Rheumatology Advance Access originally published online on July 10, 2007
Rheumatology 2007 46(9):1509; doi:10.1093/rheumatology/kem164
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Repeated B-cell depletion in clinical practice
Centre for Rheumatology, University College London, Windeyer Building, Cleveland Street, London W1T 4JF, UK.
Correspondence to: J. C. W. Edwards. E-mail: jo.edwards{at}ucl.ac.uk
We are grateful to Dr Dorner and Dr Burmester for their kind Editorial Comments [1] covering our report of repeated B-cell depletion in RA [2]. We would like to raise certain points for clarification, particularly in relation to the DAS28.
We have concerns about DAS28 being recommended for assessing the need for re-treatment. It is inherently unlikely that the best indicator of the qualitative issue of relapse, in an individual, should be a quasi-quantitative composite score designed for pooled blinded data. Our experience indicates that it would often produce false positive and negative assessments. This has serious implications because health care systems are manipulating such indices to the detriment of patients, especially in the UK. DAS28 has several weaknesses including:
- The ESR is unreliable following B-cell depletion because it may be affected by immunoglobulin populations of no relevance to need for treatment. The C-reactive protein is more relevant but since acute phase responsiveness varies change is more important than level.
- Much of DAS28 is subjective. Infusions have a major placebo effect, especially with corticosteroids. Subsequently, mechanical pain can prompt patients to seek retreatment. Reimbursement implications mean that data manipulation by patients and professionals is commonplace in the UK at least.
- There is no logical basis for adding forms of evidence for relapse. A more logical approach to the problem of ascertainment, is to use multiple thresholds, as in the ACR grading system.
- DAS28 does not reflect the assessor's confidence in the evidence (a PIP joint with equivocal long-standing swelling counts the same as a new 100 ml knee effusion) or previous behaviour of the patient's disease.
Our experience with 75 cases suggests that assessment for re-treatment is simpler. The ideal would be to re-treat just before clinical relapse, to ensure maximum immunological reconstitution and minimum distress. This will require development of pre-clinical disease measures: currently IgM rheumatoid factor is the best predictor, but that alone is not good enough. We treat on minimum evidence of relapse, which is increasingly a reasonable compromise. Dorner and Burmester point to the 5 month average delay in re-treatment in the follow-up study and we would emphasize that this is historical and we can now usually limit the delay to re-treatment to 1–2 weeks. Using the multiple threshold approach we currently re-treat, if B cells have returned and IgG levels are satisfactory, on the basis of:
- Convincing new recurrence of swelling of a joint or
- A rise in C-reactive protein to >50% above the post-treatment average together with recurrence of either inactivity stiffness or articular pain.
The need to maintain flexible clinical logic is illustrated by a swollen knee seen during a first cycle. The knee was injected with corticosteroid and re-treatment scheduled. The knee settled completely, and re-treatment postponed without problems for a further year. In other cases minimal relapse has been followed by deterioration.
After the first re-treatment assessment is different because experience shows that patients relapse at much the same interval after each treatment. A lower threshold can be used for the third treatment. Recurrence of joint pain or inactivity stiffness at around this interval, associated with two consecutive rises in IgM RF titre, we consider an adequate basis for re-treatment. Pre-emptive treatment at this interval may even be justified. Currently our impression is that monitoring of CD19+ B cells, IgG, CRP and IgM RF titre pre-treatment, 1 month post-treatment, then bi-monthly, with clinical state often assessable by telephone or email, is optimal for intelligent re-treatment.
This policy, based on 
200 episodes clearly needs further validation but is at least based on experience and logic. DAS28 adds no new information and we fear that it will be used against our patients.
A minor historical point is that the case we reported with Protheroe has little to do with the history of B-cell depletion (despite rumours in the Swedish popular press). We were told of this atypical case by Nicholas Coppard of Roche, who knew we were running a trial in RA, based on a hypothesis published in this journal in 1998 [3], with results published in the journal in 2001 [4]. We are a little surprised that Dr Dorner and Dr Burmester remain puzzled by the mechanism of B-cell depletion. Science works by generation of hypotheses, to be rejected if refuted, but retained if corroborated. The hypothesis that B-cell depletion would work by removing autoantibodies is amply corroborated by pharmacodynamics. The slower memory B-cell recovery in late relapsers is a modest statistical trend, which does not explain time of relapse. Perhaps those looking for keys to problems in dark alleys should check the pages of a journal appropriately bound in British Racing Green.
 |
References |
|---|
 Top References |
|---|
- Dörner T, Burmester GR. No B cells-no active RA? Advances in B cell depletion in RA-repeated therapy under conditions of clinical practice. Rheumatology (2007) 46:563–4.
[Free Full Text] - Popa C, Leandro MJ, Cambridge G, Edwards JC. Repeated B lymphocyte depletion with rituximab in rheumatoid arthritis over 7 yrs. Rheumatology (2007) 46:626–30.
[Abstract/Free Full Text] - Edwards JCW, Cambridge G. Rheumatoid arthritis: the predictable effect of small immune complexes in which antibody is also antigen. B J Rheumatol (1998) 37:126–30.[CrossRef]
- Edwards JCW, Cambridge G. Sustained remission in rheumatoid arthritis following a protocol designed to deplete B lymphocytes. Rheumatology (2001) 40:205–11.
[Abstract/Free Full Text]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||