Impact of sex on outcomes in patients with hairy cell leukemia. An HCL patient data registry analysis

Hairy cell leukemia (HCL) is a rare, incurable chronic hematological malignancy that makes up <1% of lymphoid neoplasms.1 Fewer than 2000 new cases of HCL are diagnosed annually in the United States.2 The median age of onset is 50–55 years and there appears to be a genetic predisposition that is sustained in people with European genetic susceptibility.1, 3 Patients often present with fatigue, pancytopenia, splenomegaly, and a markedly increased risk for infection which is the leading cause of death and serious morbidity in patients with this disease. Although HCL is incurable, significant advances have been made in the treatment of this disease over the past few decades especially the advent of purine nucleoside analogs4-6 and targeted agents7-9 with improved outcomes. For reasons that remain unknown, HCL is diagnosed in female patients at a strikingly lower rate than in male patients, with a 4:1 predominance in males.1 To date, there have been no studies aimed at characterizing HCL in the female patient population. Although substantial advances have been made in the treatment of HCL, the rarity of the disease, especially in female patients, makes it challenging to conduct meaningful research in this population. Hence, we sought to characterize the female patients with HCL using the HCL Patient Data Registry (HCL-PDR10) and evaluate their disease phenotype, response to treatment, and outcomes compared to males. The HCL-PDR10 is an international multicenter patient data registry that includes information regarding disease characteristics, treatment history, treatment response, comorbidities, other health outcomes, and molecular testing. The information is derived from source documents which are abstracted into discrete data elements in the patient data registry. The HCL Foundation has sponsored the development of the HCL-PDR to characterize the clinical features and outcomes of these diseases. The study was approved by the institutional review boards at all the participating sites and performed in compliance with the Declaration of Helsinki. This analysis included adults (aged ≥ 18 years) with HCL. Female patients with HCL were the study population, with male patients with HCL as the comparator group. The study included patients with both classical HCL (cHCL) and HCL variant (HCLv) with majority of the patients having cHCL. The primary endpoint was the time to next treatment (TTNT) in female HCL patients compared to male HCL patients. Secondary endpoints included response rate to first-line therapy and predictors of TTNT. Patient outcomes and responses were analyzed according to consensus criteria.11 Responses were categorized as complete response (CR) and partial response (PR). CR included those achieving CR, CRu (CR unconfirmed), and HR (hematologic response). PR included those achieving PR, PRu (PR unconfirmed), and pHR (partial HR). See Suppl appendix (Table S1) for the definitions of the response criteria. The TTNT was calculated from the first HCL treatment to the second HCL treatment, estimated using the Kaplan–Meier method, and compared between groups using the Log-rank test. Cox proportional hazard regression models were used to estimate the hazard ratios for the risk of TTNT. Analyses were performed using Stata software (version 16; StataCorp, College Station, Texas), and all statistical tests were two-sided with a type-1 error of 0.05. A total of 357 patients were included in the study. Of these, 265 were males and 92 were female patients. The median age at diagnosis was 52 years (19–80 years) and was comparable between male (53 years) and female patients (51 years). The study included Caucasians (98%) predominantly consistent with previously reported patient demographics,3 with a greater proportion of patients being never smokers (74%) and having had no prior chemical or radiation exposure (89%). The majority of the patients (95%, n = 337) had classical HCL with 87% of these being BRAF mutated (among those with available BRAF mutation status). Female patients had significantly higher median platelet count (92 K/μL vs. 77 K/μL, p = .04) but lower median hemoglobin (11.5 gm/dL vs. 12.5 gm/dL, p = .01) compared to male patients. The most common therapy administered was cladribine (61%) followed by pentostatin (17%). There were no differences in the immunophenotypic features (CD11c, CD25, CD103, CD123, and CD200) among those with available data between male and female patients with HCL. The median follow-up time from the start of first HCL treatment was 6.3 years (overall), 6.2 years in male, and 6.3 years in female patients. Table S2 shows pertinent baseline characteristics stratified by sex. Among the patients who had disease status assessed after the first treatment, n = 224 (males 169, females 55), there was no significant difference in response rates (see Table S3). Female patients had a significantly longer median TTNT (17.6 years, 95%CI = 7.6 to not reached [NR]) compared to males (8 years, 95% CI = 6.3–14.1), p = .02 (Figure 1). Patients who achieved CR after first-line therapy had significantly longer median TTNT (14.1 years, 95% CI = 10.4-NR) compared to those who achieved PR (6.9 years, 95% CI = 2.9-NR) and SD (0.7 years, 95% CI = 0.5-NR), p < .01 (Figure S1). On subgroup analysis based on age at diagnosis, female patients with HCL had significantly longer median TTNT compared to males (NR vs. 7.2 years, p = .01, Figure S2) for those who were ≤ 60 years, however, for patients who were older than 60 years, there was no difference in the median TTNT (NR vs. NR, p = .28, Figure S3). We further analyzed the data among those who were ≤ 60 years in two groups, ≤50 years and 51–60 years. Among patients who were ≤ 50 years, females had a significantly longer median TTNT compared to males (NR vs. 6.5 years, p = .04, Figure S4). However, for those who were 51–60 years, although the median TTNT was longer in female compared to male patients, this did not reach statistical significance (NR vs. 7.6 years, p = .05, Figure S5). Age at diagnosis, sex, smoking status, BRAF mutation status, median hemoglobin, median platelet count, bone marrow HCL percent, and response to treatment were evaluated to determine the predictors of TTNT in the study. After adjusting for factors associated with TTNT in the univariate analysis (age at diagnosis, sex, median hemoglobin, median platelet count, and response to treatment), female sex (HR = 0.32, 95% CI = 0.11–0.97, p = .04) and CR to first-line therapy (HR.0.26, 95% CI = 0.11–0.57, p = .001) remained associated with significantly lower risk of initiating next line therapy on multivariable analysis (Table S4), indicative of significantly longer TTNT. As the patients in the HR and pHR did not have confirmatory bone marrow after first-line treatment, we did a sensitivity analysis excluding these patients to assess the TTNT between males and female patients. The median TTNT was significantly longer in female patients (NR, 95% CI = 6.6-NR) compared to male patients (8 years, 95% CI = 6.0–14.1, p = .03, Figure S6), which was in line with the main analysis. When we analyzed the TTNT based on the response to treatment after excluding HR and pHR, the results were similar to the main analysis (Figure S7). We report the findings of what is to our knowledge the largest HCL registry analysis to date comparing the characteristics and outcomes of female HCL patients with male patients and make several important observations. First, the distribution of male to female patients was 3:1 in the study. Second, female patients who were 60 years and younger had a significantly longer median TTNT compared to males. Third, patients who achieved CR to first-line therapy had significantly longer median TTNT compared to those who achieved PR or SD. Lastly, female sex and CR to first-line therapy were predictive of longer TTNT. In diffuse, large B-cell lymphoma (DLBCL) and follicular lymphoma (FL), female sex hormones have been shown to confer a protective effect including conferring a survival advantage in these lymphomas, which was limited to premenopausal women in some studies.12, 13 In our study, the analysis of different age groups showed that the advantage in female patients with HCL occurred specifically in the younger (≤60 years) patients. In an exploratory analysis, we divided the cohort of patients aged ≤60 years into two cohorts (≤50 years and 51–60 years), hypothesizing a protective hormonal effect. Our data show that female patients aged ≤50 years have significantly longer TTNT than males, while those aged 51–60 years do not, suggesting a diminishing protective effect over time which could be consistent with the impact of female sex hormones. Sex differences in immunological surveillance that make men more vulnerable to proto-oncogenic mutations but also chronic, potentially oncogenic infections, have been suggested.14 In our study, there was no difference in smoking status and exposure history between male and female HCL patients.In conclusion, we show for the first time that there are differences in outcomes based on sex in HCL patients. Sex was an independent predictive factor in determining TTNT, in that female patients with HCL had significantly longer TTNT compared to male patients after adjusting for other significant variables in the multivariable analysis. Although hormonal factors seem to play a role, there is probably not a single universal explanation behind the observed sex differences for either incidence or outcome, which may be related. A better understanding of the underlying mechanisms behind sex differences is important to improve the outcomes in HCL patients. We acknowledge the patients who participated in the Hairy Cell Leukemia Patient Data Registry and the Hairy Cell Leukemia Foundation for providing the funding support for Registry (GR110873). We acknowledge The OSU Department of Research Information Technology for their technical support provided for the registry. KAR is a Scholar in Clinical Research of the Leukemia and Lymphoma Society (CDP 2331-20). Narendranath Epperla: Research funding: Beigene, Speakers Bureau for Incyte and Beigene, Honoraria/consulting/ad boards for TG Therapeutics, Pharmacyclics, BeiGene, Seattle Genetics, and Novartis. James S Blachly: Consultant and or on advisory boards for AbbVie, AstraZeneca, Astellas, Innate Pharma, Kite Pharma, MingSight Pharma; has a patent on a leukemia diagnostic device. Kerry A. Rogers: Research funding from Genentech, AbbVie, Novartis, and Janssen, consulted for Genentech, AbbVie, Pharmacyclics, AstraZeneca, Innate Pharma, Janssen, and Beigene, and received travel funding from AstraZeneca. Clive S. Zent: Research funding from Acerta/AstraZeneca and TG Therapeutics. Versha Banerji: Research funding: Hairy Cell leukemia foundation, Lymphoma Canada, Canadian Institutes of Health Research, Leukemia Lymphoma society of Canada, CancerCare Manitoba Foundation; Honoraria: Abbvie, Astra Zeneca, Beigene, Janssen; Royalties: Biogen. Michael Grever: Consultant: Astra Zeneca, Pharmacyclics, Ascerta, Axio, Inc; Research Funding: Hairy Cell Leukemia Foundation for Patient Data Registry; Travel Expenses: Hairy Cell Leukemia Foundation; Scientific Board: Chair, Hairy Cell Leukemia Foundation Scientific Board (no reimbursement); and Scientific Honorarium: University of Pittsburgh. Leslie Andritsos: Research funding from HCLF; consulting with innate pharma. Qiuhong Zhao, Mirela Anghelina, Jasmine Neal, Gerard Lozanski, Christopher C. Oakes, Seema A. Bhat: no conflict of interest. Written informed consent was obtained from the patient to publish this report in accordance with the journal's patient consent policy. The data that support the findings of this study are available from the corresponding author upon reasonable request Data S1. Supporting information. 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.