Mixed Type Autoimmune Hemolytic Anemia: Clinical Severity and Treatment Patterns in a Large Observational Study

Mixed type autoimmune hemolytic anemia (mAIHA) is a rare form of AIHA characterized by the presence of both warm reacting IgG autoantibodies and IgM cold agglutinins. Diagnosis requires a second level direct antiglobulin test (DAT) with monospecific antisera, positive for IgG and complement (C3), spontaneous autoagglutination at 20°C and cold agglutinin titer ≥ 1:64. The clinical picture may vary, resembling more that of warm AIHA (wAIHA) or cold agglutinin disease (CAD) but is often severe, with high degree hemolysis. The mixed positivity of the DAT implies uncertainty in the treatment algorithm and management is guided by the presenting predominant features on a case-by-case basis. Corticosteroids are usually given at high doses, and early rituximab is often necessary. However, in an expanding armamentarium of new drugs in study for wAIHA or CAD, mAIHAs are often excluded from clinical trials [1-3]. To characterize clinical features and treatment patterns of mAIHAs, we performed a retrospective observational study of mAIHA patients evaluated at two Italian Hematologic reference Centers from September 1999 to June 2025. Clinical and hematological features at diagnosis, treatments and responses, infections, thrombotic complications, and survival were systematically collected. Key clinical outcomes of mAIHA cohort were thereafter compared with those of a cohort of wAIHA and CAD patients (Supporting Information). The cohorts of wAIHA and CAD were monocentric (Milan), while those of mAIHA came from two institutions (see above). We report a total of 91 mAIHA cases, followed for a median of 40 months (range 1–217), Table 1. There was a slight female predominance (57.1%), and median age was 57 years (9–89). Median hemoglobin (Hb) at diagnosis was 7.1 g/dL (2.9–11.5), 51 (56%) had Hb < 8 g/dL, and 25 (27.5%) < 6 g/dL. Median LDH levels were 1.99 × upper limit of normal (ULN), (0.6–11), and 55% of patients displayed inadequate reticulocyte compensatory response as per (bone marrow reticulocyte index [BMRI] < 121). Median agglutinin titer was 1:384 (2–4096), five patients presented with DAT positivity for IgA + C, and eight had history of immune thrombocytopenia prior to mAIHA diagnosis (Evans syndrome). Those four patients with cold agglutinin titer < 1:64 had positive autoagglutination. Twenty (22%) patients had splenomegaly, while in 10 (11%) bone marrow was positive for clonal B lymphocytic infiltrate, mostly CAD like, chronic lymphocytic leukemia-like and marginal zone lymphoma-like in two cases (all patients received bone marrow evaluation). An IgM monoclonal component was evident in 2/3 of patients. Table 2 displays treatment strategies: transfusions were required in 32% of cases, and median number of relapses was 1 (0–4) at a median of 12.2 months from first line (1–123). Patients received a median of 2 (0–6) treatments, 60% (55) more than one, 28% (25) more than two. These included steroids alone in 71 patients (78%) with an overall response rate (ORR) of 80%, complete response (CR) in 44% only, and rituximab (375 mg/sqm weekly for 4 weeks) plus steroids frontline in nine patients (10%), all responding (CR 44%). Rituximab was also used as an early second or as the third line of choice in 35% (17/48) and 63% (12/19) of cases respectively with responses in 94% of patients (CR 70%) in the second line, and in 83% (CR 42%) in the third line. Cytotoxic immunosuppressants were used in 25 cases (27%), mostly as second or further line (64%). ORR reached 90% (CR 56%) when used as first line while was lower 74% (CR 26%) when used as second or further. The most common choice was azathioprine in 13 patients, followed by cyclophosphamide in 11, cyclosporine in 3 and mycophenolate in 2. Daratumumab and bortezomib were used as late-lines in one patient each, leading to a partial and no response, respectively. Twelve patients (13%) underwent splenectomy with good response rate (ORR 91%, all except one complete), at a median of 12.1 months (5–31) from diagnosis. Of note, in 18 cases (23%) intravenous immunoglobulins were used as concomitant therapy (ORR 60%), while only four patients (4%) received recombinant erythropoietin (rEPO). Finally, in all cases, during the acute phase, steroids were used in combination with rituximab or immunosuppressants that have a known longer time to response, and then rapidly tapered. Therapy choice was overall not influenced by the presence of an identifiable B-cell infiltrate. During follow up, 10 patients (11%) had a thrombotic event, while 18 (20%) faced an infectious complication. Infection rate was comparable in patients undergoing rituximab therapy as a first line or second or further lines (33%; 3/9, 9/27 respectively), cytotoxic immunosuppressive therapy (20%, 5/25), or splenectomy (33%, 4/12). Overall, 19 patients (21%) died at a median of 39 months from diagnosis, although median OS for the entire cohort was not reached (133–non reached). In the univariate Cox regression analysis (Table S1), older age was associated with poorer survival, whereas female sex and the use of steroids as first-line therapy were protective. In the multivariate model (Table S2) no independent predictors of survival were identified; however, there was a trend towards worse survival in patients receiving cytotoxic therapy as second-line treatment (hazard ratio 5.21, 95% CI 0.94–28.96; p = 0.06). Of the 19 deaths, 6 were attributable to AIHA-related complications such as multiorgan failure and infections (pneumonia in one patient and sepsis in one patient who underwent splenectomy). Compared to historical data from patients with wAIHA or CAD (Table S3), those with mAIHA more frequently presented very severe anemia (Hb < 6 g/dL) as compared to CAD, but similar to wAIHA (27% vs. 26% in wAIHA but 8% in CAD, p = < 0.001): despite this, need for transfusions was similar across groups (32%; 38%; 40%), even in the CAD cohort, probably due to the more chronic nature of the disease, as well as with the less reported risk of alloimmunization reported in CAD. In contrast, patients with mAIHA more frequently required two or more lines of therapy, particularly when compared with wAIHA (60% vs. 37%; 50% in CAD; p = 0.001). They also experienced higher rates of infectious complications (20% vs. 15% and 11%, respectively) and AIHA-related mortality (6% vs. 2.5% and 2%), whereas the incidence of thrombotic events was comparable across groups (11% vs. 14%). The need for transfusions was similar across groups (32%; 38%; 40%), as was the rate of thrombotic events (11%; 14%). This study represents the first comprehensive analysis of patients with mAIHA, a rare and poorly characterized subtype of AIHA, difficult to diagnose and with a more severe clinical presentation, higher treatment need, more frequent infectious complications and poorer outcome, as compared with isolated wAIHA and CAD. This may be due to the double nature of mAIHA with IgG and IgM autoantibodies, thus involving both antibody-dependent cellular cytotoxicity and complement activation. Regarding diagnosis, mAIHA should not be confounded with wAIHA with complement activation representing 1/3 of cases driven by IgG1 and 3 subclasses. Quantification of cold agglutinin titers (> 1:64), and detection of spontaneous agglutination at 20°C are critical for differential diagnosis [1, 2, 4]. In our cohort, six patients had cold agglutinin titer < 1:64, however associated with spontaneous erythrocyte agglutination at 20°C. This evidence adds to the accepted concept that lower titers of cold agglutinins should not be disregarded in the case of clinically evident complement mediated hemolysis. As expected, a bone marrow monoclonal B cell infiltrate was present in two third of patients, stressing that serum electrophoresis, bone marrow evaluation, and imaging for organomegalies are advised at mAIHAs diagnosis as recommended for CAD. Regarding therapy, patients required a median of two treatment lines (with some receiving up to six) with more frequent and severe relapses than wAIHA and CAD alone, in line with previous findings [5-9]. Notably, corticosteroids and rituximab used frontline yielded a CR in about 40% of patients only lasting about 12 months more in line with CAD behavior. Cytotoxic immunosuppressants, generally considered a third-line option for wAIHA, were commonly used in earlier lines with CRs in about half of patients, and no apparent increase of infectious risk as compared to other treatments. Additionally, splenectomy was performed in 13% of cases, yielding an ORR greater than 90% and mostly complete, higher than what is typically seen in wAIHA [1, 6]. These data are likely influenced by the long observational period dating back to 1999 and would require confirmatory prospective studies. In our series, rEPO was underutilized, being administered in 4% of cases only, despite the presence of inadequate reticulocytosis, a marker of severity and inappropriate marrow compensation [10], in over half of the patients. Finally, none of our patients received complement inhibitors that are effective in decreasing hemolysis and transfusion need in CAD. These drugs may be highly effective on the CAD component of mAIHA, which were however excluded from most trials [11-14]. All in all, our experience supports steroids plus rituximab as first line, based on the CAD-like features of mAIHA, followed by repetition of rituximab if the prior response lasted at least 24 months, as suggested from data on other forms of AIHA, or immunosuppressants in case of early failure, given the lack of other approved treatments and the good outcome observed in the present study. However, the off-label use of plasma cell targeting agents and complement inhibition in patients with evidence of either plasma cell infiltrates or complement mediated hemolysis might fulfill the unmet need of multi-relapsed patients in the near future. Concerning complications, thrombotic events in our mAIHA cohort were similar as other AIHA subtypes, while infectious complications were higher [15, 16] likely due to the higher number of therapies and greater use immunosuppressive treatment in this setting. Notably, one patient who underwent splenectomy died of sepsis, underscoring the importance of vigilance notwithstanding vaccinations. Infections were the leading cause of AIHA-related mortality, which was more frequent in mAIHA compared to other subgroups [1, 8]. In conclusion, mixed AIHA emerges as a particularly severe form of AIHA, characterized by refractory hemolysis requiring multiple therapy lines, and increased susceptibility to infections. Early and precise diagnosis is essential to guide therapy, which should be tailored to the dominant pathogenic mechanism. The inclusion of mAIHA patients in the clinical trials with novel drugs for AIHAs might finally fulfill the critical unmet need of this rare condition [17-20]. The study was partially funded by the Italian Ministry of Health, Current Research Grant. The study was conducted in accordance with the Declaration of Helsinki and approved by the local Ethics Committee. All Authors approved present submission. The authors declare no conflicts of interest. The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions. Data S1: ajh70321-sup-0001-Supinfo.docx. Table S1: Univariate Cox Regression analysis for survival. Table S2: Multivariate Cox Regression analysis (for p < 0.15 in univariate analysis). Table S3: Key clinical features and outcome figures across AIHA types. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.