Sirolimus as Long-term Graft-versus-host-disease Prophylaxis in Haploidentical Hematopoietic Stem Cell Transplant Recipients for Non-malignant Disorders is Associated with High Incidence of Acneiform Lesions.

BACKGROUND: Sirolimus has provided the option for calcineurin inhibitor (CNI)-free immunosuppressive therapy in both solid organ transplant (SOT) and hematopoietic stem cell transplantation (HSCT). However, long-term use of sirolimus has been reported to be associated with a high incidence of cutaneous side effects in SOT, particularly acneiform lesions.
MATERIALS AND METHODS: We studied the incidence of acneiform lesions and the risk factors in 41 HSCT recipients between ages 4 and 64 years, undergoing HSCT for malignant (n = 29) and non-malignant diseases (n = 12) from haploidentical family donors.
RESULTS: Seven patients developed acneiform lesions at the median of 85 days (range, 45-105 days). Acneiform lesions occurred in 6/11 patients on sirolimus and in only 1/30 patients not receiving sirolimus (P = 0.001). This was more frequent in patients with non-malignant disorders (5/12 versus 2/29, P = 0.01) and those receiving graft from female donors (7/23 versus 0/18, P = 0.01).
CONCLUSIONS: Despite being frequently reported in SOT, this is the first such report in HSCT. Our study suggests that prolonged use of sirolimus might be associated with high incidence of acneiform lesions in haploidentical HSCT recipients with non-malignant diseases, particularly in those receiving graft from a female donor. We discuss the possible reasons for these findings and the putative mechanism of acneiform lesions in these patients.

What was known?

Sirolimus, an M-TOR pathway inhibitor is associated with high incidence of acneiform lesions in recipients of solid organ transplantation.

Introduction

Sirolimus has been used in solid organ transplantation for over a decade, but has made a delayed entry in the field of hematopoietic stem cell transplantation (HSCT) for graft-versus-host-disease (GVHD) prophylaxis.[12] Cutaneous side-effects of sirolimus are much more pronounced compared to other immunosuppressive agents.[3] The incidence of acneiform lesions in renal transplant recipients has been reported to be as high as 46%.[4] Notwithstanding the limited use of sirolimus, no study has yet reported of cutaneous lesions or acne in HSCT recipients.

Matched family donor for HSCT is available to only 20% of patients. We initiated a haploidentical family donor HSCT in India since 2011 to address this issue. In this program, sirolimus was introduced for GVHD prophylaxis over the last 24 months. A sharp increase in acneiform lesions was noted in this period. Hence, we analyzed the data on 41 consecutive patients undergoing haploidentical HSCT for the incidence, pattern and response to treatment of acneiform lesions and its risk factors.

Materials and Methods

Forty-one patients received a haploidentical HSCT with peripheral blood stem cell (PBSC) graft from March 2011 to October 2014. Written informed consent and ethical approval were obtained for all of the patients in accordance with the Declaration of Helsinki.

Conditioning regimen and GVHD prophylaxis

The conditioning consisted of fludarabine and alkylating agents with or without 2 Gy Total Body Irradiation as previously reported.[5] GVHD prophylaxis consisted of post-transplant cyclophophamide 50 mg/kg on days 3 and 4, followed by cyclosporine (CSA) and mycophenolate mofetil from day 5. If the patient was intolerant to CSA or had significant toxicity, this was replaced by sirolimus. In seven patients, undergoing haploidentical HSCT for non-malignant diseases, in addition to the above, sirolimus was initiated on day 7 with trough levels of 8–14 ng/ml on day 0 and maintained thereafter until 9–12 months and then tapered. In these patients, CSA was stopped after day 100 if there was no GVHD.

Supportive care

All patients were treated in protective isolation rooms provided with high efficiency particle air filters. Antimicrobial prophylaxis was instituted as per the departmental guidelines. Co-trimoxazole was administered to patients following stable engraftment for 3 months or until achievement of CD4+ cell count >200 cells/μl. For patients allergic or intolerant to co-trimoxazole, levofloxacin or azithromycin was used.

Monitoring of cutaneous side-effects

All patients were examined weekly for the first 100 days and fortnightly thereafter. The patient was reviewed by the dermatologist on appearance of any skin lesions other than acute GVHD. Acneiform lesions were defined as appearance of inflammatory or non-inflammatory papules, pustules, or nodules in the seborrheic areas. Acneiform lesions were graded using the “Global Acne Grading System” at diagnosis and follow-up.[6] All patients were administered topical retinoid either in combination with benzoyl peroxide or topical clindamycin as first-line therapy. Oral doxycycline was added if the lesions progressed or the disease was severe on the scoring system. Oral retinoids were not used as per departmental guidelines.

Statistics

An outcome was determined to be significantly different if the observed P value was < 0.05. Binary variables were compared between the groups using the chi square test and the continuous variables were analyzed using an independent sample t test taking into account Levenes test for equality of variances. Analyses were performed using statistical software IBM SPSS Statistics Version 20.

Results

Patient characteristics

The patient characteristics and transplant outcomes are detailed in Table 1. The overall survival was 56.8% (CI 48.5–65.1) at 2 years.

Table 1

Characteristics of patients with and without post-transplant acneiform lesions

Acneiform lesions and risk factors [Tables 1 and 2]

Seven patients developed acneiform lesions at the median of 85 days (range, 45–105 days). Acneiform lesions tended to be predominant in males (6/28 versus 1/13 in females, P = 0.27).

The significant risk factors for the development of acneiform lesions were: (1) the use of sirolimus as GVHD prophylaxis (6/11 versus 1/30 patients not receiving Sirolimus, P = 0.001), (2) HSCT for non-malignant diseases (5/12 versus 2/29 in malignant diseases, P = 0.01), and (3) HSCT from a female donor (7/23 versus 0/18 from male donors, P = 0.01).

The median trough level of sirolimus was 8.2 ng/ml at the onset of acne compared to 8.9 ng/ml, 30 days earlier.

The clinical characteristics of acneiform lesions in these seven patients have been described in detail in Table 2. Acneiform lesions in four patients were graded as very severe, one as severe and the other two as moderate. The lesions were persistent in five of seven patients and resolved only after tapering and/or withdrawal of the drug. The only female patient (Patient 3) who experienced the lesions on sirolimus subsequently developed GVHD and received several other immunosuppressants and antibiotics, with resolution of acneiform lesions. The single patient in the non-sirolimus group (Patient 1) with acneiform lesions had prompt resolution with treatment. The other probable sirolimus-associated toxicities in these patients were mucosal dryness in four patients and limb edema in three. These resolved on tapering or stoppage of sirolimus.

Table 2

Characteristics of the patients with acneiform lesions

The median duration of sirolimus therapy was 12 months (4–14 months). The duration of immunosuppression was longer for non-malignant diseases (median of 12 months versus 4 months in malignant diseases). None of the patients had recurrence of acneiform lesions after discontinuation of sirolimus.

Discussion

Immunosuppressive drugs are used for protracted periods in solid organ transplants and for a few months to years in HSCT depending on the disease and post-transplant complications.[789] Sirolimus promotes induction of regulatory T cells which hastens tolerance,[10] which in malignant diseases might affect the graft-versus-leukemia effect.[11] However, this is not the primary concern in HSCT for non-malignant diseases and we opted to introduce sirolimus as GVHD prophylaxis in our protocol for non-malignant diseases and in malignant diseases, when CNI could not be administered.

Our study of 41 patients undergoing haploidentical HSCT showed a high incidence of acneiform lesions (64%) among those receiving sirolimus for prevention of GVHD as compared to 8% in the non-sirolimus group. In addition, a higher incidence of acneiform lesions was noted in patients undergoing HSCT for non-malignant disease than those for malignant disease. It is possible that this is simply because of the fact that sirolimus was used for longer periods in patients with non-malignant diseases targeting a higher drug level. However, there was no correlation with trough sirolimus levels. Indeed, the onset of acneiform lesions occurred at a median of 3 months and most patients with malignant diseases were on a tapering mode of the drug by 100 days. Acneiform lesions occurred in almost all patients after stoppage of prophylactic antibiotics. It is possible that the anti-acne effect of broad spectrum antibiotics could have played a role in the occurrence of sirolimus-induced acneiform lesions after their withdrawal. The only other report on long-term use of sirolimus in HSCT was in 10 patients with sickle cell disease.[12] The study mentions arthralgia, pneumonitis and edema as sirolimus-associated toxicities, but acneiform lesions were not mentioned. It remains unclear whether acneiform lesions did not occur in these patients or they were ignored.

The pathogenesis of the acneiform lesions in patients receiving sirolimus is not well understood. The epidermal growth factor (EGF) pathway is thought to be of definite importance.[13] Sirolimus has been shown to inhibit EGF through the mTOR signaling pathway.[14] Patients treated with monoclonal antibodies against EGFR or EGFR tyrosine kinase inhibitors show a range of cutaneous toxicities similar to that seen in patients on sirolimus, the most frequent being acneiform lesions.[15] Taking the various studies into consideration, it would appear that an imbalance in the EGF pathway regulation by sirolimus results in the development of acneiform lesions. The male preponderance in the development of acneiform lesions may be explained on the basis of the EGF pathway regulation. Sirolimus down-regulates testosterone synthesis, whereas testosterone up-regulates EGF receptor synthesis.[1617]

Finally, the role of HLA-mismatched female donors in the genesis of acne remains interesting but ambiguous. Both HLA mismatch and female to male HSCT contribute to alloreactivity, the latter being induced by H-Y antigen in males[18] but none of our patients with acne had acute GVHD. Indeed, levels of EGF are reduced before and during GVHD[19] and it might be speculated that the same might happen with subclinical alloreactivity induced by HLA-mismatched female donors.

However, in contrast to other reports, our patients did not respond significantly to doxycycline or topical anti-acne therapies. It is possible that these patients would have improved with low-dose isotretinoin as has been recommended,[420] which was not administered in this cohort to avoid lipid and liver abnormalities. Interestingly, we did not notice any major non-dermatological toxicity in the patients with sirolimus-associated acneiform lesions.

This is the first study to report on acneiform lesions in HSCT recipients on prolonged prophylaxis with sirolimus. Our data raises several interesting issues pertaining to the possible pathogenesis of acneiform lesions in HSCT recipients in relation to the nature of the underlying disease, duration of sirolimus exposure and donor gender in the context of HLA-mismatched HSCT. Further studies are required to explore and expand on these findings.

What is new?

Despite its use in HSCT recipients, no association of acneiform lesions and sirolimus has been published in the past.

We have documented increased incidence of acneiform lesions in patients on long-term sirolimus as GVHD prophylaxis after haploidentical HSCT.

Our study also suggests that this phenomenon is probably more common in those receiving HSCT for nonmalignant diseases.

Our study also suggests that transplantation from a female donor to a male recipient in the setting of haploidentical HSCT might increase the risk of acneiform lesions in those receiving long-term sirolimus.