Outcomes and treatment patterns in primary and secondary plasma cell leukemia: insights from a large US cohort study

Abstract

Plasma cell leukemia (PCL) is a rare and aggressive plasma cell disorder with a short survival duration despite the use of novel therapeutics. PCL remains an understudied disease for which there is no current standard of care treatment. Knowledge on optimal novel drug sequencing, including the role and timing of hematopoietic stem cell transplant as primary and salvage therapy is needed. We conducted a large retrospective analysis of 153 PCL patients, using the new definition of ≥5% circulating plasma cells, including clinical features and treatment outcomes across seven academic centers in the United States. This is, to the best of our knowledge, the largest multicenter study conducted in the US of this rare disease. Disease presentation, clinical and genetic characteristics, and treatment patterns of 93 patients with primary PCL and 57 with secondary PCL are described. Additionally, associations between patient characteristics and mortality were investigated using Cox proportional hazards regression models. Secondary PCL was associated with a worse prognosis, with a median overall survival of 3.2 months compared to 36.6 months for primary PCL (P<0.001). Receipt of a transplant was associated with a survival advantage in both primary PCL (Hazard Ratio [HR] 0.16, P<0.001) and secondary PCL (HR 0.20, P=0.001). No significant difference in outcomes was observed between proteasome inhibitor-based triplet regimens and the VTD-PACE like regimen. The presence of extramedullary disease and high-risk cytogenetics was not associated with survival in the primary PCL group.

Introduction

Plasma cell leukemia (PCL) is a rare and aggressive malignancy, constituting only 0.1-0.6% of all plasma cell disorders in the United States.¹ There are two variants of PCL: primary plasma cell leukemia (pPCL) arises de novo, and secondary plasma cell leukemia (sPCL) develops from multiple myeloma (MM).² Both subtypes carry a poor prognosis, but sPCL is often more aggressive.³ Clinically, PCL patients present with more severe symptoms and biological abnormalities as compared to MM. Disease manifestations include thrombocytopenia, renal impairment, concomitant extramedullary disease (EMD), and high-risk cytogenetic abnormalities contributing to poor performance and a high mortality rate.^3,4

Until recently, a PCL diagnosis required ≥20% circulating plasma cells or an absolute plasma cell count of ≥2,000/ mm³, as defined by Kyle et al.⁵ The International Myeloma Working Group (IMWG) expanded the definition of PCL in 2021 to include patients with ≥5% circulating plasma cells, after several studies found these patients to have similarly poor outcomes.⁶ PCL remains an understudied disease for which there is no current standard of care treatment. Novel treatments for PCL have been introduced in recent years, including CD38 targeted antibodies, proteasome inhibitors (PI), immunomodulatory drugs used during induction and maintenance phases, hematopoietic stem cell transplantation (HSCT), and cellular therapies such as chimeric antigen T-cell receptor (CAR-T) therapies.^7,9 Despite these advancements, overall survival (OS) has remained limited, with only incremental gains observed between 2005 and 2020, ranging from 13 months to 18.3 months in pPCL and 1.2-3.5 months in sPCL.¹⁰,¹¹ One study reported one-month, one-year, and three-year survival rates as 97.0%, 68.9%, and 36.1% for pPCL and 57.6%, 7.6%, and 3.8% for sPCL, respectively.¹¹ Elucidating optimal novel drug sequencing, including the role and timing of HSCT as primary and salvage therapy, is critically important.

One of the primary challenges in treating PCL is the limited data and lack of large, well-designed clinical trials. Most studies to date have been small, retrospective, or single center in nature, limiting the generalizability of their findings. The bulk of the literature has focused on pPCL, with comparatively fewer studies exploring the more aggressive sPCL type. The heterogeneity in clinical course and treatment responses between pPCL and sPCL further complicates the development of standardized treatment protocols.^3,4,12

This large, multicenter retrospective analysis aims to examine patient characteristics, disease features, and treatment-associated survival outcomes for both pPCL and sPCL patients. This is, to the best of our knowledge, the largest multicenter study conducted in the US of this rare disease. By identifying key factors that influence prognosis, this study seeks to inform future treatment strategies to improve survival outcomes for this rare but deadly malignancy.

Methods

Data from patients diagnosed with pPCL and sPCL between January 2010 and January 2021 were collected from electronic medical records at seven academic centers in the United States and entered into a study-specific REDCap database. The study protocol received approval from the Institutional Review Board of each participating institution. The inclusion criteria required a diagnosis of PCL defined by the presence of ≥5% circulating plasma cells, as per the 2021 IMWG consensus guidelines,⁶ and sPCL patients were required to have a prior diagnosis of MM. High-risk genetic features, as defined by the revised 2016 IMWG consensus guidelines, were defined by t(4;14), t(14;16), Del17p13, a non-hyper diploid karyotype and gain (1q).¹³ Response data were determined by each treating institution.

Patient characteristics were summarized descriptively. Kaplan-Meier survival curves were generated to assess OS, defined as the time from diagnosis to last follow-up or death. Patients that were lost to follow-up were censored at the date they were last known to be alive. Cox proportional hazards regression models were employed to test the associations between patient characteristics and survival, producing hazard ratios (HR) with 95% confidence intervals (CI), adjusted for PCL type and transplant status. All statistical analyses were performed using Stata 18, with two-sided P values and an alpha level of 0.05.

Results

Demographics and clinical characteristics

A total of 157 patients were initially identified, with 4 excluded due to unknown survival times, leaving 153 patients for analysis (96 with pPCL and 57 with sPCL). The median follow up-time was 13.1 months. The median time from initial MM diagnosis to sPCL progression was 21.2 months (range: 1-194.5 months). All patients in the data set met the inclusion criteria of ≥5% circulating plasma cells. However, the data on actual percentage of circulating plasma cells were incomplete (Table 1). The median age at diagnosis was similar for pPCL and sPCL patients at 60.0±10.8 years and 60.7±9.2 years, respectively (P=0.684). Average creatinine levels were higher in pPCL patients (3.3±0.8 mg/dL) compared to sPCL patients (1.7±1.4 mg/dL) (P=0.315), while beta-2 microglobulin levels were significantly elevated in pPCL (12.6±14.6 mg/L) compared to sPCL (7.0±6.7 mg/L) (P=0.001). Conversely, LDH levels were higher in the sPCL cohort (838.8±937.1 U/L) compared to pPCL (342.9±235.2 U/L) (P=0.144). Albumin levels were similarly low in both groups, averaging 3.3±0.8 g/dL in pPCL and 3.3±0.6 g/dL in sPCL (P=0.754). EM disease was observed more frequently in sPCL patients (34.0%, 16/47) than in pPCL patients (20.5%, 17/83) (P=0.098). Among high-risk genetic features, del(17p) was more common in sPCL (42.9%, 12/26) compared to pPCL (34.4%, 21/61) (P=0.484), and t(4;14) was also slightly more frequent in sPCL (24.0%, 6/25) than in pPCL (21.1%, 12/57) (P=0.778). However, t(14;16) was observed more among pPCL patients (22.4%, 13/58) than in sPCL patients (4.3%, 1/23) (P=0.059). The t(11;14) translocation was observed more frequently at diagnosis (54.4%, 31/62) or at any time during the disease course (50.7%, 36/71) in pPCL patients compared to sPCL patients (40.9%, 9/22, and 48.3%, 14/29, respectively) (P>0.05).

In terms of staging, pPCL patients were primarily classified as International Staging System (ISS) stage III (81.2%, 26/32), with smaller proportions at stages I (6.2%, 2/32) and II (12.5%, 4/32). In contrast, sPCL patients showed a more balanced distribution across ISS stages III (48.0%, 12/25), II (36.0%, 9/25), and I (16.0%, 4/25) (P=0.027). Revised (R)-ISS staging also reflected higher proportions of advanced disease in pPCL, with 64.0% (16/25) in stage III, 32.0% (8/25) in stage II, and 4.0% (1/25) in stage I, compared to 50.0% (9/18) in stage III, 33.3% (6/18) in stage II, and 16.7% (3/28) in stage I for sPCL patients (P=0.421). However, staging analysis was limited due to missing data.

Table 1.Patient characteristics.

Transplantation rates differed significantly between the groups, with 69.9% (65/93) of pPCL patients undergoing transplantation compared to only 23.8% (10/42) of sPCL patients. Among pPCL patients, 50.0% (45/90) received a single autologous (auto) transplant, 4.4% (4/90) a single allogeneic (allo) transplant, 1.1% (1/90) tandem auto-auto, and 11.1% (10/90) tandem auto-allo at induction. At relapse, transplantation included 2.2% (2/90) receiving a single autologous transplant, 4.4% (4/90) a single allogeneic transplant, 1.1% (1/90) tandem auto-auto, and 1.1% (1/90) tandem auto-allo. For sPCL patients, transplantation at induction included 14.3% (6/42) receiving a single autologous transplant, 2.4% (1/42) a single allogeneic transplant, and 2.4% (1/42) tandem auto-allo, while at relapse only 5.0% (2/40) underwent a single autologous transplant.

Treatment regimens also varied between the groups. PI-based triplet regimens were more commonly used in pPCL patients (72.7%, 40/55) compared to sPCL patients (33.3%, 9/27) (P=0.003), whereas VTD-PACE-like (bortezomib-thalidomide-dexamethasone-cisplatin-doxorubicin-cyclophosphide-etoposide) regimens were more frequently utilized in sPCL patients (40.7%, 11/27) compared to pPCL patients (12.7%, 7/55) (P=0.003). Daratumumab-containing quadruplets were infrequently used in both groups (3.6% in pPCL vs. 3.7% in sPCL; P=0.003). Clinical outcomes showed that more pPCL patients achieved a complete response (CR) or better (62.5%, 25/40) compared to sPCL patients (41.7%, 10/24) (P=0.126). Minimal residual disease (MRD) negativity was observed in 13.3% (10/75) of pPCL patients, compared to 2.6% (1/24) of sPCL patients (P=0.109).

Survival

The median OS for the entire cohort was 20.4 months; 68.0% of patients died during the study period. Notably, patients with pPCL demonstrated a significantly longer median OS of 36.6 months compared to sPCL, whose median OS was only 3.2 months (P<0.001). An exploratory analysis was conducted to evaluate high-risk genetic features as defined by the revised 2016 IMWG consensus guidelines, which include the presence of t(4;14), t(14;16), Del17p13, a non-hyper diploid karyotype, and a gain (1q) (13). However, the small sample size limits definite conclusions. In this exploratory analysis, high-risk genetic features were not significantly associated with worse survival across the entire cohort (HR 1.10, P=0.749). Furthermore, the presence of the t(11;14) translocation did not demonstrate a significant survival benefit (HR 0.65, P=0.203) (Online Supplementary Tables S1, S2). EM disease, as defined by the treating institution (it is uncertain whether true EM disease or para-medullary disease), was non-significantly associated with worse survival as well (HR 1.57, P=0.919) (Table 2, Figures 1, 2).

Receipt of any transplant, irrespective of type (auto, allo or tandems) significantly improved survival across the entire cohort and both PCL subtypes (P<0.001). The median OS for the entire cohort was 48.7 months for those who underwent transplantation, compared to 5.9 months for those who did not (P<0.001). For pPCL patients, the median OS with transplantation was 49.5 months, compared to 12.5 months without (P<0.001). In sPCL patients, transplantation (N=10/42) resulted in a median OS of 16 months compared to 3.4 months without any transplant (P=0.005). Allogeneic transplantation did not show a superior survival over autologous HSCT (49.5 vs. 37.9 months, P=0.348). Patients who underwent transplantation were generally younger, with a median age of 56.4±9.0 years compared to 64.8±10.0 years for non-transplanted patients (P<0.001). pPCL patients were more likely to receive transplants than sPCL patients (P<0.001). MRD negativity was more frequent in the transplant group (16.1%, 9/38) compared to the non-transplant group (2.5%, 1/76) (P<0.001). However, there were no significant differences between the transplant and non-transplant cohorts in baseline levels of LDH, creatinine, beta-2 microglobulin, albumin, or the prevalence of EM disease. Similarly, no significant differences were observed in the frequency of high-risk genetic features, t(11;14), type of chemotherapy or novel regimens, and CR rates between the two groups (Online Supplementary Table S3).

Among treatment modalities, no significant differences were observed between PI-based triplet therapies and VTD-PACE-like regimens across the entire cohort or by PCL type. The median OS for the overall cohort with PI-based triplet therapy was 28.2 months compared to 12.6 months with VTD-PACE like regimen (P=0.212). In pPCL patients, the median OS with PI-based triplet therapy was 36.6 months and that with VTD-PACE like regimen was 31.8 months (P=0.734). While for sPCL patients, median OS with PI-based triplet therapy was 7.4 months compared to 6.6 months with VTD-PACE like regimen (P=0.654) (Figure 3).

Table 2.Associations with survival.

Discussion

Plasma cell leukemia remains a highly aggressive malignancy with poor survival, characterized by early relapse and high mortality rates, even in the current era of novel therapeutic agents and HSCT.^10,11 Survival outcomes in PCL have been reported to be influenced by various factors, including PCL subtype, high-risk genetic features, clinical parameters, and response to treatment.^3,14-18

The median OS for pPCL reported in prior studies varies widely, from 18 months to over 30 months.^19,20 This variability is likely attributable to the lack of standardized diagnostic criteria and treatment protocols for this rare malignancy and differences in median follow-up durations, which complicate direct comparisons of outcomes among novel agents and transplantation. Given the high tumor burden in PCL, there is an urgent need for rapid disease control, and both PI-based therapies as well as HSCT have demonstrated deeper response rates compared to conventional therapies such as alkylating agents (melphalan, cyclophosphamide, cisplatin) and anthracyclines (mitoxantrone, cytarabine, adriamycin) in combination with vincristine and etoposide.^18,21 In our study, the median OS of pPCL patients was 36.6 months. Among this cohort, 47.0% received PI-based treatment, and 69.9% underwent any transplantation, with 50.0% receiving single autologous transplantation. These findings are comparable to the 36.3-month median OS observed in a small prospective study by Royer et al., where 64.1% (25/39) underwent transplantation and 89.0% (35/39) of patients received PI-based induction.²⁰ Pagano et al.¹⁸ noted that receiving any transplant improves survival in pPCL,¹⁵ consistent with our findings (HR 0.16, P<0.001). Pagano et al. also further added that transplant showed a significant decrease in risk of death by 69.0% and risk of relapse by 88.0%.¹⁸

Figure 1.Overall survival for patients with plasma cell leukemia. (A) Death occurred in 104 of 153 (68.0%) plasma cell leukemia (PCL) patients. Median (95% Confidence Interval [CI]) survival was 20.4 (13.1-26.8) months. (B) Stratifying by PCL type, median (95% CI) survival was 36.6 (24.7-61.6) months for primary PCL (solid line) and 3.2 (1.3-5.2) months for secondary PCL (dashed line) (P<0.001).

Our pPCL cohort demonstrated a considerably longer median OS of 49.5 months with any transplant. Previous literature suggests that tandem auto-allo transplant may contribute to prolonged survival in pPCL. For instance, Lebovic et al. reported that pPCL patients undergoing tandem auto-allo achieved OS beyond 40 months.²² Lawless et al.²³ also identified a significant 100-day progression-free survival (PFS) benefit with tandem auto-allo, with a median OS of 34.6% for allogeneic transplant recipients and 31.3% for autologous transplant recipients at 60 months. In their cohort, 16.2% of patients received tandem auto-allo, closely aligning with our cohort’s rate of 14.6% for tandem auto-allo.²³ Mahindra et al. further reported a 3-year median OS of 84.0% in pPCL patients treated with tandem auto-allo.²⁴ However, a larger study by Dhakal et al. involving only 0.08% of patients undergoing tandem auto-allo transplantation demonstrated a lower median OS of 28.0% with autologous and 31.0% with allogeneic transplantation at four years.²⁵ Therefore, additional prospective studies are warranted to elucidate the potential benefits of upfront tandem autologous-reduced intensity conditioning allogeneic transplantation in pPCL. Consistent with previous literature, our study did not suggest a superior survival advantage of initial allogeneic transplantation over autologous transplantation, likely due to high rate of non-relapse mortality (NRM) ranging from 17% to 41.0% with allogeneic transplantation in various studies,^23-25 perhaps due to lack of power to find such a difference. Furthermore, the numbers were too small to identify a potential benefit among patients undergoing allogeneic transplant from a match-related donor, although this population is known to experience lower rates of NRM.

While PI-based therapies have shown superior responses in pPCL compared to conventional chemotherapy combinations,^21,26 our study did not observe a significant survival difference between PI triplet regimens over conventional chemotherapy combinations such as VTD-PACE, potentially due to missing data in 43.0% of pPCL cohort. However, Katodritou et al. reported improved survival with the bortezomib-lenalidomide-dexamethasone (VRd) regimen compared to bortezomib-dexamethasone-cyclophosphamide (VCd), bortezomib-anthracycline-dexamethasone (PAD), vincristine-doxorubicin-dexamethasone (VAD), and melphalan prednisone (MP).²⁷ Similarly, a single center study noted survival advantage of bortezomib or carfilzomib over the VTD-PACE regimen.¹⁰ The influence of treatment response may be modulated by immune type or cytogenetics, as bortezomib has been shown to be effective in patients with CD27-positive plasma cells, a subset that comprised 37.0% of the Katodritou study cohort.^27,28 While immune modulatory drugs like lenalidomide are involved in cyclin D1 dysregulation,²⁹ which is coded by CCND1 and was noted to be a commonly associated 14q32 translocation frequently present among pPCL.³⁰ Therefore, missing data regarding immune-phenotype and immunoglobulin high chain (IGH) translocation data in our study may have limited our ability to draw conclusions in this regard. Survival improvements for sPCL with novel agents and transplantation have been modest. In our sPCL cohort, the median OS of 16 months with transplantation observed here mirrors the 17.5 months reported by Wang et al.³¹ Despite these efforts, sPCL survival remains significantly lower than that of pPCL, and the underlying mechanisms remain poorly understood.

Figure 2.Overall survival for plasma cell leukemia patients with or without a transplant after diagnosis. Plasma cell leukemia (PCL) patients with (solid line) or without (dashed line) a transplant after their PCL diagnosis. (A) Median (95% Confidence Interval [CI]) survival for entire cohort of PCL patients was 48.7 (34.3-67.7) months with transplant and 5.9 (3.5-8.8) months without transplant (P<0.001). (B) Median (95% CI) survival for primary PCL was 49.5 (34.3-72.7) months with transplant and 12.5 (5.7-24.7) months without transplant (P<0.001). (C) Median (95% CI) survival for secondary PCL was 16.0 (0.8- not reached) months with transplant and 3.4 (1.3-6.3) months without transplant (P=0.005).

This disparity is likely driven by differences in the treatment responses, molecular pathways, heavy disease pre-treatment during MM therapy, and drug resistance.¹² Skerget et al. identified high-risk proliferative gene expression in 83.3% of their sPCL cohort compared to 26.1% of pPCL cases, potentially explaining the more aggressive clinical behavior of sPCL at the molecular level.³² Moreover, response rates to PI-based therapy are lower in sPCL (25-60%) as compared to >80.0% in pPCL.^16,31 In addition, quality of response to PI-based regimens varies between pPCL and sPCL.¹⁹ The poor performance status of sPCL patients, often a consequence of extensive pre-treatment during the MM phase, further limits their ability to undergo salvage therapies.¹⁶

Despite advancement in treatment regimens, survival in PCL patients remains poor, with a median OS of the entire cohort in the current study of 20.4 months and with over 60% of patients succumbing within a few years of diagnosis. While deep initial responses are important, sustained remission remains the key goal. In the early 1990s, Sica et al. observed that, despite achieving clinical and laboratory remission in a PCL case, MRD often persists, which can hinder longer-term remission.³³ Kun-Huei Yeh et al. reported a case of PCL that remained in complete remission for up to 59 months with negative MRD.³⁴ Moreover, Kaiser et al. are currently exploring the impact of MRD on survival outcome in a prospective study. This study involves high-risk MM and PCL patients undergoing intensive induction therapy with daratumumab, cyclophosphamide, bortezomib, lenalidomide, and dexamethasone (Dara-CVRd), followed by high-dose melphalan/autoSCT, bortezomib-augmented therapy, and Dara-VRD consolidation for 18 cycles, followed by Dara-R maintenance. The study has shown a decrease in MRD positivity from 41.0% post-induction to 14.0% 100 days post-ASCT, although survival outcomes have yet to be reported.³⁵ In our study, MRD status was unavailable for the majority of patients. However, treatments guided by MRD status, which can be reliably assessed via multi-parameter flow cytometry and next generation sequencing, could significantly influence survival outcomes.³⁶

Cytogenetic aberrations such as del1p21, t(4;14), and MYC translocations, are all associated with inferior survival outcomes.^3,37,38 While the prognostic significance of t(14;16), amp1q21 remains unclear, t(11;14) has been associated with longer survival in some studies.³⁹ In the current pPCL cohort, we observed a non-significant adverse impact of del17p. However, this is to be interpreted with caution as there were few del17p events noted and 36.4% of the cohort had data missing, limiting the study’s statistical power. Similarly, in independent studies, Tiedemann et al. reported no significant impact of del17p on pPCL survival.³ In contrast, Deng et al. and Li et al. observed significant survival associations with del17p.^10,40 The negative prognostic implications of del17p are possibly attributed to TP53 inactivation.^3,39 The discrepancies could stem from variations in cytogenetic assessment methods in retrospective studies, as conventional techniques may fail to detect sub-chromosomal mutations or RNA splicing variations. Further use of gene expression profiling and whole-genome sequencing is highlighting the presence of enriched MAF translocations, frequent copy number alterations, biallelic inactivation of TP53, driver gene mutations in KRAS, ERAS, and NRAS, high-risk gene signatures, such as the GENE70 and GENE80 models, and an altered transcriptome involving lipid metabolism and adhesion molecules in pPCL compared to MM, conferring its aggressiveness.^41,42 As high-throughput genomic technologies advance, we anticipate more precise molecular stratification of PCL, which may contribute to personalized treatment approaches. Similar to the current pPCL cohort, the substantial missing data on high-risk genetic features in the sPCL cohort (49.0%) limited our ability to draw definitive conclusion about its role on this sub-type. The t(11;14) translocation has been associated with improved survival outcomes in PCL patients, with a reported median OS of 39.2 months compared to 17.9 months in patients without this translocation.^37,39 This favorable prognosis may stem from a more advantageous transcriptomic profile, a lower frequency of double-hit events, and the differential expression of BCL2 family members linked to this translocation.³⁹ However, in the current analysis, no significant survival benefit was observed for patients with the t(11;14) translocation, likely due to missing data in over 40.0% of the cohort.

Figure 3.Associations with survival. Forest plot of plasma cell leukemia (PCL) patients with (solid line) or without (dashed line) a transplant after their PCL diagnosis. (See Table 2).

Clinical and biological parameters, such as elevated LDH, hypoalbuminemia, elevated beta-2 microglobulin, ISS stage III, and thrombocytopenia, have also been shown to negatively impact survival in PCL patients.^14-18,26 In our study, we investigated the role of EMD on survival outcomes in PCL, recognizing that the evaluation of EMD at diagnosis and its response to treatment are critical components of PCL management, as per IMWG consensus guidelines.⁶ Extramedullary disease is observed with greater frequency in PCL compared to MM, with reported incidences in one study of 20.0% in pPCL, 28.0% in sPCL, and 8.0% in MM.⁴ The prevalence of EMD also varies between PCL subtypes. Tiedemann et al. documented high incidence of EMD in pPCL, while Pagano et al. observed lesser EMD involvement in pPCL, a finding corroborated by Papadhimitriou et al. and Li et al.^3,4,10,18 The immature immune-phenotype of PCL plasma cells, often characterized by absence of CD56 and CD20 overexpression, likely contributes to the increased propensity for spread beyond the bone marrow microenvironment.^14,43 In our cohort, EMD did not demonstrate an independent correlation with worse survival outcomes in either type. However, this may be attributed to incomplete data for 15.0% of patients, potentially limiting the significance of our findings. Nevertheless, Li et al. identified worse survival in both PCL types with EMD, with 47.0% of their cohort presenting EMD.¹⁰ Additionally, Jiminez-Zepada et al. suggested TP53 mutations and chromosome 1 alteration may drive EMD and its associated poor prognosis.⁴⁴ A dreadful complication of PCL is CNS involvement which is associated with significantly shorter survival.⁴⁵ Although the current cohort lacks sufficient data on CNS involvement, small scale case studies indicate a higher prevalence in sPCL, particularly among Caucasians and those treated with cyclophosphamide-dexamethasone.^10,45 In our study, 3 patients had documented CNS disease: 2 with primary PCL and one with secondary PCL. The patients had missing data on most disease characteristics, but all patients had elevated LDH, and one had EMD. The 2 affected individuals were aged 70 and 88 years of age, while the younger patient, aged 45, had undergone a single autologous transplant at induction. Neuro-meningeal relapse has been noted in some studies involving transplantation of PCL patients, highlighting consideration of incorporating prophylactic intrathecal treatment.²⁰ Our study has several limitations. First, its retrospective design may introduce inherent biases. Additionally, there was a significant amount of missing data across key variables, including high-risk genetic features (missing in 39.0% of cases), first-line chemotherapy (46.0% missing), and treatment response (58.0% missing), which limits our ability to draw definitive conclusions. Moreover, the study population was predominantly white (90.0%), restricting our ability to explore racial and ethnic differences. Selection bias is a key limitation, as patients who were too ill or had uncontrolled disease were less likely to undergo transplantation or may have preferentially received PACE-based or other intensive regimens, confounding treatment outcome comparisons. Also, prior studies have demonstrated improved response rates, prolonged PFS, and enhanced OS with daratumumab-based quadruplets, but the low utilization in our cohort (approx. 4%) limited our ability to assess its impact on survival.^10,30 Additionally, EM disease classification varied by institution, making it unclear whether cases reflected true EM or para-medullary involvement, limiting analytical precision. Furthermore, it is possible that t(11;14) did not appear to confer a survival benefit because the BCL-2 inhibitor venetoclax was not widely used during the study period, limiting its potential therapeutic impact.

Despite these limitations, this is, to the best of our knowledge, the largest multicenter study conducted in the US of this rare disease. The multicenter design also enhances the generalizability of our findings. Importantly, we incorporated the updated IMWG criteria, which define PCL by the presence of at least 5% circulating plasma cells, broadening the scope of our analysis compared to the more restrictive 20% threshold used previously.⁶

In conclusion, although significant progress has been made in the understanding and treatment of PCL, continued research is essential to further refine therapeutic strategies and address ongoing challenges. Transplantation remains a key treatment option, offering survival advantage across PCL subtypes. Future investigations should focus on novel agents to improve response rates, particularly in sPCL. Promising new therapies such as BCL-2 inhibitor venetoclax, have been shown to have potential in inducing responses in pPCL cases with t(11;14), a translocation more commonly observed in this subtype.^3,4,46-48 Improved responses with venetoclax have also been noted in relapsed sPCL cases harboring t(11;14) FISH abnormalities.⁴⁸ Additionally, bispecific T-cell engagers targeting B-cell maturation antigen (BCMA), CD 19, and other targets represent promising avenues for future investigation.^8,49 While the results of CAR-T therapy in PCL have been modest, continued research in this area is warranted. Integration of MRD-directed maintenance approaches may help to ascertain the minimum treatment durations necessary to achieve optimal survival. Moreover, our study lacks data on risk factors for secondary PCL, prior transplant history, and clonal evolution, as the dataset includes only patients already diagnosed with PCL or those who have progressed to sPCL. Future studies should investigate these risk factors to better understand disease progression and optimize treatment strategies. Finally, a concerted international research effort is essential for making meaningful progress in the treatment of rare diseases such as PCL.

Footnotes

• Received December 12, 2024
• Accepted April 9, 2025

Correspondence

*SS and SP contributed equally as first authors

K. Gowin kgowin@coh.org

Disclosures

No conflicts of interest to disclose.

Contributions

SS and SP wrote the manuscript. BW carried out the statistical analysis, and reviewed and edited the manuscript. AL, IC, DS, DG, ML, MO, MW, ON, LS, DD, VV, DC and KG supervised the study, and reviewed and edited the manuscript.