Wednesday, December 8, 2010

Lupus - persistent low complements

Consider a patient with long-standing SLE who has biopsy proven WHO class IV lupus nephritis, previously treated with cyclophosphamide and steroids, which induced remission. Due to recurrence of proteinuria and desire to remain steroid-free, MMF was commenced with good effect. Other maintenance treatment from our Rheumatology colleagues included hydroxychloroquine for extra-renal manifestations of lupus.

A subset of patients, despite having minimal proteinuria, no haematuria and no flares requiring steroid treatment, have persistent hypocomplementaemia. This got me thinking that these patients may possibly have persistent sub-clinical activity – the question being, what do we do – treat the numbers or treat the patient?

A retrospective review was recently published looking at a similar patient group. Examination of over 28 years of data identified 56/924 patients who had clinically quiescent disease for two years, during which they had no steroid or immunosuppressant treatment (other than anti-malarials) but had persistently elevated anti-dsDNA and/or hypocomplementaemia. These patients had a lower baseline disease activity score and a lower need for steroids or immunosuppression use up front than those not meeting these criteria.

In this ‘clinically quiescent serologically active’ group 62.5% had both elevated anti-dsDNA and hypocomplementaemia, 23.2% had elevated anti-dsDNA only, while 14.3% had hypocomplementaemia only. During follow-up, 58.9% developed a flare at a median of 155 weeks, 10.7% developed normal anti-dsDNA and complement levels (clinically quiescent serologically quiescent) and 30.4% had persistent elevations in anti-dsDNA and complement levels (remained clinically quiescent serologically active).

To compare those who flared with the other two groups, they reviewed levels of anti-dsDNA and complement levels from their previous visits prior to the flare. They found that absolute levels or change in levels of anti-dsDNA or complements did not predict flares.
Also of note, in comparing the original 56 patients with the remainder of the group there was no statistically significant difference in mortality during follow-up between the two groups (14.2% v 5.4% p=0.06).

The authors’ conclusions were that the clinically quiescent serologically active subgroup certainly warrant close monitoring (as 59% ultimately develop a lupus flare), but that treatment should be based on clinical findings, not the numbers.

3 comments:

Anonymous said...

I'd question the author's conclusions re: no significant mortality difference (14.2% v 5.4% p=0.06; n=56). This looks like a power issue, and I think a few more patients would push this p-value across the magical 0.05. Great post, keep it up!

The Doctors' Rheum said...

Thanks for posting. I may use this for an upcoming journal club.

Anonymous said...

Same goes for ANCA levels and disease activity. As clinicians, we do always do follow the clinical picture.