How Can We Help?
Systemic lupus erythematosus
Overview
Previous diagnostic criteria required finding four or more of the ARA criteria to make the diagnosis quite likely, but simply because you can’t find four did not exclude the diagnosis.
- Skin – malar rash, painless oral ulcers, photosensitivity, discoid lupus (each counts as one criterion)
- Joints – non-deforming arthritis
- CNS – fits and psychosis
- Serous cavities – pleural effusion and pericardial effusion
- Renal – proteinuria (>0.5g/d) and casts associated with a range of glomerulonephritides.
- Haematological – anaemia, thrombocytopaenia, lymphopenia
- Positive ANF
The revised criteria are:
The 2019 criteria were developed to encourage earlier diagnosis and effective treatment, and in the initial work seem to have achieved this. Subsequent articles focusing on adolescents and young adults, and with new datasets, show less impressive (if any) improvements in sensitivity. However the diagnosis is now based on a positive ANA as an entry criterion, with accrual of 10 further points from the different categories (only one item can be used from each category.)
Prognosis.
In studies published in the 1990s, 5 year survival exceeded 90% as opposed to studies from earlier decades where survival was closer to 70%. There are now some concerns that treatment side-effects may be limiting further improvements in survival.1
Work-up
If you suspect the illness, remember that the diagnosis is primarily clinical, so put some effort into looking for the features listed above. Useful tests in all such patients are an ANF and an HIV test. Drilling down with further serology is fascinating, but expensive and rarely if ever changes management.
Classify as having life-threatening disease or not: skin rash and a bit of arthritis is not, renal failure and polyserositis with psychosis is. Clearly this is nearly always a bit subjective, but the key criterion should be to imagine what would happen to the patient if the particular feature were to worsen.
Note that even the presence of four criteria does not give a probability of disease of much more than 25% when the prevalence is 1%, which is not an unreasonable estimate for a teaching hospital. In rural areas the prevalence is likely to be even lower.
Treatment of SLE
Mild non-life-threatening skin and joint disease can be managed with non-steroidal anti-inflammatories and chloroquine 150 mg/d (omit on weekends)
Serositis usually responds promptly to prednisone 0.5–1 mg/day, as do most of the haematological changes. Renal and CNS lupus is often treated the most aggressively, usually with a combination of prednisone 2 mg/kg/d and cyclophosphamide 1 to 2 mg/kg/d orally, although there is currently a vogue for bolus once a month intravenous therapy using the same medications. Which is better is still unclear.
In order to allow a reduction in the dose of steroids, azathioprine and prednisone may be combined (in patients not on cyclophosphamide!)
HIV and SLE in combination.
Manage each on its own merits. Chloroquine does no harm in HIV, and there is even a bit of evidence to suggest it might be beneficial. If corticosteroids are needed for SLE, then use them, but limit course duration (wean earlier than you might do otherwise). Renal disease, and especially nephrotic syndrome, can be troublesome, as both conditions may cause focal segmental glomerulosclerosis.
SLE and poor medication adherence.
The neuropsychiatric manifestations of the illness do result in there being perhaps a higher proportion of individuals with reduced ability to take medication. This can be enormously frustrating on the one side, and a source of considerable gratification if you and the patient do eventually get it right, because increasing adherence may lead to improved cognition and thus better adherence. In the non-adherent state, getting assistance from a relative or hospitalising for a short time may be necessary. Parenteral monthly therapy also has a lot to offer in this situation.
Renal biopsy in SLE
Indications are always a bit controversial. It is of lesser value in the patient who is flaring in other ways and whom you plan to treat aggressively in any case (i.e. regardless of the biopsy result). Patients who are not particularly active but who have heavy proteinuria or deteriorating GFR probably warrant biopsy, the idea being that in this group the finding of very active disease would lead to more aggressive treatment for the renal pathology in its own right.
Over-treated patients with mild SLE.
Some patients with SLE seem to get lots of steroids for a long time, even when they only have mild skin disease. It is worth considering whether such patients wouldn’t be better off on nothing, or a bit of chloroquine.
Perspective – interpreting results of ‘collagen screens’
The first principle is that some of the tests are very sensitive and are useful for screening for auto-immune disease; if positive they don’t take you much further, but if negative they quite dramatically reduce the probability of disease. Other tests are specific – they aren’t often positive but when they are it means that it is quite likely that the patient has the associated disease. An example of a sensitive test is an ANF. An example of a specific test is an Scl-70 (systemic sclerosis). A further complication is that the tests are seldom performed in unselected patients, so the generalisability of test results is sometimes difficult to judge2.
| Test | Sens | Spec | LR+ | LR– |
|---|---|---|---|---|
| ANF for identifying3 | 0.94 | 0.97 | 31.3 | 0.1 |
| Combined ANF + RF* | 0.75 | 0.93 | 10.7 | 0.3 |
| Anti-dsDNA for SLE | 0.73 | 0.98 | 36.5 | 0.28 |
(*When used to differentiate autoimmune rheumatic diseases (such as RA, SLE and Sjogren’s) from non-autoimmune rheumatic diseases such as osteoarthritis and fibromyalgia)4
Perspective – how likely are false positives
The chance of normal persons having a positive ANA is highly dependent on the cut-off titre chosen 5, with 31.7% of normal people having a titre of 1:40, 13.3% a titre of 1:80, 5% a titre of 1:160, and only 3.3% a titre of 1:320. Stated differently, even a titre of 1:160 still has a one in twenty chance of being a false positive.
Perspective – intravenous or oral treatment for lupus nephritis
“The literature is burdened with clinical anecdotes and suboptimally designed clinical studies6.”
In spite of a great deal of rhetoric, there is probably little to choose between intravenous and oral therapy with cyclophosphamide/prednisone. The former has possibly adherence advantages, although some patients find it quite rigorous treatment; the latter is quick and simple in terms of prescribing and hospital time.
Ugarte A, Ruiz-Irastorza G. SLE: the changing prognosis. Lupus. 2016 Oct;25(12):1285-7. doi: 10.1177/0961203316652948. Epub 2016 Jun 15. PMID: 27307448. ↩
Ulvestad E. Modelling autoimmune rheumatic disease: a likelihood rationale. Scandinavian J Immunol. 2003;58:106–11 ↩
SLE Loong T. Understanding sensitivity and specificity with the right side of the brain. BMJ. 2003;327:716–9. ↩
Ulvestad E. Modelling autoimmune rheumatic disease: a likelihood rationale. Scandinavian J Immunol. 2003;58:106-11 ↩
Tam EM, Feltkamp TE, Smolen JS, et al. Range of antinuclear antibodies in ‘healthy’ individuals. Arthritis ad Rheumatism. 1997;40:1601–11 ↩
Lewis EJ. The treatment of lupus nephritis: revisiting Galen. Ann Intern Med. 2001;135:296–7 ↩