A phase II study of tacrolimus and thymoglobulin as graft-versus-host-disease prophylaxis in related donor allogeneic hematopoietic cell transplantation

Clinicaltrials.gov identifier:

NCT01246206. The study was approved by Wayne State University Institutional Review Board.

To the Editor:

Acute graft versus host disease (aGVHD) is a frequently occurring complication following allogeneic hematopoietic stem cell transplantation (AHSCT). Commonly used GVHD prophylaxis regimens consist of combination of a calcineurin inhibitor with methotrexate (MTX).¹ However, severe mucositis and delayed engraftment are major limitations of MTX in this setting.² Attempts at substituting MTX with sirolimus or mycophenolate have led to less mucositis, but essentially no improvement in aGVHD incidence.³ In recent years, the use of thymoglobulin in combination with other immunosuppressive agents has demonstrated efficacy in reducing acute and chronic GVHD without increasing relapse rates in matched related and unrelated AHSCT.^4,5 We conducted a phase II study to evaluate the safety and efficacy of tacrolimus and thymoglobulin combination in patients undergoing matched- related peripheral blood AHSCT, based on the hypothesis that the combination would have additive effect on preventing GVHD with a more favorable side effect profile.

Eligible patients undergoing matched related AHSCT for hematologic malignancies were enrolled in the study. All patients received intravenous rabbit thymoglobulin (0.5 mg/kg on day −3, 1.5 mg/kg on day −2, 2.5 mg/kg on day −1) and tacrolimus 0.03 mg/kg/day starting on day −3. Patients were switched to oral tacrolimus once they had adequate oral intake, with goal therapeutic target range of 5–15 ng/mL for the first 60 days. Tacrolimus was subsequently tapered off by day + 180.

The primary end points included incidence and severity of aGVHD and the occurrence of any study combination related toxicity in the first 6 months. Secondary end points were relapse rate, non-relapse mortality (NRM), overall survival (OS), relapse free survival (RFS), and GVHD-free relapse-free survival (GRFS) at one-year.

A single arm Simon’s two-stage design was used. Assuming the rate of aGVHD is 28.8% (a 40% drop) compared to the published 48% with tacrolimus alone, the trial was designed with 80% power and 0.05 type I error with 19 evaluable patients in stage one and 22 patients in stage two. Assuming 15% of patients will be non-evaluable, a total of 48 patients were approved from IRB. Descriptive and competing risk analyses were performed. P values less than .05 will be considered significant. All statistical analyses were performed in R.

Out of 20 evaluable patients enrolled in stage one, 12 patients developed aGVHD (3 had grade 1, 7 had grade 2, 2 had grade 4), which exceeded the study threshold of nine. Therefore, the study was terminated after the first stage based on pre-specified design.

The median time to neutrophil and platelet engraftment was 11 (range, 7–12) and 17 days (range, 0–25), respectively. One year cumulative incidence of grade II-IV aGVHD was 45% (95% CI, 22.3%- 65.4%), with grade III-IV aGVHD incidence being 10% (95% CI, 1.6%- 27.8%). One year cumulative incidences of cGVHD and extensive cGVHD were 50% (95% CI, 26.0%−70.0%) and 45% (95% CI, 22.2%- 65.5%), respectively (Figure 1). Cytomegalovirus (CMV) and Epstein- Barr virus (EBV) reactivations were detected in six (30%) and two (10%) patients, respectively, with one developing GI CMV disease. One year relapse rate was 25% (95% CI, 8.7%−45.5%) and median time to relapse was 317 days (range, 31–994). Of the eight relapsed patients, six had active disease at the time of transplant. Three patients received donor lymphocyte infusions (DLI) and one received a second AHSCT. Three of the eight patients remained alive at time of study discontinuation.

FIGURE 1.

Top left: Grade II-IV aGVHDwith relapse or NRMas competing risks. Top right: Grade III-IV aGVHDwith relapse or NRMas competing risks. Middle left: cGVHDwith relapse or NRMas competing risks.Middle right: Relapse ratewith NRMas competing risk. Bottomleft:Non-relapse mortalitywith death due to relapse as competing risk. Bottom right: OS, RFS, GRFS

At a median follow-up of 54.3 months (95% CI, 43.6—NR), median RFS was 28.5 months (95% CI, 11.4—NR) and median OS was not reached. One year NRM, RFS, GRFS, and OS were 10% (95% CI, 1.6%−27.8%), 65% (95% CI, 47.1%−89.7%), 30% (95% CI, 15.4–58.6), and 85% (95% CI, 70.7%−100%), respectively (Figure 1). Eight patients died at a median of 444 days (range, 63–1448). The causes of death were recurrence or progression of underlying malignancy (62.5%), cGVHD (25%), and cardiac failure (12.5%). The baseline characteristics and results are summarized in Supporting Information Table S1.

In the stage one portion of this study, 12 of the 20 patients developed aGVHD, which required termination based on pre-specified threshold of nine. Despite negative outcome, the combination may warrant further study. Only 2 of the 12 patients had grade IV aGVHD with the rest having grade I or II aGVHD, which is often clinically insignificant without any adverse effects. This study has also shown that toxicity profile of tacrolimus and thymoglobulin combination is favorable with lower rates of severe mucositis and rapid neutrophil engraftment, which makes thymoglobulin an intriguing potential alternative to MTX.

In addition, the long-term outcomes, especially NRM and survival, are comparable to other commonly used regimens. Previous studies have reported higher rates of relapse with thymoglobulin in AHSCT with reduced intensity conditioning, most likely explained by higher doses of thymoglobulin (7 mg/kg).⁶ In our study, despite the use of T- cell depleting agent and high-risk patient population, the relapse rate was lower. Additionally, we did not observe increased incidence of CMV or EBV reactivations or other opportunistic infections. We believe that a favorable survival and NRM outcomes with the combination of tacrolimus and thymoglobulin may have resulted from reduced rates of severe aGVHD, lower incidence of opportunistic infections, early neutrophil engraftment, and absence of complications such as sinusoidal obstruction syndrome which is typically associated with sirolimus-based regimens. One of the strengths of our study is >1500 days long median follow-up. cGVHD is a time-dependent event and it takes at least 2–3 years to observe its detrimental effects on NRM and quality of life.

In conclusion, the combination of tacrolimus and thymoglobulin did not meet our primary endpoint in reduction of grade II-IV aGVHD below a pre-determined threshold. However, severe GVHD (III-IV) was low and the combination had a favorable toxicity profile, while providing comparable relapse rate, NRM, and long-term survival. Further studies may be warranted using larger sample size or using grade III-IV GVHD as an endpoint.