Allogeneic hematopoietic cell transplantation for peripheral T-cell NHL results in long-term disease control

Abstract

The study analyzed outcomes of a consecutive case series of 37 patients with peripheral T-cell non-Hodgkin lymphoma, from related and unrelated donors, using allogeneic hematopoietic cell transplantation (allo-HCT), between the years 2000 and 2007. All patients were pretreated; the majority had either relapsed or progressive disease (n=25, 68%), 13 had cutaneous histologies (CTCL), and all were ineligible for autologous transplant. Fully ablative conditioning regimens were used in 13 patients while 24 patients underwent reduced intensity conditioning (RIC). At five years the overall survival (OS) and progression-free survival (PFS) probabilities were 52.2% and 46.5%, respectively. At the time of analysis, 9 (24.3%) patients had either relapsed (n=6) or progressed (n=3) post allo-HCT. The cumulative incidences of relapse/progression and non-relapse mortality at 5 years were 24.3% and 28.9%. No statistically significant variables for survival or relapse were discovered by univariate Cox-regression analysis of disease and patient characteristics; differences between CTCL and other histologies were not significant. The median follow-up of 64.0 months (range: 16.4–100.4) indicates a mature data-set with probable cure in the survivors. The relapse/progression curves reached and maintained plateaus after 1 year post-transplant, demonstrating that long-term disease control is possible after allo-HCT in PTCL patients with advanced disease.

INTRODUCTION

Mature T-cell non-Hodgkin Lymphoma (NHL) and NK-cell neoplasms, collectively called peripheral T cell lymphomas (PTCL), comprise about 12% of all NHL and 15–20% of aggressive lymphomas worldwide [1]. PTCLs exhibit great morphological diversity and histological variation even within individual disease entities [2]; the current 2008 WHO classification of lymphoid neoplasms recognizes over 20 types of PTCLs [3,4]. The most common histologies include: peripheral T-cell lymphomas not otherwise specified (PTCL-nos), anaplastic large-cell lymphomas (ALCL), and angioimmunoblastic T cell lymphomas (AITL). The most aggressive histologies include: hepatosplenic T-cell NHL, gamma/delta T-cell malignancies and NK T-cell lymphomas [5]. Cutaneous T-cell neoplasms (CTCL) are classified separately and are considered a clinically unique entity [6], generally with a more indolent behavior, excepting a few subtypes including Sézary syndrome and transformed mycosis fungoides.

For aggressive lymphomas, a T-cell phenotype confers a worse clinical outcome compared to B-cell lymphomas, with the exception of ALK-positive ALCL [7,8]. Long-term survival at 5 years remains at 10–30% for most histologies [9]. Advanced disease stage, high prognostic index at presentation [10] and inherent chemomoresistance [11] contribute to this dismal outcome [12]. For CTCL patients, Sézary syndrome and advanced mycosis fungoides are often preceded by prolonged disease courses with multiple treatment regimens, ultimately resulting in chemoresistance [13]. Current therapeutic strategies for T-cell NHL remain poorly defined and tend to be extrapolated from treatment paradigms for B-cell NHL. As a result, relapsed and chemo-refractory disease remains a significant clinical dilemma in the care of these patients.

High dose chemotherapy followed by autologous stem cell transplantation (ASCT) is the standard therapeutic option for patients with aggressive B-cell lymphomas failing primary therapy [14,15]. However, many patients with relapsed PTCL are not candidates for ASCT due to chemo-refractory/progressive disease, extensive bone marrow involvement, or failure to adequately mobilize stem cells [16,17]. Furthermore, 25–30% of patients who achieve a complete response to ASCT will relapse [18]. For those PTCL patients who fail, or are ineligible for ASCT, allogeneic (related or unrelated) hematopoietic cell transplantation (allo-HCT) has been offered in an attempt to harness a potential graft-versus-lymphoma (GVL) effect [18,19]. In these studies, the GVL effects are inferred from response to reduced intensity conditioning regimens, donor lymphocyte infusions (DLI) and withdrawal of immunosuppression at the time of relapse (10 responding patients in all 3 studies), and the correlation of GVHD with decreased relapse rate [19–25] The specific factors related to disease biology that impact survival outcomes, disease control, mortality and morbidity in the setting of allo-HCT for PTCL remain poorly defined.

Due to the rarity of these diseases, most single center studies of allo-HCT in T-cell NHL are limited by small numbers of patients, heterogeneity of histologies and treatment regimens and short follow-up. Three multicenter studies [20, 23,26] have reported consolidated data on a relatively large number of patients in an attempt to understand the efficacy of allogeneic stem cell therapy in PTCL and CTCL, in general, and in specific subtypes where the numbers permit. Here we report the results of allo-HCT performed in 37 patients with PTCL at the City of Hope between 2000 and 2007. We demonstrate an overall survival of 52.2% at 5 years post allo-HCT and long-term disease control in a population with advanced disease.

PATIENTS AND METHODS

The City of Hope (COH) prospective longitudinal transplant database identified 37 consecutive patients with PTCL as defined by the 1999 WHO criteria [27], that were treated with allo-HCT, from HLA matched related and unrelated donors, between the years of 2000 and 2007. This use of data for retrospective analysis was approved by the COH Institutional Review Board (IRB) and anonymity of all patient information was maintained. Pathology review of biopsy specimens was conducted by the COH hematopathology department to confirm the diagnosis of PTCL prior to transplant as per institutional policy. Disease status at time of transplant was confirmed by clinical assessment including physical examination and laboratory evaluation, imaging by CT scans and nuclear imaging, bone marrow biopsies as well as other tissue biopsies and photo documentation (per institutional standard operating procedures). Post transplant evaluation of disease status with imaging studies, bone marrow biopsies and engraftment analysis occurred at 30 days, 60 days, 1 year post transplant and yearly thereafter or as clinically indicated. International Working Group criteria (IWG) criteria [28] were used to define disease response post transplant.

Patient Characteristics

There were 37 patients with a diagnosis of PTCL including: PTCL nos (n=8); AILT (n=4); ALCL (n=6: 3 ALK⁺, 2 ALK⁻, 1 ALK unknown); rare histologies (n=6) including NK/T cell lymphomas both nasal and extra-nasal, enteropathy-type-T-cell NHL, and hepatosplenic T-cell lymphoma. There were 13 patients with cutaneous T-cell lymphomas including mycosis fungoides (MF) and Sézary syndrome; these patients were analyzed as a separate subgroup called CTCL, with all other patients grouped together as Other PTCL. The analysis excluded patients with a diagnosis of T-cell lymphoblastic leukemias/lymphomas and HTLVI/II associated T-cell lymphoma/leukemia. Patient and transplant characteristics are summarized in Table I for the entire population and also for the CTCL and Other PTCL subgroups. The median age at transplant was 40 years (range: 7–72) and there were 27 males and 10 females. Disease status at the time of transplant was as follows: 17 patients had failed primary chemotherapy and were classified as induction failure, 1 patient had progressed on primary therapy and was considered to have primary progressive disease, 7 patients were in complete remission (CR) following therapy (including 3 patients in CR1 and 4 in CR2), and 12 patients were either in partial remission after relapsing (n=5) or in first relapse (RL1, n=7). The distribution of PIT scores, for the Other PTCL subgroup, as defined by Gallamini et al. [10] is shown in Table I. Patients were treated with a median of three prior regimens for the entire cohort. The median number of prior regimens was 3 (1–8) in patients from the Other PTCL group and 6 (4–9) in the CTCL subset. The median number of months from diagnosis to transplant was 17 (4–112) for the entire cohort, 38 (9–88) for the CTCL group and 11 (4–112) for the Other PTCL group. Only one patient had received a prior ASCT in this cohort. The median KPS at time of transplant was 90 (range: 60–100).

Table I.

Patient, Disease and Transplant Characteristics

	Total Population	Other PTCL	CTCL)
Characteristic	N (%) or Median (Range)	N (%) or Median (Range)	N (%) or Median (Range)
Number of patients (N)	37	24	13
Patient Gender
Male	27	18	9
Female	10	6	4
Age at Transplant (years)	40 (7–72)	39.5 (7–72)	50 (18–61)
Donor Type
Sibling	26 (70.3)	19 (79.2)	7 (53.8)
Matched Unrelated Donor	11 (29.7)	5 (20.8)	6 (46.2)
Disease Status at Transplant
1st Complete Remission	3	2	1
2nd Complete Remission	4	4	0
Partial Remission	5	5	0
1st Relapse	7	5	2
Primary Progressive	1	0	1
Primary Induction Failure	17 (46.0)	8	9
PIT Groups
1 (PIT Score = 0)	NA	13 (54.2)	NA
2 (PIT Score = 1)	NA	2 (8.3)	NA
3 (PIT Score = 2)	NA	6 (25.0)	NA
Unknown	NA	3 (12.5)	NA
Prior Regimens	3 (1–9)	3 (1–8)	6 (4–9)
Months from Diagnosis to Transplant	17 (4–112)	11 (4–112)	38 (9–88)
Stem Cell Source
Bone Marrow	5 (13.5)	4 (16.7)	1 (7.7)
Peripheral Blood	31 (83.8)	19 (79.2)	12 (92.3)
Cord Blood	1 (2.7)	1 (4.2)	0
Conditioning Intensity
Reduced	22 (59.4)	13 (54.2)	9 (69.2)
Non-myeloablative	2 (5.4)	1 (4.2)	1 (7.7)
Myeloablative	13 (35.1)	10 (41.7)	3 (23.1)
Conditioning Regimen
Bu | Cy	1	0	1
FTBI | Cy	7	5	2
FTBI | Cy | ATG	1	1	0
FTBI | VP-16	4	4	0
Fludarabine | TBI	2	1	1
Fludarabine | Melphalan	22 (59.5)	13	9
GVHD Prophylaxis
CSA/MMF Based	20 (54.1)	12 (50.0)	8 (61.5)
Tacro/Siro Based	17 (45.9)	12 (50.0)	5 (38.5)

MF–mycosis fungoides, SS–Sézary Syndrome, CTCL–cutaneous T-cell lymphoma, DX–diagnosis, Bu–busulfan, CTX–cyclophosphamide, FTBI–fractionated total body irradiation, ATG–anti-thymocyte globulin, VP-16–etoposide, CSA–cyclosporine, MMF–mycophenylate mofetil

Eligibility Criteria

Patients with very aggressive histologies such as gamma/delta T-cell NHL were offered allo-HCT early in the disease course (CR1/PR1) due to the known poor prognosis and aggressive nature of these diseases. Patients who relapsed after conventional chemotherapy or had primary chemo-refractory disease were offered allo-HCT after they were found to be ineligible for high-dose therapy and ASCT. Patients with CTCL were not offered ASCT as it provides poor long-term disease control [29].

Donor Selection

Donor selection was based on the availability of an HLA-identical or single-antigen mismatched family donor, or alternatively from an HLA matched unrelated donor based on molecular typing of HLA A, B, C, DR, and DQ loci. A total of 26 (70%) patients were transplanted using a sibling donor, and 11 (30%) patients had a matched unrelated donor. Source of stem cells was as follows: bone marrow (n=5), growth factor mobilized stem cells (n=31), cord blood (n=1).

Conditioning Regimens and GVHD Prophylaxis

A total of 24 patients received reduced-intensity conditioning regimens (RIC) and 13 received fully ablative regimens. Myeloablative conditioning regimens included: 1) fractionated total body irradiation (FTBI) / cyclophosphamide (Cy) – TBI at 1320 cGY given in 11 fractions over 4 days plus IV Cy at 60 mg/kg ideal bodyweight for 2 days, 2) FTBI/etoposide (VP-16) – FTBI at 1320 cGY given in 11 fractions over 4 days plus IV VP-16 at 60 mg/kg adjusted bodyweight for one day, and 3) busulfan (Bu)/Cy – IV Bu adjusted to AUC between 1000 and 1200 plus IV Cy at 60 mg/kg ideal bodyweight for 2 days. Reduced intensity conditioning regimens included: regimen 4) fludarabine (Flu)/ total body irradiation (TBI) – IV Flu at 25 mg/m² for 5 days plus a single dose of TBI at 200 cGY given on day -1, and regimen 5) Flu/melphalan (Mel) – IV Flu at 25 mg/m² for 5 days plus IV Mel at 140 mg/m² for one day. Specific GVHD prophylactic regimens are shown in Table I.

Statistical Methods

Survival estimates were calculated using the Kaplan-Meier product-limit method; 95% confidence intervals were calculated using the logit transformation and the Greenwood variance estimate [30]. Differences between Kaplan-Meier curves were assessed by the log-rank test. Patients who were alive at the time of analysis were censored at the last contact date. Overall survival (OS) was measured from the day of stem cell infusion to death from any cause. Progression-free survival (PFS) was defined as time from stem cell infusion to recurrence, progression or death from any cause, whichever occurred first. The relapse/progression incidence (RP) was defined as time from stem cell infusion to recurrence or progression. Non-relapse mortality (NRM) was measured from stem cell infusion to death from any cause other than disease relapse or disease progression. Non-relapse-related mortality and relapse-related mortality were considered competing risks for mortality. The cumulative incidence of NRM and relapse-related mortality was calculated using the method described by Gooley et al [31]. Differences between cumulative incidence curves in the presence of a competing risk were tested using the Gray method [32]. The significance of demographic, disease, and treatment features was assessed using either univariate Cox regression analysis [33] or its competing-risks analogue [34]. Univariate models were used to model time to event endpoints (e.g., OS, PFS, RP, and NRM), as a function of the prognostic variables. The list of prognostic variables was determined from a literature review that identified factors associated with survival and/or disease relapse/recurrence in patients treated with allo-HCT. These variables were: histopathological subtype, patient age at allo-HCT (<40 years, ≥40 years), disease status at the time of allo-HCT (CR or PR; relapse/PD/IF), donor type (sibling, unrelated), and conditioning regimen (ablative, reduced-intensity). The impact of GVHD (acute, chronic, any GVHD) was analyzed as a time-dependent covariate [35,36]. For the time-dependent covariate analyses, GVHD evaluation began after the infusion of stem cells. The time-dependent covariate took on the value of ‘1’ if the GVHD assessment was positive and ‘0’ otherwise. The value of the time-dependent covariate remained the same until the next GVHD assessment. All calculations were preformed using SAS 9.2 (SAS Institute, Cary, NC) and R (version 2.4.1; http://www.r-project.org). Statistical significance was set at the P <0.05 level; all P values were two-sided. The data were locked for analysis on February 1, 2009 (analytic date).

RESULTS

Outcomes

Study outcomes are summarized in Table II, both for the entire population and for the Other PTCL and CTCL subgroups. The median follow up was 20.3 months (range: 0.7–100.4 months) for all patients and 64.0 months (range: 16.4–100.4 months) for surviving patients. The median follow-up was 49.2 months (range: 16.4–100.4) for surviving Other PTCL patients and 86.4 (range: 31.9–95.3) for CTCL patients. At last contact, 20 patients (54.1%) were alive and 17 (45.9%) had expired. The OS probability at 5 years was 52.2% (95%CI: 43.0–60.5) (Figure 1). The primary causes of death included: disease progression (n=6), infection (n=6), acute GVHD (n=1), chronic GVHD (n=1), secondary malignancy (duodenal cancer, n=1), and multi-organ failure (n=2). There were 9 deaths prior to day 100 mainly due to transplant-related causes, 4 in the Other PTCL group and 5 in the CTCL group. The proportion of patients who either relapsed or progressed following transplant was 24.3% including 6 relapses and 3 progression events (defined as disease progression without remission); 7 of these patients died from their disease and 2 relapsed patients remained alive at the analysis date, 1 in CR and 1 with active disease. No relapsed/progressed patients received donor lymphocyte infusions, and all patients received additional treatments, except for 2 progressives who died rapidly. The only relapsed or progressed patient who was alive and disease-free at the analytic date, had relapse only in the skin and achieved remission via topical steroids and retinoids.

Table II.

Summary of Outcomes

	Total	Other PTCL	CTCL
Outcome	N (%)	N (%)	N (%)
Number of patients (N)	37	24	13
aGVHD
Yes	19 (51.4)	12 (50.0)	7 (53.9)
Grade I	5 (13.5)	4 (16.7)	1 (7.7)
Grade II	8 (21.6)	6 (25.0)	2 (15.4)
Grade III	2 (5.4)	1 (4.2)	1 (7.7)
Grade IV	4 (10.8)	1 (4.2)	3 (23.1)
No	18 (48.6)	12 (50.0)	6 (46.2)
cGVHD
Yes	23 (62.2)	16 (66.7)	7 (53.9)
Limited	3 (8.1)	3 (12.5)	0
Extensive	20 (54.1)	13 (54.2)	7 (53.9)
No	5 (13.5)	4 (16.7)	1 (7.7)
Expired < 100 days post-HCT	9 (24.3)	4 (16.7)	5 (38.5)
Any GVHD (aGVHD/cGVHD)
Yes	29 (78.4)	19 (79.2)	10 (76.9)
No	8 (21.6)	5 (20.8)	3 (23.1)
Relapse/ Progression post-HCT
Yes	9 (24.3)	6 (25.0)	3 (23.1)
No	28 (75.7)	18 (75.0)	10 (76.9)
Vital Status post-HCT
Alive	20 (54.1)	15 (62.5)	5 (38.5)
Dead	17 (45.9)	9 (37.5)	8 (61.5)
Cause of Death
Disease Progression	6 (16.2)	4 (44.4)	2 (25.0)
Infection	6 (16.2)	3 (33.3)	3 (37.5)
multi-organ failure	2 (5.4)	1 (11.1)	1 (12.5)
aGVHD	1 (2.7)	0	1 (12.5)
cGVHD	1 (2.7)	0	1 (12.5)
secondary malignancy (duodenal)	1 (2.7)	1 (11.1)	0

aGVHD–acute graft-versus-host disease, NA–not applicable, cGVHD–chronic graft-versus-host disease, HCT–hematopoietic cell transplant

Figure 1. Outcomes for total population.

All curves are for the 37-patient population with a median follow-up of 64 months for survivors. Panel A shows the Kaplan-Meier estimate of survival probability. Panel B shows progression-free survival, defined as time from stem cell infusion to recurrence, progression, or death from any cause, whichever occurred first. Panel C shows cumulative incidence of relapse/progression, defined as time from stem cell infusion to recurrence or progression. Panel D shows non-relapse mortality, measured as time from stem cell infusion to death from any cause other than disease relapse or disease progression. In panels C and D, relapse/progression and non-relapse mortality were treated as competing risks.

Curves for OS, PFS, relapse incidence and NRM are shown for the entire population in Figure 1, and are stratified based on CTCL or Other PTCL subgroups in Figure 2. The probability of PFS was 46.5% (95%CI: 38.4–54.1) at 5 years for the entire population. For CTCL and Other PTCL, 5-year PFS was 38.5% (95%CI: 28.4–48.5) and 49.7% (95%CI: 38.3–60.1) respectively, with no statistically significant difference (P=0.34). A plateau in the cumulative incidence of relapse/progression at 24.3% was achieved 1 year post allo-HCT. The cumulative incidence of non-relapse mortality at 100 days, 1 year and 5 years was 16.2%, 18.9% and 28.9% respectively for the total population, and was not statistically significantly different for the CTCL versus Other PTCL subgroups. Patients with active disease at the time of transplant showed a trend toward poorer overall survival, with 5-year OS estimates of 43.2%, for patients not in CR/PR, compared to 72.9% for patients in CR/PR (p=0.07) indicating that remission status could impact the outcome of the transplant.

Figure 2. Outcomes stratified by histological subgroup.

All curves are stratified for histology: CTCL (n=13, black lines) versus Other PTCL (n=24, gray lines), with a median follow-up of 86 months for the CTCL surviving patients and 49 months for the Other PTCL group. P-values for log-rank comparison of curves are given in the bottom right hand corner of each graph. Panel A shows the Kaplan-Meier estimate of survival probability. Panel B shows progression-free survival, defined as time from stem cell infusion to recurrence, progression, or death from any cause, whichever occurred first. Panel C shows cumulative incidence of relapse/progression, defined as time from stem cell infusion to recurrence or progression. Panel D shows non-relapse mortality, measured as time from stem cell infusion to death from any cause other than disease relapse or disease progression. For data in panels C and D, relapse/progression and non-relapse mortality were treated as competing risks.

Graft-versus-host disease

Of the 37 evaluable patients, acute GVHD was seen in 19 (51.4%) patients (Table II). Of these, 13 had grade I/II acute GVHD, the remaining 6 experienced grade III/IV acute GVHD. Chronic GVHD developed in 23/28 evaluable patients (82.1%); extensive chronic GVHD in 20 (71.4%). In the final analysis, 29/37 patients (78.4%) experienced some form of GVHD.

Univariate Analyses

The univariate analyses results for OS, PFS, and RP are summarized in Table III. There were no statistically significant differences in OS, PFS, and RP in patients who had CTCL versus Other PTCL. Similarly, there were no statistically significant differences in outcome between patients who received fully ablative conditioning and those conditioned with a reduced intensity regimen., For patients who underwent transplantation with advanced disease, there was a trend toward a 3-fold increased risk of death (HR: 3.1, 95%CI: 0.9–10.7; p=0.08). Prognostic Index for PTCL-unspecified (PIT score) [10] was also analyzed as a potential influence on hazard risk for OS, PFS and RPR in the Other PTCL group only. The analysis was performed on the following separate groupings: PIT group 1/2 vs. 3/4, PIT group 1 vs. 2/3/4, and PIT Level: 1 vs. 2 and 1vs. 3/4. There was no significant difference found by any grouping method. Table III shows the results for PIT group 1/2 versus 3/4.

Table III.

Univariate Analysis of Outcomes

Overall Survival
Variable	Reference Group	Total n=	Events	Censored	Hazard Ratio	Confidence Interval	p-Value
Age at HCT	Age < 40	37	17	20	0.96	0.37–2.48	0.93
PTCL versus CTCL	PTCL	37	17	20	2.00	0.77–5.20	0.16
Disease Status at HCT	1CR/2CR/PR	37	17	20	3.06	0.87–10.69	0.08*
PIT Group	1 / 2	21	7	14	1.06	0.21–5.50	0.94
Donor Type	Sibling	37	17	20	0.96	0.34–2.74	0.94
Conditioning Regimen	RIC	37	17	20	0.69	0.25–1.87	0.46
aGVHD Max Grade	None-II	37	17	20	2.17	0.49–9.62	0.31
cGVHD Max Grade	Limited	23	7	16	0.68	0.16–2.88	0.60
Progression-Free Survival
Variable	Reference Group	Total n=	Events	Censored	Hazard Ratio	Confidence Interval	p-Value
Age at HCT	Age < 40	37	19	18	0.94	0.38–2.32	0.90
PTCL versus CTCL	PTCL	37	19	18	1.56	0.62–3.89	0.34
Disease Status at HCT	1CR/2CR/PR	37	19	18	1..80	0.65–5.02	0.26
PIT Group	1 / 2	21	9	12	0.74	0.15–3.55	0.70
Donor Type	Sibling	37	19	18	0.76	0.27–2.10	0.59
Conditioning Regimen	RIC	37	19	18	0.58	0.22–1.54	0.27
aGVHD Max Grade	None-II	37	19	18	1.70	0.39–7.43	0.438
cGVHD Max Grade	Limited	23	9	14	1.46	0.37–5.79	0.59
Relapse/Progression
Variable	Reference Group	Total n=	Events	Censored	Hazard Ratio	Confidence Interval	p-Value
Age at HCT	Age < 40	37	9	18	0.67	0.18–2.49	0.55
PTCL versus CTCL	PTCL	37	9	18	1.10	0.27–4.42	0.89
Disease Status at HCT	1CR/2CR/PR	37	9	18	1.28	0.32–5.15	0.73
PIT Group	1 / 2	21	5	12	0.64	0.07–5.74	0.69
Donor Type	Sibling	37	9	18	0.27	0.03–2.16	0.22
Conditioning Regimen	RIC	37	9	18	0.69	0.17–2.75	0.59
aGVHD Max Grade	None-II	37	9	18	1.00	0.00-UND	0.99
cGVHD Max Grade	Limited	23	5	14	1.98	0.28–14.17	0.50

Univariate Cox proportional hazards analysis:

^**significance: p ≤0.05;

^*trend: p ≤ 0.1; Confidence intervals shows 95% confidence ranges. HCT–hematopoietic cell transplant, aGVHD–acute graft-versus-host disease, cGVHD–chronic graft-versus-host disease, PTCL–peripheral T-cell lymphoma, 1CR–first complete remission, 2CR–second complete remission, PR–partial remission, RIC-reduced-intensity conditioning, UND-undetermined.

Time-dependent covariate analysis: In this analysis because the timing of GVHD is considered as part of the risk calculation, the model assesses benefit or harm after GVHD onset for the endpoint of interest. This analysis showed no significant impact on survival, relapse/progression or mortality hazard risk for any of the following variables: 1) time to onset of any GVHD (acute or chronic), 2) time to onset of aGVHD (any grade), 3) time to onset of acute GVHD ≥ grade II, 4) time to onset of aGVHD ≥ grade III, 5) time to onset of chronic GVHD (limited or extensive).

DISCUSSION

T-cell phenotype in aggressive lymphomas confers a poor prognosis as reported by the International T-cell lymphoma project [9], a collaborative epidemiological study involving over 22 centers and 1500 patients worldwide. This study reports 5-year overall survival of the most common types of T-cell lymphoma as follows: PTCL-NOS and AILT (32%), ALCL ALK⁺ (70%), ALK⁻ (49%), NK/T cell (nasal type 42%, extra nasal type 9%) hepatosplenic and enteropathy type T-cell NHL (7–10%). Relapse and death rates for T-cell lymphomas are unacceptably high with standard therapies.

The suggestion of a graft-versus-lymphoma effect was first introduced in the early 1990s after demonstration that an allogeneic transplant for lymphoma was associated with a lower risk of relapse as compared to autologous transplant [37,38]. These initial studies included both B- and T-cell lymphomas and did not examine graft versus T-cell lymphoma independently. The high transplant-related mortality associated with allo-HCT offset any benefit obtained from the decreased relapse rate thus negating any survival advantage. Over the last decade the introduction of reduced-intensity conditioning regimens and improved supportive care, including prophylaxis and treatment of GVHD and infections, has reduced the upfront morbidity and mortality of allogeneic transplant, making it more accessible to an aging population [39,40]. The RIC therapeutic modality relies primarily on a graft-versus-tumor effect and it is becoming routine to offer RIC transplants for histologies such as low-grade B-cell lymphomas [41,42] where the graft-versus-lymphoma effect has been well defined. These studies have also demonstrated improved disease control in the presence of GVHD and documented treatment of post-transplant relapse by withdrawal of immunosuppression and administration of DLI as evidence of a graft-versus-tumor effect.

The role of allogeneic transplant in patients with T-cell NHL is evolving but systematic prospective studies for this group of patients are lacking. Le Gouill et al [23], in a large multi-center retrospective study of 77 patients with T-cell NHL, report a 5-year OS and event-free survival (EFS) of 57% and 53% respectively with a TRM of 33%. In multivariate analysis, chemo-resistant disease and the occurrence of grade 3/4 acute GVHD are associated with poor outcome. The presence of CR or PR at the time of transplant is associated with a statistically significant improvement in OS at 5 years as compared to stable, progressive or refractory disease: 69% vs. 29% (p=0.04). This study fails to show an association between GVHD and improved outcome; rather, the presence of acute GVHD was a poor prognostic factor for survival. They propose a GVL effect by demonstrating that 2 patients receiving DLI achieve sustained remission. Similarly, Kim et al. [20] report 5-year survivals of 70% with PTCL and 30% with NK/T-cell lymphomas in 54 patients. Their analysis shows that chronic GVHD was associated with a poor outcome, but they also report that 3 patients receiving DLI for relapsed disease achieve a sustained response. Wosserman et al. [43] and Molina et al. [44] focus only on the specific histologies of ALCL and CTCL respectively and report impressive long-term survival in these patients after allogeneic stem cell transplant. Corradini et al. [25] report a positive GVL effect in a prospective phase II trial of 17 PTCL patients using a RIC regimen of thiotepa/Cy/Flu. Overall survival of 80% at 3 years and PFS of 64% at 3 years points to long-term disease control via allogeneic transplant but most notable is their NRM of only 6%. Fifteen of these 17 PTCL patients were in CR or PR at the time of transplant, which likely contributes to the excellent outcomes. Providing further evidence for GVL, this group administered DLI to 2 relapsed patients and also withdrew immunosuppression from one patient, successfully inducing remission in all three. A recently published retrospective study by Kyriakou et al. of allogeneic transplant for AITL [45] shows a decrease in relapse rate associated with development of cGVHD, lending further support to a GVL effect in T-lymphoma. In this study, the 45 patients, 44% of whom received RIC, exhibit a 3-year PFS of 53%, OFS of 64% and relapse rate of 20%. In summary, the literature demonstrates that it is possible to achieve survival rates of 40–60% at 3–5 years with allogeneic transplants for T-cell lymphoma. High rates of NRM (up to 50%) are reported with fully ablative conditioning regimens but the use of RIC is associated with lower toxicity [40] and there is some evidence for a graft-versus-T-cell lymphoma effect.

Since 35% of patients in this study had mycosis fungoides or Sézary syndrome, we investigated possible differences between the CTCL patient subgroup and other subtypes of PTCL. CTCL has a clinical course distinct from other forms of PTCL, with a tendency for multiple treatment regimens over many years, as evidenced by the fact that our CTCL subgroup had double the median number of prior regimens as did the Other PTCL group. Once patients progress to advanced phase disease the median survival is between 1.5 to 4 years.[46,47]. The documentation in the literature of transplant outcomes for CTCL, consists primarily of case studies and series, including a previous City of Hope report [44]. A recent review of 60 CTCL patients from the EBMT registry [26] estimates an OS of 54% at 3 years; however, the extensive use of T-cell depletion (42%) in this study was associated with a reduced PFS.

Here we report the results of a single institution experience using allogeneic stem cell transplant in 37 patients with PTCL followed for over 5 years. We obtained 52% OS and 46.5% PFS at 5 years, with best results seen in patients transplanted in a state of a complete or partial remission compared to patients who had active disease at the time of transplant. Patients were offered allogeneic in lieu of autologous stem cell transplant due to progressive disease, bone marrow involvement or histology. Sixty-five percent of the patients received a reduced intensity conditioning regimen (RIC) with no statistically significant differences in OS or PFS compared to the ablative therapy patients. Histology subtype CTCL versus Other PTCL also made no significant difference in any survival outcome measure. While CTCL tends to have a less aggressive clinical course, the more indolent disease manifestation is offset by the larger number of prior regimens, higher median age, and extended time between diagnosis and transplant (much of it spent on immunosuppressive therapy) in our CTCL group, resulting in higher NRM.

Presence or absence of any form of GVHD, was not found to impact any survival outcomes, based on a time-dependent analysis. As our study population is relatively small, the lack of statistical associations is not surprising. The proportion of patients who relapsed or progressed was 24.3% with relapses occurring within the first 1 year post transplant. Overall mortality was 46% with only 5 deaths attributed to disease relapse while remaining deaths were secondary to infection or complications of GVHD. These data indicate that allogeneic stem cell transplant is capable of conferring long-term disease control even in relapsed patients, with improved overall outcomes attainable through reduction of NRM. The low risk of relapse, and the plateaus in survival and relapse curves, are encouraging and indicate that a cure may be possible in over half the patients, who have now been followed for over 5 years.

Although patients in remission tend to have better overall survival than those with active disease at transplant (73% vs. 43% at 5-years), a 5-year PFS of 43% for patients with active disease is still a chance that many would consider worth taking, given the lack of alternatives. Achieving a remission state prior to transplant typically improves the outcome of allogenic stem cell transplant, emphasizing the importance of more effective T-lymphoma salvage therapies.

In this analysis, only 6/37 patients died of disease progression (5-yr RP cumulative incidence of 24.3%), again emphasizing the excellent disease control that can be achieved by the modality of allogeneic stem cell transplant, even in these heavily pretreated patients. Our 5-yr overall survival of 52% and NRM of 29%, while promising, highlight the need for improved remediation of infections, GVHD and other complications of allo-HCT. The advent of novel agents with activity against T-cell lymphomas including HDAC inhibitors [48–52], antifolates like pralatrexate [53], antibodies such as alemtuzumab (Campath) [54], and anti CD4 (Humax) [55] may change the paradigms in T-cell lymphoma therapy. Several of these agents, although still in clinical trials, show strong indications of activity against T-cell lymphoproliferative disorders, potentially enabling refractory patients to attain remission as a bridge to more effective stem cell transplantation. We have shown that allo-HCT can provide effective long-term disease control for PTCL, even in a population with advanced disease status. A follow-up period of over five years for surviving patients suggests that this long-term disease control is actually cure. The future combination of allogeneic hematopoietic cell transplant with emerging low-toxicity therapies for T-cell lymphoma will further enhance this curative potential.