetry is of some value).

There may be bilateral fine inspiratory crackles, or even no auscultatory signs.

There should not be adenopathy on the CXR. Disease is usually diffuse and bilateral. Pleural effusions are very uncommon and should lead one to think of an alternative diagnosis.

Direct microscopy of induced sputum^[79],^[80],^[81] allows the diagnosis to be made rapidly although it is often necessary to treat on ‘spec’ initially.

Treat with co-trimoxazole 16 single strength tablets (4 tabs 4x/d) per day orally or 12 tablets per day if less than 60 kg. You need to treat for three weeks, and then strongly consider life-long prophylaxis after that. If the patient is started on ARVs, consider stopping prophylaxis once the CD4 count has been above 200 for more than 6 months.

Adjuvant prednisone 40 mg 12 hrly may be of benefit if significantly hypoxaemic. (Taper: 40 mg 12 hourly for 5 days, then 40 mg daily for 5 days, then 20 mg daily for 10 days^[82]. Some people then stop abruptly, but if twitchy about this after 20 days of steroids wean to 10 mg/d for 3 days and 5 mg/d for 3-4 days before stopping.)

Remember that these high doses of co-trimoxazole are nauseating, so consider adding metoclopramide 10 mg 3x/d early in the course of the treatment.

Pneumothorax is relatively rare complication, but its presence supports the diagnosis.

Trimethoprim inhibits distal tubular sodium channels and this can precipitate dangerous hyperkalaemia if used in the sorts of doses need for pneumocystis – check the [K⁺] at least weekly while on the high dose.

If the patient is not improving after 4-5 days, consider^[83]:

ARDS
Co-infection with CMV
Abscess formation by another organism.
Tuberculosis.
Non-responsive pneumocystis disease.

In patients who are allergic to co-trimoxazole or who are thought to have non-responding disease, a suitable alternative is clindamycin plus primaquine, although the latter may be difficult to source in South Afica^[84]. There is only anecdotal evidence for clindamycin and dapsone.

Perspectives – pneumocystis pneumonia.

Diagnosis

A common problem is to differentiate pneumocystis from atypical tuberculosis, or even pneumocystis in an individual with CXR features of ‘old’ tuberculosis. Purulent sputum, pleuritic lateralising chest pain, and unequivocal hilar or upper mediastinal adenopathy favour tuberculosis. Tests such as LDH and desaturation with exercise are likely to be positive in both conditions. The table below reflects the real-world situation of trying to differentiate pneumocystis from either tuberculosis or bronchopneumonia. From a single study^[85] of 229 patients.

Almost half of patient with pneumocystis report having a cough for more than a week, 40% have loss of weight, and more than one quarter (28%) have purulent sputum. Effusions are rare (3%.)

Pleuritic pain, night sweats, crackles, a RR >20, and wheeze performed uniformly poorly as discriminators, although a wheeze had a LR+ of 3.8 for pneumonia (sens only 0.26).

In a nutshell, it is easier to make a positive diagnosis of TB or lobar pneumonia and use the presence of one of these to make the diagnosis of pneumocystis less likely. Pleuritic pain and purulent sputum do not exclude pneumocystis, but make it less likely. LOW suggests TB.

It is worthwhile developing increasing familiarity with CXR appearances, because hilar adenopathy has a LR+ of 7.15 for TB, and a specificity of 98%. It is relatively insensitive (11%) Cavitation performs even better in making the diagnosis of TB (LR+ 10.7) but is also uncommon in HIV positive persons (sensitivity 17%). LDH is f some value if the differential is a viral LRTI (LR+ 2.14, LR- 0.13); it’s not even worth requesting if the alternative diagnosis is tuberculosis.

Sputum induction with hypertonic saline and an ultrasound nebuliser has varying sensitivities ranging from 20% to 60% presumably depending to some extent on the enthusiasm of the investigators^[86]. It is certainly a very reasonable alternative to bronchoscopy in situations of resource constraints.

Perspective – efficacy of co-trimoxazole

There is a belief that treatment for pneumocystis with co-trimoxazole is invariably successful; the cure rate is however related to severity of disease at presentation, severity of immunosuppression, and prior use of co-trimoxazole prophylaxis. Mortality overall even with timeous treatment is 10-15%. Patients sick enough to require ICU admission have mortality rates of 35-80%.

Perspective – renal impairment and co-trimoxazole

Co-trimoxazole is relatively contra-indicated in patients with renal impairment, but in suspected pneumocystis there are not many available alternatives. One option is to modify both the dose and the dose intervals (the numbers in the table refer to tablets of co-trimoxazole of standard strength – 80/400, and 2.5 means two and a half tablets):

Perspective – co-trimoxazole and hypoglycaemia

A postulated sulphonylurea-like action has been proposed to explain the severe hypoglycaemia which has been reported occasionally in patients on high-dose co-trimoxazole^[87]. The majority of these patients had renal impairment, or liver disease, and some were on medications associated with lowering of blood glucose (beta-blockers, ACE-inhibitors and propoxyphene.)

Chest radiology in HIV

HIV has led to a number of relatively new patterns on chest X-ray^[88].

Pneumocystis

The CXR may be normal in 5-10% – just dyspnoeic and hypoxic, however there are usually bilateral, symmetrical, fine reticular opacities. (Rare differential – lymphocytic interstitial pneumonitis, which responds to steroids.) Variations: unilateral/focal; occasionally lobar consolidation. Rarely miliary. Cavitation is rare but the presence of a pneumatocoele is highly suggestive.

Complications: pneumothorax.

Prophylactic therapy may alter appearance – predominantly upper lobe medium reticular densities identical to reactivation tuberculosis.

The presence of lymphadenopathy or effusion should make one strongly suspect an alternative diagnosis.

Tuberculosis

Early in the course of HIV this looks like conventional reactivation disease (upper lobes and apical parts of lower lobes).

Later in HIV progression, TB presents atypically^[89] with diffuse coarse interstitial densities often in the middle or lower lobes, with adenopathy in 30 to 80%. Cavitation becomes rare (18%). Pleural effusions occur in 10-20% of the late onset group.

Cryptococcosis

Variable – single or multiple well-defined nodules with or without cavitation

Occasionally reticular interstitial infiltrates, effusions, adenopathy. Rarely miliary.

Bronchitis

Remember all pyogenic infections occur with increased frequency.

Peribronchial thickening and ‘tram-tracking’.

Kaposi’s sarcoma.

Coarse poorly defined nodules throughout the lungs, with associated perihilar coarse linear densities. Rapid nodule enlargement is usually due to haemorrhage.

Pleural effusion occur in 35-50%, but adenopathy is rare (<10%)

Non-Hodgkin’s lymphoma.

Pleural effusion occurs in 50%, adenopathy in 25%, reticulonodular or alveolar consolidation in 25%. Occasionally there are large rapidly growing nodules.

Chemoprophylaxis – co-trimoxazole.

Trials in Africa have demonstrated definite survival benefit from co-trimoxazole, but not necessarily due to reduced pneumocystis infection.

If you have a patient who understands the principle of prophylaxis, and doesn’t think co-trimoxazole is a cure for AIDS, then use it in any patient with an absolute lymphocyte count of less than 1200 or CD4 count less than 200 or WHO stage III or IV.

If patients demonstrate intermittent compliance, then re-think its value in that individual.

If you do decide to use it, the dose is 800 mg of SMX and 160 mg of TMP (two single strength tablets of co-trimoxazole, or one double strength tablet) DAILY in one dose. There is little evidence in favour of twice daily regimens. An alternative regimen^[90] (perhaps not effective for concurrent toxoplasmosis prophylaxis) is one single strength tablet daily.

The routine prescription of co-trimoxazole prophylaxis on admission to patients with HIV and an as yet poorly characterised infection is not advisable – co-trimoxazole is a reasonable broad-spectrum antibiotic, and the dose used may be sufficient to cause confusion by partially treating other conditions. It is also worth remembering that the absolute benefit of the strategy is quite small. Leaving aside for a moment the issue of prevention of septicaemic episodes (as perhaps demonstrated in the Ivory Coast trials), consider the issue of pneumocystis prophylaxis over a two week admission:

Index
Pneumocystis risk per year no prophylaxis	18%
Pneumocystis risk over 2 weeks	0.69%
RRR with prophylaxis	50%
ARR over 2 weeks	0.35%
NNT for 2 weeks	285

Perspectives – co-trimoxazole for prophylaxis:

Two trials^[91],^[92] from Abidjan in the Ivory Coast are helpful. The first trial (HIV-infected adults) demonstrated a hazard ratio of 0.57 (0.43-0.75) of remaining ‘event free’ if on Rx (46% versus 64%) compared with placebo. There was no demonstrated mortality advantage. Social worker home visits were undertaken to aid compliance. The second trial looked at HIV-infected patients with tuberculosis. The same dose was used. There was reduced mortality overall (due to less enteritis and septicaemia) with a relative risk reduction of 46% (CI 23-62%).

These two trials were combined with a third pilot study^[93] (stopped early) to produce a meta-analysis^[94] showing a 6.8% absolute reduction in mortality with a NNT of 15 (9 – 82). The meta-analysts made the point that the data does not allow conclusions to be drawn about the use of co-trimoxazole in patients who do not have tuberculosis.

A further observational study^[95] at four sites in Uganda and Zimbabwe found a 5% absolute risk reduction in mortality at 5 years (87% versus 82%, NNT 20) in patients with pre-ART CD4 counts of less than 15 cells per microlitre. With CD4 counts at initiation of greater than 150 at initiation, the ARR was 2% over five years. Most of the benefit (OR 0.5) occurred in the first year and was largely independent of CD4 count (insufficient patients in the group with CD>200 to reach statistical significance, but similar effect size.) After being on ART for more than 72 weeks, there was no additional mortality benefit, regardless of current CD4 count.

A recent systematic review^[96] identified no further RCTs on efficacy, but gathered some information on discontinuation – in Africa, there is weak evidence that it continues to provide protection against malaria, and perhaps diarrhoea, and pneumonia, when CD4 is >350, but no mortality benefit. Once the viral load was suppressed, it provided little evidence of benefit in preventing pneumocystis, regardless of CD4 count (studies in Europe and Asia.) There is clearly need for more information in this area.

Perspective – co-trimoxazole desensitisation:

Allergy to sulphonamides is relatively common, and in view of the benefits of co-trimoxazole, various desensitisation regimens have been described. Validation of their efficacy is scanty, although some small trials are available^[97].

Patients were also prescribed anti-histamines orally daily at the discretion of the doctor looking after them. In this placebo-controlled RCT of 191 individuals, patients either took the regimen as above plus a placebo tablet each day, or took a genuine co-trimoxazole tablet each day (direct re-challenge) plus titrated doses of placebo syrup in the volumes as above.

At six months, 75% of the dose escalation group as opposed to 57% (p= 0.014) of the direct re-challenge group were still on co-trimoxazole, but this was long after the trial was unblinded.

Desensitisation should be distinguished from provocation testing. The aim of the former is to allow the individual to develop tolerance to the agent, whereas the aim of the latter is to determine if an individual truly does have an allergy. In both situations drugs are given in escalating doses, but in the latter they are escalated quite rapidly^[98], and if no hypersensitivity is demonstrated, then it may be appropriate to assume that the product can be given in full dosage. This situation might apply when an allergy was reported on history in a sick patient requiring that drug.

A speculative regimen (doses given every hour orally) for co-trimoxazole might be:

Time(hrs)	Dose of co-trimoxazole
0	1ml of dilute solution (use an insulin syringe to mix 0.1 ml of the 40/200 mg per 5 ml syrup in 200 ml 5% dextrose water)- i.e. give 0.004/0.02 mg
1	10 ml of the above solution
2	0.05 ml of 40/200 mg/5ml syrup (use an insulin syringe to draw up, and dilute in 5ml of 5% dextrose water)
3	0.5 ml of 40/200 mg per 5 ml syrup
4	5 ml of 40/200 mg per 5 ml syrup
5	2 single strength tablets

All such regimens are largely empiric and poorly investigated. If attempted, the patient should be under direct doctor supervision throughout the process, and hydrocortisone, adrenalin and an injectable antihistamine should be available in the ward, although anti-histamine should NOT be given prophylactically.

Note also that the regimen suggested above parallels that for other antibiotics in the quoted paper, but is not actually listed among their examples. It should be emphasized that a far safer procedure is to use an alternative drug if one is available.

Pleural effusions in HIV

HIV positive individuals with pleural effusions are managed similarly to those who are HIV negative. The differential contains a few extra diagnoses such as Kaposi’s (commonly mistaken for tuberculosis) and primary effusion lymphoma due to HHV-8. In the setting of disseminated Kaposi’s don’t label someone as having tuberculosis simply because they have an exudative pleural effusion, and remember that Kaposi’s may bleed excessively after attempted pleural biopsy, so the diagnosis is often made on balance of probability. Like Kaposi’s elsewhere, pleural disease will settle with ART and chemotherapy.

Perspective – diagnostic strategies in patients with pleural effusions

Some studies in resource constrained environments have shown the prevalence of TB in patients presenting with pleural effusions to be 95%^[99]. In this situation, pleural fluid protein > 50g/l has a sensitivity of 82%, spec 0.83, LR+ 4.82, LR- 0.22.. An ADA of > 10 U/L has a sensitivity of 97%, specificity of 0.50, LR+ 1.94 and LR- 0.06.) In other words, a low ADA with a protein of < 50g/l makes the diagnosis very unlikely.

Tuberculosis in HIV positive individuals.

There are many ways in which tuberculosis can present in individuals infected with HIV, and a familiarity with most of them is essential for individuals working in resource constrained environments where both diseases are endemic.

Possibilities include:

Lower or right middle lobe infiltrate with associated adenopathy – so-called ‘primary’ TB…
A miliary picture or diffuse haziness.
Pleural effusion, pericardial effusion or ascites. Late pericardial constriction
Asymmetrical cervical adenopathy.
Splenic hypodensities and retroperitoneal adenopathy on ultrasound.
Cold abscesses.
Meningitis or tuberculoma or tuberculous brain abscess.
Arthritis or osteitis.
Eye, skin, or renal involvement.
Orchitis.

Because of the many different presentations, it should correctly be considered in any sick patient with HIV; the corollary should however be addressed with care – not all ill patients with HIV have TB!

Diagnosis is clearly tissue specific – if you find an enlarged node or a subcutaneous cold abscess, aspirate it. Pleural fluid microscopy has a somewhat higher yield than in HIV negative individuals, but is still not particularly rewarding. In situations of suspected disseminated disease, apart from the obvious culture sources (sputum, marrow aspirate, node aspirate) also consider sending urine for culture – it is easy to obtain and performs quite well as a culture source.

Perspective – co-infection with HIV and tuberculosis.

The occurrence of another important infection in patients with HIV and tuberculosis occurring during the course of the same admission is well recognised. Reported frequencies in one study^[100] (N=58) were oesophageal candidiasis in 33%, toxoplasmosis in 5%, and pneumocystis and cryptococcosis each in 2%. Equally sobering was the finding that active tuberculosis had been immediately preceded by the finding of oesophageal candidiasis in 40% of patients.

Do not diagnose oesophageal candidiasis without also looking for tuberculosis.

HIV and MDRTB

Multi-drug resistant tuberculosis in HIV is often associated with prior adherence difficulties, so planning therapy in this situation can be very challenging. MDR is conventionally defined by the presence of resistance to both INH and rifampicin. Even in the non-HIV situation, treatment is expensive, prolonged and not all that successful. Key features of management are formal isolation until sputum are smear negative, and then prolonged therapy for a year to eighteen months with ethambutol, ethionamide, ofloxacin and pyrazinamide, with kanamycin or amikacin being added intramuscularly for the first three months. Unless the patient is very ill, do not start therapy for MDR TB simply because the patient isn’t getting better; poor immunity is a more likely cause. Do not add a single new drug to a failing regimen, but rather plan a new regimen containing at least three new drugs. In general, try to refer all such patients to an MDR management centre. Note that it is preferable to avoid tenofovir while the patient is on treatment with an aminoglycoside, so consider zidovudine, stavudine or abacavir in its place.

A further challenge is the emergence of extensively drug-resistant tuberculosis (XDR TB), where resistance is present to the conventional second line agents. (It is defined as resistance to INH, rifampicin, quinolones, and either aminoglycosides or capreomycin or both.) This version has a very high mortality in HIV positive individuals^[101], and nosocomial transmission to other HIV positive patients is a real possibility.

HIV and mycobacteria other than tuberculosis

This is not a particularly common condition locally, but it should be thought of in patients with unexplained pyrexia, unexplained pulmonary findings, atypical abdominal pain, and in those with AFB-positive ‘tuberculosis’ who fail to respond to conventional anti-tuberculous therapy. It may also be under diagnosed in late presenting patients with apparently near-end stage disease. It usually occurs in individuals with low CD4 counts (<50) – the lower the count, the more likely it becomes. The most common candidate is Mycobacterium avium complex.

Diagnosis – blood or bone marrow aspirate for culture (i.e. sites that should normally be sterile). A single isolate from sputum is often a commensal, but persistent isolation of the same organism in an ill patient warrants discussion with a specialist.

Therapy. The main principle is to try to get the patient onto ART. Prognosis for sustained remission is not that good otherwise – probably less than 50% even with appropriate therapy. A drug combination of clarithromycin 500mg 2x/d and ethambutol 1.2g/d (15-20 mg/kg/d) are usually recommended. Treatment should continue for at least a year and until the CD4 count is > 100 on ART. (Other options are rifampicin, streptomycin, ciprofloxacin, and amikacin. Good quality comparative studies are lacking, but at least try to get the macrolide.)

Empiric antituberculous therapy in HIV

There is a very major temptation for over-worked staff to use this modality, especially in the light of evidence that tuberculosis in the developing world is often under diagnosed. This section should be read in conjunction with the notes on ‘trials of therapy’, but in general there are a number of reasons why trials of empiric TB treatment are a bad thing.

You may be missing/masking another diagnosis.

What do you do when the patient perhaps starts to improve, but then develops a life threatening reaction to the TB treatment?

You may be condemning a person with already limited quality and duration of life to a treatment that will make him or her feel even worse.

Despite this, there are patients in whom the clinical suspicion of disseminated tuberculosis is high, but making the diagnosis is difficult. This may be due to urgency (the patient is very ill and needs treatment now) or because of logistic or other impediments to formal tissue or microbiological confirmation.

Management:

Consider consulting a specialist, even if only telephonically.
Try hard to make a diagnosis. Consider aspiration biopsy of nodes, sputum induced with hypertonic saline, biopsy of suspicious skin lesions, urine TB culture, and possibly lumbar puncture or bone marrow/ liver biopsy.
Discuss with the patient, if possible, and decide on a time course for the trial, with pre-specified outcome measures. (c.f. ‘trials of therapy’).
At the very least, send sputum for TB culture.
It is important not to delay TB treatment unnecessarily, and it is worth remembering the WHO guideline on sputum smear negative TB in this context:

*Any one of: resp rate > 30, temp > 39, pulse > 120, or unable to walk.

Perspective – effect of WHO guideline on outcomes in smear negative disease

The WHO guideline alluded to above was published in 2007, and subsequently its use has been compared to ‘standard care’. In a South African study^[102] of 338 patients, ongoing hospital admission at one week was reduced from 38% to 27%, and mortality at eight weeks was 17% versus 32%. Although this was an observational cohort rather than a RCT it does illustrate that earlier TB treatment may be advisable when the clinical suspicion is high.

INH prophylaxis for tuberculosis in HIV

There is reasonable evidence that, in patients with a tuberculin skin test of more than 5 mm, six months of standard dose INH reduces the chance of contracting tuberculosis. Two meta-analyses differed somewhat as to whether there was a mortality advantage. It is clearly essential to exclude active tuberculosis (symptoms, CXR) before contemplating this therapy, and it is important to ensure that the basic TB services are performing adequately. Use INH 300mg daily for six months (adjust dose by weight if very thin.)

The case for secondary prophylaxis (after treatment) is less clear, although there are proponents for this^[103]. In HIV positive South African gold miners, the risk of developing tuberculosis was 19 per 100 patient years^[104] – the authors of this cohort study (not a good design for proving efficacy) suggested that INH would prevent one case of recurrent TB for every 5 patients with CD4 counts of less than 200 treated for a year. However the baseline risk in miners is higher than in many other communities.

A local trial^[105] demonstrated benefit in giving INH to a group in Cape Town where 70% were already on ART. There was no mortality advantage, and the reduction in TB incidence only demonstrated statistical significance when ‘possible/probable TB cases were combined with those with definite TB: HR 0.63 (95% CI 0.41 to 0.94, control event rate 3.6/100 person years) – NNT 25, and NNH 100 (withdrawals due to raised liver enzymes.) The study also found that this was independent of tuberculin skin test, and benefit (may) last up to three years.

Rheumatology – HIV

Patients who are HIV infected can obviously be affected by all the same rheumatic conditions that affect HIV negative individuals, and these will not be discussed here. Joint symptoms in HIV positive individuals often tend to be managed symptomatically without a great deal of attention being paid to the aetiology; because many of the conditions are self-limiting (resolving within weeks), this strategy often pays off. However some are more confusing or lead to lengthier disability and are discussed below in more detail.

Reactive arthritis

This is probably the commonest rheumatological condition in HIV positive individuals that comes to medical attention. It presents similarly to the illness in HIV negative individuals:

Asymmetrical involvement of a few joints at a time, and a mixture of large and small joints.
There may be an associated enthesitis, which is very suggestive of the condition. Enthesitis is inflammation of an enthesis, which is the site where a tendon joins onto bone – e.g. the Achilles tendon insertion into the calcaneus.
There are the usual associations with chlamydia, salmonella and various other infections that are relatively common in HIV positive individuals. A characteristic presentation is a few weeks after an episode of severe diarrhoea.
Initial therapy is with adequate doses of NSAIDS – e.g. 400 mg 8 hourly of ibuprofen, if tolerated.
Some authorities advocate treating a presumed chlamydial infection with doxycycline 100 mg daily for a couple of weeks; the evidence that this does much is tenuous.
If it doesn’t settle and seems to be moving into a more chronic phase, consider adding chloroquine base 150 mg/day (omit on weekends); the addition of prednisone 10 mg/d for a month or so may help.
Patients who clearly aren’t settling should be considered for sulphasalazine, but this is expensive, so before embarking on this consider whether you have the correct diagnosis.

HIV-associated arthritis.

Some authors consider that this is due to direct involvement of the joint by HIV; for practical purposes it is very similar to a self-limiting reactive arthritis without the enthesitis. Treatment is with NSAIDS.

Psoriatic arthritis.

Psoriatic arthritis is not that common, and when it does occur, it is sometimes missed. It also presents like a reactive arthritis, but check carefully for any evidence of psoriasis on the skin, and specifically check for nail pitting. Patients with preserved CD4 counts (>200/microlitre) can be considered for methotrexate 10-15 mg once weekly, with folate 5 mg daily. Avoid prednisone, although intra-articular injection of corticosteroids is an option for isolated joint flares.

Myopathies

Zidovudine is associated with a condition of generalised myalgia and proximal muscle weakness associated with a rise in creatine kinase, although this is less common with the lower doses used nowadays. It responds to stopping the zidovudine. Septic infection of muscle (pyomyositis) is discussed under its own heading. DILS (see next section) may also present with myopathic symptoms.

Diffuse infiltrative lymphocytosis syndrome (DILS)

This is basically Sjogren’s in HIV positive individuals. Occasionally patients present with prominent parotidomegaly, which sometimes responds to radiotherapy, but the full spectrum includes a lymphocytic pneumonitis, mononeuropathies, aseptic meningitis (which is usually labelled as tuberculous) an interstitial nephritis and arthritis. Patients who are severely affected may respond to prednisone in a dose of 0.5-1 mg/kg/day.

Femoral head osteonecrosis

Probably under-recognised and its clinical significance is still unclear. Can be confused with CMV radiculopathy as it presents with intermittent groin pain of gradual onset, and could also be confused with a monoarthritis. Pain is worse on weight bearing and internal rotation. Prior to the radiologically characteristic femoral head collapse there may be osteopenia and subchondral sclerosis with occasional cyst formation^[106]. A case-control study from the US^[107] suggested prevalence as high as 4.4% of HIV positive individuals. The aetiology is unclear.

Vasculitic syndromes in HIV.

Poorly characterised rather distressing syndromes with peripheral gangrene due to larger vessel occlusions are described. Apart from cold agglutinins and hyperviscosity, there also seems to be a real vasculitic process rather similar to Takayashu’s, as well as an SLE-like vasculitic syndrome. Diagnosis and management is difficult as most individuals seem to present at a time when the HIV infection is very far advanced. Therapy with prednisone 1 mg/kg is reasonable but it is important to consider the dangers of further immunosuppression in such patients.

SLE and rheumatoid arthritis.

There are some suggestions that these two are more common in HIV positive individuals, but this may just reflect referral bias. Diagnosis and management is similar to HIV negative individuals, although caution is need in individuals with low CD4 counts when considering immunosuppression.

Pyrazinamide and ethambutol cause joint pain, usually thought to be associated with gout. Zidovudine has been mentioned previously.

Special care issues – HIV

Surgery in HIV positive individuals

Patients presenting with surgical emergencies should in general be operated on regardless of HIV status, unless that status is such to have rendered them in too poor a state to tolerate surgery or anaesthesia, or if they have a condition that would render surgery futile – e.g. massive cerebral infarction.

Elective surgery is more problematic, particularly when the proposed operation is highly resource intensive, either in terms of theatre and staff time, or in terms of technology required – e.g. valve replacement surgery. If the surgery can wait and the patient is a candidate for anti-retroviral therapy, it would make some sense to establish the patient on this first to reduce viraemia and improve immunity.

Palliative care in HIV.

The management of “terminal” patients with HIV infection presupposes a number of things:

The patient does in fact have AIDS.
There is no disease “worth” reversing.
Any proposed curative therapy is likely to fail.

Many clinicians argue that even in these circumstances it is worth trying ART, but there will still be those patients in whom for various reasons this is just not feasible.

Management:

This involves discussing the issues with the patient, and if appropriate, the relatives (obviously with the patient’s consent). After that, the secondary priorities revolve around alleviation of discomfort for the patient and distress for the family. Remember that the best way of managing the discomfort of things like oral thrush is to treat them properly – nystatin is better than morphine for this condition!

Care should as far as possible revert to the family/friends or other community care-givers, with support from the health system in the form of basic education, nursing supplies such as gloves and linen savers, and most importantly, the establishment of a channel of communication whereby questions and problems which the family may encounter can be addressed without the need for an expedition to the hospital.

Analgesics should be used as required, as should anti-motility drugs for chronic diarrhoea that has proven untreatable by other methods.

Cautions:

All too often the patient is seen for the first time by a busy practitioner, is noticed to be thin and apathetic, and is summarily started on morphine. This is NOT the correct way to go about things. Anecdotally, there are very many people who have been labelled as terminal but who are now well and productively employed some years after that sort of diagnosis has been made. The usual cause of this is a slippery slope of oral thrush leading to poor intake, which causes malnutrition, which may or may not be complicated by tuberculosis. Food and TB drugs are cheap, and their lack should not lead to patient death. There are very few patients where it is possible to unequivocally exclude significant reversible pathology based on a single visit.

Common genuinely terminal illnesses might be:

Tuberculous or cryptococcal meningitis in a patient remaining obtunded some weeks after starting treatment.
Severe disseminated CMV infection.
Non-responding toxoplasmosis.

Home-based care in HIV

The rationale for this is that hospitals in high-prevalence areas cannot hope to cope with all the potential HIV positive individuals who possess very plausible reasons for admission. It is worthwhile supporting such initiatives, and making contact with those that are active in your area.

Needle stick injuries

There are a numerous reviews of the basics of needlestick injury management^[108].

Perspective – seroconversion risk.

The chance of seroconverting after a percutaneous injury from a needle or instrument contaminated with blood from an HIV positive person is 0.18 – 0.46%^[109]. Factors accounting for varying rates are inoculum size, injury depth and index patient illness severity –the blood of someone with a high viral load (shortly after seroconversion or end-stage) is more dangerous than that of someone who is seropositive but not ill.

The procedure.

First-aid: Allow to bleed and then wash thoroughly with soap and water

Testing:

The patient: In the first first hour or so you should be taking treatment rather than rushing around with tubes of blood and multiple forms. Once you’ve taken one dose, the incremental risk (and cost) of taking further doses till results are available is minimal. Test the patient later.
The injured person. From the point of view of workmen’s compensation, it is important to take a sample within a day or so of the injury (the burden of proof that the injury occurred at work lies with the injured party). After that, testing at six weeks, three months and six months is probably adequate.

Zidovudine (AZT) is the only drug shown to reduce transmission. Because of the difficulty in researching this area more satisfactorily, rather than because the trial gave a definitive answer, the CDC and other advisory agencies support the use of at least dual therapy and in some situations triple therapy. In the case-control trial in question^[110] there was an 80% reduction in expected seroconversions.

Caveats. There are already 11 documented cases of seroconversion having occurred inspite of prompt zidovudine monotherapy.

The details. Potentially infectious fluids include: blood, CSF, semen, vaginal secretions, and synovial, pleural, pericardial, ascitic and amniotic fluids.

Splashes on intact skin require no ARV treatment.

Splashes on mucosa, non-intact skin. Lamivudine plus zidovudine (dual therapy.) Needlestick or other percutaneous injury: add lopinavir/ritonavir as well (‘triple therapy’)

PEP is often difficult because of GIT intolerance. Dual TDF and 3TC may be best tolerated, and arguably better than taking nothing because of side-effects.

Basic regimen: TDF 300mg/d + FTC 200 mg/d + ATV/r 300/100 mg/d, all for 4 weeks. Llopinavir/ritonavir 200/50 mg 2 tabs 12 hourly can be used as the PI, but may be less well tolerated.

If renal impairment or if the patient is known to have virological failure on a TDF based regimen, then: zidovudine 300mg2x/d plus lamivudine 150 mg 2x/d for 28 days plus the PI (either ATV/r or LPV/r).

Therapeutics – HIV and ARVs

Introduction

Starting ART is exciting and can be very rewarding – it is marvellous to watch somebody gain fifteen kilograms in a few months and go from being wheelchair bound to back at work. There are a number of caveats:

Ensure that the patient is genuinely ready to start. Initiation of life-long therapy should seldom be seen as an emergency. If in doubt, wait a few weeks. Possible exceptions to this are patients with extremely low CD4 counts (e.g. less than 50).

Try to make sure that the patient doesn’t have any other smouldering infections waiting to explode as immunity starts to reconstitute. If the patient has been losing weight and has a chronic cough, deal with the tuberculosis first. Patients with new tuberculosis should normally get at least the first two months of TB therapy before starting ARVs as well.

Make sure that adherence counselling and understanding of therapy is completed before starting treatment – don’t make the patient find out the hard way.

Don’t start treatment immediately after making a diagnosis of HIV – give patients a chance to settle down for at least a few weeks and get their minds round the idea of a life-long chronic illness before inflicting life-long chronic treatment.

Many patients experience quite uncomfortable side effects on initiating therapy – mention that this may occur beforehand, and plan an avenue of re-attendance so that these problems can be dealt with as they occur.

Perspective – delays in starting ARVs

Undue delays in starting ARVs are potentially very problematic – about 15% of patients with Stage 4 HIV infection may die in the first three months after enrollment in readiness/treatment programs, as opposed to about 3% with Stage 3^[111]; the pre-treatment mortality rate may be 36 per 100 person-years overall, with two thirds of deaths in the 90 days from enrolment occurring in those not yet started on ARVs.

Antiretroviral therapy

Indications for starting treatment with ARVs:

HIV positive and pregnant.
HIV positive and tuberculosis or other severe opportunistic infections such as cryptococcal meningitis. (WHO stages III and IV)
Any HIV positive patient in the new ‘test and ‘treat’ approach

In general, if there is another active disease, settle the treatment for that first, but avoid long delays if the patient is severely immunocompromised:

1. First line Rx:

Tenofovir 300 mg daily

Lamivudine 300 mg daily

Dolutegravir 50 mg nocte ( add another 50 mg mane if on TB regimens containing rifampicin)

All patients starting tenofovir should have serum creatinine measured. If the estimated GFR (see table inside back cover) is less than 60, this agent should be avoided. All patients on it should have serum creatinine repeated at one, three and at six months, and then yearly. Prescribing: this combination of three agents is available as a fixed dose combination Do not write ‘FDC’ because many medicines are fixed dose combinations: cotrimoxazole, rif/isoniazid, amoxyclav. Try ‘TDF/3TC/DTG 300/300/50 mg’ at the very least.

First line regimen failure is usually due to adherence difficulties and warrants careful attention to trying to resolve this issue; rapid regimen changes in the face of poor adherence achieve little.

For regimens containing ATV/r use double dose LPV/r if on rifampicin; if LPV/r not tolerated then use ATV/r but replace rifampicin with rifabutin

Side-effects

Abacavir can cause a severe hypersensitivity reaction (malaise, fever, rash, GIT upset, cough, sore throat, dyspnoea) which tends to occur in the first 1 to 6 weeks of use.

Emtricitabine has a relatively benign safety profile, but has been associated with hyperpigmentation of the palm and soles and headache: like its twin, lamivudine, it has very occasionally been associated with lactic acidosis, hepatitis, skin rashes and anaemia.

Stavudine causes peripheral neuropathy, fatty liver and occasional lactic acidosis

Lamivudine can cause anaemia, diarrhoea and pancreatitis but is generally well tolerated.

Efavirenz causes skin rash (continue Rx unless fever, abnormal LFTs or eosinophilia), GIT upsets and neuropsychological changes such as dizziness, dysphoria, drowsiness, and vivid dreams. This latter may be of particular concern in cultures were communication with ancestors by dream is seen as commonplace, and may sometimes prove a disincentive to ongoing therapy. A randomised double blind trial^[112] which had a side-arm looking at neuropsychological side-effects of efavirenz found that discontinuation rate was 6% for this effect versus 0% in the non-efavirenz arm. Beyond one month, there was no statistically significant difference in scores for sleep, depression, anxiety, or neuropsychological performance. There have been a couple of case reports linking it to epilepsy. A more recent and concerning association is with quite severe liver disease^[113]:

nonspecific hepatitis: normal bilirubin and INR, moderate elevation of both transaminases and canalicular enzymes
mixed cholestasis-hepatitis: jaundiced, and canalicular enzymes more markedly elevated than transaminases
submassive necrosis: jaundice with markedly elevated transaminases and raised INR, more common in young women with CD4 count > 350, and (apart,obviously, from stopping the efarirenz) possibly warranting consideration of low dose prednisone (0.25mg/kg/d) in view of the severe immune reaction on histology.

Nevirapine causes skin rash (manage same as efavirenz), GIT upsets and hepatitis, which can be very severe, and is possibly common in women with higher CD4 counts, although there is some observational evidence that this is not so^[114].

Zidovudine is associated with anaemia – check FBC at baseline and then after the first, second, third, and sixth months.

Tenofovir is associated with renal dysfunction, particularly a Fanconi-type syndrome characterised by glycosuria, proteinuria, and phosphaturia. Avoid if if estimated GFR < 60, and avoid, or monitor renal function closely, if on other nephrotoxic drugs.

Second line Rx:

Treatment failure is defined as a new VL (HIV RNA) > 1000 after adequate suppression , or failure to ever suppress after adequate treatment, followed by intensive counseling and compliance assessment such that treatment is considered to be being taken, and then a repeat VL 2 months later still >1000.

In patients failing a NNRTI based regimen (e.g. tenofovir, emtricitabine or lamivudine and either efavirenz or nevirapine) change to:

Zidovudine 300 mg 2x/d (or abacavir 300mg 2x/d if anaemic)

Lamivudine 150 mg 2x/d.

Dolutegravir 50 mg nocte

Two years of failure on a DTG based regimen would lead to consideration of a PI (e.g. AZT + 3TC + LPV/r.) Remember that in patients with HepbsAg positivity, continuing TDF and either 3TC or FTC is appropriate.

Monitoring: Baseline cholesterol and triglycerides, then after 3 months if on lopinavir/ritonavir. If develop TB on this regimen, increase the dose of lopinavir/ritonavir to 3 tablets 2x/d one week after starting TB treatment. After a further week, increase the dose to 4 tablets 2x/d.

The lopinavir and ritonavir are available as a combination capsule containing 200 mg lopinavir and 50 mg ritonavir. (i.e. 2 tablets 2x/d)

Side-effects of lopinavir/ritonavir

GIT upsets are relatively common on starting treatment, and lipid abnormalities and lipodystrophy, are described. Ritonavir can cause dysgeusia (funny taste) and raised liver enzymes. There are important drug interactions, especially with TB medication and anticonvulsants. Patients who develop hyperlipidaemia on this agent should be started on diet, and if this doesn’t work, options are either atorvastatin 10 mg daily (not simvastatin) for hypercholesterolaemia or bezafibrate 400 mg nocte for significant hypertriglyceridaemia. Don’t forget the CVS risk profile – it may not be necessary to start a statin on every patient with slightly raised cholesterol.

Third line regimens

These are now available to adherent patients failing the first two regimens after at least two years on ARVs, but are managed centrally in consultation with an expert panel. One option might be raltegravir/darunavir/etravirine.

Fixed dose combination antiretrovirals

Fixed dose combination antiretrovirals offer advantages in terms of cost, patient convenience, and handling (fewer pills to transport and dispense.) They may reduce medication errors, and may over time lead to a reduction in stock-outs. They are no more effective than individual pills given in combination, assuming good adherence with both regimens.

There are a number of different fixed dose combinations, and there is a tendency to prescribe them by trade name. There are circulating perceptions that some brands are better than others, and on occasion prescribers have been known to swop erroneously from one brand to another in the belief that this constitutes a beneficial change in a failing ARV regimen.

It is advisable to prescribe using the letters “FDC” and at least the abbreviations if not the full generic name for each constituent, e.g.: “TDF/FTC/EFV 300/200/600mg (FDC)” would be an acceptable abbreviation for the fixed dose combination tablet containing tenofovir, emtricitabine and efavirenz. If this seems like a lot of writing, remember that it is a lot less than writing out each medication with its dose on a separate line of the script. Not putting the FDC in brackets can also cause confusion with emtricitabine which is abbreviated as FTC.

Patients often only know the trade names of their medications which can cause confusion, so a selection of these are listed below. Please do not prescribe by trade name!

FDC	Constituents	Dose	Proprietary names
EFV/FTC/TDF	Efavirenz 600 mg Emtricitabine 200 mg Tenofovir 300 mg	1 tab daily	Atripla Odimune Tribuss
LPV/r	Lopinavir 200 mg Ritonavir 50 mg	2 tab 12 hrly	Aluvia Kaletra
Other combinations, but not routinely used in the public sector
ABC/3TC	Abacavir 600mg Lamivudine 300 mg	1 tab daily	Kivexa
FTC/TDF	Emtricitabine 200 mg Tenofovir 300 mg	1 tab daily	Truvada Didivir

Perspective – efficacy of antiretrovirals

Trials of antiretrovirals in less-developed environments are still mostly only able to report on relatively short follow-up; one cohort of patients in the Cameroon^[115] given a fixed dose nevirapine, stavudine, and lamivudine combination showed that after 6 months 80% of patients had an ‘undetectable’ viral load (less than 400 copies per ml) and the mean rise in CD4 count was 83 cells per microlitre. The 2NN study^[116] also featured this combination in an open label study and found treatment failure at 48 weeks in about 40% of patients.

The importance of adherence in ART

Resistance to all antiretroviral classes is well described, with resistance emerging relatively rapidly in some developing world situations as well^[117].

Many texts talk about the missing of three doses a month as if this was some kind of threshold, above which resistance development is negligible. It would be appropriate for the individual patient (as opposed to a cohort in a trial) to regard ANY missed doses as potentially detrimental. The goal to aim for is 100% compliance, not 95%.

Strategies to improve adherence

There is very little evidence on which to base recommendations, but a few principles apply:

Patients who understand what they are taking, why they are taking it and when they should take it are more likely to do so than those for whom their medication is just an amorphous mush in terms of understanding. Communicate.

Do not overdo the side dishes. Patients on antiretrovirals have enough to do concentrating on getting this right without a whole bunch of other medications being given for the most tenuous of reasons. Although other medication may seem innocuous enough, it may have opportunity costs if it leads to a patient forgetting to take a dose or two of ARVs.

As with ‘home DOTS’ try to get somebody else in the home to help the patient remember.

Use situational reminders – e.g. pills next to toothbrush, or associated with any activity that takes place regularly around about the time doses are due.

Patients with cellphone alarms can set these to remind themselves. Most people can afford the sort of alarm available on even the cheapest digital watch.

Tick charts and pillboxes may help patients remember, and will certainly help highlight missing doses.

Not all patients will be able to achieve perfect adherence – don’t get so hung up about this that the patient gets scared to tell you of adherence failures – if you are still communicating, you may be able to improve things.

Perspective – adherence to ARV regimens.

Several studies have reported strong associations between adherence rates of less than 95% of all doses and virological failure, and studies mainly from the US suggest that suboptimal adherence is very common. Reasons frequently listed for failure to take a dose are^[118]:

Forgot
Too busy/Separated from medication
Asleep
Too depressed
Too ill
Too many side effects.

This list is limited and to some extent remediable, and if one looks at the factors associated with poor adherence, the situation is even clearer:

Don’t trust the doctor
Alcohol abuse
Depression or other mental illness
Inadequate patient education about medication
Inadequate access to medication

Addressing all of these issues prior to starting medication, and continuing to build on relationships of trust and open access to care are thus essential to ongoing adherence and therapeutic success. If you don’t care, the patient probably won’t.

Perspective – efficacy of adherence interventions.

Adherence interventions should be seen as multi-faceted – the clinical benefit of single interventions is modest, and although there are some studies showing that you can improve adherence, one systematic review^[119] only found 16 studies showing that adherence interventions improved clinical outcomes, and in most of them the size of benefit was relatively small.

Other therapies in HIV.

Multivitamins have recently been associated with a modest advantage in terms of the composite outcome of progression to WHO Stage 4 or death^[120] (NNT 16) although there was no actual survival advantage in this group of women given multivitamins for an average of five years.

Perspective – multivitamin trials in HIV

An underpowered study from Thailand^[121] used a combination of multivitamins and minerals, and the abstract notes that the death rate was lower in the micronutrient arm. In fact the hazard ratio for mortality only reached statistical significance for the single subgroup with CD4 <100. This subgroup had 40 members in each arm, and the CI on the HR of 0.03 was a sweeping and virtually meaningless 0.07 to 0.97. The drop-out ratio in the trial overall was over 20%, which in general means the results should be treated with circumspection. There was no effect on CD4 count or viral load.

The Tanzanian study by Fawzi et al demonstrated that in the multivitamins only group there was no AIDS related mortality advantage at two, four, or five years. The composite of death from AIDS related causes or progression to stage 4 was significant at two and five but not four years. The difference in progression to stage 4 was significant at all 3 time periods, but with startlingly different relative risks – 15%, 17% and then 50% at 5 years. The very small number of deaths in the active group (2, 3 and then 18) resulted in very wide confidence intervals. The higher CD4 counts (by about 50 ) were significant at all time intervals. This trial supports giving multivitamins to anaemic malnourished peripartum women, where they may slow progression but probably do not influence mortality. Parallel evidence from further trials would be welcome. A Cochrane review^[122] was a little more forthright, stating that there is as yet no convincing evidence of either a mortality or morbidity advantage from the use of these products as supplements in adults, although replacing deficiencies to RDA levels is sensible.

Dealing with the sick patient on ART

Look hard in the usual way for a source of infection – tuberculosis, other chest infections, cryptococcal meningitis, and acute diarrhoeal illness. If one of these is identified, treat it.

The immune reconstitution syndrome can be quite characteristic – see separate entry.

Next, consider a drug reaction.

Pointers about the timing of reactions due to immune reconstitution versus those due to drug reactions are unreliable.

Immune reconstitution disease (IRD/IRIS).

Immune reconstitution disease (also called immune reconstitution inflammatory syndrome, or IRIS^[123], immunorestitution disease and immune restoration disease), describes a complex of different situations arising where patients deteriorate clinically soon after starting ART when the strengthening immune system ‘discovers’ and starts to react to previously unrecognised infections.

Manifestations range from a flu-like illness to polyserositis to specific features of tuberculosis or cryptococcosis, or whichever underlying infection is responsible. Presentations vary from the dramatic (e.g. tuberculosis associated ARDS) to just fever and malaise. There are even suggestions that the prevalence of pre-eclampsia in obstetric patients started on ART may return to that of HIV negative controls.^[124] (Untreated HIV positive women are said to be less likely to develop eclampsia.)

Diagnosis

Consider the diagnosis of IRD if^[125]:

Paradoxical deterioration. There is unexpected worsening of TB or another known opportunistic infection.

Temporal association to ART initiation. The deterioration happens soon after starting ART. (For TB, median time to deterioration is 4 weeks, usually between 2 and 10 weeks, but anywhere between 1 and 44 weeks has been described.)

Early stages of treatment of opportunistic infection. Patients with TB are usually only on treatment for about two months. (i.e. ART started less than two months after starting TB treatment, usually because of CD4<100.)

Immune reconstitution. There is evidence of immune restoration such as a rise in CD4 count or an impressive granulomatous response or caseous necrosis in tissue samples from TB patients. Note that occasionally IRD develops before the CD4 rises, presumably reflecting improvement in other aspects of immune function. Also, only about half of patients will have full viral load suppression at the time IRD is diagnosed, although most will show evidence of an impressive fall in load.

No other cause. No obvious alternative explanation such as non-adherence to TB medication, or presence of MDR TB.

Management

Adequate treatment of the underlying condition, once identified

Continue with ART.

General anti-inflammatory measures such as NSAIDS and simple analgesia.

On rare occasions with life-threatening reactions, it may be appropriate to give a short course of corticosteroids (e.g. prednisone 2 mg/kg/day) but there is as yet very little controlled trial information on efficacy.

Paradoxical reactions to TB treatment in HIV negative patients.

A syndrome of deterioration TB, beginning a month or so after initiation of therapy, was well recognised prior to the HIV era^[126], and serves as a reminder that treatment of any severe chronic infection may lead to restoration of immune function, exposure to more antigen, cytokine release, and lymphocyte activation. The best characterised example was neurological deterioration due to new oedema surrounding an often occult CNS tuberculoma in patients recently initiated on therapy for pulmonary tuberculosis. Consider using high dose corticosteroids (e.g. prednisone 2 mg/kg/day, or dexamethasone) in patients on TB treatment presenting with new onset neurological deficits, where the CT scan demonstrates a ring-enhancing lesion with surrounding oedema.

ART in patients with epilepsy

Many of the conventional anti-epileptics interact with ART, resulting in lower drug levels and possible failure of ARV therapy. Assuming there is a genuine need for ongoing anti-epileptics then phenytoin, carbamazepine and phenobarbitone should be avoided. The agents of choice are either lamotrigine or valproate, and if possible the dose of these should be titrated up to therapeutic levels while slowing weaning the previous agent(s). Once a therapeutic dose with valproate or lamotrigine monotherapy has been established, it is reasonable to introduce the ART.

Lactic acidosis on ART

All NRTIs, but particularly didanosine and stavudine, are associated with the development of lactic acidosis, by causing mitochondrial toxicity through an effect on DNA polymerase gamma, which is an enzyme needed for mitochondrial DNA replication. The problem is usually only seen after some months of ART use. For every 250 patients treated for a year, one is likely to develop lactic acidosis, but the exact figures will vary with population profile. It is said to be more common in obesity, in women, and in pregnancy.

It presents with rather non-specific features such as dyspnoea, abdominal pain and nausea, unexplained tachycardia, and weight loss (often >2 kg in 3 months).

The problem with diagnosis is that it is a disease in evolution – patients with mild symptoms and non-diagnostic tests may develop florid disease a few days or weeks later.

Diagnosis is based on finding an increased anion gap metabolic acidosis due to a raised lactate (found either on fingerpick testing with a clinic machine or else on laboratory testing – grey top (fluoride) tube). The lactate will usually be more than 4.0 mmol/l in significant disease.

	One or more of: unwell; resp rate >20, serum bicarbonate <20, low pH	Normal resp rate, bicarbonate and pH
Lactate > 2.5 mM	Lactic acidosis	If clinically well, asymptomatic hyperlactaemia
Lactate < 2.5 mM	Either incubating disease or something else	If unwell, either something else or very early disease

Once the diagnostic possibility is raised, there is no test or combination of tests which effectively excludes the diagnosis at a single time – ALL such patients need early meticulous follow-up by with a carefully communicated review plan.

(A common error is to label the patient as not lactic acidosis and refer back with the expectation that a primary caregiver will continue to monitor carefully. All too often, review happens too late, or not at all.)

1. Patients with suspected very mild disease (normal respiratory rate and fingerprick lactate <2.5 mM) can be observed with a repeat lactate in 3 or 4 days and then again at one week.

2. All symptomatic patients with fingerprick lactate >2.5 need acidosis excluding by the demonstration of a normal respiratory rate and depth, normal bicarbonate (>20 mM) and normal venous blood pH. Send off liver enzymes and lipase. If in doubt about the patient’s ability to return for follow-up, admit. If there is no acidosis and the repeat lactate is still <4.0 mM a day or two later change stavudine to zidovudine, and re-check lactate after 3 days, then weekly till normal. (Check FBC as per protocol for zidovudine initiation.) If lactate doesn’t settle by day 3, stop all ARVs immediately.

3. If the lactate is >4 mM, or if it is >2.5 mM but with a respiratory rate >20 or bicarbonate <20 or pH < 7.3 on venous blood gas, admit the patient, stop all ARV immediately, do LFT amylase and lipase plus blood culture and look for clinical signs of sepsis.

Ensure that the patient does not continue to take ARVs from a bedside locker supply by carefully explaining the reason for stopping the medication.

IV Vitamin B Complex 1 amp 12 hourly and thiamine 100 mg IV 12 hrly.

IV fluids – 5% dextrose 1 litre 8 hourly, but watch for fluid overload.

An intravenous third generation cephalosporin such as ceftriaxone 1g 12 hourly, because sepsis can be clinically indistinguishable and also causes a lactic acidosis.

IV bicarbonate is commonly recommended for severe acidosis, in the absence of firm evidence of efficacy. Give slow boluses of 0.5 mM/kg – e.g. 25 mM at a time – and assess response clinically and biochemically.

Some patients may require ventilation.

There are theoretical grounds for the use of pyridoxine, but little empiric evidence of benefit.

Restarting ARV after an episode of lactic acidosis.

Ask. The options are a NNRTI (efavirenz or nevirapine) with lopinavir-ritonavir 4 caps 2x/d (safest) or else for milder disease, monthly lactate monitoring for 3 months with zidovudine, lamivudine and a NNRTI. A properly validated plan for patients who contract tuberculosis on the lopinavir-ritonavir NNRTI regimen is still awaited; initial tentative suggestions might be to increase the lopinavir-ritonavir to 6 capsules 2x/d^[127], although this study looked at rifampin and not rifampicin.

Perspective: Stability of lactate specimens over time

The lactate in transported specimens rises, leading to universal recommendations that it be transported on ice. This is technically difficult and time consuming. In a study on normal volunteers^[128], the magnitude of the rise in specimens in fluoride tubes was relatively small (0.2 mmol/l) and is probably not clinically important. The same concept (measurable rise leading to a technically difficult recommendation of unlikely clinical value) applies to prolonged tourniquet application – mean increase of 0.12 mmol/l at five minutes^[129].

Other specific drug reaction syndromes

Skin rashes

Minor rashes may be due to something else (scabies, papular urticaria) but could still be drug induced.

Check temperature, LFTs and eosinophil count.

If the patient is unwell, febrile, or has a transaminase rise, stop all three drugs and ask for help.

The rashes due to nevirapine or efavirenz usually start within a month or two of commencing treatment. Up to 16% of patients given nevirapine may develop rash if the drug is started without dose escalation. The proportion is lower if the drug is started slowly. A severe rash ascribed to nevirapine should warrant a swap to efavirenz, but if it was severe enough to qualify as a nevirapine-associated Stevens-Johnson, consider switching to lopinavir-ritonavir. Patients developing hypersensitivity to abacavir should not be re-challenged with this agent.

Nausea and vomiting

Consider the possibility of lactic acidosis (usually only after about 6 months of treatment) pancreatitis, and hepatitis, but in general a minor degree of nausea occurring early after starting treatment usually settles with metoclopramide and continuing the drugs. Remember that very thin young patients can develop oculogyric crises after treatment with metoclopramide – it goes away quickly anyway, but responds to biperiden or orphenadrine.

Hepatomegaly

This may be due to hepatic steatosis (often stavudine, but actually any NRTI). Ultrasound is helpful in suggesting the diagnosis.

Hepatitis

About 6% of patients given ART will develop some evidence of hepatic injury^[130] although clinical symptoms may only occur in 1.3 patients per 1000 person years.

In the case of the NRTIs it is presumed to be related to mitochondrial injury, and typically, but not always, happens after more than 6 months of treatment, and is accompanied by LDH elevation and lactic acidosis as well as the transaminase rise. There is also an association with peripheral lipodystrophy (face and arms, legs).

Nevirapine is an exception in that the liver injury can occur during the first 8 weeks, and may be associated with a leukocytosis and eosinophilia. Drug reactions due to nevirapine are said to be more severe once immunity has reconstituted, so think very carefully about changing to this agent once the CD4 is above 200.

The prevalence of PI associated hepatotoxicity is proportional to cumulative dose, and there may be associated hyperlipidaemia. It is unclear how well various reputations are deserved, but stavudine is considered a common cause of hepatotoxicity amongst the NRTIs, whereas lamivudine is thought to be relatively safe.

In all cases of suspected ART associated liver disease, it is important to consider the potential role of Hep B and C, where immune reconstitution may lead to a flare of the underlying viral hepatitis. Differentiating the two is not always easy, as the hepatotoxicity due to the drugs also often occurs some months after starting treatment, at a time when immune reconstitution might be expected to be occurring. In one study^[131], with a baseline Hepatitis C prevalence of 48%, about 20% of patients on nevirapine or efavirenz developed a greater than 5 times upper limit of normal elevation in AST or ALT.

Management

Be alert and pick up problems early: liver enzymes should be repeated every two weeks initially when starting nevirapine, and at one month then three monthly in all others where there is increased risk of liver disease (e.g. on TB treatment)

If the enzyme rise is less than 5x ULN, and there is no fever, skin rash or leukocytosis, continue all therapy but check enzymes weekly initially; more than 80% will eventually normalise, although on average this takes nine months.

If the liver injury appears more severe, stop all therapy and re-evaluate once settled; re-introduction should be done under close biochemical supervision, and particularly hepatotoxic combinations should be avoided – e.g. higher dose didanosine, or didanosine in combination with stavudine. If on nevirapine, stop this and use efavirenz instead (except in pregnancy).

Pancreatitis.

Think of this in patients presenting with abdominal pain, and do an amylase and lipase. HIV opportunistic infections may be to blame, and don’t forget alcohol, but the key consideration is medication. Didanosine and stavudine are listed^[132], with lamivudine also being a possibility, especially in children. Hypertriglyceridaemia due to protease inhibitors has also been implicated, as have other medications such as co-trimoxazole, hydroxyurea and TB drugs (rifampicin). Management is supportive and withdrawal of all potentially offending agents.

Other cause of abdominal pain.

Bloating, flatulence and diarrhoea have been associated with didanosine.

Lipodystrophy.

All PIs and most NRTI may be associated with this. Differentiate lipodystrophy (central obesity, buffalo hump, insulin resistance, dyslipidaemia, usually due to protease inhibitors) from lipoatrophy (loss of fat on limbs and face, usually ascribed to stavudine and zidovudine). True gynaecomastia in males (palpable subareolar breast tissue) is often ascribed to either efavirenz or nevirapine, and appears to resolve spontaneously within a year in about 75% of cases.

Dyslipidaemia on ARVs

This can occur in patients on lopinavir/ritonavir, and if not controlled with diet, switching to atazanavir/ritonavir (300/100mg daily) is possible. If hypertriglyceridaemia persists on this, consider bezafibrate 400mg daily (poorly tolerated.) For hypercholesterolaemia, use atorvastatin 10mg daily (not simvastatin.)

Peripheral neuropathy.

Often due to stavudine, or the combination of stavudine and anti-tuberculous therapy. Will have started after starting the antiretrovirals, so take a history. (And ideally have the absence of peripheral neuropathy documented in the notes prior to starting therapy!) If severe, switch stavudine to zidovudine.

Myopathy.

Usually ascribed to zidovudine – muscle pains/aches and a raised creatine kinase. Screen for lactic acidosis.

Dysphoria.

Vivid dreams, dizziness, and other ‘soft’ CNS symptoms have been associated with efavirenz. Switch to nevirapine if severe.

Dysgeusia.

Ritonavir – depending on how much effect on quality of life the alteration of taste is causing, it may be possible to carry on with therapy.

Anaemia or neutropenia.

This is usually blamed on zidovudine especially if occurring in the first three months of treatment; lamivudine is occasionally implicated. If on zidovudine, attempt a dose reduction first – e.g. to 200 mg 2x/d, but if Hb is less than 6.0 or WCC < 0.5, the drug should be stopped and replaced with another agent. Blood transfusion may be required initially.

Replacing non-tolerated drugs in a succeeding regimen

The changes mentioned under side effects are appropriate if the regimen is otherwise successful. If it is failing or if it is unclear which of several agents may be responsible, it may be necessary to move on to a new regimen

Moving on – failed regimens

There should be a log fall, or greater, in viral load after one month, and to less than 50 copies per ml by six months. The CD4 count should show a significant upward trend and there should be evidence of clinical improvement – weight gain and wellbeing. If the above has not happened by six months, the reasons for failure should be analysed (even earlier if possible!) with a view to swapping to a second line regimen.

Analysing failure on ART.

Treatment failure is usually due to a combination of factors, which may include:

Poor treatment adherence

Poor choice of regimen (e.g. dual therapy or monotherapy)

Intrinsic multi-drug resistance of the organism the patient acquired.

Drug interactions.

Take a very careful history including inspecting the packets/containers and asking to see all other medication the patient may be taking – the most bizarre things can crop up (e.g. a patient on ART taking activated charcoal as a ‘purifier’!) Testing for resistance is expensive and rarely feasible, so after careful discussion with the patient, it is usually necessary to move on to a second regimen once all potential adherence issues have been addressed. If you still have concerns, address them first before starting the new regimen – as salvage therapy is of limited success and is quite costly, it is better to get the more standard regimens to work than to keep on going on a quest for increasingly sophisticated and costly drugs.

Interpreting CD4/VL patterns

The CD4 and viral loads do not always behave as they should, and some patterns may be disconcerting:

	Viral load suppressed	Viral load static	Viral load rising
CD4 rising	Ideal	If already 2-log fall in titre, just repeat in 6 months. If not 2-log fall, fix adherence	Possible lab error – repeat CD4. Usually means failed regimen, and CD4 drop will happen soon. Fix adherence
CD4 low and not rising	Wait a bit and repeat if patient well or clinically improving.	Adherence and repeat at six months	Failed regimen – fix adherence and consider new regimen if not already on PI; if are, continue same regimen with careful attention to adherence
CD4 falling	Treatment failure or new opportunistic infection – look for the latter and fix adherence. Then consider new regimen	Review adherence while considering new regimen

ARV dosing in patients with renal impairment.

Medication		GFR
		<50	<25	<10
Didanosine	<60 kg	125 mg/d	125 mg/d	125 mg/d
	> 60 kg	200 mg/d	125 mg/d	125 mg/d
Lamivudine		150 mg/d	100 mg/d	50 mg/d
Emtricitabine		200mg2nd day	200mg 3^rd day	200mg 4^th day
Stavudine	<60 kg	15 mg 2x/d	15 mg/d	15 mg/d
	> 60 kg	20 mg 2x/d	20 mg/d	20 mg/d
Tenofovir		300 mg alt days	300 mg 2x/wk	300 mg/wk
Zidovudine				100 mg 3x/d
Abacavir	No dose adjustment
Efavirenz	No dose adjustment
Nevirapine	No dose adjustment
Lopinavir/ritonavir	No dose adjustment

These doses are a representative guide only – it is always preferable to calculate the GFR properly and then consult the individual package insert for each medication.

Wilson D, Naidoo S, Bekker L, et al. Handbook of HIV medicine. Oxford University Press. 2002. ↑
South African National Policy on HIV Testing. Department of Health, 1 December 1999. ↑
Merchant RC, Waxman MJ. HIV screening in health care settings. JAMA 2010;304:348-9. ↑
Bor R, Miller R, Johnson M. A testing time for doctors: counselling patients before an HIV test. BMJ 1991;303:905-7. ↑
Evian C. Primary AIDS Care. Jacana Press. p163 -176 ↑
Bem C. Human immuodeficiency virus positive tuberculous lymphadenitis in Central Africa: clinical presentation of 157 cases. Int J Tubercl Lung Dis. 1887;1:215-9. ↑
Malin A, Ternouth I, Sarbah S. Epitrochlear lymph nodes as marker of HIV disease in sub-Saharan Africa. BMJ 1994;309:1550-1. ↑
Arango CA, Midani S, Alvarez A, et al. Usefulness of acute phase reactants in the diagnosis of acute infections in HIV-infected children. South Med J. 1999;92:209-13. ↑
Storgaard M, Laursen AL, Andersen PL. The C-reactive protein responses in HIV-infected patients with pneumonia. Scand J Infect Dis.1993;25:305-9. ↑
Redd AD, Eaton KP, Kong X, et al. C-reactive protein levels increase during HIV-1 disease progression in Rakai, Uganda, despite the absence of microbial translocation. JAIDS 2010;54:556-9. ↑
Lala SG, Madhi SA, Pettifor JM. The discriminative value of C-reactive protein levels in distinguishing between community-acquired bacteraemic and respiratory virus-associated lower respiratory tract infections in HIV-1-infected and -uninfected children. Ann Trop Paediatr 2002; 22:271-9 ↑
Hopstaken RM, Stobberingh EE, Knottnerus JA, et al. Clinical items not helpful in differentiating viral from bacterial lower respiratory tract infections in general practice. J Clin Epidemiol. 2005;58:175-83. ↑
Korppi M. Non-specific host response markers in the differentiation between pneumococcal and viral pneumonia: what is the most accurate combination? Pediatr Int.2004;46:545-50. ↑
Drain PK, Mayeza L, Bartman P et al. Diagnostic accuracy and clinical role of rapid CRP testing in HIV-infected individuals with presumed tuberculosis in South Africa. Int J Tubercl Lung Dis. 2014;18:20-26. http://dx.doi.org/10.5588/ijtld.13.0519 ↑
Rerkpattanapipat P, Wongpraparut N, Jacobs LE, et al. Cardiac manifestations of acquired immunedeficiency syndrome. Arch Intern Med. 2000;160:602-8. ↑
Limsukon A, Saeed AI, Ramasamy V, et al. HIV-related pulmonary hypertension. Mount Sinai J Med 2006;73:1037-44. ↑
Petrosillo N, FloresSC, Almodovar S, et al. Pulmonary hypertension associated with HIV-Taskforce. PVRI Review 2010;2:42-4. ↑
Mehta NJ, Khan IJ, Mehta RN et al. HIV related pulmonary hypertension. Chest 2000;118:1133-41. ↑
Engels EA, Rosenberg PS, Biggar RJ. Zoster incidence in human immunodeficiency virus-infected hemophiliacs and homosexual men. J Inf Dis. 1999;180:1784-1789. ↑
Khost RG, Straus SE. Postherpetic neuralgia – pathogenesis, treatment and prevention. N Engl J Med. 1996;335:32-40. ↑
Onunu AN, Uhunmwangho A. Clinical spectrum of herpes zoster in HIV-infected versus non-HIV infected patients in Benin City, Nigeria. West Afr J Med. 2004;23(4):300-4. ↑
Jackson JL, Gibbons R, Meyer G, et al. The effect of treating herpes zoster with oral acyclovir in preventing postherpetic neuralgia. A meta-analysis. Arch Intern Med. 1997;157(8):909-12. ↑
Dubinsky RM, Kabbani H, El-Chami, et al. Practice parameter: treatment of postherpetic neuralgia. Neurology. 2004;63:959-65. ↑
Thuluvath PJ, Connolly GM, Forbes A, et al. Abdominal pain in HIV infection. QJM. 1991;287:275-85. ↑
O’Keefe EA, Wood R, Van Zyl A, et al. Human immunodeficiency virus-related abdominal pain in South Africa. Aetiology, diagnosis and survival. Scand J Gastroenterol.1998;33:212-7 ↑
Bek B, Boeckh M, Lepenies J, et al. High level sensitivity of quantitative pp65cytomegalovirus (CMV) antigenaemia assay for diagnosis of CMV disease in AIDS patients and follow up. J Clin Microbiology 1996;34:457-9. ↑
Kerr AR, Ship JA. Management strategies for HIV-associated aphthous stomatitis. Am J Clin Dermatol. 2003;4:669-80. ↑
Vasquez JA. Invasive oesophageal candidiasis: current and developing treatment options. Drugs. 2003;63:971-89. ↑
Nyst MJ, Perriens JH, Kimputu L, et al. Gentian violet, ketoconazole and nystatin in oropharyngeal and oesophageal candidiasis in Zairian AIDS patients. Ann Soc Belg Med Trop.1992;72:45-52. ↑
Khosla S, Wolfson JS, Demerjian Z, et al. Adrenal crisis in the setting of high-dose ketoconazole therapy. Arch Intern Med.1989;149:802-4. ↑
Vasquez JA. Options for the management of mucosal candidiasis in patients with AIDS and HIV infection. Pharmacotherapy. 1999;19:76-87. ↑
Albougy HA, Naidoo S. A systematic review of the management of oral candidiasis associated with HIV/AIDS. SADJ. 2002;57:457-66. ↑
Powderley WG, Mayer KH, Perfect JR. Diagnosis and treatment of oropharyngeal candidiasis in patients infected with HIV: a critical reassessment. AIDS RES HUM Retroviruses. 1999;15:1405-12. ↑
Fitchenbaum C, Zackin R, Rajicic N, et al. Amphotericin B oral suspension for fluconazole-refractory oral candidiasis in persons with HIV infection. AIDS 2000;14:845-52. ↑
Wilcox C. Short report: time course of clinical response with fluconazole for Candida oesophagitis in patients with AIDS. Aliment Pharmacol and Therapeutics. 1994;8:347-50. ↑
Itescu S, Brancato LJ, Buxbaum J, et al. A diffuse infiltrative CD8 lymphocytosis syndrome in HIV infection: a host immune response associated with HLA-DR5. Ann Intern Med. 1990;112:3-10. ↑
Parry E, Godfrey R, Mabey D et al. Principles of medicine in Africa. Cambridge University Press. 2004 ↑
Brook MG, Ayles H, Harrison C, et al. Diagnostic utility of bone marrow sampling in HIV positive patients. Genitourinary Medicine. 1997;73:117‑21 ↑
Anderson D, Ali K, Blanchette V, et al. Guidelines on the use of intravenous immune globulin for hematologic conditions. Transfusion Med Reviews. 2007; 21(2), S1, S9-S56. ↑
Vamvakas EC, Blajchman MA. Deleterious clinical effects of transfusion-associated immunomodulation: fact or fiction. Blood. 2001;97:1180-95. ↑
Sacktor NC, Nakasujja N, Skolasky R, et al. Antiretroviral therapy improves cognitive impairment in HIV+ individuals in sub-Saharan Africa. Neurology. 2006;67:311-14. ↑
Sacktor NC, Wong M, Nakasujja N, et al. The International HIV Dementia Scale: a new rapid screening test for HIV dementia. AIDS 2005;19:1367-74. ↑
Dismukes WE, Pappas PG, Sobel JD. Clinical mycology. Oxford University Press. 2003 p40. ↑
Larsen RA, Leal MA, Chan LS. Fluconazole compared with amphotericin B plus flucytosine for cryptococcal meningitis in AIDS. Ann Int Med. 1990;113:183-7. ↑
Saag MS, Powderly WG, Cloud GA, et al. Comparison of amphotericin B with fluconazole in the treatment of acute AIDS-associated cryptococcal meningitis. The NIAID Mycoses Study Group and the AIDS Clinical Trials Group. N Engl J Med. 1992;326:83-9. ↑
Brouwer AE, Rajanuwong A, Chierakul W, et al. Combination antifungal therapies for HIV-associated cryptococcal meningitis: a randomised trial. Lancet. 2004;363:1764-7. ↑
Day JN, Chau TT, Wolbers M, et al. Combination antifungal therapy for cryptococcal meningitis. N Engl J Med 2013;368:1291-302. ↑
Boulware DR, Meya D, Muzoora C, et al. ART initiation within the first 2 weeks of cryptococcal meningitis is associated with higher mortality: A multisite randomized trial. Abstract 144. Conference on retroviruses and opportunistic infections (CROI), Atlanta, USA, 6 March 2013 ↑
Levay PF, Gryffenberg H. Do normal C-reactive protein levels exclude the diagnosis of cryptococcal meningitis? SA Fam Pract. 2006;48:14. ↑
Modi M, Mochan A, Modi G. Management of HIV-associated focal brain lesions in developing countries. Q J Med. 2004;97:413-21. ↑
Tipping B, de Villiers L, Wainright H, et al. Stroke in patients with human immunodeficiency virus. J Neurol Neurosurg Psychiatry. 2007;78:1320-4. ↑
Gendelman HE, Lipton SA, Epstein L, et al. The neurology of AIDS. Chapman & Hall 1998 p276. ↑
Skopelitis EE, Kokotis PI, Kontos AN, et al. Distal sensory polyneuropathy in HIV-positive patients in the ART era: an entity underestimated by clinical examination. Int J STD & AIDS. 2006;17:467-72. ↑
Brew BJ, Tisch S, Law M. Lactate concentrations distinguish between nucleoside neuropathy and HIV neuropathy. AIDS 2003;17:1094-6. ↑
Lasso M, Perez J, Noriega LM, et al. Polyradiculopathy caused by cytomegalovirus in AIDS patients: successful treatment with highly active antiretroviral therapy (ART.) Rev Med Chil. 2001;129:1061-4. (Abstract only.) ↑
Bradley WG, Verma A. Painful vasculitic neuropathy in HIV-1 infection: relief of pain with prednisone therapy. Neurology. 1996;47:1446-51. ↑
Kallenberg CG. Overlapping syndromes, undifferentiated connective tissue disease, and other fibrosing conditions. Curr Opin Rheumatol. 1995;7:568-73. ↑
Hughes RA, van Der Meche FG. Corticosteroids for treating Guillain-Barre syndrome. Cochrane Database Syst Rev. 2000;(3):CD001446. ↑
Mehndiratta MM, Hughes RA. Corticosteroids for chronic inflammatory demyelinating polyradiculoneuropathy. Cochrane Database Syst Rev. 2002;(1):CD002062. ↑
Hughes RA. Management of chronic inflammatory demyelinating polyradiculoneuropathy. Drugs. 2003;63:275-87. ↑
Ropper AH. Current treatments for CIDP. Neurology. 2003;60:S16-22. ↑
Hernandez-Albujar S, Arribas JR, Royo A, et al. Tuberculous radiculomyelitis complicating tuberculous meningitis: case report and review. Clin Infect Dis. 2000;30:915-21. ↑
Delobel P, Signate A, El Guedj M, et al. Unusual form of neurocysticercosis associated with HIV infection. Eur J Neurol. 2004;11:55-8. ↑
Kane SL, Dasta JF. Clinical outcomes of critical illness polyneuropathy. Pharmacotherapy. 2002;22:373-9. ↑
Marzocchetti A, Di Giambenedetto S, Cingolani A, et al. . Reduced Rate of Diagnostic Positive Detection of JC Virus DNA in Cerebrospinal Fluid in Cases of Suspected Progressive Multifocal Leukoencephalopathy in the Era of Potent Antiretroviral Therapy. J. Clin. Microbiol. 2005, 43(8):4175. ↑
Wang Y, Kirby JE, Qian Q. Effective use of JC virus PCR for diagnosis of progressive multifocal leukoencephalopathy. J Med Microbiol 2009;58:253–255 ↑
Noorbakhsh F, Johnson RT, Emery D, et al. Acute disseminated encephalomyelitis: clinical and pathogenesis features. Neurol Clin 2008;26:759–780. ↑
Bhigjee AI, Patel VB, Bhagwan B, et al. HIV and acute disseminated encephalomyelitis. South Afr J Med 1999;89:283-4. ↑
Titulaer MJ, McCracken L, Gabilondo I, et al. Treatment and prognostic factors for long-term outcome in patients with anti-NMDA receptor encephalitis: an observational cohort study. Lancet Neurol 2013; 12: 157–65. ↑
Leport C, Chene G, Moriat P, et al. Pyrimethamine for primary prophylaxis of toxoplasmic encephalitis in patients with human immunodeficiency virus: a double-blind randomised trial J Infect Dis 1996;173:91-97. ↑
Torre D, Cari S, Speranza F, et al. Randomised trial of trimethoprim-sulphamethoxazole versus pyrimethamine-sulphadiazine for therapy of toxoplasmic encephalitis in patients with AIDS. Antimicrobial Agents and Chemotherapy. 1998;42:1346-1349. ↑
Rao TK, Friedman EA. Outcome of severe acute renal failure in patients with acquired immunodeficiency syndrome. Am J Kidney Dis. 1995;25:390-8. ↑
Daugas E, Rougier JP, Hill G. ART-related nephropathies in HIV-infected patients. Kidney Int. 2005;67:393-403. ↑
Gerntholtz TE, Goetsch SJ, Katz I. HIV-related nephropathy: a South African perspective. Kidney Int. 2006;69:1885-91. ↑
Haas M, Kaul S, Eustace JA. HIV-associated immune complex glomerulonephritis with “lupus-like” features: a clinicopathologic study of 14 cases. Kidney Int. 2005;67:1632-3. ↑
Kaplan LD, Straus DJ, Testa MA, et al. Low-dose compared with standard-dose m-BACOD chemotherapy for non-Hodgkin’s lymphoma associated with human immunodeficiency virus infection. N Engl J Med. 1997;336:1641-8. ↑
Lidman C, Berglund O, Tynell E, et al. CD4 + cells and DC4+ percent as risk markers for Pneumocystis carinii pneumonia (PCP): implications for primary PCP prophylaxis. Scand J Infect Dis 1992; 24:157-160. ↑
Centers for Disease Control. Update: acquired immunodeficiency syndrome: United States. Morb Mortal Wkly Rep 1986;35:757-766. ↑
Zaman, Muhamed K., et al. Rapid Noninvasive Diagnosis of Pneumocystis carinii from Induced Liquified Sputum. Annl Int Med. 1988; 109:7-10. ↑
Pitchenik, Arthur E., et al. Sputum Examination for the Diagnosis of Pneumocystis carinii Pneumonia in the Acquired Immunodeficiency Syndrome. Am Rev Resp Dis. 1986;133: 226-229. ↑
Bustamante, Enrique A., et al. Sputum Induction Compared with Bronchoalveolar Lavage by Ballard Catheter to Diagnose Pneumocystis carinii Pneumonia. Chest. 1994;105:816-822. ↑
Thomas CF, Limper AH. Pneumocystis pneumonia. N Engl J Med. 2004:350:2487-98. ↑
Fishman JA. Treatment of infection due to Pneumocystis carinii. Antimicrobial Agents and Chemotherapy. 1998;42:1309-1314. ↑
Smego RA, Nagar S, Maloba B, et al. A meta-analysis of salvage therapy for Pneumocystis carinii pneumonia. Arch Intern Med. 2001;161:1529-33. ↑
Selwyn PA, Pumerantz AS, Durante A, et al. Clinical predictors of Pneumocystis carinii pneumonia, bacterial pneumonia and tuberculosis in HIV infected patients. AIDS 1998;12:885-93. ↑
Schlomer DD. Induced Sputum may be an Acceptable Initial Investigation in the Diagnosis of Pneumocystis Carinii Pneumonia. Oxford Centre for Evidence Based Medicine CAT 1999. ↑
Strevel EL, Kuper A, Gold WL. Severe and protracted hypoglycaemia associated with co-trimoxazole use. Lancet Infect Dis. 2006;6:178-82. ↑
Sande MA, Volberding PA. The medical management of AIDS. 3rd Ed. 1992, p433. ↑
Geng E, Kreiswirth B, Burzynski J, et al. Clinical and radiographic correlates of primary and reactivation tuberculosis. A molecular epidemiological study. JAMA. 2005;293:2740-5. ↑
USPHS/IDSA Prevention of Opportunistic Infections Working Group. 1999 USPHS/IDSA Guidelines for the prevention of opportunistic infections in persons infected with human immunodeficiency virus. Ann Intern Med. 1999;131:873-908. ↑
Anglaret X, Chene G, Attia A, et al. Early chemoprophylaxis with trimethoprim-sulphamethoxazole for HIV-1-infected adults in Abidjan, Cote d’Ivoire: a randomised trial. Lancet. 1999;353:1463-68. ↑
Wiktor SZ, Sassan-Morokro M, Grant AD, et al. Efficacy of trimethoprim-sulphamethoxazole prophylaxis to decrease morbidity and mortality in HIV-1-infected patients with tuberculosis in Abidjan, Cote d’Ivoire: a randomised controlled trial. Lancet. 1999; 353:1469-75. ↑
Maynart M, Lievre L, Sow P S, et al. Primary prevention with cotrimoxazole for HIV-1-infected adults: results of the pilot study in Dakar, Senegal.. J Acquir Immune Defic Syndr. 2001;26:130-6. ↑
Grimwade K, Swingler, G. Cotrimoxazole prophylaxis for opportunistic infections in adults with HIV. The Cochrane Database of systematic reviews 2003, Issue 3. Art. No.: CD003108. ↑
Walker AS, Ford D, Gilks CF, et al. Daily co-trimoxazole prophylaxis in severely immunosuppressed HIV-infected adults in Africa started on combination antiretroviral therapy: an observational analysis of the DART cohort. Lancet 2010;375:1278-86. ↑
Suthar AB, Vitoria MA, Nagata JM et al. Co-trimoxazole prophylaxis in adults, including pregnant women, with HIV: a systematic review and meta-analysis Lancet HIV 2015; 2: e137–50 http://dx.doi.org/10.1016/S2352-3018(15)00005-3 ↑
Straatmann A, Bahia, F, Pedral-Sampaio D, et al. A randomized, pilot trial comparing full versus escalating dose regimens for the desensitization of AIDS patients allergic to sulphonamides. Braz J Infect Dis. [online]. 2002;6:276-280. (http://www.scielo.br/scielo.php?script=sci_arttext&pid=S1413-86702002000600002&lng=en&nrm=iso). ↑
Messaad D, Sahla H, Benahmed S, et al. Drug provocation tests in patients with a history suggesting an immediate drug hypersensitivity reaction. Ann Intern Med. 2004;140:1001-6. ↑
Richter C, Perenboom R, Swai AB, et al. Diagnosis of tuberculosis in patients with pleural effusion in an area of HIV infection and limited diagnostic facilities. Trop Geogr Med. 1994;46:293-7. ↑
Llibre JM, Tor J, Manterola JM, et al. Risk stratification for the dissemination of tuberculosis in HIV-infected patients. J Med 1992;298:149-57. ↑
Gandhi NR, Moll A, Sturm AW, et al. Extensively drug-resistant tuberculosis as a cause of death in patients co-infected with tuberculosis and HIV in a rural area of South Africa. Lancet. 2006;368:1575-80. ↑
Holtz TH, Kabera G, Mthiyane T, et al. Use of a WHO-recommended algorithm to reduce mortality in seriously ill patients with HIV infection and smear-negative pulmonary tuberculosis in South Africa: an observational cohort study. Lancet Infect Dis. 2011;11:533-40. ↑
Grant AD, Charalambous S, Fielding KL, et al. Effect of routine isoniazid preventive therapy on tuberculosis incidence among HIV-infected men in South Africa. JAMA. 2005;293:2719-25. ↑
Churchyard GJ, Fielding K, Charalambous S, et al. Efficacy of secondary isoniazid preventive therapy among HIV-infected Southern Africans: time to change policy? AIDS 2003;17:2063-70. ↑
Rangaka MX, Wilkinson RJ, Boulle A, et al. Isoniazid plus antiretroviral therapy to prevent tuberculosis: a randomised double-blind, placebo-controlled trial. Lancet. 2014;384:682-690. http://dx.doi.org/10.1016/S0140-6736(14)60162-8 ↑
Watson RM, Roach NA, Dalinka MK. Avascular necrosis and bone marrow oedema syndrome. Radiol Clin N Am. 2004;42:207-19. ↑
Miller KD, Masur H, Jones EC, et al. High prevalence of osteonecrosis of the femoral head in HIV-infected adults. Ann Intern Med. 2002;137:17-24. ↑
Henderson DK. Postexposure chemoprophylaxis for occupational exposures to the human immunodeficiency virus. JAMA. 1999;281:931-936. ↑
Gerberding JL. Prophylaxis for occupational exposure to HIV. Annl Int Med 1996;125:497-501. ↑
Case-control study of HIV sero-conversion in health care workers after percutaneous exposures to HIV infected blood. MMWR 1995;44:929-33. ↑
Lawn SD, Myer L, Orrell C, et al. Early mortality among adults accessing a community-based antiretroviral service in South Africa: implications for programme design. AIDS. 2005;19:2141-8. ↑
Clifford DB, Evans S, Yang Y, et al. Impact of efavirenz on neuropsychological performance and symptoms in HIV-infected individuals. Ann Intern Med. 2005;143:714-21. ↑
Sonderup MW, Maughan D, Gogela N, et al. Identification of a novel and severe pattern of efavirenz drug-induced liver injury in South Africa. AIDS 2016;30:1483-1485. ↑
Manfredi R, Calza L. Nevirapine versus efavirenz in 742 patients: no link of liver toxicity with female sex and a baseline CD4 count greater than 250 cells per microlitre. AIDS. 2006;20:2233-6. ↑
Laurent C, Koouanfack C, Kaulla-Shiro S, et al. Effectiveness and safety of a generic fixed-dose combination of nevirapine, stavudine and lamivudine in HIV-1-infected adults in Cameroon: open-label multicentre trial. Lancet. 2004;364:29-34. ↑
Van Leth F, Phanuphak P, Ruxrungtham K, et al. Comparison of first-line antiretroviral therapy with regimens including nevirapine, efavirenz, or both drugs, plus stavudine and lamivudine: a randomised open-label trial, the 2NN study. Lancet 2004;363:1253-63. ↑
Hira SK, Panchal K, Parmar PA, et al. High resistance to anti-retroviral drugs: the Indian experience. Int J STD and AIDS. 2004;15:173-7. ↑
Dybul M, Fauci AS, Bartlett JG, et al. Guidelines for using antiretroviral agents among HIV-infected adults and adolescents. Ann Intern Med. 2002;137:381-433. ↑
Haynes RB, McDonald H, Garg AX, et al. Interventions for helping patients to follow prescriptions for medications. (Cochrane Review). The Cochrane Library, issue 3, 2003. Oxford: Update Software. ↑
Fawzi WW, Msamanga GI, Spiegelman D, et al. A randomized trial of multivitamin supplements and HIV disease progression and mortality. N Engl J Med. 2002;351:23-32. ↑
Jiamton S, Pepin J, Suttent R, et al. A randomized trial of the impact of multiple micronutrient supplementation on mortality among HIV infected individuals living in Bangkok. AIDS 2003; 17: 2461-9. ↑
Irlam JH, Visser ME, Rollins N, et al. Micronutrient supplementation in children and adults with HIV infection. Cochrane Database Syst Rev 2005 Oct 19;(4):CD003650 ↑
Shelburne SA, Hamill RJ. The immune reconstitution inflammatory syndrome. AIDS Reviews. 2003;5:67-79. ↑
Wimalasundera RC, Larbalestier N, Smith JH, et al. Pre-eclampsia, anti-retroviral therapy and immune reconstitution. Lancet. 2002;360:1152-4. ↑
Lawn SD, Bekker L, Miller RF. Immune reconstitution disease associated with mycobacterial infections in HIV-infected individuals receiving antiretrovirals. Lancet Infect Dis. 2005;5:361-73. ↑
Venkataramana A, Pardo CA, McArthur JC, et al. Immune reconstitution inflammatory syndrome in the CNS of HIV-infected patients. Neurology 2006;67:383-8. ↑
La Porte CJ, Colbers EP, Bertz R, et al. Pharmacokinetics of adjusted-dose lopinavir-ritonavir combined with rifampin in healthy volunteers. Antimicrob Agents Chemotherapy. 2004;48:1553-60. ↑
Astles R, Williams CP, Sedor F. Stability of plasma lactate in vitro in the presence of antiglycolytic agents. Clin Chem. 1994;40:1327-30. ↑
Bany-Mohammed FM, Macknin ML, van Lente FL, et al. The effect of prolonged tourniquet application on serum bicarbonate. Cleve Clin J Med. 1995;62:68-70. ↑
Spengler U, Lichterfeld M, Rockstrh K. Antiretroviral drug toxicity – a challenge for the hepatologist? J Hepatol 2002;36:283-94. ↑
Ena J, Amador C, Benito C, et al. Risk and determinants of developing severe liver toxicity during therapy with nevirapine-and efavirenz-containing regimens in HIV-infected patients. Int J STD & AIDS. 2003;14:776-81. ↑
Manfredi R, Calza L, Chiodo F. A case-controlled study of HIV-associated pancreatic abnormalities during the ART era. Focus on emerging risk factors and specific management. Eur J Med Res. 2004;9:537-44. ↑

Vidal, J E, Peixoto de Miranda, E J F, Gerhardt, J, Croda, M, & Boulware, D R. (2017). Is it possible to differentiate tuberculous and cryptococcal meningitis in HIV-infected patients using only clinical and basic cerebrospinal fluid characteristics?. SAMJ: South African Medical Journal, 107(2), 156-159. https://dx.doi.org/10.7196/samj.2017.v107i2.11162 ↩
Rutakingirwa MK, Kiiza TK, Joshua Rhein J.“False negative” CSF cryptococcal antigen with clinical meningitis: Case reports and review of literature.
Medical Mycology Case Reports. 2020;29:29-31.https://doi.org/10.1016/j.mmcr.2020.06.003.(https://www.sciencedirect.com/science/article/pii/S2211753920300403 ↩
community-based early adherence support in people with advanced HIV infection starting antiretroviral therapy in Tanzania and Zambia: An open-label, randomised controlled trial. Lancet. 2015;385(9983):2173–2182. https://doi.org/10.1016/S0140-6736(15)60164-7 ↩
Tenforde MW, Shapiro AE, Rouse B, Jarvis JN, Li T, Eshun-Wilson I, Ford N. Treatment for HIV-associated cryptococcal
meningitis. Cochrane Database of Systematic Reviews 2018, Issue 7. Art. No.: CD005647. DOI: 10.1002/14651858.CD005647.pub3. ↩

Created On20/01/2022

Last Updated On24/02/2022

0 out Of 5 Stars

5 Stars		0%
4 Stars		0%
3 Stars		0%
2 Stars		0%
1 Stars		0%

Was this article helpful?

0 out Of 5 Stars

5 Stars		0%
4 Stars		0%
3 Stars		0%
2 Stars		0%
1 Stars		0%

How Can We Help?

HIV

0 out Of 5 Stars

General

HIV testing

Counselling checklist

Procedure for HIV testing

Assessment:

Examination:

Poor appetite:

Difficulty eating

Anabolic medication

HIV and adenopathy.

Perspective – HIV adenopathy.

Perspective – epitrochlear nodes in HIV.

Initial management of adenopathy.

When not to do a needle aspiration node biopsy.

Management of nodes where FNAB is unhelpful

HIV and C-reactive protein (CRP)

Cardiology – HIV

HIV and arrhythmias.

HIV and pulmonary hypertension/right heart failure.

HIV and cardiomyopathies.

HIV and pericardial effusions.

HIV and ischaemic heart disease.

HIV and rheumatic heart disease.

HIV and blood pressure.

Dermatology – HIV

Overview

Itchy rashes.

Drug hypersensitivity reactions

Vesicular eruptions

Herpes zoster in HIV

Histoplasmosis

Kaposi’s sarcoma

Kaposi’s mimics

Gastrointestinal disorders in HIV

Abdominal pain in HIV

CMV infection

Mouth problems associated with HIV

Oral candidiasis

Perspective – oral candidiasis

Oesophageal candidiasis.

Parotid disease in HIV.

Peri-anal pathology in HIV

Diarrhoea in HIV

Perspective – treating the treatable.

Hepatitis and hepatic pain in HIV

Non-hepatitic RUQ pain in HIV

HIV and perineal disease

Approach to suspected abdominal tuberculosis in HIV.

Perspective – alternative diagnoses in patients with RUQ pain.

Haematological – HIV.

ITP

TTP (thrombotic thrombocytopenic purpura)

Chronic anaemias and pancytopenias associated with HIV

Haemolytic anaemias.

Hyperviscosity syndromes and vasculitides in HIV

Metabolic syndromes in HIV.

Diabetes and HIV.

HIV and possible Addison’s disease.

HIV and severe cachexia.

Neurology – HIV.

HIV associated dementia.

Perspective – screening for HIV associated dementia

Chronic headache in HIV positive individuals

Management:

Cryptococcal meningitis in HIV

HIV and the slightly abnormal CSF that is India Ink negative

HIV and seizure disorders

Suspected meningitis in HIV positive patients with focal signs: can the LP wait?

HIV-associated focal brain lesions

Perspective – referral bias

HIV and stroke

HIV and cranial neuropathies

HIV and paraplegias and pseudo-paraplegias

HIV and peripheral neuropathy

AIDS-associated neuropathy

Antiretroviral neurotoxicity (‘nucleoside neuropathy’)

Perspective – distinguishing DSPN and nucleoside neuropathy