Impact of recipient statin treatment on graft-versus-host disease after allogeneic hematopoietic cell transplantation

Abstract

We retrospectively analyzed outcomes among 1206 patients with hematologic malignancies who had hematopoietic cell transplantation (HCT) from HLA-identical siblings (n=630) or HLA-matched unrelated donors (n= 576) at a single institution between 2001 and 2007 for a correlation between recipient statin use and risk of graft-versus-host disease (GVHD). Among recipients with cyclosporine-based postgrafting immunosuppression (n=821), statin use at the time of transplant (6%) was associated with a decreased risk of extensive chronic GVHD (multivariate hazard ratio [HR], 0.62; 95% confidence interval [CI], 0.4–1.0; p=0.05) and an increased risk of recurrent malignancy (HR, 1.75; 95% CI, 1.0–3.0; p=0.04). Recipient statin use, however, had no apparent impact on the risks of chronic GVHD and recurrent malignancy among recipients given tacrolimus-based immunosuppression (n=385; 8% statin-treated). Risks of acute GVHD, non-relapse mortality and overall mortality were not significantly affected by recipient statin use. Hence, recipient statin treatment at the time of allogeneic HCT may decrease the risk of chronic GVHD in patients with cyclosporine-based immunosuppression but at the expense of a compromised graft-versus-tumor effect.

INTRODUCTION

In recent years, statins have been shown to affect immune responses through a variety of mechanisms, including induction of T-cell hyporesponsiveness by interference with intracellular signaling pathways, expansion of regulatory T-cells (Treg), polarization towards an anti-inflammatory T-cell phenotype (Th2), and down-regulation of antigen-presenting cell function [1-7].

More recent studies have shown that statins protect against graft-versus-host-disease (GVHD) after hematopoietic cell transplantation (HCT) [7-10]. In an MHC-mismatched murine HCT model, for example, donor or recipient treatment with statin resulted in significantly reduced acute GVHD lethality [7]. These findings were corroborated by Wang et al. [11], who found that lovastatin, but not pravastatin, prevented GVHD by virtue of interfering with LFA1-mediated homing of T-cells to GVHD target organs. In a retrospective analysis of 67 patients who had received HLA-matched or mismatched related or unrelated allografts, Hamadani et al. described a trend towards decreased risk of grade 2-4 acute GVHD among the 10 recipients who were treated with statins at the time of HCT, but definitive conclusions were limited by the small sample size [8].

In a retrospective analysis of 567 consecutive recipients of allografts from HLA-identical sibling donors, we recently reported that donor statin use was associated with profound protection against grades 3-4 acute GVHD [9]. Since our analysis of the potential impact of recipient statin use on GVHD was limited by the sample size in the previous study, we now analyzed data from both related and unrelated transplants. A similarly broad examination of donor statin use on GVHD was not possible because medication data for unrelated donors are not available. We found that, among recipients with cyclosporine-based immunosuppression, statin use was associated with an approximately 40% reduced risk of chronic GVHD without affecting the risk of acute GVHD.

PATIENTS AND METHODS

We retrospectively analyzed data from consecutive patients who had allogeneic HCT from HLA-identical sibling or HLA-matched unrelated donors for treatment of hematologic malignancies between January 2001 and December 2008. All recipients and donors had given written informed consent to treatment and use of medical records for research according to protocols approved by the Institutional Review Board at FHCRC. To be included in the final analysis, recipients had to be at least 18 years old. Medication data were obtained by retrospective review of medical records and independently confirmed by a pharmacist. Transplant recipients were considered “on statin treatment” if they were treated with any statin drug at any dose for at least 3 consecutive months prior to transplant. Statin treatment could be discontinued during the preparative regimen and resumed immediately after transplant. Data on recipient statin treatment were collected until day 100 after transplant. Characteristics of recipients, donors and treatment variables are shown in Table 1.

Table 1.

Recipient, donor and transplant characteristics according to recipient statin use at the time of transplant.

	Recipients without statin treatment, n=1130	Recipients with statin treatment, n=76	p
Recipient characteristics	n (%)	n (%)
Median age, range (years)	52, 20-79	60, 31-72	<0.001
Age > 50 years	627 (55)	68 (89)	<0.001
Sex, female	513 (45)	16 (21)	<0.001
Disease
AML	417 (37)	24 (32)	ns
MDS	284 (25)	21 (28)
NHL	112 (10)	9 (12)
CML	88 (8)	7 (9)
MM	69 (6)	3 (4)
ALL	77 (7)	4 (5)
CLL	47 (4)	6 (8)
HD	17 (1)	0 (0)
Other	19 (2)	2 (3)
Disease risk *
Standard	541 (48)	31 (41)	ns
High	589 (52)	45 (59)
Donor characteristics
HLA - Identical sibling	602 (53)	28 (37)	0.006
- Matched unrelated	528 (47)	48 (63)
Median age, range (years)	43, 18-80	42, 19-83	ns
Age > 50 years	331 (29)	26 (34)	ns
Sex, female	509 (45)	34 (45)	ns
Female donor with male recipient	113 (10)	10 (13)	ns
Transplant characteristics
Conditioning
Myeloablative	689 (61)	31 (41)	<0.001
Nonmyeloablative	441 (39)	45 (59)
Stem cell source
Mobilized blood cells	1031 (91)	73 (96)	ns
Marrow	99 (9)	3 (4)
GVHD prophylaxis—n (%)
Cyclosporine with MTX or MMF†	773 (71)	48 (63)	ns
Tacrolimus with MTX or MMF‡	357 (29)	28 (37)
ATG	78 (7)	6 (8)	ns

^*Disease risk: “Standard” refers to chronic myeloid leukemia in chronic phase, myelodysplastic syndromes without excess blasts, and leukemia and lymphoma in remission. “High” refers to all other hematologic malignancies.

^†Seven patients received additional sirolimus.

^‡10 patients received additional sirolimus.

Abbreviations: ALL, acute lymphatic leukemia; AML, acute myeloid leukemia; ATG, anti-thymocyte globulin; BM, bone marrow; CLL, chronic lymphatic leukemia; CML, chronic myeloid leukemia; GVHD, graft-versus-host disease; HD, Hodgkin disease; MDS, myelodysplastic syndrome; MM, multiple myeloma; MMF, mycophenolate mofetil; MTX, methotrexate; NHL, non-Hodgkin lymphoma; PBSC, peripheral blood stem cells.

Multivariate analyses of time-to-event endpoints were performed using Cox regression. Death was treated as a competing risk for analysis of relapse. Death and relapse were treated as competing risks for analysis of acute and chronic GVHD. Relapse was treated as a competing risk for the analysis of non-relapse mortality. The landmark analysis was restricted to patients still at risk to develop chronic GVHD at day 100 (excluding patients who had died, relapsed, or developed chronic GVHD prior to day 100). All p-values are 2-sided.

RESULTS AND DISCUSSION

Among the 1206 recipients included in this analysis, 76 (6%) were treated with a statin at the time of transplant (atorvastatin, 53%; simvastatin, 21%; pravastatin, 13%; lovastatin, 8%; and rosuvastatin, 5%). Fifty-two percent and 48%, respectively, had grafts from HLA-identical siblings and HLA-matched unrelated donors (Table 1). Compared to patients not treated with a statin, those treated with a statin were older (median age 60 vs. 52 years, p<0.001), more frequently male (79% vs. 55%, p<0.001), recipients of unrelated grafts (63% vs. 47%, p=0.006), and prepared with a reduced-intensity conditioning regimen (59% vs. 39%, p<0.001). There were no significant differences between the two groups with respect to distributions of disease risk, median donor age, donor gender, stem cell source, and type of immunosuppression after transplant.

After adjustment for donor and recipient age (≤ 50 vs. >50 years) and gender (female-into-male vs. other), donor type (related vs. unrelated), conditioning intensity (high-dose vs. reduced-intensity), disease risk (standard vs. high), stem cell source (mobilized blood cells vs. marrow), type of calcineurin inhibitor used after transplant (cyclosporine vs. tacrolimus) and year of transplant in multivariate analysis, the risks of grade II-IV and grade III-IV acute GVHD, non-relapse mortality and overall mortality were similar between recipients with and without statin treatment (Table 2). The risk of extensive chronic GVHD, however, was 38% lower in recipients with statin treatment given cyclosporine-based immunosuppression after transplant compared to those without statin treatment (n=48 vs. n=773; hazard ratio [HR], 0.62; 95% confidence interval [CI], 0.4-1.0; p=0.05). In contrast, among patients with tacrolimus-based immunosuppression after transplant, the risk of chronic GVHD was not significantly affected by recipient statin use (n=28 vs. n=357; HR, 1.15; 95% CI, 0.7-2.0; p=0.61). The incidence of de novo onset chronic GVHD was similar between statin-treated compared to not statin-treated recipients (16% vs. 13%). Importantly, the decreased risk of chronic GVHD among statin-treated recipients with cyclosporine-based immunosuppression was associated with a significantly increased risk of recurrent malignancy (multivariate HR, 1.75; 95% CI, 1.0-3.0; p=0.04).

Table 2.

Multivariate analysis for association of recipient statin use with outcomes after allogeneic hematopoietic cell transplantation from HLA-matched related and unrelated donors.^*

	All Recipients n=1206			Recipients receiving postgrafting cyclosporine n=821			Recipients receiving postgrafting tacrolimus n=385
Outcome	Hazard Ratio	95% CI	p	Hazard Ratio	95% CI	p	Hazard Ratio	95% CI	p
Grade 2-4 GVHD	0.93	0.7-1.3	0.63	0.87	0.6-1.3	0.47	1.11	0.7-1.8	0.68
Grade 3-4 GVHD	1.17	0.6-2.1	0.62	1.01	0.5-2.2	0.99	1.55	0.6-4.1	0.37
Chronic GVHD	0.78	0.6-1.1	0.17	0.62	0.4-1.0	0.05	1.15	0.7-2.0	0.61
Relapse / disease progression	1.26	0.8-2.0	0.34	1.75	1.0-3.0	0.04	0.60	0.2-1.7	0.33
Non-relapse mortalty	0.90	0.6-1.3	0.60	0.86	0.5-1.4	0.54	0.97	0.5-1.8	0.93
Overall mortality	0.98	0.7-1.3	0.90	1.07	0.7-1.6	0.71	0.86	0.5-1.5	0.59
Landmark Analysis** (among recipients alive, on statin treatment, in remission, and without chronic GVHD at day 100)
		n=863			n=560			n=303
Chronic GVHD	0.63	0.4-1.0	0.06	0.40	0.2-0.9	0.02	0.98	0.5-1.9	0.96
Relapse / disease progression	1.76	0.9-3.3	0.08	2.53	1.3-5.1	0.009	0.84	0.2-3.6	0.81

^*Outcomes among recipients treated with a statin medication are compared with those among recipients who were not treated with a statin medication after HCT. Specifically: All recipients (n=76 vs. n=1130); recipients receiving postgrafting cyclosporine (n=48 vs. n=773); recipients receiving postgrafting tacrolimus (n=28 vs. n=357).

^**All recipients (n=38 vs. n=863); recipients receiving postgrafting cyclosporine (n=22 vs. n=560); recipients receiving postgrafting tacrolimus (n=16 vs. n=303).

Results were adjusted for sex-mismatch (F→M), conditioning-intensity, donor age (>50 years), patient age (>50 years), tacrolimus-based immunosuppression, disease risk, stem cell source, donor type, and year of transplant.

Abbreviations: CI, confidence interval; GVHD, graft versus host disease.

At day 100 after transplant, 87% of statin-treated patients (66 of 76) were still alive. Of those, 67% continued to receive statin treatment beyond day 100. In a landmark analysis restricted to recipients alive without chronic GVHD, and on statin treatment on day 100 after transplant (n=38), the association between statin use and protection against chronic GVHD (multivariate HR, 0.40, p=0.02) or increased risk of recurrent malignancy (multivariate HR, 2.53, p=0.009) among recipients with cyclosporine-based immunosuppression was even more pronounced (Table 2).

The interaction between statins and cyclosporine observed in the current study was statistically significant (p=0.04). This was consistent with our previous analysis among recipients of related grafts, where protection against acute GVHD mediated by donor statin use was strongly associated with cyclosporine-based immunosuppression [9]. This apparent effect-association between statins and cyclosporine was unexpected and could reflect synergistic impairment of T-cell mitochondrial function by both drugs, leading to compromised T-cell responsiveness to alloantigen and establishment of tolerance [12,13]. Alternatively, in particular since the statin/cyclosporine effect-association was seen in statin-treated recipients, it might be explained by the fact that statin serum concentrations are considerably higher in patients treated with cyclosporine than in those treated with tacrolimus [14,15]. This appears to be related to the fact that cyclosporine, in contrast to tacrolimus, has profound inhibitory effects on hepatocyte membrane-transporters known to affect statin metabolism, including organic anion transporting peptides (OATP) [15-17]. OATP(1B1)-inhibition by cyclosporine but not tacrolimus appears to be more relevant with respect to explaining differential effects of the two calcineurin inhibitors on statin levels than inhibition of cytochrome P450(3A4) or the p-glycoprotein efflux pump [15].

Our previous analysis of transplant outcomes according to donor and recipient statin use after related HCT did not show a significant association between recipient statin use and risk of chronic GVHD. This lack of association was possibly related to sample size limitations and a differential effect size according to donor type, since the reduction in risk of chronic GVHD associated with statin use in the current study was less pronounced among related recipients (adjusted HR, 0.69; 95% CI, 0.4-1.2; p=0.21) than unrelated recipients (HR, 0.44; 95% CI, 0.2-1.1; p=0.07).

The main statin-associated toxicities described in the literature are muscular and hepatocellular injury [18]. The risk of statin toxicity is considerably increased by the concurrent use of calcineurin inhibitors that are thought to increase statin levels by competing with the hepatic disposition of statins [19,20]. Therefore, the increased risk of statin-related toxicity among allograft recipients requiring calcineurin inhibitor-based immunosuppression is a legitimate concern. Among the 76 recipients treated with statins in our study, 12 (16%) experienced predominantly mild-to-moderate statin-related toxicity within 90 days after transplant. We observed ≥ 3-fold increases in serum aminotransferase enzyme levels (n=7), myalgias without evidence of rhabdomyolysis (n=2), and documented rhabdomyolysis (n=3) with serum creatinine kinase levels >10,000 U/L and serum creatinine >2.0 mg/dL. Discontinuation of statin treatment resulted in prompt and complete resolution of symptoms.

In conclusion, our data suggest that recipient statin use, which appears to be associated with manageable toxicity, may confer protection against extensive chronic GVHD, but not against acute GVHD, in patients given postgrafting cyclosporine. In contrast, at least in related donor HCT, donor statin use confers profound protection against grade III-IV acute GVHD without an apparent effect on chronic GVHD [9]. These findings can be interpreted to suggest differential mechanisms in the pathophysiology of acute and chronic GVHD. If confirmed in well-designed prospective clinical trials, GVHD-prevention through donor statin-“conditioning” may become an important addition to the armamentarium of approaches for prevention of acute GVHD. The benefit of reducing the risk of chronic GVHD by virtue of treating recipients with statin, however, needs to be carefully weighted against the increased risk of recurrent malignancy.