Ocular Graft-Versus Host Disease Prevention with Topical Cyclosporine in Non-Myeloablative Allogeneic Stem Cell Transplant

Introduction

Ocular Graft versus host disease (oGVHD) is a common complication in receptors of allogeneic or haploidentical stem cell transplants (allo, haplo-SCT); incidence has been reported between 37-55%. Manifestations can range from mild (dry eye) to very severe (corneal perforation). Topical (eye drops) cyclosporine (CsA) has been reported to be an effective treatment of oGVHD once the disease has been established, but long before clinical manifestations are evident, lymphocyte infiltration and damage to the lacrimal gland have taken place. In this study, we evaluate if topical CsA administered from the day of graft success and over one year is effective for prevention of oGVHD

Methods

We conducted a prospective, single arm, phase 2 study. We included patients that received an allo-SCT or haplo-SCT for malignant and non-malignant hematologic diseases who previously didn´t have dry eye syndrome (primary or secondary to other diseases) or corneal surgeries. All patients received systemic CsA for one year aiming serum at levels of 250-350 ng/mL and metothrexate 10 mg/m² at days+3,+5 and +7 post-transplant (PT). Patients with haplo-SCT also received PT cyclophosphamide 50 mg/kg at days +3 and +4 and mofetil mycophenolate 1 gram/day for one month. We administered CsA in eye drops immediately at the day of graft success (demonstrated with > than 500 X 10⁹ /L neutrophils and >20,000 10⁹ /L platelets for two days unsupported with transfusions) and for one year PT. Systemic steroids 2nd line rituximab and 3rd line bortezomib were allowed in case of systemic GVHD. All patients were evaluated by an ophtalmologist before entering the trial for adequate tear quality with Schirmer test, integrity of corneal surface, eye pressure, and tear breakup time. After grafting, all patients were evaluated monthly with Schirmer test and a complete opftalmologic evaluation was made in months 3,6,9 and 12 post-transplant. Patients were evaluated and diagnosed with oGVHD according NIH criteria. We compared our results with an historical cohort of patients allografted in our center. This study was registered in Clinical Trials.gov NCT02144025

Results

We included 21 patients (15 HLA identical (71.4%) and 6 haploidentical (28.6%) transplants). Median age was 42(17-62) years, the causes of transplants were hematologic malignancies in 19 patients and aplastic anemia in 2; median CD34 cell dose was 5.8 X 10⁶/Kg (0.70-18.76 X 10⁶/Kg). Median follow up of the whole cohort was 254 days (76-583). Systemic CsA target levels were met in 17 patients at all measurements, in the other 4 median levels were of 164 ng/mL (159-179 ng/mL). Seven patients (33.3%) received CsA+prednisone (PDN), 6 CsA+PDN+rituximab and 1 CsA+PDN+rituximab+bortezomib for systemic GVHD. OS at 365 days was 58% and EFS was 38%. Systemic acute GVHD (aGVHD) incidence grade II-IV was 57% (n=12) and grade III-IV 9.5%(n=2). Chronic systemic GVHD (cGVHD) incidence was 57% (n=12) and extensive (EcGVHD) 14.3% (n=3). We documented 1 case of definite oGVHD at last follow-up of the cohort. Systemic aGVHD grade II-IV (P=0.57), grade III-IV (P=0.48), cGVHD (P=0.52) or EcGVHD (P=0.13) did not have association with oGVHD development. Median of first Schrimer measurement (at study entry) of all cohort was 7.5 mm and 11.3 mm (P=0.005) at last follow-up. Comparing incidence of oGVHD of our study group (n=1, 5%) with an historical cohort of 26 patients allografted in our center (n=6, 23%) (P=0.087) difference is not statistically significant.

Conclusion

Allografted patients that received topical CsA augmented the quantity and quality of the tear. Although not statistical significant, there is a clear tendency of less oGVHD incidence in the stydy cohort Tear quality improved, which might be a sign of preservation of lacrimal gland function. A randomized, blinded and placebo-controled trial is needed to clearly determine if topical CsA is effective for oGVHD prevention.