Prevalence and clinical significance of small PNH clones in patiens with primary autoimmune hemolytic anemia

Background Autoimmune hemolytic anemia (AIHA) is caused by autoantibodies directed against autologous erythrocytes, diagnosed by the direct antiglobulin test (DAT), and classified as warm (IgG+ or IgG and C+), cold (C+), mixed (IgG+high titer C+), and atypical (DAT-, IgM warm, and IgA+). Recently, autoimmunity against marrow precursors, increased presence of fibrosis/dyserythropoiesis, and possible evolution to bone marrow failure (BMF) have been reported and correlated with poor outcome. PNH clone is a typical finding in up to 20-40% of patients with BMF and MDS, and its presence correlates with favorable prognosis and predicts response to immunesuppression. Aims To evaluate the prevalence of PNH clones in patients with primary AIHA, and assess its prognostic/predictive significance. Methods AIHA patients followed at two tertiary hematology centers in UK and Italy were followed-up from December 2007 until February 2018. The PNH clone was assessed at the time of AIHA diagnosis by FLAER flow-cytometry technique. Clinical characteristics, including therapy lines and marrow features, were also collected. Continuous and categorical variables were assessed by Student’s t-test and chi-square or Fisher test, respectively. Kaplan Meyer method was used to evaluate and compare relapse free survival (RFS) and overall survival (OS). Results Table 1 shows clinical and morphologic characteristics of the 35 patients enrolled, altogether and divided according to PNH clone positivity. Twelve patients (34%) showed a positive clone, with a median size of 0.2% on granulocytes (range 0.1-16%). Considering AIHA type, 2 atypical forms (DAT negative) were PNH+, warm AIHA mainly PNH-, and cold and mixed forms equally distributed. PNH+ cases were more anemic, although not significantly, and displayed higher hemolytic pattern compared to PNH- ones, both considering absolute LDH levels and categorization (92% of cases versus 47% showing LDH levels above 1.5x ULN). As expected, Hb levels positively correlated with bone marrow responsiveness index (r=0.40, p=0.008) and negatively with LDH levels (r=-0.31, p=0.03). Considering marrow characteristics, PNH+ patients showed lower prevalence of hypercellularity (28 vs 43%), MF-1 fibrosis (57 vs 71%, p=0.07), dyserythropoiesis (14 vs 57%, p=0.01), and displayed a smaller lymphoid infiltrate (p=0.05), compared to PNH- ones. Lymphoid infiltrate was CD3+ in 1 PNH+ patient, CD20+ in 2 PNH- cases, and mixed in all the others. Regarding therapy, PNH+ and PNH- cases received 1st and 2nd lines (steroids, rituximab, immunesuppressors and splenectomy) with equal rate. However, the former less frequently required a 3rd treatment course (8% versus 35%, p=0.09) and displayed a significantly higher response to rituximab therapy (100% vs 70%, p=0.04). AIHA related complications occurred in 18 cases, with thrombosis observed more commonly in PNH+ patients (25% vs 13%), acute renal failure in PNH- only (21%), and infections equally in both groups (17 and 21%). Finally, 7 patients died during the observation period, mostly because of infections (N=5). The small number of events precluded any statistical analysis. Conclusion Here we firstly report the prevalence of a small PNH clone in about 30% of primary AIHA patients. PNH+ cases show an increased hemolytic pattern and a better response to first and second therapy lines, suggesting a prevalent autoimmune habitus.