Allotransplantation for patients age 40 years and greater with non-Hodgkin Lymphoma (NHL): encouraging progression-free survival

Abstract

Non-Hodgkin lymphomas (NHL) disproportionately affect older patients who uncommonly receive allogeneic hematopoietic cell transplantation (HCT). We analyzed CIBMTR data on 1248 patients ≥40 years receiving reduced-intensity conditioning (RIC) or non-myeloablative (NMA) HCT for aggressive (n=668) and indolent (n=580) NHL. Aggressive lymphoma was more frequent in the oldest cohort [(age 40–54) 49% vs. (55–64) 57% vs. (≥65) 67% p=0.0008]; fewer patients ≥65 had prior autografting [26% vs. 24% vs. 9%; p=0.002)]. Rates of relapse, acute and chronic GVHD and non-relapse mortality (NRM) at one year were similar [22%, 95% confidence interval (CI) 19–26%; 27%, 95% CI 23–31%; 34%, 95% CI 24–44%]. Progression-free (PFS) and overall (OS) survival at 3 years was slightly lower in older cohorts [OS:54%, 95% CI 50–58%; 40%, 95% CI 36–44%; 39%, 95% CI 28–50%; p<0.0001]. Multivariate analysis revealed no significant effect of age on acute or chronic GVHD or relapse. Age ≥55 years, Karnofsky performance status <80, and HLA-mismatch adversely impacted NRM, PFS, and OS. Disease status at HCT, but not histologic subtype, worsened NRM, relapse, PFS and OS. Even for patients ≥55 years, OS still approached 40% at 3 years suggesting HCT effects long-term remissions and remains underutilized in qualified older patients with NHL.

Introduction

Allogeneic hematopoietic cell transplantation (HCT) for patients with non-Hodgkin lymphoma (NHL) is increasingly used for patients with high-risk and relapsed/refractory disease¹. As over one-half of cases are diagnosed in those older than 65 years, this represents a growing population of patients for whom allogeneic HCT could provide long-term disease free survival and improve outcomes². It is postulated that conventional myeloablative conditioning prior to HCT is not feasible for the vast majority of older patients due to limited physiologic resilience and accompanying comorbidities. Hence non-myeloablative (NMA) and reduced intensity conditioning (RIC) strategies have made HCT available to less fit individuals who have relapsed or poor-risk hematologic malignancies amenable to allogeneic HCT. Recent reports show acceptable non-relapse mortality (NRM) rates of 10–20% and 2–3 year progression-free survival reported from 25–75% depending on the NHL subtype. However, data specific to older patients with NHL remains limited^3–7.

We recently examined the influence of age on outcomes in older patients with acute myeloid leukemia (AML) or myelodysplastic syndrome (MDS) in first complete remission (CR) and found similar outcomes compared to younger patients when given RIC HCT regimens⁸. In this analysis, we examined the same question in those receiving a RIC or NMA allogeneic HCT for NHL of aggressive or indolent histologies to define post-HCT outcomes in older patients and to evaluate patient, disease and treatment characteristics influencing these outcomes.

Patients and methods

Data for this analysis were submitted to the Center for International Blood and Transplant Research (CIBMTR), a voluntary working group of more than 450 transplant centers worldwide who contribute data on consecutive allogeneic HCT to a statistical center housed at the Medical College of Wisconsin in Milwaukee and the National Marrow Donor Program (NMDP) in Minneapolis. Patients are followed longitudinally with yearly follow-up. Computerized checks for errors and on-site audits of participating centers ensure data quality. Physician review of data and additional requested information from reporting centers are included. Observational studies conducted by the CIBMTR are performed with a waiver of informed consent and in compliance with Health Insurance Portability and Accountability Act (HIPAA) regulations as determined by the Institutional Review Board and the Privacy Officer of the Medical College of Wisconsin.

Patient selection

Patients 40 years or older receiving a RIC or NMA HCT between 2001–2007 for aggressive [(diffuse large B cell (n=202), mantle cell (279), immunoblastic/anaplastic B/T cell (52), peripheral T cell (60), peripheral T cell lymphoma NOS (25), Burkitt (4), other (46)] and indolent [small lymphocytic lymphoma (SLL)/chronic lymphocytic leukemia (CLL) (156), follicular (387), marginal zone (13), other (24)] NHL were included. Patients were classified as being in first (n=87) or second (231) complete remission, first (478) or second (304) partial remission (CR1/2 or PR1/2), resistant (RD, 304) disease as known prior to HCT. Grafts were from a related or unrelated donor (URD) and cord blood grafts were not studied. Patients receiving prior autografts were included.

A total of 1248 cases were identified; 668 patients with aggressive and 580 with indolent NHL were treated at 165 centers. There were 1119 patients with B-cell and 106 patients with T-cell histology (3 patients were not classifiable). Ages ranged from 40 to 75 years and were divided into 3 age cohorts for analysis: 40–54 years (n= 614); 55–64 years (n= 552) and ≥ 65 years (n= 82). Previously established criteria for donor:recipient HLA matching were used to define well-matched, partially-matched or mismatched categories⁹. Preparative regimens were classified as either RIC or NMA. RIC included any regimen consisting of: 1) ≤500 cGy total body irradiation (TBI) as a single fraction or ≤800 cGy if fractionated; 2) ≤9 mg/kg busulfan oral (or intravenous equivalent); 3) <140 mg/m2 melphalan; 4) <10mg/kg thiotepa; or 5) BEAM regimen (carmustine, etoposide, cytarabine, and melphalan)^10–11. Other regimens were classified as NMA where prompt hematopoietic recovery could reasonably be expected without transplantation within 28 days¹². T-cell depletion accomplished via ex vivo or in vivo methods was included.

Study end points and definitions

Primary outcomes were overall (OS) and progression-free (PFS) survival defined as survival from allogeneic HCT without death or without disease progression or relapse. NRM was defined as any death in the first 28 days post-transplant or any death after day 28 without documented NHL progression or relapse. All data were censored at date of last reported follow-up. Secondary endpoints included neutrophil recovery defined as time to absolute neutrophil count of ≥ 500 neutrophils/mcL sustained for three consecutive days and the cumulative incidence of acute (grades II–IV) and chronic graft-versus-host disease (GVHD) as defined by consensus criteria^13–14.

Statistical analysis

Patient-, disease-, and transplant-related variables were compared between age cohorts using the chi-square test for categorical variables and the Kruskal-Wallis test for continuous variables. Univariate probabilities of PFS and OS were calculated using the Kaplan-Meier estimator with variance estimated by Greenwood’s formula. Probabilities of neutrophil recovery, acute and chronic GVHD, NRM and relapse were calculated using cumulative incidence curves to accommodate competing risks. Ninety-five percent confidence intervals (95% CI) for all probabilities and p-values of pairwise comparisons were derived from pointwise estimates and calculated using standard techniques.

Age groups’ influence on neutrophil recovery was compared using logistic regression. Cox proportional hazards regression models were used for all other outcomes. The proportional hazards assumption was tested for all variables. Variables were stratified in the model when the proportionality assumption did not hold. Patient-, disease-, and transplant-related variables were considered in the model building procedure, but as recipient age was the main interest of this study, it was included in all steps of model building. Separate analyses of CR1 vs. CR2 and PR1 vs. PR2 patients identified no significant differences in major endpoints and these groups were combined for subsequent analysis. Patient-related variables considered were gender and Karnofsky performance status (KPS) of < 80 vs. ≥ 80. Disease-related variables included histology subtype (aggressive vs. indolent), disease status at time of HCT (CR1/2+ vs. PR1/2+ vs. refractory) and time from diagnosis to HCT (<2 years vs. ≥2 years). International prognostic index (IPI) was considered, but LDH was not reported for 69% of the patients and thus it could not be assigned. Extranodal sites and number of pre-transplant therapies were considered as well. Transplant-related variables included: year of transplant (2001–2004 vs. 2005–2007), donor/recipient CMV serostatus (−/− vs. −/+ vs. +/− vs. +/+), HLA matching (HLA matched sibling vs. well-matched URD vs. partially matched URD vs. mismatched URD), stem cell source [bone marrow vs. peripheral blood (PBSC)], GVHD prophylaxis [cyclosporine (CSA) ± methotrexate (MTX) ± other vs. tacrolimus (Tac) ± MTX ± other vs. T-cell depletion], donor-recipient gender match (M-M vs. M-F vs. F-M vs. F-F); and conditioning regimen intensity (RIC vs. NMA). Risk factors with p < 0.05 were included in the models. The potential interaction between the main effect of age and all significant covariates were examined. To analyze any possible impact of specific histology on outcomes, we divided patients into 5 histological groups (diffuse large B cell, mantle cell, SLL/CLL, follicular, and other and included them in the models for relapse/progression, PFS and OS. All computations were performed using the statistical package SAS version 9.1.

Results and outcomes

Patient Characteristics

Table 1 summarizes patient-, disease-, and transplant-related variables across age groups. Of 1248 patients, 614 (49%) were age 40–54, 552 (44%) were age 55–64 years and 82 (7%) were ≥ 65 years. Most patients were transplanted for aggressive NHL subtypes which were most often seen in the oldest cohort of patients (49% 40–54 years vs. 57% 55–64 vs. 67% ≥65, p=0.0008). The oldest group also had the lowest incidence of prior autologous HCT (26% 40–54 vs. 24% 55–64 vs. 9% ≥65 years, p=0.002). 225 patients with aggressive and 73 patients with indolent NHL had a previous autologous HCT. Though not statistically differing across the age groups, 23–30% of all patients were transplanted with refractory disease including 30% in the oldest group. Use of an HLA-matched related sibling (MRD) or URD was similar across the age groups. The remaining variables of gender, performance status, interval from diagnosis to transplant, donor-recipient sex match, donor-recipient CMV serostatus, and donor-recipient HLA match were balanced across age cohorts. GVHD prophylaxis most often included a calcineurin inhibitor with or without methotrexate, but 37–46% of patients received T-cell depletion, particularly those in the oldest cohort. Median follow-up for the 3 cohorts ranged from 47–56 months.

Table 1.

Characteristics of 1,248 patients aged ≥ 40 years receiving RIC or NMA allogeneic HCT for NHL

Patient Characteristics	40–54 N (%)	55–64 N (%)	≥ 65 N (%)	p-value
Number of patients	614	552	82
Number of centers	120	112	45
Age, median (range)	49 (40–54)	59 (55–64)	67 (65–75)	<0.0001
Male sex	406 (66)	359 (65)	64 (78)	0.06
Karnofsky score prior transplant ≥80%	512 (83)	481 (87)	71 (87)	0.24
Missing	38 (6)	30 (5)	2 (2)
Prior autologous transplant	160 (26)	131 (24)	7 (9)	0.002
NHL Histology*				0.0008
Aggressive	299 (49)	314 (57)	55 (67)
Indolent	315 (51)	238 (43)	27 (33)
Disease status at transplant				0.79
CR1/CR2+	173 (28)	160 (29)	19 (23)
PR1/PR2+	233 (38)	218 (39)	32 (39)
Resistant disease	155 (25)	126 (23)	25 (30)
Unknown or untested sensitivity	53 (9)	48 (9)	6 (7)
Interval from diagnosis to transplant,				0.12
< 2 years	217 (35)	167 (30)	31 (38)
≥ 2 years	397 (65)	385 (70)	51 (62)
Donor-recipient sex match				0.04
M-M	262 (43)	230 (42)	51 (62)
M-F	122 (20)	109 (20)	9 (11)
F-M	144 (23)	129 (23)	13 (16)
F-F	86 (14)	84 (15)	9 (11)
Donor-recipient CMV serostatus				0.41
−/−	193 (31)	142 (26)	19 (23)
−/+	171 (28)	171 (31)	24 (29)
+/+	173 (28)	170 (31)	31 (38)
+/−	60 (10)	55 (10)	6 (7)
Unknown	17 (3)	14 (3)	2 (2)
Graft source				0.53
Bone marrow	84 (14)	78 (14)	15 (18)
Peripheral blood	530 (86)	474 (86)	67 (82)
HLA Match				0.29
HLA-identical sibling	262 (43)	208 (38)	29 (35)
Unrelated donor, well-matched	222 (36)	234 (42)	39 (48)
Unrelated donor, partially matched	95 (15)	76 (14)	9 (11)
Unrelated donor, mismatched	15 (2)	17 (3)	1 (1)
Unrelated donor, missing	20 (3)	17 (3)	4 (5)
Year of transplant				<0.0001
2001–2004	363 (59)	243 (44)	36 (44)
2005–2007	251 (41)	309 (56)	46 (56)
Conditioning regimen intensity				0.05
Reduced intensity	333 (54)	260 (47)	43 (52)
Nonmyeloablative	281 (46)	292 (53)	39 (48)
Conditioning regimen				0.25
TBI≤200cGy	111 (18)	129 (23)	17 (21)
TBI>200cGy	21 (3)	22 (4)	4 (5)
Alkylator only (Cy, Mel, Bu, Thio), no TBI	443 (72)	359 (65)	57 (70)
Other	39 (6)	42 (8)	4 (5)
GVHD prophylaxis				0.43
CSA±MTX±other	205 (33)	162 (29)	19 (23)
Tac±MTX±other	179 (29)	172 (31)	25 (30)
T-cell depletion**	230 (37)	218 (39)	38 (46)
Median follow-up of survivors, months	56 (3–111)	47 (2–111)	47 (2–86)

Abbreviations: CR:complete remission; PR:partial remission; HLA:human leukocyte antigen; M:male; F:female; CMV: cytomegalovirus; Cy:cyclophosphamide; Mel:melphalan; Bu:busulfan; Thio:thiotepa; TBI: total body irradiation; CSA: cyclosporine A; MTX: methotrexate; Tac:tacrolimus.

^*Detailed in Methods: Patient selection

^**Includes in vivo (anti-thymocyte globulin or alemtuzumab) and ex vivo T-cell depletion.

Neutrophil Recovery and GVHD

Neutrophil recovery by day 28 was similar across age groups (Table 2). Multivariate analysis demonstrated that older patients (Odds ratio (OR) 0.32, 95% CI 0.14–0.75; p=0.0085, KPS<80% (OR 0.28, 95% CI 0.21–1.85; p=0.0002), and refractory disease status at HCT (OR 0.25, 95% CI 0.10–0.60; p=0.002) were associated with lower likelihood of prompt engraftment. Use of PBSC grafts vs. marrow was associated with faster neutrophil recovery (OR 3.09, 95% CI 1.64–5.85; p=0.0005).

Table 2.

Univariate analysis of outcome for all patients aged ≥ 40 years receiving RIC or NMA allogeneic HCT for NHL

Age group	40–54		55–64		≥ 65
Outcome Event	N	Prob (95%CI)	N	Prob (95%CI)	N	Prob (95%CI)	P-valuea
Neutrophil Engraftment	614		552		82
@ 28 days		96 (95–97)%		96 (94–97)%		89 (82–95)%	0.09
Acute GVHD Grades II–IV	614		552		82
@ 100 days		35 (31–39)%		34 (30–38)%		33 (23–44)%	0.92
Chronic GVHD	614		551		82
@ 3 years		56 (52–60)%		54 (49–58)%		48 (37–59)%	0.39
Non-relapse mortality	605		552		82
@ 100 days		13 (10–16)%		17 (14–20)%		18 (11–28)%	0.11
@ 1 year		22 (19–26)%		27 (23–31)%		34 (24–44)%	0.05
Progression/Relapse	605		552		82
@ 3 years		28 (24–32)%		33 (29–37)%		33 (23–44)%	0.22
Progression-free Survival	605		552		82
@ 3 years		44 (39–48)%		32 (28–36)%		27 (17–37)%	<0.0001
Overall Survival	614		552		82
@ 3 years		54 (50–58)%		40 (36–44)%		39 (28–50)%	<0.0001

^aPointwise p-value.

The incidence of grades II–IV acute GVHD (33–35% by day 100, p=0.92) and chronic GVHD (48–56% by three years, p=0.39) was similar across all age subgroups (Table 2). Multivariate analysis demonstrated no impact of age on acute or chronic GVHD. Multivariate analysis did, however, show a higher risk of acute GVHD in those not in CR at time of HCT (PR1/PR2 RR 1.37, 95% CI 1.08–1.75, p=0.01; Resistant RR 1.46, 95% CI 1.12–1.91, p=0.0005) and when using a RIC vs. a NMA regimen (RR 1.31, 95% CI 1.08–1.59; p=0.007). Use of a non-CSA GVHD prophylaxis regimen was associated with less acute GVHD (Tac±MTX±other RR 0.76, 95% CI 0.60–0.96, p=0.02; T-cell depletion RR 0.45, 95% CI 0.35–0.57, p<0.001). Compared to matched sibling HCT, URD and HLA mismatch (URD well matched 1.45, 95% CI 1.20–1.75, p=0.002; URD partially matched 1.75, 95% CI 1.35–2.26, p<0.001) and RIC conditioning regimens (RR 1.29, 95% CI 1.32–3.36; p<0.002) were all associated with higher incidence of chronic GVHD.

Non-relapse Mortality

Cumulative incidence of NRM did not differ among the age cohorts either at day 100 (13–18%; p=0.11) or 1 year (22–34%; p=0.05) (Table 2). Univariate analyses stratified by disease histology (aggressive vs. indolent) also showed no impact of age on NRM at these same time points (Table 3). As too few patients in the oldest cohort (n=7) had prior autologous HCT, we performed a separate analysis in the two younger cohorts comparing outcomes in those with or without prior autologous HCT. We observed higher 3 year NRM if patients had received an autograft before their allogeneic HCT (prior autograft 42%, 95% CI (25–31) vs. 28%, 95% CI (37–48), p<0.0001). In multivariate analysis, older age was associated with worse NRM (55–64 years RR 1.52, 95% CI 1.24–1.86, p<0.001; ≥65 years RR 1.57, 95% CI 1.08–2.29, p=0.02), although NRM was similar in the two older age cohorts (Table 4). Lower KPS, advanced disease at HCT, and less favorable HLA match also adversely impacted NRM.

Table 3.

Univariate outcomes by disease histology for patients aged ≥ 40 years receiving RIC or NMA allogeneic HCT for NHL

	40–54		55–64		≥ 65
Outcome Event	N	Prob (95%CI)	N	Prob (95%CI)	N	Prob (95%CI)	P-valuea
Aggressive NHL
Non-relapse mortality	297		314		55
@ 100 days		13 (9–17)%		16 (12–20)%		24 (13–36)%	0.15
@ 1 year		22 (17–27)%		26 (22–32)%		38 (26–51)%	0.05
Progression / Relapse	297		314		55
@ 3 years		34 (28–39)%		36 (31–42)%		29 (18–42)%	0.55
Progression-free Survival	297		314		55
@ 3 years		38 (32–44)%		30 (25–35)%		25 (14–37)%	0.04
Overall Survival	299		314		55
@ 3years		50 (44–56)%		37 (31–43%)		34 (21–47)%	0.003
Indolent NHL
Non-relapse mortality	308		238		27
@ 100 days		13 (9–17)%		18 (13–23)%		8 (1–20)%	0.10
@ 1 year		23 (18–27)%		27 (22–33)%		24 (10–42)%	0.46
Progression / Relapse	308		238		27
@ 3 years		23 (18–27)%		28 (22–34)%		41 (23–61)%	0.11
Progression-free Survival	308		238		27
@ 3 years		49 (43–55)%		35 (29–41)%		30 (13–50)%	0.003
Overall Survival	315		238		27
@ 3 years		58 (52–63)%		44 (38–51)%		50 (30–79)%	0.013

^aPointwise p-value.

Table 4.

Multivariate analysis of NRM, relapse/progression, progression-free and overall survival after RIC or NMA allogeneic HCT for NHL

		N	Relative Risk (95% CI)	p-value*	Overall p-value
Non-relapse Mortality
Age
	40–54	614			0.0002
	55–64	552	1.52 (1.24–1.86)	<.0001
	≥ 65	82	1.57 (1.08–2.29)	0.0189
Significant covariates
KPS
	≥ 80	1064			0.0001
	< 80	114	1.87 (1.40–2.50)	<.0001
NHL disease status
	CR1/CR2+	352			<.0001
	PR1/PR2+	483	1.27 (0.99–1.63)	0.0576
	Resistant disease	306	1.90 (1.45–2.49)	<.0001
HLA match
HLA identical sibling		499			<.0001
URD well matched		495	1.36 (1.07–1.71)	0.0116
URD partially matched		180	2.30 (1.74–3.03)	<.0001
URD mismatched		33	2.9 (1.76–4.77)	<.0001
Relapse
Age
	40–54	614			0.059
	55–64	552	1.29 (1.05–1.59)	0.0176
	≥ 65	82	1.18 (0.78–1.77)	0.4321
Significant covariates
NHL disease status
	CR1/CR2+	352			<.0001
	PR1/PR2+	483	1.67 (1.27–2.19)	0.0002
	Resistant disease	306	2.73 (2.04–3.64)	<.0001
GVHD prophylaxis
CSA+/−MTX+/−other		386			<.0001
Tac+/−MTX+/−other		376	0.88(0.67–1.16)	0.3607
T-cell depletion		486	1.52 (1.18–1.95)	0.0011
Progression Free Survival
Age
	40–54	614			0.0001
	55–64	552	1.37 (1.18–1.59)	<.0001
	≥ 65	82	1.34 (1.01–1.76)	0.0397
Significant covariates
KPS
	≥ 80	1064			<.0001
	< 80	114	1.63 (1.30–2.05)	<.0001
NHL disease status
	CR1/CR2+	352			<.0001
	PR1/PR2+	483	1.45 (1.21–1.75)	<.0001
	Resistant disease	306	1.45 (1.88–2.78)	<.0001
GVHD prophylaxis
CSA+/−MTX+/−other		386			0.0001
Tac+/−MTX+/−other		376	0.87 (0.72–1.06)	0.1655
T-cell depletion		486	1.26 (1.05–1.50)	0.0129
HLA match
HLA identical sibling		499			<.0001
URD well matched		495	1.13 (0.96–1.34)	0.1468
URD partially matched		180	1.39 (1.12–1.72)	0.0029
URD mismatched		33	2.28 (1.56–3.32)	<.0001
Overall Survival
Age
	40–54	614			<.0001
	55–64	552	1.47 (1.25–1.72)	<.0001
	≥ 65	82	1.47 (1.09–1.98)	0.0127
Significant covariates
NHL disease status
	CR1/CR2+	352			<.0001
	PR1/PR2+	483	1.29 (1.06–1.58)	0.0113
	Resistant disease	306	1.97 (1.60–2.44)	<.0001
KPS
	≥ 80	1064			<.0001
	< 80	114	1.87 (1.48–2.37)	<.0001
HLA match
HLA identical sibling		499			<.0001
URD well matched		495	1.30 (1.09–1.56)	0.0043
URD partially matched		180	1.90 (1.52–2.38)	<.0001
URD mismatched		33	2.21 (1.48–3.30)	0.0001
Conditioning Regimen
Nonmyeloablative		612			0.03
Reduced Intensity		636	1.19 (1.02–1.39)	0.03

Abbreviations: KPS:Karnofsky performance status; CR:complete remission; PR:partial remission; HLA:human leukocyte antigen; URD:unrelated donor; GVHD:graft versus host disease; CSA:cyclosporine A; MTX:methotrexate; Tac:tacrolimus.

^*compared to reference group

Relapse and Progression

Relapse and progression rates at 3 years were similar across all age groups and were confirmed in a separate analysis stratified by disease histology (Tables 2, 3; Figures 1A, 1C). In the younger two cohorts, relapse incidence was similar for patients receiving prior autologous HCT (both groups had 30% relapse at 3 years, p=0.88). Multivariate analysis demonstrated no significant impact of age, but the risk of progression or relapse worsened with advanced disease at HCT and by GVHD prophylaxis. T-cell depletion was associated with a significantly increased risk of progression or relapse (RR 1.52, 95% CI 1.18–1.95; p=0.001). The number of extramedullary sites involved and number of pre-transplant therapies did not influence the risks of relapse (p>0.1).

Figure 1. Three year relapse and progression-free survival based on histology.

(A) Progression/Relapse-Aggressive NHL

(B) Progression-free Survival-Aggressive NHL

(C) Progression/Relapse-Indolent NHL

(D) Progression-free Survival-Indolent NHL

Progression-Free and Overall Survival

In univariate analysis, PFS at 3 years was highest in the youngest age group, but there was no difference in PFS between the two older cohorts (Table 2, Figure 2A). PFS also differed between age cohorts when patients were stratified by aggressive/indolent disease histology (Table 3, Figures 1B, 1D, 2A). Multivariate analysis (Table 4) showed that older age (≥55 and >65 years), lower KPS (<80%), disease status other than CR1/CR2, GVHD prophylaxis and greater HLA mismatch were all associated with inferior PFS.

Figure 2. Three year progression-free survival and overall survival following allogeneic HCT for NHL.

(A) Progression-free Survival

(B) Overall Survival

Three year OS also differed significantly across the age groups (Table 2, Figure 2B) though 39% of even the oldest cohort survive beyond 3 years. In the two younger cohorts, a prior autologous HCT was associated with lower 3 year OS (prior autograft 35%, 95% CI (30–41) vs. 52% (48–55), p<0.0001). In multivariate analysis, older age was associated with worse OS as was disease status at allogeneic HCT, KPS<80, HLA mismatch and use of a RIC regimen (Table 4). Following HCT, the primary causes of death were similar in the 3 age cohorts: relapse (33–37%), infection (17–21%), GVHD (14–17%), and organ failure (13–15%).

To examine the effect of histology on outcomes, we assessed the impact of five specific histologic subgroups on the incidence of relapse/progression, PFS and OS. In multivariate analysis we found no significant influence of these histologic subgroups on any outcome (all p values >0.1). Additionally, we saw no significant differences in any outcome whether patients were transplanted for a B or a T-cell NHL subtype (all p values>0.3). Neither extramedullary sites involved or number of pre-transplant therapies influence the risks of PFS or OS (p>0.1).

Discussion

We examined the impact of age in a large group of older patients receiving RIC or NMA allogeneic HCT for NHL. In this sizeable cohort, we found that only a modest number of patients over the age of 65 years received an allogeneic HCT for their disease. While age had a modest adverse effect in patients older than 55 compared to those 40–54 years, outcomes of patients age 55–64 and ≥65 years were equivalent with no significant differences in NRM, relapse, PFS or OS. Older age also did not influence the incidence of acute or chronic GHVD, major complications which might be tolerated poorly by older patients.

It is not surprising that HLA disparity, poorer performance status, T-cell depletion and advanced disease status at time of transplant adversely affected major HCT outcomes as each have been reported to have prognostic implications^9,16–20. Aggressive NHL is also generally associated with worse outcomes^21–22. Although histology subgroup had no significant effect on any HCT outcome in multivariate analysis, 67% of all patients in the ≥65 years and 57% in the 55–64 years age group had aggressive NHL. Small numbers of patients in the oldest age groups may have limited these analyses. The oldest cohort also rarely had prior autologous HCTs which precluded their inclusion for the effect of prior autologous HCT. In the two younger cohorts, however, we observed an adverse effect of prior autologous HCT on both NRM and OS. Other reports also demonstrate a modest negative influence of prior autologous HCT on subsequent allogeneic HCT. In two small series, Thomson et al. showed no influence of prior autologous HCT on outcomes²³, while Rodriguez et al. showed that prior autologous HCT was associated with worse relapse and OS with a trend toward worse NRM²⁴. Since autologous HCT remains the initial HCT option for most patients with NHL, the impact of prior autografting directly influences the selection of patients for allografting.

Even for patients with advanced NHL, the observed outcomes of RIC or NMA allogeneic HCT were encouraging. This is important as patients who relapse after autologous HCT have very poor median survivals reported between 3–8 months depending on histologic type of NHL^25–27. In mantle cell lymphoma, a median survival of just 23 months in patients relapsed post-autologous HCT was seen with survival of only 6 months if relapse occurred within 1 year after HCT.²⁸ In another report, relapsed/refractory diffuse large B cell lymphoma patients, 36% of patients receiving second line salvage therapy failed to respond and responses lasted only a median duration of only 4 months.²⁹ Survival after relapse following autografting using an age-adjusted international prognostic index (IPI) showed PFS and OS rates of only 16% and 18% at 4 years even in those with chemo-sensitive disease²⁷. In the current analysis, we observed 3 year survival of 39%, even in the oldest age group, suggesting that select older patients can benefit from allogeneic HCT and can have extended survival unattainable with other salvage approaches.

We were surprised at how few patients ≥65 years underwent allogeneic HCT even though over half of all NHL cases occur in patients ≥60 years³⁰. Similar to observations in AML/MDS, most patients eligible for allogeneic HCT procedures are not referred and never receive such therapy³¹. Contemporary registry analyses from the EBMT and the CIBMTR examining relapsed high grade lymphomas emphasize the curative potential of this procedure^32–33, yet recognize the rarity of its use. Thus HCT is uncommonly applied despite recent reports of improved HCT outcomes for patients in the modern era^34–37.

We note the limitations of our study due to heterogeneity of pre-HCT therapies and no direct data to clarify the medical decision-making in selecting patients for allogeneic HCT. Detailed comorbidity information other than KPS was not available for this study though it may directly inform patient selection by physicians considering allogeneic HCT. Although some appropriate clinical selection bias exists in choosing only the fittest older patients for an allograft, these promising outcomes suggest that careful pre-transplantation judgments can identify patients able to tolerate this curative therapeutic approach.

Though our results suggest that outcome differences are modestly influenced by age, even the oldest patients had encouraging outcomes. Attention to disease stage, HLA disparity, and performance status prior to HCT could further improve these results by identification of NHL patients most likely to benefit from a RIC allogeneic HCT. Our analysis supports the referral of older patients with NHL for allogeneic transplantation.