难治性和复发性骨髓衰竭综合征的挽救性免疫抑制疗法
Salvage therapy for immune-mediated bone marrow failure
研究回顾了多种免疫抑制疗法在难治性和复发性骨髓衰竭综合征中的疗效,包括 alemtuzumab、abatacept、sirolimus 和 tofacitinib。结果显示,这些疗法在不同类型的骨髓衰竭综合征中均显示出一定疗效,为临床治疗提供了实证依据。
Introduction
Acquired bone marrow failure syndromes (BMFS) encompass immune-mediated hematopoietic disorders including aplastic anemia (AA) and diseases resulting in lineage-restricted cytopenias such as idiopathic neutropenia (IN), pure red cell aplasia (PRCA) and cytopenias occurring in the context of large granular lymphocyte leukemias (LGLL). These conditions share various similarities in their pathogenesis, mostly related to their common immune pathogenesis and clinical overlap [1]. The specific trigger antigens and their distribution within hematopoietic hierarchy account for diverse presentations, such as lineage-restricted neutropenia, reticulocytopenia or pancytopenia, propensity to clonal progression common in AA and other clinical associations. Similarly, the distinct contribution of various effector mechanisms may account for clinical differences, e.g., direct immune attack, effector cytokines or the presence of concomitant autoantibody responses. Moreover, BMFS may coexist with similar clinical features, e.g., PRCA, IN, in the context of LGLL, plasma cell dyscrasias, or hemophagocytic lymphohistiocytosis (HLH) in the context of AA. Distinct diseases can also transition into each other, e.g., single lineage cytopenia evolving to pancytopenia or vice versa upon partial response in AA and clonal neoplastic complications seen with AA or LGLL. The heterogeneity in hematological findings translates into significantly varying clinical presentations, such as acute-onset severe pancytopenia commonly seen in AA, leading to abrupt bruising and dyspneic symptoms, while a more indolent disease course seen with LGLL can be captured asymptomatically at routine workups. Significant differences in comorbidities are present between immune-mediated BMFS, such as co-occurrence of hemolytic anemia in AA secondary to paroxysmal nocturnal hemoglobinuria (PNH) [2], which is absent in single-lineage cytopenia or response to immunoglobulins or anti-B cell therapies occasionally seen in patients with single-lineage cytopenias and not seen in AA [3]. Nevertheless, the common feature in these seemingly idiopathic conditions, diagnosed after exclusion of known causes of acquired BMFS [4,5,6,7,8,9,10,11,12], is the significant response to immunosuppressive therapies (IST) targeting immune effector cells [13, 14]. While the exclusion of non-immune causes is important to maximize the response rate [15,16,17,18,19], a significant minority of patients remain refractory or suffer from a relapse [2, 20, 21].
Clinical trials have studied common conditions such as AA and LGLL in standard circumstances [13, 22], but systematic trials are difficult for rare conditions such as PRCA and IN or the more narrowly defined settings such as specific cytopenias or relapsed/refractory disease. The existing literature led to the establishment of a standard of care for BMFS in the treatment-naïve setting, but medical evidence identifying the best standard of care for relapsed and refractory disease is lacking, and the few existing studies in the refractory setting often have limited external validity given the heterogeneity of biology of disease states [23,24,25,26]. Hence, advances in therapy have to be derived from empiric experiences recorded from retrospective case series and meta-analyses [27,28,29,30,31]. One common factor unifying the refractory cases of BMFS is the overlapping spectrum of potentially effective therapeutics. For instance, ATG is an effective therapy for AA and LGL in the refractory setting [13, 32, 33], and tacrolimus has been used in cases with intolerance to cyclosporine (CsA) [27, 31, 34]. Sirolimus is widely utilized in the post-transplant setting and has been occasionally used in single lineage cytopenias [28, 30, 33, 35]. Alemtuzumab has also been studied in a clinical trial setting in AA and in a series of refractory patients treated with various administration regimens in the salvage setting in LGLL [36,37,38]. However, its use is limited by availability, its perceived high complication rate stemming from high-dose regimens in chronic lymphocytic leukemia, and lack of clinical trials in chronic low-dose subcutaneous administration. Abatacept is a recombinant fusion protein of CTLA-4 extracellular domain and the IgG Fc region, blocking CD80/86 engagement of CD28, and no data exist on its efficacy in BMFS as a first- or subsequent-line IST [33]. Tofacitinib is a Janus kinase inhibitor, inhibiting JAK1 and JAK3, and similarly to abatacept, no large-scale study has been conducted showing its efficacy in BMFS.
Over the years, we have treated a sufficiently large volume of BMFS patients at our institution who, for various reasons, have received these agents. In this framework, the clinical utility of our observations is enhanced by a uniform approach to the workup, standardized care pathway for selection of next therapy and a diversified and rational approach to salvage therapy considering comorbidities and financial toxicity. As a result, we have compiled a large cohort of patients with these conditions, which can provide clues to the appropriate choice of agent to maximize potential benefits in difficult and rare clinical scenarios. Here, we describe a cohort of patients with BMFS that had extensive refractory or multiply relapsed disease ( ≥ 4 median lines of therapy) from a single institution that were treated with salvage IST. Through expansion of real-world evidence on these less utilized regimens, clinicians left in a clinical conundrum with treatment-resistant BMFS can turn to real-world data to help make informed decisions for their patients.
Methods
Patients and therapeutics
This retrospective cohort study was conducted under Cleveland Clinic Foundation IRB 15-1278 clinical registry, in accordance with the Declaration of Helsinki. Informed consent was obtained from all subjects before enrolling in the registry. We screened our institutional records for patients with diagnoses of AA, PRCA, immune thrombocytopenia, autoimmune anemia, and LGLL. Clinical and laboratory data were collected, and patients were annotated with the diagnosis that best fit their condition, with underlying hematological conditions (hemolytic anemia, plasma cell dyscrasia, LGLL, MDS, etc.) included as hematological comorbidities (for diagnoses and clinical definitions see Supplementary Table 1). Of a total cohort of 280 patients fulfilling diagnosis criteria, 129 patients who were not on therapy long enough to assess response ( < 1 month), those with unavailable follow-up monitoring, and those using combination IST were excluded, leaving 151 patients on salvage IST (Supplementary Table 2). Alemtuzumab was used subcutaneously, first with an initial test dose of 3 mg, followed by 10 mg weekly doses. Sirolimus was initially started at 2 mg daily and dose-titrated to achieve therapeutic serum levels. Tofacitinib had two regimens, 5 mg twice daily or an 11 mg extended-release form used once daily, both given orally. Abatacept was used subcutaneously, at 50, 87.5, or 125 mg doses. All patients started on alemtuzumab were screened for latent TB and received anti-Pneumocystis jirovecii prophylaxis with inhaled 300 mg pentamidine solution once monthly.
Response criteria
Treatment responses were measured at 3 months after initiation of therapy, and response at 1 month after start of treatment, 6 months, and 12 months are also shown for patients with available data. Reported response rates include both partial and complete response, and response criteria for each pathology can be found in Supplementary Table 1. In disease-specific response calculations, at least two lineage responses were considered response for AA and LGLL.
Statistical analysis
Statistical differences between groups for continuous variables were tested with the Mann–Whitney U test. Differences between groups in categorical variables was tested with Fisher’s exact test. A p-value below 0.05 was assumed to be statistically significant.
Results
Clinical characteristics of study cohort
After initial screening, we identified 280 patients fulfilling study criteria. The required clinical data were available for a subset of patients, and patients without clear diagnoses, intermittent cytopenias or those on multiple IST concomitantly were filtered out; the remaining number of patients investigated for response was 151 (Supplementary Fig. 1). Of these, LGLL (n = 71), severe AA (sAA) (n = 19), moderate AA (mAA) (n = 19) and PRCA (n = 30) were the most common diagnoses. The median age of all patients was 62.5 years (interquartile range 53–77), and 56% of patients were male (Table1). The highest frequency of refractory cases was documented in AA (61%), compared to 33% and 35% in PRCA and LGLL, respectively. Hemolytic anemia was a common comorbidity, observed in 27% of patients. M-protein was observed in 16%; 4% of M-proteins were IgA, 23% were IgM, and 77% were IgG type, and 62% of these were Kappa. 13% of patients were diagnosed with MGUS, and the rest were MM (n = 2), CLL/SLL (n = 1), and lymphoplasmacytic lymphoma (n = 1). Among these patients, sirolimus (n = 73), abatacept (n = 40), alemtuzumab (n = 39), and tofacitinib (n = 33) were the most commonly used salvage immunosuppressants (Table 2). Overall, 133 of 151 patients had anemia, 83 patients had thrombocytopenia, and 99 patients had neutropenia at the time of initiation of salvage therapy. As of completion of this study, 2 sAA, 2 mAA, 2 PRCA, and 3 LGLL patients underwent allogeneic hematopoietic stem cell transplants due to relapsed or refractory disease.
Response rates by disease
Of patients with AA, overall response rate to salvage treatment was 44% (n = 11/25), 43% (n = 6/14) for severe AA and 46% (n = 5/11) in mAA. In sAA, response rates were 46% (n = 6/13) in anemia, 39% (n = 5/13) in thrombocytopenia, and 56% (n = 5/9) in neutropenia, respectively. Similarly, in mAA, response rates were 25% (n = 2/8) in anemia, 40% (n = 6/15) in thrombocytopenia, and 75% (n = 6/8) in neutropenia, respectively (Fig. 1A). For PRCA, overall salvage rate was 66% (n = 17/26) (Fig. 1B). In LGLL, overall response rate was 57% (n = 26/46). Response rate in anemia was 44% (n = 17/39), in thrombocytopenia 76% (n = 16/21), and in neutropenia 76% (n = 22/29) (Fig. 1C). Sirolimus was utilized in 60% of AA patients, 86% of PRCA patients, and 41% of LGLL patients. Abatacept was utilized in 5% of AA patients and 31% of LGLL patients. Tofacitinib was utilized in 3% of AA patients, 10% of PRCA patients, and 35% of LGLL patients. Alemtuzumab was utilized in 36% of AA patients, 30% of PRCA patients, and 27% of LGLL patients. Specific response rates for each treatment option and the number of treated patients with each therapy are in Fig. 1D. No statistically significant difference was observed in response rates between therapeutic agents.
A Response rates by lineages in patients with aplastic anemia. B Response rate in pure red cell aplasia. C Response rates by lineages in large granular lymphocytic leukemia. D Response rates for each medication in each affected lineage. The numerator is the number responded, and the denominator is the number treated.
Response rates by affected lineage
In 133 patients with immune-mediated hypoproliferative anemia at the start of salvage IST, regardless of diagnosis, combined overall response rate was 45% at 3 months (Fig. 2A) for all studied drugs. Similarly, in 83 patients with thrombocytopenia, overall salvage rate at 3 months was 58% (Fig. 2B). In 99 patients with neutropenia, overall salvage rate was 73% (Fig. 2C). Highest response rate in patients with anemia was achieved with sirolimus at 59% in refractory cases, and 54% in those who responded to prior IST but relapsed (Fig. 2D). In patients with thrombocytopenia at the start of salvage IST, highest rate of response in terms of platelet numbers was observed with tofacitinib at 75%, 67% of refractory and 80% of relapsed cases responded (Fig. 2E). In patients with neutropenia at the start of IST, highest response rate was again detected with abatacept (80%). In patients with refractory neutropenia, the highest response was seen with abatacept (83%), while in those with neutropenia that previously responded to IST, the highest response rate was detected with sirolimus (83%). (Fig. 2F) No statistically significant difference was observed in response between agents (Supplementary Fig. 1).
Overall response rates of each effected lineage regardless of diagnosis. Numbers in colored squares show how many patients were treated with each medication. A Patients with anemia. The red dashed line is the combined IST response in all patients. B Patients with thrombocytopenia. C Patients with neutropenia. D Patients with anemia, left: refractory disease; right: relapsed on prior IST. Blue dashed line: combined response in all IST in all treatment-refractory patients. Dark red dashed line: combined response in all IST in all treatment-relapsed patients. E Patients with thrombocytopenia. F Patients with neutropenia.
Acquired BMFS are frequently relapsing diseases [2, 21], and waning immunosuppression carries a risk of relapse. In our cohort, the highest response rate was achieved in 1 month at 68% for immune neutropenia, 48.6% for thrombocytopenia, and 27% for anemia. The 3-month response rates were 44% for anemia, 62% for thrombocytopenia, and 73% for neutropenia. Response rates at 6 months were 53%, 75%, and 74%, and response rates at 12 months were 70%, 80%, and 74% for anemia, thrombocytopenia, and neutropenia, respectively (Fig. 3A). Response rates for each lineage of cytopenia, with number of patients on therapy can be found in Fig. 3B, C, D. In patients who had a response, median time to response (TTR) in AA was 2 months with sirolimus and abatacept, and 1 month with alemtuzumab, in PRCA, median TTR was 1 month with sirolimus and alemtuzumab, and in LGLL, TTR was 1 month with sirolimus, abatacept, and alemtuzumab, and 3 months with tofacitinib. Median duration of response in AA was 3 months with sirolimus and alemtuzumab and 7.5 months with abatacept; in PRCA, it was 7.5 months with sirolimus and 12 months with alemtuzumab; and in LGLL, it was 3 months with tofacitinib and alemtuzumab and 7.5 months with sirolimus and abatacept.
A Combined response rates at 1, 3, 6, and 12 months of therapy for all studied IST. B Response rates at each time point for individual therapies in anemia. C Response rates at each time point for individual therapies in thrombocytopenia. D Response rates at each time point for individual therapies in neutropenia. Risk tables show number of patients treated and responded at each time point.
Regardless of the diagnosis, hemoglobin was the slowest to respond to a specific treatment at 1 month, with response rate in AA 23% (Fig. 4A), 13-47% in LGLL, and 33–57% (Fig. 4B) in PRCA (Fig. 4C), even though some patients achieved durable response by later months (Fig. 4C). Patients with thrombocytopenia and neutropenia generally responded well to salvage therapy faster than patients with anemia, and some patients retained durable responds later in treatment, though no statistical difference was observed between response rates in lineages and different therapeutics. (Fig. 4D–G). The number of patients on each therapy throughout the duration of the observation period is in Fig. 4H.
Response rates by duration of therapy, affected lineage, and diagnosis. A Response rates in aplastic anemia patients with anemia. B Response rates in large granular lymphocytic leukemia patients with anemia. C Response rates in pure red cell aplasia patients. D Response rates in aplastic anemia patients with thrombocytopenia. E Response rates in large granular lymphocytic leukemia patients with thrombocytopenia. F Response rates in aplastic anemia patients with neutropenia. G Response rates in large granular lymphocytic leukemia patients with neutropenia. H Top rows denote the number of patients who responded and bottom rows denote the number of patients treated with each regimen. Al alemtuzumab, purple dots in figures, T tofacitinib, brown dots in figures, Ab abatacept, orange dots in figures, S sirolimus, green dots in figures. Thromb thrombocytopenia, Neutro neutropenia.
Indications for utilization, associated toxicities, treatment duration, and indication to stop treatment
The most common indication to start salvage therapy was refractory disease (n = 164/184), followed by nephrotoxicity from prior regimen (n = 12). Hepatotoxicity (n = 3) and recalcitrant rash (n = 1) were the next most common indications for starting salvage regimens. The most common toxicity observed in general was high-risk infections requiring hospitalization (n = 16); pneumonia (n = 7) was the most common documented severe infection requiring hospitalization, seen in patients taking abatacept (n = 1), sirolimus, alemtuzumab, and tofacitinib (n = 2 each). In cases when the pathogen was identified, CMV was the most common culprit; viremia was the most common presentation (n = 3), followed by colitis (n = 1), presenting in patients receiving alemtuzumab (n = 3) and tofacitinib (n = 1). After infections, the most common toxicity was non-infectious skin changes (n = 12), most common of these being an erythematous, pruritic rash (n = 7), all seen in patients taking sirolimus. Oral sores were seen in 2 patients taking tofacitinib.
An important aspect to consider for long-term IST is the total financial burden, including overhead costs associated with monitoring, prophylaxis and management of toxicities, in addition to the direct cost of medicine and office visits. In total, 20 patients were observed to have financial toxicity, in which a delay or change in treatment had to be made due to costs attached to the studied regimen. This was most common in patients taking abatacept (n = 9), followed by sirolimus (n = 6).
For the 29 patients (40%) who stopped sirolimus due to any reason other than no response to therapy and 15 patients who were still on therapy at last follow-up, median duration of treatment was 13 months (range 1–128). The most common reason to stop therapy was death (n = 12). For the 14 patients (35%) who stopped abatacept for indications other than no response to therapy and 14 patients on therapy at last follow-up, the median duration of treatment was 22 months (range 1–201), and the most common reason for termination of therapy was clinical change in rheumatological comorbidity (n = 3). For 13 patients (39%) who stopped tofacitinib for reasons other than non-response and 9 patients receiving treatment at last follow-up, the median duration of treatment was 19 months (range 1–115); the most common reason for stopping therapy was death (n = 4). For 12 patients who stopped alemtuzumab treatment for any reason other than non-response and 2 patients with ongoing therapy on follow-up, the median duration of treatment was 4 months (range 1–30), and the most common reason for stopping treatment was complete remission of disease (n = 4).
Discussion
Single- or multi-lineage cytopenias due to BMFS can lead to significant morbidity, but the most common forms of acquired cytopenias are fortunately responsive to IST. [22, 39,40,41] The frontline standard-of-care therapies were established in clinical trials, but beyond standard of care, various salvage regimens need to be used, with their utility empirically established and documented in case reports and series. Here, we analyzed the efficacy of salvage IST regimens used in a single high-volume center, in refractory and relapsed cases of BMFS in 151 patients. The top 4 novel salvage IST regimens in this setting included alemtuzumab, abatacept, tofacitinib, and sirolimus. The results from the analysis of response rate in this study may inform a more rational application of salvage IST in these conditions.
We demonstrated a robust response rate to alemtuzumab in refractory and multiply relapsed AA, PRCA, and LGLL. Alemtuzumab has the largest body of evidence in BMFS, including prospective trials [3, 36,37,38, 42,43,44]. In addition to previous smaller series from our center [3, 45], it has been used in AA, LGLL, and PRCA in primary and refractory settings. Results from this study showing a 3-month ORR of 50% in PRCA and an ORR of 73% at 3 months in LGLL are comparable to those described in prospective studies: a cohort of 13 treatment-naïve and relapsed PRCA patients had a reported ORR of 69% at 3 months [44], and a cohort of 25 LGLL patients had a 3-month ORR of 56% [38]. However, even when used in similar settings, it is difficult to compare results between studies, as the previously mentioned trials utilized a significantly higher loading dose (73–103 mg in 5–10 days), followed by CsA maintenance, which is in comparison to 10 mg alemtuzumab maintenance dose once monthly we describe here. Differences in administration route are also likely to introduce bias, as even though similar efficacy between IV and SC routes was established when used as consolidation therapy in CLL and in AA no similar comparison exists for LGLL [42, 46]. Interestingly, there is a significant difference in the rate of detected adverse events: in PRCA patients treated with alemtuzumab, infections were observed in only 28% of our cohort compared to 54%, and no neutropenia was observed in our cohort compared to 31%. In LGLL patients, we observed severe lymphopenia in only 11% of patients, compared to grade 3–4 lymphopenia in 88%, and the infection rate was only 21%, compared to 64% of those treated. Similarly, in severe AA, alemtuzumab was investigated in treatment-naïve, relapsed, and refractory settings.[36, 37, 42] In the treatment-naïve setting, the trial was stopped early [37] due to lower ORR in comparison to standard of care. In the refractory setting, 3- and 6-month response rates were 19% and 33%, respectively. The most common adverse effect was neutropenic fever, in 48% of patients, and CMV reactivation was detected in 100% of those who received alemtuzumab. In comparison, we have shown a 33% ORR at 3 months, and no patients in this setting were still on alemtuzumab monotherapy at 6 months to be evaluated. The most common side effects in our cohort were severe lymphopenia and bacteremia, seen in 25% of patients each. These differences show the necessity for stringent clinical evaluation, considering the risk of significant adverse events and possibly a better opportunity to achieve response with different protocols compared to those that have been historically studied, using condensed and higher-dose therapy. Overall, the numerically similar response rates in our study compared to established literature and lower adverse event rates suggest intermittent subcutaneous administration of alemtuzumab is an avenue that requires further research, with the potential to retain the IST efficacy while potentially improving the safety of this underutilized IST regimen.
Sirolimus has previously been investigated as an IST agent in BMFS, specifically PRCA and AA [30, 35, 47,48,49]. In addition to case reports, 2 different studies have independently demonstrated efficacy in AA and acquired PRCA, respectively. In PRCA, the studied cohort composition was mixed, relapsed, refractory, and intolerant to CsA or CsA/cyclophosphamide, and ORR was 50% and 84.4% at 3 and 6 months, respectively. In contrast, we demonstrated 74% and 92% ORR at 3 and 6 months, respectively. In sAA, sirolimus was added to the standard of care (SOC) ATG/CsA, and ORR was 37% in the sirolimus group compared to 57% in SOC, and 51% vs 62% at 6 months. However, as an add-on to up-front SOC, results of this study are not directly comparable to our findings. Given no directly comparable data in the literature, our findings showing 60% and 67% ORR at 3 and 6 months, respectively, bring real-world evidence to a novel setting and imply that sirolimus could be a viable salvage therapy option in immune-mediated BMFS.
Ruxolitinib, a JAK1/2 selective inhibitor, previously demonstrated to have 20% true RR (n = 1/5) at 6 months in LGLL [50]. However, another 3 patients were documented to have stable disease in the same study, increasing the reported ORR to 80%. Another study [51] also demonstrated efficacy of ruxolitinib in a treatment refractory/relapsed cohort of 21 LGLL patients, with a 72% (n = 13/18) ORR within 3 months, and 86% (n = 18/21) at 12 months. Tofacitinib, which is mainly used as IST in rheumatoid arthritis, a common comorbidity with LGLL, has a similar but not identical mechanism of action to ruxolitinib, with more selective inhibition of JAK3, and has a higher potential to inhibit T-cell-restricted proliferation and differentiation compared to ruxolitinib [52,53,54]; however, this was not previously shown in a head-to-head comparison in human trials. We observed a 57% ORR (n=8/14) at 3 months and 50% (n=5/10) at 6 months in LGLL, which are numerically lower than previously described evidence for ruxolitinib; however, different response criteria and disease settings make unbiased direct comparisons challenging. Although abatacept, through its blockage of costimulatory signaling in T-cells, could be effective in BMFS, in a similar mechanism of action to its efficacy in graft-versus-host disease [55] and rheumatoid arthritis [56], which is a common comorbidity with LGLL, no data on its efficacy in BMFS or neutropenia in rheumatoid arthritis exist. Given the high rate of LGLL and rheumatoid arthritis overlap, many patients in our cohort were exposed to abatacept as an IST, and its demonstrated efficacy in BMFS is novel and needs to be validated in further studies.
Our findings show interesting and novel trends, first of which was response rate differences observed in affected lineages, regardless of underlying disease pathology, with anemia having the lowest response rate in both LGLL and AA, while neutropenia having the highest response rate. This was true for both early response achieved in 1 month after starting therapy and throughout the duration of treatment, with response rates only converging toward 12 months, after many patients with refractory disease had been taken off therapy. This finding warrants further research, perhaps incorporating molecular investigations such as cytokine signaling and immune activation assays involving lineage-restricted progenitors, to better delineate the differences in immunomodulatory effects of T-cells targeting erythroid and other progenitors. A second interesting finding from our data is the observed lineage-specific time to response. Patients with neutropenia and thrombocytopenia achieved a 68% and 49% response rate at 1 month, respectively, a shorter time period than was utilized in the clinical trials. Even in this multiply relapsed and refractory cohort, this rapid-onset salvage response suggests alternating forms of IST are not futile, and rapid improvement in clinical condition could be possible. Finally, even though this study did not have the necessary power to statistically compare the differences in response rates between different IST modalities, all treatments showed significant response, with at least 44% of patients with cytopenia of any lineage by 3 months, and response rates as high as 80% in some cases. In addition to these, we have demonstrated efficacy of different IST modalities in settings where scarce or no prior evidence existed.
All the limitations of a retrospective single-institute study notwithstanding, inherent bias arises from selection of patients in a non-randomized manner, especially in a heterogenous group of BMFS patients with a variety of hematological phenotypes. Investigating a relapsed/refractory cohort of these rare conditions results in limited cohort sizes, in turn, limiting the power of statistical comparisons between analyzed therapeutics. In addition, treatment by different physicians at different timelines with different clinical criteria for response and without following a unified protocol inherently leads to treatments being stopped for reasons not captured in this study, preventing an intention-to-treat analysis, most likely introducing survivorship and attrition biases resulting in elevated response rates at later timepoints points compared to earlier ones. However, our results indicate general expected utility of salvage modalities not utilized in any randomized controlled trials in these settings, providing important hints to the application of salvage IST regimens in immune-mediated BMFS. Confirmation of the findings in this real-world report should be pursued in prospective randomized clinical trials, although doing such studies in these rare populations is unlikely to be completed. Thus, medical evidence to support clinical decisions in the relapsed setting is an area of unmet need and needs further attention.
Data availability
All data generated or analyzed during this study are included in this published article and its supplementary information files.
References
Young NS, Maciejewski J. The pathophysiology of acquired aplastic anemia. N Engl J Med. 1997;336:1365–72.
Young NS. Aplastic anemia. N Engl J Med. 2018;379:1643–56.
Balasubramanian SK, Sadaps M, Thota S, Aly M, Przychodzen BP, Hirsch CM, et al. Rational management approach to pure red cell aplasia. Haematologica. 2018;103:221–30.
Mortimer PP, Humphries RK, Moore JG, Purcell RH, Young NS. A human parvovirus-like virus inhibits haematopoietic colony formation in vitro. Nature. 1983;302:426–9.
Shade RO, Blundell MC, Cotmore SF, Tattersall P, Astell CR. Nucleotide sequence and genome organization of human parvovirus B19 isolated from the serum of a child during aplastic crisis. J Virol. 1986;58:921–36.
Green R, Datta Mitra A. Megaloblastic anemias: nutritional and other causes. Med Clin North Am. 2017;101:297–317.
Hoffbrand AV, Herbert V. Nutritional anemias. Semin Hematol. 1999;36:13–23.
Wickramasinghe SN. Diagnosis of megaloblastic anaemias. Blood Rev. 2006;20:299–318.
Eichner ER, Hillman RS. The evolution of anemia in alcoholic patients. Am J Med. 1971;50:218–32.
Lindenbaum J, Roman MJ. Nutritional anemia in alcoholism. Am J Clin Nutr. 1980;33:2727–35.
Chang YH. Myelodysplastic syndromes and overlap syndromes. Blood Res. 2021;56:S51–s64.
Karantanos T, DeZern AE. Biology and clinical management of hypoplastic MDS: MDS as a bone marrow failure syndrome. Best Pract Res Clin Haematol. 2021;34:101280.
Frickhofen N, Heimpel H, Kaltwasser JP, Schrezenmeier H, German Aplastic Anemia Study G. Antithymocyte globulin with or without cyclosporin A: 11-year follow-up of a randomized trial comparing treatments of aplastic anemia. Blood. 2003;101:1236–42.
Rosenfeld SJ, Kimball J, Vining D, Young NS. Intensive immunosuppression with antithymocyte globulin and cyclosporine as treatment for severe acquired aplastic anemia. Blood. 1995;85:3058–65.
Tanaka TN, Bejar R. MDS overlap disorders and diagnostic boundaries. Blood. 2019;133:1086–95.
Kyle RA, Gertz MA, Witzig TE, Lust JA, Lacy MQ, Dispenzieri A, et al. Review of 1027 patients with newly diagnosed multiple myeloma. Mayo Clin Proc. 2003;78:21–33.
Blain H, Hamdan KA, Blain A, Jeandel C. Aplastic anemia induced by phenytoin: a geriatric case with severe folic acid deficiency. J Am Geriatr Soc. 2002;50:396–7.
Hill QA, Stamps R, Massey E, Grainger JD, Provan D, Hill A, et al. Guidelines on the management of drug-induced immune and secondary autoimmune, haemolytic anaemia. Br J Haematol. 2017;177:208–20.
Savage SA, Dufour C. Classical inherited bone marrow failure syndromes with high risk for myelodysplastic syndrome and acute myelogenous leukemia. Semin Hematol. 2017;54:105–14.
Chattopadhyay S, Lionel S, Selvarajan S, Devasia AJ, Korula A, Kulkarni U, et al. Relapse and transformation to myelodysplastic syndrome and acute myeloid leukemia following immunosuppressive therapy for aplastic anemia is more common than with allogeneic stem cell transplantation, with a negative impact on survival. Ann Hematol. 2024;103:749–58.
Patel BA, Groarke EM, Lotter J, Shalhoub R, Gutierrez-Rodrigues F, Rios O, et al. Long-term outcomes in patients with severe aplastic anemia treated with immunosuppression and eltrombopag: a phase 2 study. Blood. 2022;139:34–43.
Loughran TP Jr, Zickl L, Olson TL, Wang V, Zhang D, Rajala HL, et al. Immunosuppressive therapy of LGL leukemia: prospective multicenter phase II study by the Eastern Cooperative Oncology Group (E5998). Leukemia. 2015;29:886–94.
Wu Z, Gao S, Feng X, Li H, Sompairac N, Jamshidi S, et al. Human autoimmunity at single-cell resolution in aplastic anemia before and after effective immunotherapy. Nat Commun. 2025;16:5048.
Park S, Kim HK. Epidemiologic sequential analysis of pure red blood cell aplasia and T-cell large granular lymphocyte leukemia in Korea. Ann Hematol. 2025;104:2881–7.
Magnano L, Rivero A, Matutes E. Large granular lymphocytic leukemia: current state of diagnosis, pathogenesis and treatment. Curr Oncol Rep. 2022;24:633–44.
Cheon H, Dziewulska KH, Moosic KB, Olson KC, Gru AA, Feith DJ, et al. Advances in the diagnosis and treatment of large granular lymphocytic leukemia. Curr Hematol Malig Rep. 2020;15:103–12.
Du Y, Huang Y, Zhou W, Liu X, Chen F, Yang C, et al. Effective tacrolimus treatment for patients with non-severe aplastic anemia that is refractory/intolerant to cyclosporine A: a retrospective study. Drug Des Devel Ther. 2020;14:5711–9.
Long Z, Yu F, Du Y, Li H, Chen M, Zhuang J, et al. Successful treatment of refractory/relapsed acquired pure red cell aplasia with sirolimus. Ann Hematol. 2018;97:2047–54.
Chen Z, Chen M, Yang C, Han B. Immunosuppression therapy is effective for both acquired tumor-associated and primary pure red cell aplasia: a matched pair case-control study. Ann Hematol. 2020;99:1485–91.
Chen Z, Liu X, Chen M, Yang C, Han B. Successful sirolimus treatment of patients with pure red cell aplasia complicated with renal insufficiency. Ann Hematol. 2020;99:737–41.
Alsultan A, Goldenberg NA, Kaiser N, Graham DK, Hays T. Tacrolimus as an alternative to cyclosporine in the maintenance phase of immunosuppressive therapy for severe aplastic anemia in children. Pediatr Blood Cancer. 2009;52:626–30.
Zhu Y, Gao Q, Hu J, Liu X, Guan D, Zhang F. Clinical features and treatment outcomes in patients with T-cell large granular lymphocytic leukemia: a single-institution experience. Leuk Res. 2020;90:106299.
Sanikommu SR, Clemente MJ, Chomczynski P, Afable MG, Jerez A, Thota S, et al. Clinical features and treatment outcomes in large granular lymphocytic leukemia (LGLL). Leuk Lymphoma. 2018;59:416–22.
Zhu X, Guan J, Xu J, Wei J, Jiang L, Yin J, et al. Pilot study using tacrolimus rather than cyclosporine plus antithymocyte globulin as an immunosuppressive therapy regimen option for severe aplastic anemia in adults. Blood Cells Mol Dis. 2014;53:157–60.
Huang Y, Chen M, Yang C, Ruan J, Wang S, Han B. Sirolimus is effective for refractory/relapsed/intolerant acquired pure red cell aplasia: results of a prospective single-institutional trial. Leukemia. 2022;36:1351–60.
da Fonseca AR, Justino CC, Campos de Molla V, Yamakawa PE, Rabelo IB, Arnold L, et al. Outpatient subcutaneous alemtuzumab is feasible and safe for aplastic anemia and is associated with high response rates. Blood Adv. 2025;9:906–12.
Scheinberg P, Nunez O, Weinstein B, Scheinberg P, Wu CO, Young NS. Activity of alemtuzumab monotherapy in treatment-naive, relapsed, and refractory severe acquired aplastic anemia. Blood. 2012;119:345–54.
Dumitriu B, Ito S, Feng X, Stephens N, Yunce M, Kajigaya S, et al. Alemtuzumab in T-cell large granular lymphocytic leukaemia: interim results from a single-arm, open-label, phase 2 study. Lancet Haematol. 2016;3:e22–9.
Scheinberg P, Nunez O, Weinstein B, Scheinberg P, Biancotto A, Wu CO, et al. Horse versus rabbit antithymocyte globulin in acquired aplastic anemia. N Engl J Med. 2011;365:430–8.
Frickhofen N, Kaltwasser JP, Schrezenmeier H, Raghavachar A, Vogt HG, Herrmann F, et al. The German Aplastic Anemia Study Group. Treatment of aplastic anemia with antilymphocyte globulin and methylprednisolone with or without cyclosporine. N Engl J Med. 1991;324:1297–304.
Loughran TP Jr, Kidd PG, Starkebaum G. Treatment of large granular lymphocyte leukemia with oral low-dose methotrexate. Blood. 1994;84:2164–70.
Aggarwal N, Manley AL, Shalhoub R, Durrani J, Rios O, Lotter J, et al. Alemtuzumab in relapsed immune severe aplastic anemia: long-term results of a phase II study. Am J Hematol. 2023;98:932–9.
Risitano AM, Schrezenmeier H. Alternative immunosuppression in patients failing immunosuppression with ATG who are not transplant candidates: Campath (Alemtuzumab). Bone Marrow Transpl. 2013;48:186–90.
Risitano AM, Selleri C, Serio B, Torelli GF, Kulagin A, Maury S, et al. Alemtuzumab is safe and effective as immunosuppressive treatment for aplastic anaemia and single-lineage marrow failure: a pilot study and a survey from the EBMT WPSAA. Br J Haematol. 2010;148:791–6.
Mohan SR, Clemente MJ, Afable M, Cazzolli HN, Bejanyan N, Wlodarski MW, et al. Therapeutic implications of variable expression of CD52 on clonal cytotoxic T cells in CD8+ large granular lymphocyte leukemia. Haematologica. 2009;94:1407–14.
Byrd JC, Peterson BL, Rai KR, Hurd D, Hohl R, Perry MC, et al. Fludarabine followed by alemtuzumab consolidation for previously untreated chronic lymphocytic leukemia: final report of cancer and leukemia group B study 19901. Leuk Lymphoma. 2009;50:1589–96.
Scheinberg P, Wu CO, Nunez O, Scheinberg P, Boss C, Sloand EM, et al. Treatment of severe aplastic anemia with a combination of horse antithymocyte globulin and cyclosporine, with or without sirolimus: a prospective randomized study. Haematologica. 2009;94:348–54.
He G, Zhang X, Wu D, Sun A, Wang X. Relapse of aplastic anemia responsive to sirolimus combined with cyclosporine. Pediatr Blood Cancer. 2011;56:1133–5.
Niu H, Qi W, Wang Y, Xing L, Fu R, Shao Z, et al. Successful sirolimus therapy of an aplastic anemia patient with chronic kidney disease: a case report. Medicine. 2020;99:e20669.
Moskowitz AJ, Ghione P, Jacobsen E, Ruan J, Schatz JH, Noor S, et al. A phase 2 biomarker-driven study of ruxolitinib demonstrates effectiveness of JAK/STAT targeting in T-cell lymphomas. Blood. 2021;138:2828–37.
Marchand T, Pastoret C, Damaj G, Lebouvier A, Herbaux C, Moignet A, et al. Efficacy of ruxolitinib in the treatment of relapsed/refractory large granular lymphocytic leukaemia. Br J Haematol. 2024;205:915–23.
Roskoski R Jr. Janus kinase (JAK) inhibitors in the treatment of inflammatory and neoplastic diseases. Pharm Res. 2016;111:784–803.
O’Shea JJ, Holland SM, Staudt LM. JAKs and STATs in immunity, immunodeficiency, and cancer. N Engl J Med. 2013;368:161–70.
Traves PG, Murray B, Campigotto F, Galien R, Meng A, Di Paolo JA. JAK selectivity and the implications for clinical inhibition of pharmacodynamic cytokine signalling by filgotinib, upadacitinib, tofacitinib and baricitinib. Ann Rheum Dis. 2021;80:865–75.
Watkins B, Qayed M, McCracken C, Bratrude B, Betz K, Suessmuth Y, et al. Phase II trial of costimulation blockade with abatacept for prevention of acute GVHD. J Clin Oncol. 2021;39:1865–77.
Cope AP, Jasenecova M, Vasconcelos JC, Filer A, Raza K, Qureshi S, et al. Abatacept in individuals at high risk of rheumatoid arthritis (APIPPRA): a randomised, double-blind, multicentre, parallel, placebo-controlled, phase 2b clinical trial. Lancet. 2024;403:838–49.
Funding
CBP, VV, and JPM disclose funding support from the Aplastic Anemia and MDS International Foundation. The rest of the authors declare no relevant funding.VeloSano 10 Visionary Award (to J.P.M).
Author information
Authors and Affiliations
Cleveland Clinic Cancer Center, Cleveland, OH, USA
Serhan Unlu, Aashray Mandala, Mark D. Orland, Carlos Bravo-Perez, Valeria Visconte, Akriti G. Jain, Alan Lichtin, Hetty E. Carraway, Carmelo Gurnari & Jaroslaw P. Maciejewski
Department of Biomedicine and Prevention, University of Rome Tor Vergata, Rome, Italy
Carmelo Gurnari
Authors
- Serhan Unlu
- Aashray Mandala
- Mark D. Orland
- Carlos Bravo-Perez
- Valeria Visconte
- Akriti G. Jain
- Alan Lichtin
- Hetty E. Carraway
- Carmelo Gurnari
- Jaroslaw P. Maciejewski
Contributions
SU designed the study, collected and analyzed data, and drafted the manuscript. AM, MO, and CBP collected and analyzed the data. VV, AGJ, AL, and HEC helped design the study. CG helped conceive the study. JPM conceived and supervised the study. All authors reviewed and edited the manuscript.
Corresponding author
Correspondence to Jaroslaw P. Maciejewski.
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Unlu, S., Mandala, A., Orland, M.D. et al. Salvage therapy for immune-mediated bone marrow failure. Leukemia (2026). https://doi.org/10.1038/s41375-026-03180-1
Received: 21 April 2026
Revised: 21 September 2026
Accepted: 12 October 2026
Published: 03 November 2026
Version of record: 03 November 2026
DOI: https://doi.org/10.1038/s41375-026-03180-1
来源:Nature 健康护理主题 · nature.com