Clinical and cytogenetic characteristics of primary and secondary plasma cell leukemia under the new IMWG definition criteria: a retrospective study

ABSTRACT Background Plasma cell leukemia (PCL) is a rare and aggressive plasma cell disorder, exhibiting a more unfavorable prognosis than multiple myeloma. PCL is classified into pPCL and sPCL. Recently, the IMWG has recommended new PCL definition criteria, which require the presence of ≥5% circulating plasma cells in peripheral blood smears. Due to its low incidence, research on pPCL and sPCL is limited. Methods We conducted a retrospective study and analyzed clinical and cytogenetic data of pPCL and sPCL patients. Overall survival (OS) and progression-free survival (PFS) were assessed by the Kaplan-Meier method, and survival distributions were compared using the log-rank test. Results This is a small cohort comprising 23 pPCL and 9 sPCL patients. Notably, sPCL patients showed a higher incidence of extramedullary infiltration and a higher percentage of bone marrow plasma cells (p = 0.015 and 0.025, respectively). Although no significant difference was found between the two groups in OS and PFS, a trend emerged suggesting a superior survival outcome for pPCL patients, with a higher cumulative 1-year PFS rate (38.3% vs. 13.3%) and a lower early mortality rate (mortality rate at 3 months: 15% vs. 33%). We also suggested that pPCL patients carrying t(11;14) may have a longer median survival time than individuals with other cytogenetic abnormalities, but this was not confirmed due to the small sample size. Conclusion Our study revealed clinical and cytogenetic features of pPCL and sPCL patients according to the new diagnostic criteria. The findings suggested a generally better prognosis for pPCL than sPCL and the likelihood of t(11;14) translocation acting as a favorable prognostic factor in pPCL. It is important to note that our study had a limited sample size, which may lead to bias. We hope well-designed studies can be conducted to provide more results.


Introduction
Plasma cell leukemia (PCL) is a rare and aggressive variant of plasma cell dyscrasia, characterized by plasmacytosis in peripheral blood (PB).PCL is classified as primary when it presents de novo in patients without previous evidence of multiple myeloma (MM) or secondary when it presents as a leukemic progression in patients with a history of MM.The crude incidence of primary plasma cell leukemia (pPCL) in Europe is 0.4 new cases per million [1].Among 75399 patients with a diagnosis of MM in the Surveillance, Epidemiology, and End Results (SEER) database from 1973 to 2009, PCL accounted for approximately 0.6% [2].From 1960 to 2008, of newly diagnosed MM patients seen at the Mayo Clinic, 1.3% were diagnosed with PCL [3].Generally, pPCL accounts for about 60-70% of all PCL cases, while secondary plasma cell leukemia (sPCL) constitutes the remaining 30-40% [4,5].
The original definition of PCL was based on the presence of an absolute count of plasma cells ≥2 × 10 9 /l and/or ≥20% circulating plasma cells (CPCs) in PB.However, the International Myeloma Working Group (IMWG) consensus noted that the criteria were overly restrictive and resulted in missed diagnoses of PCL [6].Recently, some retrospective studies have already revealed a similar adverse prognostic impact of the presence of ≥5% CPCs as previously defined PCL [7,8].Based on present evidence, the presence of 5% or greater CPCs on PB smears was considered a new criterion for pPCL diagnosis [4].However, this recommendation has not been widely used either in clinical or academic settings.Whether this revised definition applies to sPCL or not remains controversial.
Due to the relatively low incidence and prevalence of the disease, the genetic background and prognosis of the two types of PCL are not fully elucidated.Moreover, limited reported series have mainly focused on pPCL cases, and only a few sporadic sPCL patients have been reported.Herein, we compared biological characteristics, cytogenetic features, treatment response, and clinical outcomes between pPCL and sPCL under the new definition of ≥5% CPCs on PB smears.

Data source
This was a single-center retrospective study of patients diagnosed with PCL between January 2013 and December 2022 at Zhongda Hospital, China.PCL was defined by the presence of 5% or more CPCs on PB smears according to IMWG [4].Among PCL patients, 23 patients were diagnosed with pPCL and 9 patients developed sPCL.Laboratory and clinical data and treatment regimens were extracted from the electronic medical records.The following were collected: age at diagnosis of pPCL or sPCL, gender, first symptoms, paraprotein isotype, MM stage as defined by the Durie-Salmon staging system and international scoring system (ISS) [9], karyotype, cytogenetic aberrations detected by fluorescence in situ hybridization (FISH) at the time of pPCL diagnosis or sPCL transformation.Additionally, hemoglobin level, platelet count, white blood cell count, levels of calcium, serum albumin, creatinine, β2-microglobulin and serum lactate dehydrogenase (LDH), incidences of extramedullary infiltration (EMI) and lytic lesions at diagnosis of pPCL or sPCL, treatment regimens, and therapeutic responses were collected.The study was conducted in accordance with the Declaration of Helsinki and approved by the Ethics Committee of Zhongda Hospital, China (2023ZDSYLL043-P01).

Treatment and efficacy assessment
The first-line treatment regimens patients received after diagnosis of pPCL or sPCL were divided into 4 types: proteasome inhibitor (PI)-based treatment (bortezomib or ixazomib), immunomodulatory drugs (IMiDs)-based treatment (thalidomide, lenalidomide or pomalidomide), PI + IMiDs-based treatment and conventional chemotherapy.No one in this cohort received carfilzomib as the first-line treatment.Treatment outcomes classified as complete response (CR), very good partial response (VGPR), partial response (PR), and no response (NR, both stable and progressive disease) were evaluated following the IMWG criteria [13].The objective response rate (ORR) was defined as CR, VGPR, or PR.

Statistical analysis
Descriptive statistics (median and range) were calculated for all continuous variables.Fischer's exact test was performed to examine possible differences in categorical variables.Wilcoxon rank sum/Kruskal Wallis test or student's t-test were used for continuous variables.For survival analyses, overall survival (OS) was calculated from diagnosis to death resulting from any cause or last follow-up.Progression-free survival (PFS) was measured from diagnosis until disease progression, relapse, or death.Survival curves were plotted using the Kaplan-Meier method, with comparison using the log-rank test.All statistical analyses were performed with SPSS 27 software (IBM, Armonk, NY) or by using the 'survival' and 'survminer' packages in R software 4.2.2 (R Foundation for Statistical Computing, Vienna, Austria).All analyses were two-sided and statistical significance was assumed at p < 0.05.

Clinical characteristics of patients
The clinical data of 23 pPCL and 9 sPCL patients are summarized in Table 1.The median time of leukemic transformation for sPCL patients was 28 months (range: 13-92 months).There was no significant difference in baseline characteristics between pPCL patients and sPCL patients other than the percentages of BMPCs and EMI.A higher proportion of BMPCs was observed in sPCL with a median of 66.8% while the median in pPCL was 46.8% (p = 0.025), which is consistent with disease progression from MM to sPCL.The presence of EMI was more common in sPCL than in pPCL (77.8% vs. 26.1%,p = 0.015).
The median age at diagnosis of both pPCL and sPCL patients was 63 years old.In our cohort, the proportion of females among pPCL patients was 52.2%, whereas males constituted 55.6% of the sPCL group.Immunoglobulin G (IgG) was the most detected M protein type in both groups, followed by the light chain subtype.ISS stage Ⅲ was mainly found in pPCL patients (63.6% in pPCL vs. 37.5% in sPCL) while the predominance of Durie-Salmon stage Ⅲ was found in both groups (95.7% in pPCL and 77.8% in sPCL).In terms of first symptoms, 43.5% of pPCL patients represented bone pain or fracture, which was also one of the common symptoms of MM.Extramedullary tumors were most prevalent in sPCL patients, with one having pulmonary involvement and one having testis involvement, while renal dysfunction or anemia was rare.Of note, pPCL patients tended to present with decreased platelet levels and elevated β2-microglobulin levels compared to sPCL patients, although no statistical difference was found.

Prevalence of cytogenetic abnormalities
Overall, a distinct cytogenetic profile was shown between pPCL and sPCL patients (Table 2).Del(13q) was the most frequently detected cytogenetic abnormality in pPCL (50%) while 1q21+ was the most common finding in sPCL (75%).The incidence of del(17p) was 10.5% in pPCL patients and 37.5% in sPCL patients, respectively.The prevalence of t(11;14) was mainly found in pPCL patients (46.7% in pPCL vs. 28.6% in sPCL, p = 0.648).There was a significant statistical difference between patients with different numbers of HRCA in the two groups (p = 0.042).Notably, over 60% of sPCL patients were diagnosed with double-hit myeloma while its counterpart had a diagnosis rate of only  16.7%.Contrarily, no obvious differences were observed between the two types of PCL in risk stratification and karyotype, which is probably due to the limited number of patients in both groups.

Treatment and efficacy
Treatment data were available for the entire cohort, except for 2 patients who refused medical treatment.Among 30 patients, response data of first-line therapies were available for 27 patients.The most used regimen in 21 pPCL patients who received treatment was VCd (bortezomib, cyclophosphamide, and dexamethasone), followed by PAd (bortezomib, doxorubicin, and dexamethasone) (Figure 2A).Among these patients, 12 (57%) patients were treated with PIbased treatment, 4 (19%) patients received IMiDsbased therapy, 4 (19%) patients underwent PI + IMiDs treatment, and 1(5%) patient was treated with conventional chemotherapy.There were 4 patients undergoing autologous stem cell transplantation (ASCT) and none of them experience recurrence.In terms of sPCL, most patients received a minimum number of 2 MM treatment lines before sPCL diagnosis.Regimens for sPCL patients were heterogeneous, including VCd, Vd-PACE (bortezomib, dexamethasone, cisplatin, doxorubicin, cyclophosphamide, and etoposide), RCd (lenalidomide, cyclophosphamide, and dexamethasone), Pd (pomalidomide and dexamethasone) and so on (Figure 2B).Only one person underwent ASCT after being diagnosed with sPCL, which could be related to the fact that the condition of sPCL was too poor to meet the requirements for ASCT.For first-line treatment of sPCL, there were 5 (56%) patients treated with PI-based regimens, 2 (22%) patients treated with IMiDs-based regimens, and 2 (22%) patients treated with conventional chemotherapy in sPCL.No sPCL patient underwent combination therapy of PI and IMiDs.
There was no significant difference in the overall ORR between pPCL and sPCL patients who received treatment (61% vs. 33%, p = 0.059).For patients undergoing PI-based treatment, ORR in pPCL and sPCL patients were 64% and 40%, respectively.In terms of IMiDs-based treatment, ORR in pPCL and sPCL patients were 67% and 50%, respectively (Figure 2C).As is shown, the figures of ORR in pPCL patients are higher than that in sPCL, regardless of treatment regimens.Although there was no significant statistical difference between the two PCL groups due to the limited sample numbers, we supposed pPCL patients tended to have a potentially better response to antitumor therapy.
For patients treated with PI-based chemotherapy, the two kinds of PCL shared a similar prognosis.There was no significant difference in OS and PFS between the two groups (median OS: 9 months in pPCL vs.10 months in sPCL, p = 0.52; median PFS: 9 months in pPCL vs.8 months in sPCL, p = 0.39).In the IMiDs-based regimens cohort, both median OS and PFS were 8 months in sPCL patients, compared with 16 and 14 months in pPCL patients (OS: p = 0.14; PFS: p = 0.63).

Prognostic value of cytogenetic abnormalities
Cytogenetic abnormalities revealed different prognostic profiles when comparing pPCL with sPCL.The results are summarized in Table 3.The median OS and PFS for pPCL were 12 months, and the median OS and PFS for sPCL were 15 months in del (13q) patients.In pPCL, both the median OS for patients with del(17p) and those with 1q21+ was 9 months.The median PFS for patients harboring del (17p) and those with 1q21+ were 7 and 8 months, respectively.When it comes to sPCL, patients carrying del(17p) had a median OS of 12 months, while those with 1q21+ had a median OS of 6.5 months.The median PFS for these patients were 9 and 5.5 months, respectively.In addition, both the median OS and PFS in t(11;14) pPCL was 12 months whereas its counterpart exhibited a median OS of 6.5 months and a median PFS of 5 months.However, no significant difference was found in each subgroup mentioned above, possibly due to the limited cases available for analysis.There was a trend that pPCL patients carrying t(11;14) might exhibit a longer median survival time compared to individuals with other cytogenetic abnormalities (median OS: 12 months in t(11;14), 9 months in del (17p), and 9 months in 1q21+), which was consistent with the cognition that t(11;14) was a standard-risk cytogenetic marker with a more favorable outcome than other aberrations.Unfortunately, no statistical difference was observed between t(11;14) pPCL and non-t (11;14) pPCL in this small cohort (p = 0.51).The distinct presence of HRCA at diagnosis between pPCL and sPCL did not affect their prognosis (median OS: 12 vs. 8 months, p = 0.085, median PFS: 12 vs.7 months, p = 0.13).Similarly, when we compared the lifespan of high-risk patients, there was no statistical difference between pPCL and sPCL (OS: p = 0.095, PFS: p = 0.18).

Discussion
It is very important to distinguish between different diseases that share a common origin.For example, myelodysplastic syndromes (MDS) are considered as 'preleukemia' condition, as approximately 20-30% of individuals with MDS ultimately progress to acute myeloid leukemia (AML) [14].Patients with AML have conventionally received more intense therapies than those with MDS.Similarly, PCL is an aggressive plasma cell dyscrasia that possesses distinct clinical and genetic characteristics, setting it apart from MM. Establishing a clear disease definition for PCL is essential, as it allows PCL patients to receive more specific therapies and improve their overall prognosis.The level of CPCs is generally considered to be associated with survival in PCL, but the appropriate level for PCL diagnosis is challenging and remains disputed.The traditional definition of PCL (≥20% CPCs and ≥2 × 10 9 /l absolute plasma cells) may be too strict that inadvertently misclassify patients as MM when their disease is more aggressive and behaves more similarly to PCL, therefore limiting accurate reporting of outcomes to the PCL entity.Hence, reassessment of the plasma cell count defining the boundary of PCL and MM has been advocated in the latest IMWG consensus in view of patients diagnosed with a lower cutoff of ≥5% CPCs and ≥0.5 × 10 9 /l absolute plasma cells in PB have the same adverse prognosis.
Two separate studies from Mayo Clinic compared the survival of patients with 5-19% CPCs to those with traditional definitions of pPCL characterized by 20% or greater CPCs.These studies demonstrated similar OS outcomes between the two groups, thus providing validation for the feasibility of the new diagnostic criteria [7,15].In one of the studies, researchers also noted that outcomes of those with ≥5% CPCs were much poorer when compared with MM patients who did not have CPCs at diagnosis (p < 0.001) [7].Another Chinese study reported 91 pPCL patients under the revised criteria and 67 previously defined pPCL patients at a diagnosis between 2000 and 2019.Clinical characteristics were similar except for younger median age in patients with ≥20% CPCs.No statistical difference was found in survival outcomes in individuals harboring 5-19% CPCs compared to those with ≥20% CPCs (median PFS: 16 vs.20 months, p = 0.26; median OS: 30 vs. 31 months, p = 0.89) [16].
The rarity of PCL, particularly sPCL which has been experiencing an increasing incidence, has resulted in limited studies about this specific condition [17].This is a small retrospective study, aiming to display and compare clinical and genetic features, as well as disease outcomes between primary and secondary PCL under the new definition.In terms of clinical characteristics, no significant difference was found between pPCL and sPCL except that sPCL patients had higher percentages of BMPCs and EMI.Most pPCL patients were diagnosed with ISS stage Ⅲ.Interestingly, pPCL patients tended to present with lower platelet levels and elevated β2-microglobulin levels.
The EMI may be related to impaired expression of adhesion molecules in PCL tumor cells, including CD56, leukocyte function-associated antigen (LFA) 1, and very late antigen (VLA) 5 [18,19].In contrast with our results, Bezdekova et al. found an increased level of platelets at the time of pPCL diagnosis when compared with sPCL (median: 129.5 × 10 9 /l vs. 49.1 × 10 9 /l, p = 0.004) [19].We supposed that the increased platelet level and decreased β2-microglobulin level of sPCL patients in our cohort were associated with the prior MM treatment.When it comes to cytogenetic features, most pPCL patients presented with del(13q) while 1q21+ was the most frequently finding in sPCL patients.The prevalence of t(11;14) was mostly detected in pPCL while double-hit myeloma mainly occurred in sPCL.
The prognosis of both pPCL and sPCL is disappointing, even in the era of novel agents.Real-world studies from Greek, America, and Latin America implicated a median OS ranging from 18 to 21.6 months in pPCL [20][21][22].A multicenter retrospective study containing 101 sPCL patients, reported a median OS of 4.2 months.Of them, three-fourths of patients received novel agents, including PI, IMiDs, and monoclonal antibodies [23].According to our findings, the median OS was 9 months for pPCL and 10 months for sPCL, respectively.The relatively inferior median OS observed in pPCL within our cohort might be attributed to the low ASCT utilization rate.It is important to note that the enrollment of only 9 sPCL patients in this study introduces a potential for selection bias, which may have influenced the observation differences in survival time.Although no significant difference was found between the two groups in OS, there was a noticeable trend that pPCL patients might have superior survival, with a higher cumulative 1-year PFS rate and a lower early mortality rate.The results of previous studies also corroborated our assumption.Tiedemannp et al [24].and Papadhimitriou et al [25].agreed that prognostic stratification was based on cytogenetic background.On one hand, the better survival of patients diagnosed with pPCL was strongly associated with the presence of t (11;14) translocations.On the other hand, sPCL had a higher number of FISH abnormalities.
The prognosis for PCL depends on the heterogeneity of cytogenetic background.As a consequence, cytogenetic abnormalities have been included in the prognostic evaluation systems [12,26].In agreement with previous reports [15,27], our study suggested the likelihood of t (11;14) acting as a favorable prognostic factor with improved survival in pPCL.Gang An et al. also noticed that all the t (11;14) sPCL cases in their study cohort had at least one other genetic aberration when initially diagnosed as MM, which could partly explain the shorter median OS and PFS in sPCL patients with t(11;14) compared to pPCL patients with the same translocation [28].It is important to note that the number of enrolled pPCL patients is quite small, which could be responsible for the insignificant statistical difference in our study.Nowadays, novel therapy is very promising in patients carrying t (11;14) translocation.Preclinical studies revealed that t (11;14) translocation was mainly associated with sensitivity to venetoclax due to the expression of B-cell leukemia/lymphoma-2 (BCL-2) [29,30].Venetoclax singleagent showed clinical activity in patients with t (11;14) MM, with an ORR rate of 40% and a ≥VGPR rate of 27% [31].Touzeau et al. also found the high sensitivity of samples from MM patients with t (11;14) to venetoclax, including one patient diagnosed with pPCL [32].In addition, t(11;14) relapsed or refractory MM patients receiving venetoclax in combination with other antimyeloma therapies (bortezomib or daratumumab) showed promising results in achieving deep remission.The proportion of patients achieving ≥CR in these studies ranged from 54% to 58% [33,34].Currently, there are very little data from case reports regarding the efficacy of venetoclax in pPCL harboring t (11;14).All the patients mentioned in the case reports were treated with venetoclax after relapse, leading to the achievement of minimal residual disease (MRD)-negative response [35][36][37][38].A recent study of a total of 91 pPCL patients demonstrate that TP53 mutation was associated with a significantly lower PFS and OS (p < 0.05) [39].The finding from different retrospective studies was that the presence of del(13q), 1p deletion, and t(4;14) were unfavorable prognostic biomarkers in PCL [40][41][42].In contrast, some studies did not consider del(13q) and 1q21 amplification as prognostic factors [24,40].In terms of karyotype, the hyperdiploid one was more frequently observed in sPCL, which was probably due to the evolution from hyperdiploid MM [24,42].Overall, heterogeneity was observed among the study results.The inconsistency of findings between studies may be explained by the relatively small sample size.Therefore, well-designed and long-term follow-up studies are expected to further investigate the prognostic role of cytogenetic abnormalities in PCL, especially in sPCL.
In recent decades, the emergence of novel agents, such as IMiDs and PI have been breakthroughs in the treatment of MM as well as PCL, improving long-term survival [2,43].A phase II trial conducted in Italy reported an ORR of 74% and a median OS of 28 months in a pPCL cohort receiving Rd (lenalidomide and dexamethasone) treatment [44].The Greek myeloma study group retrospectively analyzed 42 HEMATOLOGY PCL patients and 29 of them received a bortezomibbased regimen.A significantly higher ORR (69% vs. 31%, p = 0.04) and prolonged OS (median OS: 13 vs. 2 months, p < 0.007) were observed in patients treated with bortezomib compared with those who received conventional therapies.This study also revealed a deeper response to bortezomib in individuals with pPCL with a median OS of 18 months in comparison with 2 months in sPCL patients (p < 0.001) [45].A multicenter study of 101 sPCL patients reported that individuals who received a PI, either alone or in combination with conventional chemotherapy, IMiD or both, had higher ORR than those treated with other regimens (p = 0.01) [23].In addition, Nandakumar et al. showed a CR rate of 40% in a retrospective cohort of patients diagnosed with PCL under the new definition and treated with PI + IMiDs.The response rate was even better in individuals undergoing ASCT after the PI + IMiDs induction regimens, with 58% of the patients reaching CR [15].Inconsistent with prior reports [45,46], no significant difference in OS and PFS between pPCL and sPCL patients treated with the same regimen was observed in our study, probably due to the limited patients enrolled in this study and the heterogeneous therapeutic protocols.
There are many other promising therapies in PCL treatment.An early study proved that pPCL patients undergoing ASCT were associated with better survival data, where the survival was longer than 3 years in 64% of patients [47].Additionally, there was evidence from 2 prospective, randomized clinical trials supporting the use of PI + IMiDs induction regimens, followed by ASCT, for newly diagnosed PCL.Significant differences were also observed in median OS and PFS between patients undergoing ASCT and those considered not eligible for ASCT [44,48].Novel agents, such as CD38 antibody (daratumumab and isatuximab) and chimeric antigen receptor (CAR) T-cell therapy have been used in several cases of PCL, achieving promising effects [49,50].Other anti-MM agents including BCL-2 inhibitor venetoclax and exportin-1 inhibitor selinexor are undergoing assessment [51].
There were several potential limitations in our study.First, this retrospective, single-center study enrolled relatively limited PCL cases and had some missing data.As a result, the small sample size may introduce bias in our results.Second, due to the limited patient number, we were unable to conduct univariate and multivariate analyses to further explore predictive biomarkers of the clinical outcome in PCL.Third, the therapeutic regimens in our study were relatively conventional.Few patients in our cohort received immunotherapy agents or targeted agents in such an era of novel agents.
In conclusion, our study is a valuable addition to the literature, as we presented the clinical and cytogenetic characteristics and their impacts on prognosis of pPCL and sPCL under the new definition.Although both pPCL and sPCL were aggressive disorders, a suggestion of superior survival in pPCL than sPCL was proposed.In addition, our findings suggested a trend of better survival outcomes in pPCL patients with t(11;14) translocation.Currently, large cohort clinical studies aiming at PCL, especially sPCL, are exceedingly scarce.With novel agents as the future of the therapeutic landscape, prospective studies in larger patient cohorts are expected to reveal the exact molecular background and further explore novel targeted drugs to improve the clinical course of PCL.

Table 1 .
Clinical and biological characteristics of patients with pPCL or sPCL.

Table 2 .
Cytogenetic characteristics of patients with pPCL or sPCL.

Table 3 .
Cytogenetic abnormalities and prognosis of pPCL and sPCL patients.