Outcomes of Allogeneic Hematopoietic Cell Transplantation in T-cell Prolymphocytic Leukemia: A Contemporary Analysis from the Center for International Blood and Marrow Transplant Research

Abstract

Background:

T-cell prolymphocytic leukemia (T-PLL) is a rare, aggressive malignancy with limited treatment options and poor long-term survival. Previous studies of allogeneic hematopoietic cell transplantation (alloHCT) for T-PLL are limited by small numbers, and descriptions of patient and transplant characteristics and outcomes after alloHCT are sparse.

Objective:

To describe outcomes of alloHCT in T-PLL and identify predictors of post-transplant relapse and survival.

Study Design:

We conducted an analysis of data using the Center for International Blood and Marrow Transplant Research (CIBMTR) database on 266 patients with T-PLL who underwent alloHCT during 2008–2018.

Results:

The 4-year rates of overall survival (OS), disease-free survival (DFS), relapse, and treatment-related mortality (TRM) were 30.0% (95% CI, 23.8–36.5%), 25.7% (95% CI, 20–32%), 41.9% (95% CI, 35.5–48.4%), and 32.4% (95% CI, 26.4–38.6%), respectively. In multivariable analyses, three variables were associated with inferior OS: myeloablative conditioning (MAC) (hazard ratio [HR] 2.18, p<0.0001); age older than 60 years (HR 1.61, p=0.0053); and suboptimal performance status defined by Karnofsky Performance Status (KPS) <90 (HR 1.53, p=0.0073). MAC also was associated with increased TRM (HR 3.31, p<0.0001), increased cumulative incidence of grade 2–4 acute graft-versus-host disease (GVHD) (HR 2.94, p=0.0011) and an inferior disease-free survival (HR 1.86, p=0.0004). Conditioning intensity was not associated with relapse; however stable disease/progression correlated with increased risk of relapse (HR 2.13, p=0.0072). Both in vivo T cell depletion (TCD) as part of conditioning and KPS <90 were associated with worse TRM and inferior DFS. Total Body Irradiation was not found to have any significant effect on OS, DFS or TRM.

Conclusion:

Our data showed that reduced-intensity conditioning without in vivo T-cell depletion (that is, without ATG or alemtuzumab) prior to alloHCT was associated with long-term disease-free survival in patients with T-PLL who were 60 or younger or who had KPS >90 or had chemo-sensitive disease.

INTRODUCTION

T-cell prolymphocytic leukemia (T-PLL) is a rare aggressive malignancy, representing approximately 2% of mature lymphocytic leukemias in adults[1,2]. Patients tend to be older, with a median age of 65 years at diagnosis. Typically, T-PLL presents with signs such as marked leukocytosis, hepatosplenomegaly, lymphadenopathy, and cutaneous lesions. Treatment options are generally limited, and outcomes are poor, with a reported median survival of 19 months[3]. Alemtuzumab, an anti-CD52 humanized monoclonal antibody, is often used in the front line in T-PLL. While complete remission (CR) rates with alemtuzumab are high (60–80%), most responses are brief, and the relapse rate remains high[4,5]. Survival of patients with relapsed T-PLL is dismal, as responses to second-line therapies are limited and generally short-lived [2,6].

Allogeneic hematopoietic cell transplantation (alloHCT) is a potential curative therapy for T-PLL and has been reported to yield durable remissions, notably in patients who are in complete remission prior to transplantation[7–12]. AlloHCT aided small subsets of patients with T-PLL, according to studies by the Center for International Blood and Marrow Transplant Research (CIBMTR) [10], European Society for Blood and Marrow Transplantation (EBMT) [7,13], the Francophone Society of Bone Marrow Transplantation and Cellular Therapy (SFGM-TC) [9], and the Japanese Society for Transplantation and Cellular Therapy (JSTCT) [14]. The benefits of alloHCT are limited by high rates of non-relapse mortality (NRM), ranging from 28–40%. In addition, there exists high risk of post-transplant relapse, many occurring within 2 years of alloHCT [10,15]. Because these studies were relatively small, researchers were unable to identify factors associated with sustained remission and improved overall survival (OS). Hence, using CIBMTR Research Database, the aim of this study is to evaluate the effectiveness of alloHCT in T-PLL and to identify predictors of post-transplant relapse and survival.

METHODS

Data sources

The CIBMTR is a nonprofit research collaboration of the National Marrow Donor Program (NMDP)/Be The Match and the Medical College of Wisconsin (MCW). More than 300 medical centers worldwide submit clinical data to the CIBMTR about HCT and other cellular therapies. Participating centers are required to report all transplantations consecutively. The CIBMTR ensures data quality through computerized checks for discrepancies, physicians’ review of submitted data, and on-site audits of participating centers. The CIBMTR complies with federal regulations that protect human research participants. The Institutional Review Boards of MCW and NMDP approved this study.

Patient selection

Adults (aged 18 and older) who underwent first alloHCT for T-PLL during 2008–2018 were included in this analysis. Graft sources included peripheral blood stem cells (PBSC) and bone marrow. Eligible donors included human leukocyte antigen (HLA)-identical sibling donors or unrelated donors (URD) matched at the allele-level at HLA-A, -B, -C, and -DRB1, and alternative donor transplantation (haploidentical, mismatched unrelated donor). Cord blood and ex vivo T-cell depleted grafts were excluded, as were patients who received syngeneic transplants. AlloHCT recipients who received in vivo T-cell depletion (TCD) with anti-thymocyte globulin (ATG) or alemtuzumab were included.

Definitions and study endpoints

Disease response was defined based on National Cancer Institute-Sponsored Working Group guidelines for chronic lymphocytic leukemia. [16] The intensity of conditioning regimens was defined using published consensus criteria.[17] The primary endpoint was OS. Death from any cause was considered an event, and surviving patients were censored at the time of last follow-up. Secondary endpoints included cumulative incidence of acute graft-versus-host disease (aGVHD), chronic graft-versus-host disease (cGVHD), treatment related mortality (TRM), progression/relapse, and disease-free survival (DFS). TRM was defined as death without preceding disease relapse/progression; relapse and progression were considered competing events. Progressive disease or recurrences of T-PLL were defined as progression after alloHCT or recurrence following CR; TRM was considered competing event. DFS was defined as survival following alloHCT without relapse or progression. Patients who survived without evidence of disease relapse or progression were censored at last follow-up. The causes of death were reported in accordance to the methodology described previously. [18]

Statistical analysis

Cumulative incidence of GVHD, relapse/progression, and TRM were calculated using the cumulative incidence estimator to accommodate for competing risks. Probabilities of OS and DFS were calculated using the Kaplan-Meier method for a univariable analysis. Multivariable regression analysis was performed using logistic regression for aGVHD, the proportional cause-specific hazards model for chronic GVHD, relapse, and TRM, and the Cox proportional hazards model for DFS and OS. The assumption of proportional hazards for each factor was tested for the proportional hazards and cause-specific hazards models, and a forward stepwise selection was used to select significant risk factors. In the final model, we retained factors with statistical significance of < 5%. We examined the interaction between the main effect and the other significant variables and found no center effect based on the score test of homogeneity[19]. The variables that were considered in the multivariable models included: recipient age, Karnofsky Performance Status (KPS), Hematopoietic Cell Transplantation Comorbidity Index (HCT-CI), disease status at transplant, intensity of conditioning regimen, use of total body irradiation (TBI) in conditioning, time from diagnosis to transplant, recipients’ cytomegalovirus (CMV) serostatus, GVHD prophylaxis, donor type, graft source, use of ATG/alemtuzumab, and year of transplant. Adjusted probabilities [20,21] were calculated based on the final regression models for OS, DFS, relapse, and TRM.

RESULTS

Baseline characteristics

The study included 266 adults who received alloHCT for T-PLL. The median follow-up was 49 months (range 3.32–116.84). The baseline patient-, disease-, and transplantation-related characteristics are described in (Tables & Figures Table 1, Supplementary Table 1). Participants’ median age at the time of alloHCT was 59.1 years (range 25.0–76.3); 53% were male; and 58% had a KPS ≥90. The majority of alloHCT recipients were white (87%). Disease status at the time of HCT was CR, partial remission (PR) and chemo-refractory disease in 56%, 30% and 11%, respectively. Most patients received PBSC grafts (89%) and calcineurin-based GVHD prophylaxis (80%). Matched related donors (30%) and 8/8 matched unrelated donors (43%) were the most common types of donors. Reduced intensity and non-myeloablative conditioning (RIC/NMA) and myeloablative conditioning (MAC) were used in 70% and 30% of cases, respectively. Commonly utilized MAC regimens included cyclophosphamide-TBI (n=33) and busulfan-fludarabine (n=20) while commonly utilized RIC/NMA regimens included fludarabine-melphalan (n=55), fludarabine-busulfan (n=33), and fludarabine-TBI (n=32). A total of 49 patients (18%) received in vivo TCD with anti-thymocyte globulin (n=47) or alemtuzumab (n=2).

Table 1.

Baseline characteristics of patients who had first alloHCT for T-PLL, 2000–2018

Characteristic	No. (%)
No. patients	266
No. centers	87
Sex
Male	140 (53)
Female	126 (47)
Age, y
Median age (range), y	59.1 (25.01–76.26)
18–29	1 (0)
30–39	7 (3)
40–49	38 (14)
50–59	98 (37)
60–69	101 (38)
≥ 70	21 (8)
Karnofsky Performance Status score
90–100	153 (58)
< 90	101 (38)
Not reported	12 (4)
HCT-CI
0	73 (27)
1–2	84 (31)
3–4	77 (25)
≥ 5	28 (11)
Not reported	4 (6)
Remission status at HCT
Complete remission	149 (56)
Partial response	80 (30)
No response/ stable/ progression	31 (11)
Not reported	6 (2)
Graft source
Bone marrow	30 (11)
Peripheral blood	236 (89)
Time from diagnosis to HCT
Median (range)	7.85 (2.07–81.74)
< 6 months	82 (31)
6–11 months	103 (39)
≥ 12 months	81 (30)
Donor type
HLA-identical sibling	80 (30)
Haploidentical	30 (11)
URD 8/8	115 (43)
URD 7/8	33 (12)
Other related	8 (3)
Conditioning regimen intensity a
Myeloablative with TBI	44 (17)
Myeloablative without TBI	34 (13)
Reduced-intensity with TBI	75 (28)
Reduced-intensity without TBI	113 (42)
GVHD prophylaxis
CNI + MMF ± others (except PTCy)	68 (26)
CNI + MTX ± others (except MMF, PTCy)	123 (46)
CNI + others (except MMF, MTX, PTCy)	20 (8)
Other prophylaxis b	55 (21)
In vivo T cell depletion (ATG/alemtuzumab) c
Yes	49 (18)
No	217 (82)
Median follow-up (range), months	49 (3.32–116.84)

Abbreviations: alloHCT, allogeneic hematopoietic cell transplantation; ATG, anti-thymocyte globulin; CNI, calcineurin inhibitor; GVHD, graft-versus-host disease; HCT, hematopoietic cell transplantation; HCT-CI, Hematopoietic Cell Transplantation Comorbidity Index; HLA, human leukocyte antigen; MMF, mycophenolate mofetil; MTX, methotrexate; PTCy, post-transplant cyclophosphamide; TBI, total body irradiation; T-PLL, T-cell prolymphocytic leukemia; URD, unrelated donor.

^aRefer to Supplementary Table 1 for full conditioning list

^bOther: CNI alone (12), CNI + PTCy + MMF (32), PTCy-MMF (1), sirolimus + PTCy (2), MTX alone (3), sirolimus-MMF-PTCy (1), monoclonal antibody + MMF (3), PTCy alone (1)

^cATG n=47, alemtuzumab n=2

Overall survival and disease-free survival

The 4-year OS and DFS were 30.0% (95% CI, 23.8–36.5%) and 25.7% (95% CI, 20–32%), respectively (Supplementary Table 2). The 4-year OS based on donor for HLA matched sibling donor (MSD), 8/8 matched unrelated donor (MUD), haploidentical donor (haplo) and 7/8 mismatch unrelated donor (MMUD) was 40.1% (95% CI, 28.9–51.8%), 24.6% (95% CI, 16.2–34.2%), 33.9% (95% CI, 15–56%) and 26.8% (95% CI, 9.6–48.9%) respectively. The 4-year DFS based on donor for MSD, MUD, haplo and MMUD was 34.9% (95% CI, 24.4–46.3%), 19.6% (95% CI, 12–28.5%), 23.4% (95% CI, 8.2–43.3%), and 28.9% (95% CI, 10.4–52.1%) respectively (Supplementary Table 3).

On multivariate analyses, RIC/NMA conditioning regimen was significantly associated with longer DFS (hazard ratio [HR] 1.86; 95%CI, 1.32–2.61; p=0.0004) and OS (HR 2.18; 95% CI, 1.53–3.09; p<0.0001) when compared with MAC. (Figures 1, 2). Performance status (KPS <90%) was associated with both inferior DFS (HR 1.51; 95% CI, 1.12–2.05; p=0.0075) and OS (HR 1.53; 95% CI, 1.12–2.08; p=0.0073), as was recipient age >60 years, which was associated with inferior DFS (HR 1.41; 95% CI, 1.03–1.93; p=0.0337) and OS (HR 1.61; 95% CI, 1.15–2.24; p=0.0053). Use of in vivo TCD resulted in inferior DFS (HR 1.50; 95% CI, 1.05–2.15; p=0.0276), but had no significant effect on OS (Table 2). Time from diagnosis to transplant did not have any significant effect on DFS or OS.

Figure 1. Adjusted overall survival by conditioning intensity (P<0.0001).

MAC, myeloablative conditioning; RIC/NMA, reduced-intensity conditioning/nonmyeloablative conditioning.

Figure 2. Adjusted disease-free survival, by conditioning intensity (P= 0.0004).

MAC, myeloablative conditioning; RIC/NMA, reduced-intensity conditioning/nonmyeloablative conditioning.

Table 2.

Multivariable regression analysis

Factors	N	OR/HR (95% CI)	P-value	Overall P-value
Overall survival
Conditioning regimen
RIC/NMA	188	1.00 (Reference)		< 0.0001
MAC	78	2.18 (1.53– 3.09)	< 0.0001
Age
≤ 60	142	1.00 (Reference)		0.0053
> 60	122	1.61 (1.15– 2.24)	0.0053
KPS
≥ 90%	153	1.00 (Reference)		0.0272
< 90%	101	1.53 (1.12– 2.08)	0.0073
Not reported	12	1.23 (0.60– 2.54)	0.573
Disease-free survival
Conditioning regimen
RIC/NMA	77	1.00 (Reference)		0.0004
MAC	187	1.86 (1.32–2.61)	0.0004
Age
≤ 60	142	1.00 Reference)		0.0337
> 60	122	1.41 (1.03– 1.93)	0.0337
KPS
≥ 90%	152	1.00 (Reference)		0.0075
< 90%	101	1.51 (1.12– 2.05)	0.0075
Not reported 4	11	1.13 (0.53–2.44)	0.7507
In vivo T-cell depletion
No	215	1.00 (Reference)		0.0253
Yes	49	1.50 (1.05–2.13)	0.0253
Treatment-related mortality
Conditioning regimen
RIC/NMA	187	1.00 (Reference)		< 0.0001
MAC	77	3.31 (2.01–5.45)	< 0.0001
Age
≤ 60	142	1.0 (Reference)		0.0108
> 60	122	1.87 (1.16– 3.04)	0.0108
KPS
≥ 90%	152	1.00 (Reference)		0.0142
< 90%	101	1.98 (1.25– 3.14)	0.0036
Not reported	11	1.18 (0.36–3.83)	0.7811
In vivo T-cell depletion
No	215	1.00 (Reference)		0.0263
Yes	49	1.79 (1.07–2.98)	0.0263
Acute GVHD
Conditioning regimen
RIC/NMA	172	1.00 (Reference)		0.0011
MAC	75	2.94 (1.54– 5.62)	0.0011
GVHD prophylaxis
CNI + MMF	65	1.00 (Reference)		0.0093
CNI + MTX	114	0.56 (0.28–1.14)	0.1077
CNI + others (except MMF, MTX, PTCy)	18	0.36 (0.11–1.17)	0.0902
PTCy ± others	33	0.26 (0.10–0.71)	0.0082
Other prophylaxis	17	2.17 (0.71– 6.60)	0.174
Chronic GVHD
GVHD prophylaxis
CNI + MMF ± others (except PTCy)	67	1.00 (Reference)		0.0015
CNI + MTX ± others (except MMF, PTCy)	121	1.06 (0.68– 1.65)	0.8045
CNI + others (except MMF, MTX, PTCy)	20	2.35 (1.31– 4.20)	0.0041
PTCy ± others	37	0.44 (0.19– 1.05)	0.0645
Other prophylaxis	17	0.65 (0.25– 1.66)	0.3677
Year of transplant
2008–2011	50	1.00 (Reference)		0.0216
2012–2015	110	0.62 (0.39–0.97)	0.0382
2016–2018	102	0.48 (0.28–0.82)	0.0069
Relapse
Disease status at HCT
CR	149	1.00 (Reference)		0.0486
PR	80	1.40 (0.91–2.17)	0.1257
No response/ SD/ PD	31	2.13 (1.23–3.71)	0.0072
Not reported	6	0.94 (0.23– 3.87)	0.932

Abbreviations: CNI, calcineurin inhibitor; CR, complete remission; GVHD, graft-versus-host disease; HCT, hematopoietic cell transplantation; HR, hazard ratio; KPS, Karnofsky Performance Status; MAC, myeloablative conditioning; MMF, mycophenolate mofetil; MTX, methotrexate; OR, odds ratio; PD, progressive disease; PR, partial remission; PTCy, post-transplant cyclophosphamide; RIC/NMA, reduced-intensity conditioning/nonmyeloablative conditioning; SD, stable disease;

TBI effect on OS and DFS was analyzed as part of conditioning intensity (Supplementary Table 7). When comparing MAC without TBI (MAC-Chemo) to MAC with TBI, TBI did not have any significant effect on OS (HR 0.83 (95% CI, 0.49–1.41; p=0.0073) or DFS (HR 1.01 (95% CI, 0.60–1.71; p=0.9628). Performing the same analysis with RIC comparing RIC with TBI to RIC without TBI (RIC-Chemo), TBI did not have any significant effect on OS (HR 1.22 (95% CI, 0.81–1.82; p=0.3437) or DFS (HR 1.17 (95% CI, 0.79–1.72; p=0.4390).

Treatment-related mortality

The 1-year and 4-year cumulative incidence of TRM were 21.5% (95% CI, 16.7–26.7), and 32.4% (95% CI, 26.4–38.6), respectively. The 4-year TRM based on donor for MSD, MUD, haplo and MMUD was 20.4% (95% CI, 11.8–30.7%), 36.6% (95% CI, 27.3–46.4%), 31.6% (95% CI, 15.5–50.3%), and 42.1% (95% CI, 23.2–62.4%) respectively (Supplementary Table 3). On multivariate analysis, MAC resulted in increased cumulative incidence of TRM (HR 3.31; 95% CI 2.01–5.45; p<0.0001) when compared to RIC (Figure 3). Additionally, performance status (KPS < 90%) (HR 1.98; 95% CI, 1.25–3.14; p=0.0036) and use of in vivo TCD (HR 1.79; 95% CI, 1.07–2.98; p=0.0263) resulted in increased incidence of TRM (Table 2).

Figure 3. Adjusted treatment-related mortality, by conditioning intensity (P<0.0001).

MAC, myeloablative conditioning; RIC/NMA, reduced-intensity conditioning/nonmyeloablative conditioning.

The effect of TBI on TRM was analyzed as part of conditioning intensity (Supplementary Table 7). When comparing MAC without TBI (MAC-Chemo) to MAC with TBI, TBI did not have any significant effect on TRM (HR 0.48 (95% CI, 0.22–1.05; p=0.0662). Comparing RIC with TBI to RIC without TBI (RIC-Chemo), TBI did not have any significant effect on TRM (HR 1.39 (95% CI, 0.74–2.64; p=0.3068).

Acute and chronic GVHD

The cumulative incidence of grades II-IV aGVHD at day 180 post alloHCT was 22.5% (95% CI, 16.8–28.9) while cumulative incidence of grades III-IV aGVHD at day 180 post alloHCT was 5.3% (95% CI, 2.8–8.6). (Supplementary Table 4). The cumulative incidence of grades II-IV aGVHD at day 180 based on donor for MSD, MUD, haplo and MMUD was 14.3% (95% CI, 6.4–24.7%), 25.7% (95% CI, 16.6–36%), 36.4% (95% CI, 17.5–57.8%) and 20% (95% CI, 5.6–40.4%) respectively (Supplementary Table 3). On multivariate analysis, MAC was predictive for increased risk of grades II-IV aGVHD (OR 2.94; 95% CI, 1.54–5.62; p=0.0011), while post-transplant cyclophosphamide (PTCy) predicted for reduced grades II-IV aGVHD (OR 0.26; 95% CI, 0.10–0.71; p=0.0082) (Table 2). In vivo TCD did not have a significant effect on aGVHD. When comparing MAC with TBI to MAC without TBI as well as RIC with TBI to RIC without TBI, TBI did not have any significant effect on aGVHD (Supplemental Table 7).

The cumulative incidences of chronic GVHD (cGVHD) at 1 year and 2 years post-transplant were 38.8% (95% CI, 32.9–44.9) and 45.5% (95% CI, 39.2–51.8), respectively. Among those with cGVHD at 1 year, 71% had extensive cGVHD and 29% with limited cGVHD, while at 2 years, cGVHD was extensive in 72% and limited in 28% of recipients with cGVHD. The cumulative incidence of cGVHD at 2 years post-transplant based on donor for MSD, MUD, haplo and MMUD was 47.5% (95% CI, 35.8–59.3%), 47.6% (95% CI, 37.9–57.4%), 33.9% (95% CI, 16.6–53.9%) and 49.1% (95% CI, 31.5–66.8%) respectively (Supplementary Table 3). Age, conditioning intensity and in vivo TCD had no significant effect on chronic GVHD. PTCy-based GVHD prophylaxis was associated with less chronic GVHD when compared to calcineurin based GVHD prophylaxis (Table 2). We also observed alloHCT performed before 2011 was associated with increased incidence of cGVHD than those performed after 2011. (Supplementary Table 6).

Relapse

The cumulative incidence of relapse/progression at 1 year and 4 years was 27.6% (95% CI, 22.3–33.2%) and 41.9% (95% CI, 35.5–48.4%). Based on the multivariate analyses (Table 2), age and conditioning intensity were not associated with rate of relapse. Stable or progressive disease at time of alloHCT was associated with increased incidence of relapse (HR 2.13; 95%CI 1.23–3.71; p=0.0072) when compared to CR. However, the depth of response at HCT (PR vs CR), in vivo TCD and TBI-based conditioning were not associated with the incidence of relapse.

Causes of death

The most common cause of death was relapse of the primary disease (52%), followed by infection (15%) and GVHD (13%). (Supplementary Table 5).

DISCUSSION

Using the CIBMTR database, we showed that long-term disease-free survival can be achieved in patients with T-PLL. We observed that RIC/NMA conditioning regimens are associated with reduced TRM and improved DFS and OS. Our analysis also found that the use of in vivo TCD strategies (ATG and/or alemtuzumab) resulted in an increased TRM and inferior DFS. Disease relapse continues to pose a challenge, with a 4-year relapse incidence of 41%. Patients with chemo-sensitive disease prior to transplant had a reduced incidence of relapse.

Data from this analysis are consistent with previous registry studies from the SFGM and the JSHCT (Table 3). The SFGM study retrospectively reported 3-year OS and DFS estimates at 36% and 26% in 27 patients with median follow-up of 33 months, while the JSHCT reported 3-year OS and PFS of 39.8% and 33.5% respectively in 20 patients with median follow-up of 51 months [9,14]. The EBMT study, a prospective observational study amongst recipients age 65 and younger with median follow-up of 50 months, reported 4-year OS and PFS of 42% and 30%, respectively[13]. However, in the EBMT series, the oldest patient was 59 years, whereas in this current CIBMTR study, 42% of patients were older than 60 years, which more closely reflects the median age of T-PLL diagnosis in the US.

Table 3.

Selected studies of alloHCT in T-PLL

Publication	Study	No. patients	Remission status at alloHCT (N)	Donor type	Regimen intensity (N)	Outcomes
Wiktor-Jedrzejczak et al. 13	EBMT	37 a	CR=22 PR=10 Other=5	MRD=15 MUD=22	MAC=13 RIC=24	4 year OS: 42% 4 year NRM: 32% 4 year relapse: 38%
Kalaycio et al.10	CIBMTR	47 (21 T-PLL)b	CR=16 PR=8 Other=21	MRD=11 MUD=19 Other: 13	MAC=19 NMA=14	1 year OS: 48% 1 year NRM: 28% 1 year relapse: 28%
Guillaume et al. 9	SFGM-TC	27	CR=14 PR=10 Other=3	MRD=10 MUD=17	MAC=10 NMA=17	3 year OS: 36% 3 year NRM: 31% 3 year relapse: 47%
Dholaria et al. 8	Moffitt Cancer Center	11	CR=9 PR=1 Other=1	MRD =5 MUD=3 Other=3	MAC=8 RIC=3	4 year OS: 56% 4 year NRM: 34% 4 year relapse: 21%
Yamasaki et al. 14	JSHCT	20	CR=6 PR=1 Other=13	MRD =5 MUD=6 Haplo=2 MMUD=7 UCB: 2	MAC=10 RIC=10	3 year OS: 39.8% 1 year NRM: 20.9% 3 year relapse: 69.6%
Murthy et al (current study)	CIBMTR	266	CR=149 PR= 80 Other=37	MRD =80 MUD=115 Haplo=30 MMUD=33 Other=8	MAC=78 RIC=188	4 year OS: 30% 4 year TRM: 32.4% 4 year relapse: 41.9%

Abbreviations: B-PLL, B cell prolymphocytic leukemia; CIBMTR, Center for International Blood and Marrow Transplant Research; CR, complete remission; EBMT, European Society for Blood and Marrow Transplantation; haplo, haploidentical donor; HCT, hematopoietic cell transplantation; JSHCT, Japan Society for Hematopoietic Cell Transplantation (now known as the Japanese Society for Transplantation and Cellular Therapy; MRD, matched related donor; MMUD, mismatched unrelated donor; MUD, matched unrelated donor; NRM, nonrelapse mortality; OS, overall survival; PR, partial response; SFGM-TC, Francophone Society of Bone Marrow Transplantation and Cellular Therapy; T-PLL, T-cell prolymphocytic leukemia; UCB, umbilical cord blood.

^aData available for 36 patients

^bB-PLL and T-PLL

The intensity of conditioning regimens across these three studies was comparable. RIC/NMA regimens were utilized in 70% patients in the current study, compared to 60% in SFGM, 50% in JSHCT and 65% in EBMT. RIC/NMA conditioning in younger patients was associated with reduced TRM and improved DFS and OS compared to younger patients receiving MAC conditioning. The survival benefit offered with RIC/NMA conditioning may be explained by graft-versus-leukemia (GVL) effect. A study by Sellner and colleagues evaluated a longitudinal quantitative minimal residual disease using clone-specific T-cell receptor (TCR)-based real-time quantitative polymerase chain reaction (PCR): They demonstrated minimal residual disease responses post-alloHCT were associated with a shift from a clonal, T-PLL-driven profile to a polyclonal signature, effectively validating GVL effect in T-PLL[22]. In our analysis, a surrogate marker of GVL, which is the impact of in vivo TCD on relapse, was not evident. The use of in vivo TCD was associated with inferior DFS due to increased risk of TRM.

High incidences of TRM have been reported in prior studies of alloHCT for T-PLL. The 4-year TRM of 32.4% is similar to reports by the EBMT (4-year NRM 32%) and SFGM (3-year TRM 31%). Predictably, we observed reduced TRM and reduced incidence of aGVHD with the use of RIC/NMA conditioning regimens. We observed that in vivo TCD was linked to increased TRM. In the current study, 18% of patients received in vivo TCD, mostly with ATG, compared to the EBMT study, in which 51% received TCD. AlloHCT with TCD has been associated with delayed immune reconstitution and increased risk of infection[23–25]. Infection was reported as the second most common cause of death. Ongoing T-cell depletion caused by pre-transplant alemtuzumab therapy might influence TRM. Additionally, one could hypothesize that ongoing T cell depletion from pre-transplant alemtuzumab therapy, in addition to the use of RIC/NMA conditioning regimens and PTCy GVHD prophylaxis, could explain the low incidence of aGVHD and severe aGVHD observed. However, we could not answer this question conclusively in this analysis, because data for time from last alemtuzumab dose to transplant nor T cell reconstitution data were available.

Outcomes were by donor type were also reviewed (supplementary table 3). Although small in numbers, it is worth mentioning that we observed both haploidentical and mismatch unrelated transplants as feasible and effective in patients with T-PLL. Haploidentical transplants in particular we found to have less cGVHD and TRM w/comparable 4 year relapse, DFS and OS when comparing to MUD and MMUD transplants. It is important to note that donor type was not found to be significant on multivariate analyses and these findings are on univariate analysis only, so it is difficult to draw significant conclusions regarding choice of ideal donor. However, with increased utilization of haploidentical transplantation [26] and feasibility and effectiveness of PTCy in allo-HCT with MMUD [27], allo-HCT should be considered for patients with T-PLL even in the absence of a HLA matched donor.

Controlling disease and preventing relapse remain difficult in patients after alloHCT. Achieving complete remission prior to alloHCT was associated with less relapse, but only when compared to stable or progressive disease and not when compared to partial remission, suggesting that chemoresponsive disease prior to alloHCT is more significant than the depth of remission. Additionally, in this analysis, we investigated the role of total body irradiation. A prospective study by the EBMT identified TBI dose of 6 Gy or more as predictive of a reduced relapse risk in a univariable analysis[13]. We looked specifically whether adding TBI to both MAC and RIC would affect OS, DFS or TRM. When comparing MAC with TBI to MAC without TBI as well as RIC with TBI to RIC without TBI, we did not appreciate any significant effect on OS, DFS and TRM. Our analysis showed differences in survival outcomes with respect to pre-transplant conditioning were more attributed to comparing conditioning intensity (MAC vs RIC) rather than use of TBI

We found that relapse rates increased over time. Incidence of relapse increased, from 27.6% at 1 year, to 41.9% at 4 years. Unfortunately, there is no standard minimal residual disease test for T-PLL, and such a test potentially could help forecast early relapse. Late relapse may reflect waning GVL effect over time. Post-transplant immune modulation strategies may help prevent late relapse. Venetoclax [28], histone deacetylase (HDAC) inhibitors [29], p53 reactivators [30,31], and Janus kinase/signal transducers and activators of transcription (JAK/STAT) inhibitors [32–35] have previously demonstrated some pre-clinical and/or clinical activity in T-PLL, and warrant further investigation for post-transplant maintenance.

This study has limitations inherent to a retrospective registry study. As this data was obtained from a transplant registry, we could not compare outcomes with patients who did not undergo alloHCT. Another limitation was the lack of pertinent pre-transplant information, such as cytogenetics, mutation data and details of therapies prior to alloHCT. Details of pre-HCT induction therapy were not available for most of our study participants, so we did not include this information in our analyses. The lack of consensus disease response criteria is a notable limitation. The CIBMTR registry defined T-PLL response criteria based on international consensus response criteria for chronic lymphocytic leukemia[16]. Only recently in 2019 were consensus T-PLL response guidelines were published[12]. Given that patients included in this analysis date back to 2008, utilizing the updated criteria was not feasible. Finally, detailed data were not available regarding the timing and severity of infections, as well as immune reconstitution.

CONCLUSION

In summary, alloHCT results in durable remissions and disease control in some patients with T-PLL. Relapse continues to remain a barrier to long-term survival. Reduced-intensity conditioning and avoidance of in vivo TCD are associated with improved outcomes. Molecular monitoring of patients for recurrence after transplant could be undertaken to identify early relapses for treatment and potentially donor lymphocyte therapy. Other novel approaches combined with alloHCT warrant investigation to further improve outcomes of alloHCT in T-PLL.

Supplementary Material

2

Supplementary Table 1: Full conditioning regimen list

Supplementary Table 2: Univariate analysis

Supplementary Table 3: Univariate analysis stratified by donor

Supplementary Table 4: Cumulative incidence of graft failure and GVHD

Supplementary Table 5: Causes of death

Supplementary Table 6: Full multivariate analysis

Supplementary Table 7: Multivariate analysis (Conditioning intensity +/− TBI)

Highlights.

• Allogeneic HCT is effective in yielding durable remissions in patients with T-PLL

• Myeloablative conditioning, age greater than 60 and KPS <90 were all associated with reduced OS

• Reduced intensity conditioning and avoidance of in vivo T cell depletion correlated with better DFS and less TRM

• TBI was not found to any significant effect on OS, DFS or TRM