Pentostatin Therapy for Steroid-Refractory Acute Graft Versus Host Disease: Identifying Those Who May Benefit

Abstract

We report outcomes of 60 patients with steroid-refractory (SR)-aGVHD treated with pentostatin. Almost half (47%) of patients had grade 4 GVHD - 22% had stage 3-4 liver GVHD and 51% had stage 3-4 lower gastrointestinal tract (LGI) GVHD. Patients received a median of 3 courses (range, 1-9) of pentostatin. Day 28 overall response rate (ORR) was 33% (n=20) (complete response 18% (n=11), partial response 15% (n=9)). Non-relapse mortality was 72% (95% confidence interval (CI) 61-84%) and overall survival (OS) was 21% (95% CI 12- 32%) at 18 months. On univariate analysis, age >60 years (HR 1.9, 95% CI 1.01-3.7, p=0.045) and presence of liver GVHD (HR 1.9, 95% CI 1.9, 95% CI 1.5-3.3, p=0.03) were significant predictors of poor OS while patients with LGI GVHD had superior OS than those without (HR 0.4, 95% CI 0.2-0.8, p=0.01). On stratified analysis, patients <60 years with isolated LGI GVHD had the best outcomes with an ORR of 48% and OS of 42% at 18 months. Among older patients, OS was 14% in those with isolated LGI aGVHD and 0% in others. Pentostatin remains a viable treatment option for SR-aGVHD, especially in patients 60 years or younger with isolated LGI involvement.

INTRODUCTION

The most common and serious complication of allogeneic hematopoietic stem cell transplant (HSCT) is acute graft versus host disease (aGVHD). Typically manifesting within the first 100 days following HSCT, multiple organs can be affected including the skin, gastrointestinal tract, and/or liver. Systemic steroids such as prednisone or methylprednisolone at a dose of 1-2 mg/kg ideal body weight are offered as initial therapy for patients with grades II-IV aGVHD.¹ However, only approximately one-third to half of these patients will respond to initial steroid therapy depending on the grade and organs involved.^2-6 Roughly half of the patients who do not respond to initial steroid therapy respond to second-line therapy but their mortality rates approximate 70%.⁶

For patients with steroid-refractory (SR) aGVHD, various additional immunosuppressive medications have been tried with no therapy of proven superiority over others.⁶ Pentostatin, a potent adenosine deaminase (ADA) inhibitor has also been studied for GVHD prevention, therapy of newly diagnosed aGVHD, and SR- acute and chronic GVHD.^7-11 However, previous reports on the role of pentostatin in SR-aGVHD have been limited by short term follow-up and small sample sizes.^{10, 12-15}

Although patients with SR-aGVHD generally have dismal outcomes, long-term survival does occur in select cases. In this study, we sought to determine the factors associated with improved responses, early mortality and long term survival after pentostatin use in patients with SR-aGVHD.

METHODS

Patient Selection

All patients who received pentostatin for the treatment of SR-aGVHD from January 2006 to December 2014 at M.D. Anderson Cancer Center (MDACC) were reviewed. Criteria for inclusion in this analysis included receiving at least one dose of pentostatin as second- or third-line treatment following progression or failure to respond to initial therapy with steroids. Patients who received pentostatin solely for GVHD prevention (prophylaxis study) or on a clinical trial for initial therapy of aGVHD were excluded from this analysis.

Variables of Interest and Data Source

Patient characteristics were retrieved from the departmental database and verified by chart review, including age, gender, indication for transplant, conditioning regimen, graft source, GVHD prophylaxis regimen, aGVHD start date, organ stage, overall grade and preceding and concurrent GVHD therapy at time of administration of pentostatin.

Objectives

Our objectives in this analysis were to (a) determine response rate to pentostatin received in the second or third line setting following failure to respond to frontline steroids for aGVHD, (b) evaluate overall survival (OS) and non-relapse mortality (NRM), and (c) determine predictors of response, early mortality and OS.

Treatment Regimen

All patients received pentostatin at a dose of at 1.5 mg/m² on days 1-3 (defined as one course), repeated every two weeks as indicated. Upon initiation of pentostatin, patients were usually continued on systemic steroids, usually at a tapered dose as clinically warranted, as well as concomitant GVHD therapy which they were already receiving. These agents included tacrolimus, cyclosporine, mycophenolate mofetil (MMF), extracorporeal photopheresis (ECP) or their combination.

Response Assessment

All patients underwent pathological evaluation for the diagnosis of GVHD. In case of discrepancy between pathological and clinical diagnosis, clinical criteria were used. Acute GVHD was graded per Glucksberg criteria¹⁶ and refers to clinical (not pathological) grading and/or staging. Steroid refractory aGVHD was defined as progression of aGVHD after 72 hours of initiation of steroids or no improvement after 7 days. The response to pentostatin was determined at 28 and 56 days from the first day of administration of pentostatin. Complete response (CR) was defined as complete resolution in all organs without development of new organ involvement. A partial response (PR) was defined as improvement in one GVHD organ by one GVHD stage without worsening in any additional organs. A mixed response was defined as improvement (PR or CR) in one organ with progression in a second organ by at least one clinical stage. No response was defined as no improvement or worsening in any organ. Progression was defined as worsening by one stage in one or more organs without improvement in any additional organ or requiring new, additional agents to control GVHD. Overall response rate (ORR) was defined as CR plus PR. Patients who died or had disease progression before day 28 were considered non-responders.

Statistical Analysis

Descriptive analyses were performed to summarize clinical and demographic characteristics of the patient population. The cumulative incidence of NRM was determined using relapse or death in relapse as competing risks. Actuarial OS was estimated using the Kaplan-Meier method. Cox’s proportional hazards regression analysis was used on univariate and multivariate analysis to evaluate predictors of survival after the initiation of pentostatin. Competing risk regression analysis (considering progression of underlying malignancy before day 28 as competing risk) was used to evaluate predictors of day 28 response after the initiation of pentostatin. Factors that were significant on univariate analysis were considered in multivariate analysis. Statistical significance was set at the 0.05 level. Statistical analyses were performed using primarily STATA 14 (StataCorp. 2015. Stata Statistical Software: Release 14. College Station, TX: StataCorp L). All outcomes were measured from the first day of administration of pentostatin.

RESULTS

Patient Characteristics

A total of 60 patients received pentostatin as second (n=22; 36.7%) or third line (n=38; 63.3%) treatment for SR-aGVHD. Table 1 demonstrates baseline characteristics for the study population. Median age at the time of start of pentostatin was 52 years (range, 2-70). Two patients were younger than 18 years and about a quarter (n=14) were over the age of 60 years. A majority of patients (63%) received myeloablative conditioning and granulocyte-colony stimulating factor (G-CSF) mobilized peripheral blood progenitor cells as the graft source (77%). Tacrolimus and methotrexate was the most common GVHD prophylaxis regimen used (75%), followed by tacrolimus plus MMF (22%). The median time to start of pentostatin was 69 days post-transplant (range, 27-595) and 15 days (range, 4-172) from the start of steroids. Patients received a median of 3 cycles of pentostatin (range, 1-9). All patients received additional concurrent GVHD therapy at the start of pentostatin, with tacrolimus (n=28) and tacrolimus plus MMF (n=21) being the most common drugs. Nine patients were also on ECP. A majority of patients had lower gastrointestinal tract (LGI) GVHD (80%, n=47) and liver was involved in 44% (n=26), which was stage 3-4 in half of the cases. Consequently, 82% (n=49) of patients had an overall grade 3-4 GVHD. Approximately half (n=31) of the population had an infection at the time of pentostatin initiation, including viral infections (n=16) such as BK cystitis, HHV-6, adenovirus and cytomegalovirus reactivation/disease, fungal sinusitis/pneumonia (n=2), bacterial infections (n=7) and others. Two patients were in the Intensive Care Unit when pentostatin was first administered, detailing the severity of illness in this patient population. The median follow-up in survivors was 19 months (range 7-77).

Table 1.

Baseline characteristics at first dose of pentostatin.

	N=60	(%)
Age at pentostatin, median (range), in years	52 (2-70)
Gender mismatched
Female-to-male	25	41.7%
Diagnosis
AML/MDS	14	23%
ALL	13	22%
Lymphoma	14	23%
CLL	10	17%
Other	9	15%
Graft source
Cord Blood	4	7%
Peripheral Blood	46	77%
Bone Marrow	10	17%
Donor type
HLA-matched related	19	31.7%
HLA-matched unrelated	35	58.3%
Haploidentical	3	5%
Cord blood	3	5%
Preparative Regimen Intensity
Non-myeloablative	22	37%
Myeloablative	38	63%
GVHD prophylaxis
Tacrolimus/Methotrexate	45	75%
Tacrolimus/MMF +/− PT Cy	13	21.7%
PT Cy	2	3.3%
Prophylactic in vivo T cell depletion
Anti-thymocyte globulin	30	50%
Alemtuzumab	5	8.33%
Time to pentostatin after transplant, median days	69
(range)	(27-595)
Time to pentostatin after first line Steroid, median days	15
(range)	(4-172)
Pentostatin Line of Therapy
Second	22	37%
Third	38	63%
2ndline treatment (in patients who received pentostatin as 3rd line)
MMF	17	45%
Photopheresis	19	50%
Infliximab	1	3%
Cyclosporine	1	3%
Cycles of Pentostatin, median	3
(range)	(1-9)
Concurrent GVHD Therapy
Tacrolimus	28	46%
Tacrolimus/MMF	21	35%
Cyclosporine	1	2%
Cyclosporine/MMF	1	2%
Tacrolimus/Photopheresis	9	15%
GVHD overall grade at time of initiation of pentostatin
2	11	18%
3	21	35%
4	28	47%
Skin GVHD Stage	N=12
1-2	9	15%
3	3	5%
UGI GVHD Stage	N=27
1	27	45%
LGI GVHD Stage	N=47
1-2	16	27%
3-4	31	51%
Liver GVHD Stage	N=26
1-2	13	22%
3-4	13	22%
GVHD organ combination
LGI only	28	46.7%
Liver only	10	16.7%
LGI + Liver	10	16.7%
LGI + Liver + Skin	6	10%
Skin + LGI	4	6.7%
Skin + Liver	1	1.7%
Skin only	1	1.7%
Follow-up among survivors, median (range), months	19 (7-77)

Abbreviations: ALL, acute lymphoblastic leukemia; AML, acute myeloid leukemia; CLL, chronic lymphocytic leukemia; HLA, Human Leucocyte Antigen; LGI, Lower Gastrointestinal tract; MDS, myelodysplastic syndrome; MMF, mycophenolate mofetil; PT Cy, Post-transplant cyclophosphamide; UGI Upper Gastrointestinal tract.

Start of pentostatin within 10 days of steroids improves responses

A total of 24 (40%) patients died (n=22) before day 28 and were considered non-responders; 2 patients had progression of the underlying malignancy (competing risk) and were considered non-evaluable. In evaluable patients, ORR at day 28 was 33.3% (n=20) including CR in 11 patients (18.3%) and PR in 9 (15%) [Table 2]. In univariate analysis assessing the predictors of overall response at day 28, we found that initiation of pentostatin within 10 days following start of steroids was the only significant factor (Hazard ratio (HR) 2.2, 95% confidence interval (CI) 1.2-4.3, p=0.02). The ORR was 50% in those who received pentostatin within 10 days of steroids (n=20) as compared to 25% in those who were started on pentostatin more than 10 days after the start of steroids (n=40). As time to start of pentostatin was the only significant factor in univariate analysis, multivariate analysis was not performed.

Table 2.

Responses to Pentostatin at day 28 and 56.

				Day 28		Day 56
				N=60	(%)	N= 60	(%)
Complete Response				11	18%	16	27%
Partial Response				9	15%	3	5%
No Response				11	18%	4	7%
Progression				5	9%	6	10%
Died prior to index day				22	37%	31	52%
Progression of malignancy				2	3%	0	0%
Day 56 response
Day 28 response		Complete Response	Partial Response	No response	Progression	N.E.	Total
	Complete Response	9	1	0	1	0	11
	Partial Response	3	2	3	1	0	9
	No response	4	0	1	2	4	11
	Progression	0	0	0	2	3	5
	N.E.	0	0	0	0	24	24
	Total	16	3	4	6	31	60

N.E., non-evaluable

Overall survival is poor and non-relapse mortality is high in patients with SR-aGVHD

The univariate estimates of NRM were 37% (95% CI 26-51) at day 30, 63% (95% CI 52-77) at 6 months and 72% (95% CI, 61-84) at 18 months. The univariate estimates of OS were 62% (95% CI 48- 73%) at day 30, 30% (95% CI 19-42) at 6 months and 21% (95% CI 12- 32%) at 18 months. Only 10 patients were alive at the end of the study period. Major causes of death included aGVHD (n= 23), chronic GVHD (n=16), infection (n=4), recurrence or persistent of underlying malignancy (n=4), second malignancy (n=1), graft rejection or failure (n=1), and multi-organ failure (n=1).

Age <60 years and presence of LGI GVHD are significant predictors of survival after pentostatin

Next, we explored factors associated with a risk of early mortality after starting pentostatin and factors associated with long-term survival. In univariate analysis, two factors that emerged as significant predictors of OS at day 30 [Supplemental table 1] and at 18 months [Table 3] were age and the type of organ involved. Patients older than 60 years had significantly poor OS compared with younger patients (HR 1.9, 95% CI 1.01-3.7, p=0.045, at 18 months). [Fig 1] Day 28 ORR and OS at 18-months in patients with isolated LGI GVHD (43% and 35%, respectively) were similar to that in patients with liver plus LGI GVHD (50% and 20%, respectively), while patients with isolated liver GVHD (10% and 10%, respectively) or any other GVHD (17% and 8%, respectively) had inferior responses and survival. [Table 4].

Table 3.

Univariate analysis for overall survival at 18 months.

	n	Hazard Ratio	95% confidence interval	P-value
First line steroids to Pentostatin, days
<=10 days	20	Reference
>10 days	40	1.1	0.6-2.0	0.8
Pentostatin used as
second line therapy	22	Reference
third line therapy	38	0.9	0.5-1.7	0.8
Age at Pentostatin
<=60 years	46	Reference
>60 years	14	1.9	1.01-3.7	0.045
Female-to-male gender Mismatch
No	35	Reference
Yes	25	0.7	0.2-1.9	0.4
Donor type
HLA-matched related	19	0.7	0.2-2.2	0.5
HLA-matched unrelated	35
Haploidentical	3
Cord blood	3
Graft source
Bone marrow	10	Reference
Peripheral blood	46	1.3	0.6-2.9	0.5
Cord blood	4	Excluded
Conditioning
Myeloablative	38	Reference
Non-myeloablative	22	1.4	0.8-2.6	0.2
Overall grade (at diagnosis)
Grade 2	11	Reference
Grade 3-4	49	1.9	0.9-4.3	0.1
GVHD organs
Skin (Reference no skin)	12	2	1.02-4.1	0.04
UGI (Reference no UGI)	27	0.95	0.5-1.7	0.9
LGI (Reference no LGI)	47	0.4	0.2-0.8	0.01
Liver (Reference no liver)	26	1.9	1.05-3.3	0.03
GVHD organ combination
LGI only	28	Reference
Liver only	10	3.3	1.4-7.3	0.004
LGI + Liver	10	1.4	0.6-3.3	0.4
LGI + Liver + Skin	6	3.1	1.02-9.6	0.05
Skin + LGI	4	2.5	0.96-6.3	0.06
Skin only	1	Excluded
Skin + Liver	1	Excluded

Figure 1:

Overall survival by age

Table 4:

Day 28 overall response rate, overall survival at day 30 and overall survival at 18 months by risk factors

	N	Day 28		Day 30				18 months
		% CR/PR	P-value	% OS	HR	95% CI	P-value	% OS	HR	95% CI	P-value
ALL PATIENTS
LGI only	28	43%	0.01 $	75%	Reference			35%	Reference
Liver + LGI	10	50%		60%	2.1	0.6-7.1	0.2	20%	1.4	0.6-3.2	0.4
Liver only	10	10%		30%	4.9	1.7-14	0.003	10%	3.3	1.5-7.4	0.004
All others *	12	17%		58%	1.7	0.5-5.5	0.3	8%	2.8	1.3-5.9	0.009
YOUNGER PATIENTS ONLY (Age ≤60 years)
LGI only	21	48%	0.04 $	86%	Reference			42%	Reference
Liver + LGI	8	50%		62%	3.6	0.7-18	0.1	25%	1.6	0.6-4.4	0.3
Liver only	7	14%		43%	6.0	1.3-27	0.02	14%	2.9	1.1-8	0.03
All Others **	10	20%		50%	4.2	0.9-18	0.05	0%	3.8	1.6-9.3	0.003
OLDER PATIENTS ONLY (Age ≥ 60 years)
LGI only	7	29%	0.50	43%	0.7	0.2-2.9	0.7	14%	0.6	0.2-1.8	0.4
All Others ***	7	14%		43%				0%

Abbreviations: CI, confidence interval; CR, complete response; HR, hazard ratio; LGI, lower gastrointestinal tract; PR, partial response

^*includes skin only (n=1), liver + skin (n=1), LGI + skin (n=4), liver + LGI + skin (n=6)

^**includes liver + skin (n=1), LGI + skin (n=4), liver + LGI + skin (n=5)

^***includes skin only (n=1), liver only (n=3), liver + LGI (n=2), liver + LGI + skin (n=1)

^$p-value denotes difference between “LGI only or Liver + LGI” versus other groups.

As age and the type of organ involved were the only factors that predicted survival, we performed further analyses stratified by these two factors. In patients 60 years or younger, OS at 18 months was 42% in those with isolated LGI GVHD, 25% in patients with liver plus LGI GVHD and 14% in those with isolated liver GVHD. There were no survivors with any other organ combination at 18 months. [Table 4, Fig 2a] Among older patients, 14% of patients with isolated LGI GVHD were alive at 18 months compared with 0% with any other GVHD. [Table 4, Fig 2b]

Figure 2:

Overall survival stratified by age and GVHD organ

Causes of death by age group are summarized in supplemental table 2. Organ toxicities and infections observed at day 56, which may be related to GVHD and/or treatment, are summarized in supplemental table 3.

DISCUSSION

We present the largest retrospective series of patients who received pentostatin for SR-aGVHD. Our findings suggest that there is a specific subset of patients that benefits from pentostatin while a specific group is destined for early mortality following its use. Younger patients (age ≥ 60 years) with isolated LGI GVHD (day 28 ORR 48%, OS 42% at 18 months) or (to a lesser extent) liver plus LGI GVHD (day 28 ORR 50%, OS 25% at 18 months) appear to benefit the most from pentostatin. In contrast, patients older than 60 years uniformly had poor survival, suggesting pentostatin should be used with caution in this population. It is likely that co-morbidities, frailty and the toxicities of prior GVHD therapies (namely steroids) in older aged patients make it challenging for positive outcomes. Even in those older patients who responded (29% of those with isolated lower GI) survival was poor (14% OS at 18 months). We also found that although early initiation of pentostatin within 10 days of steroids improved ORR, it did not impact survival. Response to treatment with early initiation of therapy and superior outcomes in younger patients may be related to the biology of the disease and the host characteristics rather than pentostatin, as is true for most GVHD therapies. Moreover, survival was not influenced by the overall GVHD grade, the type of conditioning, graft source or if pentostatin was used as a second line or third line agent, highlighting the general poor prognosis of patients who fail to respond to steroids.

Given the dismal prognosis of patients with SR–aGVHD, multiple strategies either alone or in combination have been explored in several distinct series in the past, including the use of ATG,^17-22 IL-2 receptor antagonists (daclizumab, basiliximab, inolimomab),^23-32 denileukin diftitox,^{7, 33, 34} TNF-α receptor antagonists (infliximab, etanercept), ^{31, 35-40} alemtuzumab,^41-45 sirolimus,^{46, 47} ECP, ^48-53 MMF ^{7, 54-57} and pentostatin ^{9, 11, 13-15, 58} among others. In a comprehensive recent review, Martin el at ⁶ described outcomes after various drugs used for SR–aGVHD and provided benchmarks to assess response and survival for future trials. Combining 25 to 29 different studies, the authors reported an aggregated CR rate of 32%, an aggregated ORR of 58% and an estimated 6 months weighted average survival of 49%.⁶ Of note, none of these studies used pentostatin.

The role of pentostatin as a treatment for SR-aGVHD in adults was evaluated in 5 studies [Table 5]. These studies vary remarkably not only in the patient population, but also in terms of the dose and the number of pentostatin cycles used, additional therapies for GVHD, continuation of steroids, definition of SR–aGVHD and the assessment of response. For instance, two studies that reported a CR rate of 64-70% assessed “best response”,^{10, 12} while other studies that assessed either day 28 response or durable response lasting 4 weeks reported CR of 13-33% and PR of 13-17%.^13-15 Nevertheless, one common denominator across all studies was an extremely poor prognosis of these patients with a median survival of less than 3 months. In our study, where almost half (47%) of patients had an overall grade 4 SR-aGVHD, 44% had liver involvement, over 50% had documented infection at the start of pentostatin and a majority (63%) received pentostatin as a third line treatment, we observed modest responses with CR rate of 20%, CR/PR rate of 33% and OS of 30% (95% CI 19-42) at 6 months.

Table 5:

Studies evaluating the role of pentostatin in adults with steroid refractory acute GVHD

Reference	Sampl e size	Dose of pentostatin	Number of cycles	Liver GVHD	Median time from GVHD to pentostati n use, days (range)	Median time from steroid initiation to pentostatin use, days (range)	Response assessment	Complete response	Partial response	Overall survival
Bolaños-Meade et al	23	1 mg/m2 (n= 10), 1.5 mg/m2 (n= 5), 2 mg/m2 (n= 7) 3 mg/m2 (n=1).	1 (74%) 2 (26%)	10 (43%)	34 (5–140)	-	Best response assessed weekly till day 28	64%	30%	Median 85 days from therapy
Pidala et al	12	1 mg/m2 (n=2) 1.5 mg/m2 (n=2) 2 mg/m2 (n=3) 4 mg/m2 (n=5)	2 (range; 1-8)	5 (42%)	-	42 (13–213)	durable, sustained for 4 weeks	33%	17%	Median 1.4 months from therapy
Alam et al.	15	1.4 mg/m2	1 (7%) 2 (20%) 3 (73%)	6 (40%)	33 (7–292)	-	durable, sustained for 4 weeks	13%	20%	Median 82 days from SCT
Klein et al.	23	1 mg/m2	1 (57%) 2 (22%) 3 (17%) 4 (4%)	5 (22%)	-	-	Best response assessed weekly till day 28	70%	13%	Median 85 days from therapy
Schmitt el at	24	1 mg/m2	1 (67%) 2 (23%)	6 (25%)	78 (5-211)	10 (3-42)	Day 28	21% (in GI GVHD symptoms)	17% “VGPR” (in GI GVHD symptoms)	6 months OS 25%; 2 year OS 17%
Current study	60	1.5 mg/m2	3 (range; 1–9)	26 (44%)	69 (27–5 95)	15 (4–172)	Day 28	18%	15%	Median OS: 55 days

Pentostatin inhibits cellular ADA which is an essential enzyme for the metabolism of purines. As this mechanism of action is distinct from that of other GVHD drugs, pentostatin offers a rationale non-overlapping treatment strategy for patients with SR-aGVHD. However, it is highly myelo- and lympho- suppressive, has high rates of infection and NRM. One of the main reasons for dismal prognosis in these patients is high risk of infections due to significant T- lymphopenia caused by pentostatin. However, in our study, despite the presence of infection in over half of our patients at the start of pentostatin, only 4 patients had an infection-related mortality. This is harmonious with pre-clinical data suggesting that although pentostatin can interfere with the induction of new humoral responses (after exposure to drug), it may actually enhance already established humoral responses (before exposure to drug). This can also be explained by the differential impact of pentostatin on various immune subsets –relative sparing and increased activity of natural killer (NK) cells, helper T cells, antigen presenting B-cells and increased phagocytic activity of macrophages that may improve outcomes of certain infections.^59-63 As such, presence of an active infection at the start of pentostatin should not be considered as an absolute contraindication to its use. Yet, exposure to any immunosuppressive agent in already notably immunocompromised GVHD patients does carry significant risk of infections and aggressive monitoring and treatment of infections is of paramount importance.

We acknowledge limitations of our study most of which are inherent in its retrospective nature. As anticipated in the absence of established standard of care, patients received different lines of rescue therapies before or with pentostatin. Patients with LGI plus liver GVHD had better responses (but not survival) than those with isolated liver GVHD. This finding is rather perplexing and cannot be explained from this exploratory analysis, but may possibly be related to differential mechanism of action of pentostatin on different organs. Also, given the complicated course of patients with SR-aGVHD who are commonly affected by multiple complications that contribute to NRM, it is difficult to detach the side-effects of GVHD treatment from that of GVHD itself. Moreover, despite being the largest analysis of the use if pentostatin for SR-aGVHD, this is a single center experience. Evaluating patients across multiple institutions that employ pentostatin in this setting would further elucidate characteristics and outcomes in this setting.

CONCLUSION

In summary, patients younger than 60 years with isolated LGI aGVHD had the best outcomes with pentostatin for SR-aGVHD. However, the overall long term outcome of these patients remains dismal. All patients should be encouraged to participate in clinical trials assessing the role of novel agents for treatment of SR-aGVHD. Many of these drugs have shown encouraging results in preliminary reports and are under active investigation, such as the JAK1/2 inhibitor ruxolitinib,⁶⁴ (ClinicalTrials.gov Identifier NCT02953678, NCT02913261), monoclonal antibody against CD26 on CD4+ T lymphocytes - Begelomab (NCT02411084), monoclonal antibody against integrin α4β7 (LPAM-1, lymphocyte Peyer's patch adhesion molecule 1) - Vedolizumab (NCT02993783), ultra-low dose IL-2 (NCT00529035) and the infusion of mesenchymal stem cells (NCT00603330, NCT02770430), to name a few. In the absence of clinical trial, pentostatin remains as one of the therapeutic options especially in patients 60 years old or younger with steroid-refractory isolated LGI aGVHD.