Effect of autologous hematopoietic stem cell transplantation for patients with peripheral T-cell lymphoma in China: A propensity score-matched analysis

Background The role of consolidation therapy with autologous stem cell transplantation (ASCT) in patients with peripheral T-cell lymphoma (PTCL) in first complete remission (CR1) or partial remission (PR1) remains controversial. The existing data from China are limited. Therefore, we aimed to investigate the effect of ASCT on the survival of Chinese patients with PTCL showing response to induction chemotherapy at our hospital. Methods We retrospectively reviewed the data of patients with PTCL (excluding Natural killer/T cell lymphoma) in CR1 or PR1 treated at Peking University Hospital &Institute from 1996 to 2020. Propensity score matching (PSM) was used to balance clinical characteristics between the ASCT and non-ASCT groups. The primary endpoints were event-free survival (EFS) and overall survival (OS). Results Of the 414 selected patients, 73 received ASCT consolidation and 341 did not. Over a median follow-up of 5.7 years, survival was significantly better in the ASCT group than in the non-ASCT group (median EFS, 8.1 years vs. 2.8 years, P = 0.002; median OS, 14.9 years vs. 10.2 years, P = 0.007). The 5-year EFS and OS rates were 68.4% and 77.0% in ASCT group, and 43.2% and 57.6% in non-ASCT group, respectively. The survival benefit was confirmed in the propensity score matched cohort (46 patients who received ASCT and 84 patients who did not receive ASCT): P = 0.007 for median EFS and P = 0.022 for the median OS. Cox regression analysis showed that ASCT was independently associated with better survival: hazard ratio (HR) for EFS, 0.46 (95% CI: 0.28-0.76); HR for OS, 0.50 (95% CI: 0.31-0.84). Subgroup analysis showed that ASCT was more likely to benefit higher-risk patients and those with advanced disease. Among the subtypes of PTCL, the benefit was significant in angioimmunoblastic T-cell lymphoma (HR = 0.26 [95% CI: 0.10-0.66] for EFS and 0.29 [95% CI: 0.12-0.74] for OS), but not in the other subtypes. Conclusion ASCT may improve the long-term survival of patients with PTCL in first CR or PR, especially for patients with angioimmunoblastic T-cell lymphoma. The specific groups most likely to benefit from upfront ASCT need to be clearly identified.


Introduction
Peripheral T/NK-cell lymphoma (PTCL) is a group of rare and highly heterogeneous malignant lymphoproliferative diseases originating from mature T-and NK-cells. PTCL is classified into subtypes such as angioimmunoblastic T-cell lymphoma (AITL), ALK-positive anaplastic large-cell lymphoma (ALK+ALCL), ALKnegative anaplastic large-cell lymphoma (ALK-ALCL), peripheral T-cell lymphoma not otherwise specified (PTCL-NOS), and so on (1,2). Most subtypes have a dismal prognosis (3,4). Due to the rarity and heterogeneity of PTCL, there is still no consensus on the best treatment approach. The role of autologous stem cell transplant (ASCT) consolidation continues to be debated (5-10).
Current guidelines recommended ASCT for most PTCL patients achieving remission after induction chemotherapy (11-13). The National Comprehensive Cancer Network and the American Society for Transplantation and Cellular Therapy recommend ASCT for patients with first-time complete remission (CR1) (11,12). Meanwhile, the European Society of Medical Oncology guidelines suggests that patients with first-time partial remission (PR1) could also be considered for ASCT (13). However, these recommendations were based on low-quality evidence from retrospective studies (14)(15)(16)(17)) and a few single-arm prospective studies (5, [18][19][20]. There remains no consensus on the utility of consolidative ASCT. One real-world study of lymphoma in a Swedish population showed that PTCL patients (excluding those with ALK+ALCL) receiving ASCT had better progression-free survival (HR: 0.56 [0. 39- (22).
To the best of our knowledge, no well-designed randomized controlled trial has been conducted so far to evaluate the impact of ASCT in the clinical setting. The data based on Chinese patient population is limited. Therefore, in the present study, we aimed to investigate the effect of ASCT on the survival of patients with PTCL showing response to induction therapy.

Patients
The study was granted by the ethics committee of Peking University Cancer Hospital & Institute. Informed consent was exempted because the data were analyzed anonymously. A total of 8,826 patients with lymphoma were admitted to Peking University Cancer Hospital & Institute from January 1996 to November 2020. Pathological diagnosis was based on the corresponding criteria (1,(23)(24)(25). Patients with mature T-cell lymphoma who achieved CR1/PR1 were included in this study. The main exclusion criteria included patients with natural killer/ T cell lymphoma (NK/TCL), receiving allogeneic transplantation, and missing clinical data. Assessment of response to initial treatment was evaluated using the Lugano criteria (26), and 18

Definitions
Event-free survival (EFS) was defined as the time from the initial diagnosis to the date of occurrence of disease recurrence/ progression confirmed by clinical manifestations, imaging or pathological reports, initiation of second-line salvage therapy, or death (whichever occurred first). OS was defined as the time from the initial diagnosis to death due to any cause. None of the above events that occurred at the last follow-up of the patient were censored. Follow-up data were collected via telephone interviews and a review of inpatient and outpatient medical records.

Statistical analysis
Statistical analysis was done using R 4.2.0 (https://www.rproject.org). Categorical variables were summarized as percentages and tested by the chi-square test or Fisher exact test. Propensity score matching (PSM) was performed to compare EFS and OS between ASCT and non-ASCT groups after balancing clinical parameters, which were selected according to previous literature and clinical experience; the parameters were histological subtype, age, sex, lactate dehydrogenase level (LDH), Ann Abor stage, Eastern Cooperative Oncology Group (ECOG) performance status, B symptoms, and response to the initial treatment. The caliper value was set to 0.25, and the matching ratio was 1: 2 (ASCT: non-ASCT). PSM was conducted using the "MatchIt" package in R. Survival analysis was performed before and after matching, using Kaplan-Meier curves and tested by the log-rank test. Univariate and multivariate Cox regression analyses were conducted to determine the independent effect of ASCT on survival after adjusting for other variables (sex, age, Ann Abor stage, IPI risk group, and histological subtype [ALK+ALCL vs. non-ALK+ALCL]). All significant factors from the univariate analysis were included in the multivariate analysis. Survival rates were estimated by the Cox proportional hazards model, Hazard ratios, with 95% confidence intervals, were calculated to quantify risk. Two-sided P < 0.05 was considered statistically significant.
As expected, patients with ALK+ALCL had better survival outcomes ( Supplementary Figures 2A, 2B

Survival analysis in matched data
In the propensity score matched (PSM) cohort, there was no statistically significant differences in clinical characteristics between patients in the ASCT group (n = 46) and the non-ASCT group (n = 84; Table 1). In this cohort, median EFS and OS were also better in the ASCT group than in the non-ASCT group (EFS: 8.1 years vs. 2.7 years, P = 0.007; OS: "not reached" vs. 7.7 years, P = 0.022). In addition, the survival difference remained statistically significant when excluding patients with ALK+ALCL after performing PSM method ( Supplementary  Figures 2C, 2D; Supplementary Table 4).  (Figures 2, 3). In multivariate Cox regression analysis, ASCT was confirmed as the independent risk factors for OS in PTCL patients with CR1 or PR1 (Table 2).

Discussion
PTCL has a poorer prognosis than aggressive B-cell lymphomas, with less durable remissions after anthracyclinecontaining multiagent chemotherapy (4,28,29). Treatment is particular challenging due to the relative rarity and biological heterogeneity of PTCL. The value of ASCT in PTCL remains controversial (9,10,22,30), lacking a large sample retrospective study from China. The present study provides evidence in support of the use of ASCT in Chinese chemotherapy responders.
Previous studies suggest that consolidation with ASCT may not offer survival advantages for patients with PTCL except in specific clinical scenarios (10,21,22,31). The COMPLETE study was the first large prospective cohort study to directly compare survival outcomes of patients with nodal PTCL in CR1 treated with or without consolidative ASCT. Eighty-three patients did not undergo ASCT, while 36 underwent consolidative ASCT. After a median follow-up of 2.8 years, no statistically significant difference was found in median OS between the ASCT group and the non-ASCT group (median OS: "not reached" vs. 57.6 months, respectively, P = 0.06). ASCT had no impact on the prognosis of PTCL-NOS but significantly improved OS and progression-free survival for patients with AITL; this finding is consistent with our results. Thus, it appears that patients with AITL are likely to benefit most from ASCT consolidation (16).
Our results were contrary to the findings of the large retrospective multicenter study based on the LYSA registry although both studies had similar designs. The LYSA study evaluated 269 patients with PTCL achieving CR1/PR1. The PTCL subtypes in the LYSA study included PTCL-NOS (29%), AITL (46%), and ALK-ALCL (25%). Among them, 134 patients were allocated to ASCT and 135 did not. Both Cox regression analysis and PSM analysis showed no survival advantage with ASCT consolidation (22). Our study enrolled more histologic subtypes (including ALK+ALCL, Enteropathy-associated T-cell lymphoma (EATL), and others); the differences between the study populations in the two studies may be one of the reasons for the contrary results. In addition, the performance of the intention-to-treat approach in this study may have led to more conservative estimates (32).
Although several earlier studies have reported that PTCL patients who receive upfront ASCT consolidation have a relatively good prognosis (5, 14,15,17,[33][34][35], none of the studies attempted to determine the sub-cohort of patients most likely to benefit. One pooled analysis showed that PTCL-NOS and ALK-ALCL patients without DUSP22 rearrangement could benefit from ASCT (36), indicating ALK-ALCL patients with DUSP22 rearrangement may be a potential transplant-exempt population. Thus, current evidence suggests clear benefits with ASCT consolidation in patients with AITL and uncertain benefits in patients with PTCL-NOS (probably due to the high heterogeneity in this subtype). The benefit of ASCT in different pathological subtypes needs further study.
Our results also demonstrated that upfront consolidative ASCT can improve prognosis in low-intermediate and intermediate-high risk subgroups of PTCL. Because only one patient in the high-risk subgroup underwent ASCT, we did not analyze the high-risk group. The evidence for the long-term efficacy of upfront ASCT in patients with early-stage and lowrisk IPI is insufficient (21).
Because some previous studies have suggested that the frontline chemotherapy regimen may be associated with survival outcomes (37), we compared the survival difference between patients receiving CHOP vs. CHOPE as an initial regimen after matching baseline characteristics by PSM (data not shown). No survival difference was observed in our cohort. In recent years, the discovery and clinical application of molecular biomarkers and signatures have provided unprecedented insights into PTCL (38)(39)(40)(41). Novel antibodydrug conjugates and epigenetic modifying agents have shown promising clinical efficacy (42,43). Of note, The ECHELON-2 study demonstrated a significant improvement in prognosis with frontline brentuximab vedotin (BV) plus cyclophosphamide, doxorubicin, and prednisone (A+CHP) compared with CHOP in CD30-positive PTCLs (43). Furthermore, Savage et al. recently conducted an exploratory subgroups analysis of ECHELON-2 trial supporting the role of consolidation consolidative stem cell transplant (autologous or allogeneic) in CD30-positive PTCL who achieve a CR following treatment with A+CHP. Patients who underwent stem cell transplants achieved a 64% reduction in the risk of a PFS event (HR: 0.36 [95% CI, 0.17-0.77]). The results indicate that even in the era of BV, consolidative stem cell transplant is still required (44).
Our results should be carefully interpreted because of some limitations in this study. First, this was a retrospective singlecenter study. The findings from the subgroup analysis may not be reliable because of the limited number of patients. Second, there is bound to be a discrepancy in response evaluation by PET/CT vs. CT (26). The CR rate may be underestimated for those patients who received response evaluation only by CT in the cohort. Finally, the systems of criteria for diagnosis have undergone changes in recent 20 years (1,(23)(24)(25). We enrolled patients according to the diagnostic criteria of that time, but there must be variability around the exact criteria used for diagnosis over long time span. Therefore, pathological review and molecular typing to define the subtypes of PTCL are required in future studies.

Conclusion
ASCT consolidation may prolong survival in Chinese patients with PTCL showing complete or partial response to initial chemotherapy; benefit is most likely in patients with AITL. Patients with higher-risk IPI score and advanced-stage disease may also benefit significantly from ASCT. Our findings need to be confirmed in large controlled prospective studies.

Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Ethics statement
The studies involving human participants were reviewed and approved by the ethics committee of Peking University Cancer Hospital. Written informed consent from the participants' legal guardian/next of kin was not required to participate in this study in accordance with the national legislation and the institutional requirements.

Author contributions
HG and WL designed this research. HG collected the data, performed the analysis and prepared the original draft. XJ was involved in data cleaning and analysis. MW and ND provided helpful suggestions for this study. All authors contributed to the article and approved the submitted version.