Sirolimus as Primary Treatment of Acute Graft-versus-Host Disease following Allogeneic Hematopoietic Cell Transplantation

Abstract

Glucocorticoids have gone unchallenged as an essential component of primary therapy for acute graft-versus-host disease (aGVHD) after allogeneic hematopoietic cell transplantation (HCT) despite limited complete response rates and adverse effects from this therapy. The role for alternate immunosuppressive agents as primary aGHVD treatment remains unexamined. In a series of 10 patients at high risk for corticosteroid toxicity or leukemia relapse who developed biopsy-proven grade II–III aGVHD after hematopoietic cell transplantation, we report that primary therapy with sirolimus resulted in durable complete remission of aGVHD in 5 (50%) without requirement for glucocorticoids. Mild chronic GVHD (cGVHD) developed in 4 (40%). Projected overall survival (OS) at 18 months is 79% (95% confidence interval [CI]: 38.1%–94.3%), and projected relapse-free survival (RFS) at 15 months is 70% (95% CI: 32.9%–89.2%). Sirolimus was well tolerated with mild and reversible thrombotic microangiopathy occurring in 2 patients. This experience provides preliminary evidence for the efficacy of sirolimus as a sole primary therapy in the treatment of aGVHD.

INTRODUCTION

Acute graft-versus-host disease (aGVHD) is an important complication of allogeneic hematopoietic cell transplantation (HCT). The historic first-line therapy for aGVHD has included 1–2 mg/kg of prednisone or equivalent dose of other glucocorticoids. Unfortunately, the complete response (CR) rate to this therapy is only 30% to 40% in several published series, with non-responders going on to additional immunosuppressive therapies for steroid-refractory disease [1–8]. Although previous attempts at combination therapy with additional agents added to glucocorticoids have produced mixed results [9,10], early reports from a CTN trial have suggested improved aGHVD response rates to a regimen of glucocorticoids and mycophenolate mofetil (MMF) [11]. In published series, overall survival (OS) for those with steroid responsive disease approaches 50% to 60%, with nonresponders realizing much worse outcomes because of competing threats from refractory aGVHD, toxicity, cumulative immunosuppression from additional therapies, primary disease relapse, and infectious complications. Additionally, up to 70% of patients will develop chronic GVHD (cGVHD). In the treatment of aGVHD and cGVHD, patients suffer numerous complications from glucocorticoids.

The primacy of glucocorticoids in the management of aGVHD has gone unchallenged. Conversely, alternative therapies that avoid complications of steroid exposure may offer promise for improved outcomes. Sirolimus exerts its immunosuppressive effect through inhibition of mTOR, or mammalian target of rapamycin, and by downstream effects that include inhibition of transcription and decreased kinase activity of cyclin enzymes involved in cell cycle progression; other postulated effects include inhibition of dendritic cell development and function, blockade of CD28-mediated costimulatory signaling on effector T cells, and a permissive effect on regulatory T cell expansion, proliferation, and survival [12–15]. Sirolimus has shown efficacy in the prevention [16,17] and treatment [18] of aGVHD, but some physicians find intolerable the risks of thrombotic microangiopathy (TMA) and hepatic veno-occlusive disease (VOD) in primary transplants. Sirolimus has not been examined as a sole primary therapy for aGVHD.

METHODS

A series of 10 recipients of HCT who developed aGVHD were treated with sirolimus as primary therapy; all cases of GVHD were biopsy confirmed. Primary treatment with glucocorticoids was avoided for the intolerance in older patients, and sirolimus was selected for its dual activity as immunosuppressant and anticancer drug for mitigating the exceedingly high risk for leukemia relapse in patients with active disease at the time of transplant. In all patients, tacrolimus target serum levels was decreased to 3–7 ng/mL while receiving concomitant sirolimus as an attempt to prevent TMA. Sirolimus was administered with a target serum level of 4–12 ng/mL. In the absence of ongoing aGVHD, tacrolimus was tapered with empiric dose reductions.

aGVHD was scored weekly per established consensus criteria [19]. CR was defined as sustained complete resolution of aGVHD without recurrence until death or last follow-up. Partial response was defined as an overall grade improvement of ≥1. cGVHD was scored according to the NIH consensus scoring criteria [20]. Indication, initial dose, and duration of any glucocorticoid therapy were recorded. Cumulative incidence of disease relapse, cGVHD, cytomegaolvirus (CMV) reactivation, and TMA are reported. OS, relapse-free survival (RFS), and failure-free survival were calculated by the Kaplan-Meier method, with failure defined as nonrelapse mortality (NRM) or requirement of glucocorticoid therapy after initial therapy with sirolimus. This study was approved as a retrospective review of a nonconsecutive patient series by the University of South Florida institutional review board.

RESULTS

Ten patients were treated with sirolimus as the primary therapy for biopsy-proven aGVHD at a median of 27 days after HCT (range: 15–103 days) with primary aGHVD prophylaxis consisting of either tacrolimus plus methotrexate (MTX; n = 7) or MMF (n = 3). Medications used in prophylaxis of aGVHD, namely, tacrolimus and MMF, were continued with sirolimus. Sirolimus was administered orally, and therapeutic levels were achieved in all cases, including those with gastrointestinal (GI) involvement. aGVHD was treated at the earliest possible time of onset, and therefore was grade II overall in 9/10 patients. Baseline characteristics of this series are summarized in Table 1. These patients were at high risk for primary disease relapse at time of transplant, with only 2/10 in complete remission at the time of transplant. Those not in remission included secondary acute myelogenous leukemia (AML) responsive to hypomethylating agents but with persistent blasts (range: 8%–15%) and cytogenetic abnormalities (n = 3), persistent idiopathic myelofibrosis (n = 2), multiple myeloma (MM) in very good partial remission (VGPR) with persistent low-level serum monoclonal protein (n = 1), follicular cell lymphoma in partial remission after salvage therapy with persistent hypermetabolic adenopathy on computed tomography/positron emission tomography (CT/PET) and no morphologic or molecular bone marrow involvement (n = 1), and acute lymphobalstic leukemia (ALL) s/p induction therapy with morphologic remission, but persistent immunoglobulin gene rearrangement (n = 1). At a median follow-up of 6.5 months (range: 2.4–18.3 months), 5 (50%) patients achieved sustained CR of aGVHD with sirolimus without requiring any second-line therapy with glucocorticoids or other salvage therapy (Table 2). In 1 patient, a complete remission of aGVHD was attained with sirolimus, but upon a flare of aGVHD 51 days after initial remission, a second remission was achieved with the addition of 0.5 mg/kg of prednisone. In 1 patient, 1mg/kg of prednisone was initially utilized at the onset of aGVHD, but was rapidly tapered completely off given concern for toxicity for a total duration of corticosteroid therapy of 9 days with the addition of sirolimus; after complete remission was reached, no further glucocorticoids were needed. In 3 patients (30%), glucocorticoids were used as salvage therapy after sirolimus for persistent aGVHD. Two required 0.5 mg/kg of prednisone for persistent upper (n = 1) or lower (n = 1) symptoms (aGVHD: GI grade 1, overall grade II); these were similar in baseline characteristics with those who achieved complete resolution of aGVHD with sirolimus alone. The other required 1 mg/kg of prednisone for persistent aGVHD (skin grade 3, overall grade II); this case differed in having a DRB1-incompatible, 9/10 matched unrelated donor. With persistent aGVHD after 1 mg/kg of glucocorticoids, MMF was successful in inducing durable complete remission of aGVHD. Allowing for primary therapy with sirolimus and salvage steroids, all had reached CR of aGVHD by 8 weeks after initiation of sirolimus (Figure 1).

Table 1 .

Baseline Characteristics

	Frequency
Median age	57.5 (range: 28–68)
Condition
ALL	1
AML	4
MDS	1
MM	1
MPD	2
NHL	1
Remission status
Complete remission (CR)	2
Not in CR	8
Cell source
PBSCT	10
BMT	0
Donor relation
Related donor	4
Unrelated donor	6
HLA matching
10/10	8
9/10	2
Recipient/donor sex
Female/female	3
Female/male	1
Male/female	1
Male/male	5
Conditioning regimen
Flu/Bu	5
Flu/Bu/ATG	2
Flu/Bu/Rituxan	1
Flu/Mel	1
Pento/BU/Rituxan	1
aGVHD prophylaxis
TAC/MTX	7
TAC/MMF	3
Donor/recipient CMV
Neg/neg	4
Neg/pos	3
Pos/pos	3
aGVHD onset date (median)	3.86 weeks (range: 2.14–14.71)
Overall aGVHD onset grade*
I	0
II	9
III	1
IV	0
aGVHD onset organ stage
Skin
1	2
2	1
3	2
4	0
GI
1	7
2	1
3	0
4	0
Liver
1	1
2	0
3	0
4	0

ALL indicates acute lymphoblastic leikemia; AML, acute myelogenous leukemia; MDS, myelodysplastic syndromes; MM, multiple myeloma; MPD, myeloproliferative disorder; NHL, non-Hodgkin’s lymphoma; aGVHD, acute graft-versus-host disease; cGVHD, chronic graft-versus-host disease; GI, gastrointestinal; CMV, cytomegalovirus; PBSCT, peripheral blood stem cell transplantation; BMT, bone marrow transplantation; Tac/MMF, tacrolimus/mycophenolate mofetil; Tac/MTX, tacrolimus/methotrexate; Flu/Mel, fludarabine and melphalan; Flu/Bu, fludarabine and busulfan; Flu/Bu/ATG, fludarabine, busulfan, and antithymocyte globulin.

^*All cases of aGVHD were biopsy confirmed.

Table 2.

Summary of Individual Patient Outcomes

Age	Disease	Disease remission at HCT	aGVHD prophylaxis	Overall onset grade	Skin/GI/Liver stage	CR with Sirolimus	Required steroids	Malignancy Relapse	Death
28	ALL	No	Tac/MTX	II	2/0/1	Yes	No	No	No
56	MM	No	Tac/MTX	II	0/1/0	Yes	No	No	No
68	AML	No	Tac/MTX	II	0/1/0	Yes	No	Yes	Yes
52	MPD	No	Tac/MTX	II	3/1/0	Yes	Yes 0.5 mg/kg (recurrent)	No	No
58	MPD	No	Tac/MTX	II	1/1/0	No	Yes 0.5 mg/kg (persistent)	No	No
34	NHL	No	Tac/MTX	II	0/1/0	No	Yes 0.5 mg/kg (persistent)	No	No
57	AML	Yes	Tac/MTX	II	3/0/0	No	Yes 1 mg/kg and MMF (persistent)	No	No
66	MDS	Yes	Tac/MMF	II	0/1/0	*	No	Yes	Yes
63	AML	No	Tac/MMF	II	0/1/0	Yes	No	Yes	No
67	AML	No	Tac/MMF	III	1/2/0	Yes	No	Yes	No

ALL indicates acute lymphoblastic leukemia; AML, acute myelogenous leukemia; MDS, myelodysplastic syndromes; MM, multiple myeloma; MPD, myeloproliferative disorder; NHL, non-Hodgkin lymphoma; aGVHD, acute graft-versus-host disease; cGVHD, chronic graft-versus-host disease; GI, gastrointestinal; Tac/MMF, tacrolimus/mycophenolate mofetil; Tac/MTX, tacrolimus/methotrxate; HCT, hematopoietic cell transplant.

^*1 mg/kg of prednisone utilized at the onset of aGVHD, but then rapidly tapered off after addition of sirolimus with total duration steroid treatment of 9 days; maintained CR of aGVHD with no further steroids.

Figure 1.

Number of patients with weekly overall aGVHD scores after initiation of sirolimus.

The cumulative incidence of cGVHD in this group was 40% with maximal cGVHD grade of mild, which did not require escalation of immunosuppressive therapy. The primary malignancy relapsed in 4 patients. Two died of recurrent AML, the other 2 patients are alive after relapse, 1 undergoing therapy and the other with sustained remission. With a median follow up of 6.5 months (range: 2.4–18.3 months) after transplant, projected OS at 18 months is 79% (95% CI: 38.1%–94.3%), and projected RFS at 15 months is 70% (95% CI: 32.9%–89.2%). Projected failure-free survival, as defined above, is 51% (95% CI: 16%–78%) at 18 months after transplant.

Sirolimus was overall well tolerated, with 2 cases of mild TMA that resolved without end-organ compromise after dose reduction (1) or discontinuation (1) of tacrolimus. By time of death or last follow up, 2 had immunosuppression tapered (n = 1) or entirely withdrawn (n = 1) for primary disease relapse. None otherwise had liberated from immunosuppression by median follow-up of 6.5 months (range: 2.4–18.3 months).

DISCUSSION

We report here a series of patients who were treated with sirolimus as a first-line therapy of biopsy proven aGVHD. In this group with primarily overall grade II aGVHD and skin or gut involvement, a CR rate of 50% was reached with sirolimus alone, which is comparable to that seen with glucocorticoids. Additionally, those who required glucocorticoids after primary treatment with sirolimus achieved CR with only 0.5–1 mg/kg of glucocorticoids, suggesting a potential steroid-sparing effect. Only 1 patient had aGVHD refractory to glucocorticoids, which was salvaged with MMF.

The use of sirolimus as a steroid-free primary therapy for aGVHD in these patients was driven by both concern for intolerance of steroid adverse effects, but also that of primary disease relapse after transplantation. The potential antimalignancy effect of sirolimus motivated this approach [21–24]. The subjects represented here largely had high-disease risk, as evidenced by only 2 of 10 being in complete remission at the time of transplant. In this setting, the projected 15-month RFS of 70% compares favorably with what would otherwise be expected, given the high risk nature of these patients.

Although this limited series provides early evidence to support a clinical trial of this novel approach, several questions remain. First, the effectiveness of this approach needs validation in a larger series; selection bias poses a potential threat to the internal validity of this retrospective review. Second, sirolimus was successful in inducing complete remission of aGVHD in this series largely comprised of overall grade II disease, but further work remains to be done to evaluate the effectiveness of this therapy in more advanced grade disease. Additionally, although therapeutic levels of sirolimus were reached in cases with GI involvement, more advanced vomiting and/or large volume diarrhea could preclude achieving consistent therapeutic levels, as sirolimus is only available as an oral formulation. Next, although there is a theoretic rationale that sirolimus may decrease the risk of disease relapse, this needs to be further examined in sufficiently large series. Further work also needs to be done to evaluate this approach in standard risk patients. Finally, although 50% achieved complete remission of aGVHD without glucocorticoids here, the magnitude of this steroid sparing effect would be better borne out with examination of cumulative burden of steroid exposure in larger series.