Pure Red Cell Aplasia in Systemic Lupus Erythematosus, a Nationwide Retrospective Cohort and Review of the Literature

Pure red cell aplasia in systemic lupus erythematosus, a nationwide retrospective cohort.

Background: Pure red cell aplasia (PRCA) is a cause of normochromic normocytic anaemia due to erythropoiesis failure with a profound reduction of erythroblasts precursors in bone marrow examination and low reticulocyte count in the peripheral blood. PRCA can be associated with systemic lupus erythematosus (SLE) but only isolated case reports have been reported.

Objectives: To characterize the clinical and biological course, management and response to treatments of SLE associated PRCA.

Methods: We conducted a nationwide multicenter retrospective cohort study. All participating centers (n=10) were secondary- or tertiary-care hospitals belonging to the French national network for autoimmune cytopenia. We included adult's patients (> 18 years old) presenting PRCA and SLE from 2006 to 2018 with a diagnosis of PRCA supported by bone marrow examination (bone marrow aspirate or bone marrow biopsy) and a SLE or biologic manifestations of SLE. A negative polymerase chain reaction (PCR) testing of parvovirus B19 on peripheral blood or in marrow, and/or a negative serology were mandatory to be included. SLE was defined according to the 2012 SLICC (Systemic Lupus International Collaborating Clinics) criteria or alternatively anti-DNA antibodies without clinical criterion for SLE.

Results: We enrolled 24 patients (20 women). SLE was diagnosed before PRCA for most patients (14/24 (58%), median delay 81 months). At PRCA diagnosis, mean age, haemoglobin value, reticulocyte and erythroblast count were 39.2 ± 13.2 years, 62 ± 20 g/L, 9.1 ± 7.6 x 10⁹/L and 2.8 ± 2.5 % respectively. Eleven (45%) patients experienced multiple PRCA flares (median 6). All patients but one (95%) received corticosteroids as first line therapy, resulting in only three complete sustained responses. One patient received erythropoietin combined to high dose corticosteroids and hydroxychloroquine and experienced complete response. Nineteen (79%) patients required other immunosuppressive agents with highly variable regimens. Most frequent immunosuppressive agents were cyclosporin A (n=9), rituximab (n=8), cyclophosphamide (n=8), mycophenolate mofetil (n=7), azathioprine (n=5) and everolimus (n=3). After PRCA onset, 15 (63%) patients experienced 1 to 5 SLE flares. After 76 months median of follow-up (range 13-173 months), 17 (71%) patients were in complete response for PRCA, five (21%) in partial response and two (8%) in no response. One patient who was in no response for PRCA died of gastrointestinal bleeding due to probable peptic ulcer after 8 years of PRCA course. In contrast, lupus was quiescent for all patients. Twenty-one (87%) patients required red blood cell transfusion, of whom five were diagnosed with transfusion related iron overload. Eighteen (75%) patients experienced severe infectious events requiring hospitalization. Compared to the forty-seven previously published isolated cases, our series showed more frequent PRCA relapses and use of more immunosuppressive agents that can be explained by the significantly longer follow-up duration in our study.

Conclusion: We report the first series of PRCA associated with SLE. Our results showed that it is a severe condition associated with high morbidity such as iron overload secondary to multiple transfusions and infections related to corticosteroids and other immunosuppressive agents. Corticosteroids were mainly not effective in monotherapy but most patients eventually achieved a response with other immunosuppressive agents. Multiple relapses of PRCA were observed in almost half of the cases requiring multiple lines of immunosuppressive therapy. While cyclosporine is the most widely immunosuppressive therapy used during primary PRCA, the treatments used in our study appeared to be much more varied and the best immunosuppressive treatment regimen for PRCA associated with SLE is still poorly understood.