Discussion
We report herein the first success to our knowledge of eculizumab for the treatment of wAIHA occurring in a patient with CMML.
Even if accurate data are lacking concerning the prevalence of inflammatory or autoimmune diseases in CMML patients, monocentric studies suggest that AID may occur in about 20% of cases (1,2). The outcome of patients with CMML-associated AID remains unclear, but in most cases, they exhibit rather favorable prognostic features of the underlying myeloid disorder. In a retrospective study of 377 CMML patients, those with AID were more likely to have a lower risk prognostication based on the Mayo Clinic prognostic model (1). Results were similar in a French retrospective study (3). In a third cohort, cytogenetic abnormalities seemed less frequent in CMML patients with AID (2). Finally, in a small cohort focusing on CMML-associated ITP, all cases of immune thrombocytopenia occurred in patients with low-risk CMML, and none of them experienced CMML progression or transformation into acute myeloid leukemia during follow-up (4). Considering our patient, CMML has remained stable with a follow-up of 17 years.
Among AIC, the development of ITP seems much more frequent than AIHA. Thus, the only available data reporting efficacy of treatments in CMML-associated autoimmune cytopenias concern patients with ITP. In nearly all reported cases of CMML-associated AID, first-line treatment consisted of corticosteroids, and most relapsed patients received immunosuppressive therapies. Focusing on ITP, successful treatments with corticosteroids, rituximab, splenectomy or thrombopoietin agonists have been reported, and some authors consider that CMML‐associated ITP should be treated according to current guidelines for primary ITP (4). The same rationale could apply to CMML-associated AIHA and in our patient; splenectomy led to a durable control of AIHA for about 10 years.
In AIC associated with lymphoproliferative disorders such as chronic lymphocytic leukemia, most experts believe that an optimal control of the malignant clone would represent the best option to control autoimmune manifestations. The same hypothesis could be raised in CMML-associated AID. The rationale for the use of 5-azacitidine in this setting, even in the absence of CMML progression, is based on its ability to remove or at least to lower the tumor clone. Some studies have reported efficacy of 5-azacitidine in autoimmune manifestations associated with myelodysplastic syndromes or CMML but none of them was AIHA. A prospective phase II trial, supported by French Group of Myelodysplasia (NCT02985190), is currently evaluating the efficacy and safety of 5-azacitidine in refractory autoimmune and inflammatory diseases associated with myelodysplastic syndromes, and may provide answers for clinicians. However, azacitidine cannot be considered as a salvage treatment in case of life-threatening AIC because responses are usually delayed (6,7). Moreover, its initial myelotoxicity may worsen the cytopenias.
The use of eculizumab is not approved for AIHA. A prospective phase II trial reported its efficacy in patients with CAD (8). These results can be explained by the central role of terminal complement activation in cold AIHA. The use of eculizumab has already been reported in a case of wAIHA associated with Waldenström macroglobulinemia (9). In this case, wAIHA was also refractory to steroids, rituximab and plasmapheresis. As in our patient, DAT was positive for both IgG and C3d. After four weekly doses of eculizumab hemoglobin was stabilized around 90 g/L and treatment was stopped. The patient remained in partial remission for four months before he died from refractory anemia relapse. The complement system plays an important role in the physiopathology of AIHA. Some patients, including our case, suffer from wAIHA with both IgG and C3d positive DAT. In these specific forms, autoantibodies, especially IgM, can activate the complement system in vivo , leading to the destruction of red blood cells. These IgM, which react at 37°C, contribute to hemolysis by activating the classical pathway of the complement system. Meulenbroek et al. showed that all patients with C3 on their red blood cells have an IgM component that could activate complement even if no such antibody could be detected with routine techniques (10). In wAIHA with positive DAT for C3d, complement-mediated red blood cell destruction contributes to the severity of acute hemolysis, exacerbations of chronic AIHA and poor performance of red blood cell transfusions. For these patients, use of eculizumab could prevent red blood cell transfusions, which can exacerbate hemolysis, or could restore the transfusion efficiency, while waiting for the effect of the other treatments. In addition to its cost, the suspensive effect and the risk of severe bacterial infections such as meningococcal infection could represent the main limitation of eculizumab (11). In this specific population, it is of major interest to consider the safety profile of eculizumab since previous exposure to corticosteroids or other immunosuppressive agents might increase even more the risk of severe infections. We did not observe any infectious complication during our patient’s treatment with eculizumab. In the prospective study on CAD, Röth et al , reported one case of treatment-related pneumonia among 13 patients treated with eculizumab (8). No meningococcal infection was observed.
In our case, a delayed effect of rituximab cannot be formally ruled out but the efficacy of this treatment is usually not observed beyond three months (12). The concomitant correction of thrombocytopenia is a further argument for rituximab late effect because of its proven efficacy in both primary and secondary ITP (13). Moreover, complement system has not been involved in autoimmune thrombocytopenia until now. Whether eculizumab should have been pursued such a long time in our patient is questionable. In the phase 2 study on CAD, eculizumab was given for 26 weeks (8). As our patient did not experience any infectious complication and because of the severity of AHAI presentation, eculizumab was maintained up to 56 weeks, but it may have been stopped earlier. One year after the end of treatment, no relapse has occurred so far. In case of AIHA relapse, a new course of rituximab or 5-azacitidine could be considered but since these agents have an expected delayed efficacy, eculizumab should be rechallenged if needed.