Patterns of relapse in the Rheumatoid Arthritis
cohort treated with Rituximab at University College London

Background/Purpose: Two patterns of
relapse were defined in the first studies of Rituximab (RTX) therapy in patients with Rheumatoid Arthritis (RA): either occurring coincident with
B-cell return (concordant relapse – C-R) or ‘delayed’ occurring months
after B-cell return (discordant relapse – D-R), with around 50% of
patients belonging to each subgroup in the first 24 patients studied. Later
clinical observations showed a greater frequency of C-R pattern. Our study analyzed
the current frequency of patients showing a concordant/discordant pattern of
relapse after their first cycle of RTX.

Methods: Retrospective
observational study of the RA cohort treated with RTX at University College
Hospital, selecting patients with a good response to the 1st cycle of RTX and
analyzing the pattern of relapse after that cycle. All patients fulfilled
the 1987 revised ACR criteria and had severe disease activity (DAS28 > 5.1). Data were
collected on demographics, previous and concomitant therapy, months to
repopulation and months to relapse. B cell depletion was confirmed by CD19
determination of  <5/microl of
the total lymphocyte population. B cell repopulation was considered when B
cells were again detectable in peripheral blood (CD19 count >5/microl of the
total lymphocyte population). Patients with a poor response to the therapy, or
treated with a second cycle because of persistent disease activity were taken
out of the analysis.

Results: 271 patients have received RTX at our unit between 1998 and 2012. Total
patient-years follow up was 886.05 and total number of cycles administered was
910. We selected 168 patients with a good response after 1 cycle of RTX and followed
up at least until their first relapse, which occurred 4 to 67 months after the
first cycle. Mean time to repopulation was 7.3 months (range 3-20 months) and
mean time to relapse was 10.4 months (range 3-67 months). Patients were then
divided in two patterns of relapse, that is C-R if relapse occurred less than 3
months after repopulation was first documented, and D-R if it occurred more
than 3 months after such repopulation. Seventy % (118 patients) were C-R, with
a mean time to repopulation of 7.1 months (range 4-20 months) and a mean time
to relapse of 7.7 months (range 4-20 months); Thirty % (50 patients) were D-R,
with a mean time to repopulation of 7.8 months (range 3-20 months) and a mean
time to relapse of 17.4 months (range 6-67 months).

Conclusion: there are two clear patterns of relapse after repopulation in patients
with RA treated with RTX, with a higher frequency of concordant patients. In
one third of the patients a discordant pattern of relapse has been identified.
The decision of when to retreat RA patients with RTX should be made for each
patient individually, taking into account the patterns of relapse described. This
will possibly decrease the risk of adverse events and reduce
costs for each therapy.