Background: Ocular GVHD is a known complication of chronic GVHD (cGVHD) and is considered a distinctive criterion of cGVHD. Ocular GVHD often leads to irreversible dry eye syndrome and contributes to the morbidity of cGVHD. The incidence, timing, risk factors and survival of patients (pts) with ocular GVHD are not well described in the current era of allogeneic SCT.

Methods: Ocular GVHD was defined as presence of 2 or more of the following criteria: new onset subjective symptoms of dry eye, tear film instability (TBUT ≤ 5 seconds), lacrimal insufficiency (Schirmer I ≤ 5 mm at 5 minutes), ocular surface abnormality (vital staining grade I or higher) or conjunctival inflammation (presence or absence). All patients followed in the long term transplant clinic at our center were studied retrospectively (N=128). The majority of these pts were evaluated by the same ophthalmologist (U.T.)

Results: Median age of the cohort was 45 years (range, 18–67). Fifty-one percent were males. Ablative conditioning was used in 83 pts (65%), and 78 pts (61%) underwent a related donor transplant. Ninety-one percent (N=117) of pts had a comprehensive ocular examination at approximately day 100 post SCT. Twenty pts were diagnosed with ocular GVHD at day 100 (23%). The mean tear production (in mm) on Schirmer test was significantly lower for pts with ocular GVHD compared to pts with no ocular GVHD (right eye 8.3 vs. 12.9, p=0.028; left eye 8.2 vs. 13.4, p=0.014). Schirmer test was false negative in 14/20 (60%) pts with ocular GVHD, and was false positive in 17/97 (18%) pts with no ocular GVHD. An additional 22 pts (22%) developed ocular GVHD beyond day 100 at a median of 420 days (range, 156–1504). Three-year cumulative incidence of ocular GVHD with death and relapse as competing risks was 40%. In univariate analyses, gender, preparative regimen, donor status, stem cell source and CIBMTR disease risk status were not significantly associated with ocular GVHD. Severity or pattern of organ involvement with acute GVHD (aGVHD) could not predict for ocular GVHD at day 100, except pts with skin aGVHD had a trend towards a higher incidence of ocular GVHD (21.5% vs. 9.5%, P=0.061). Pts who developed ocular GVHD beyond day 100 had significant extra-ocular GVHD prior to the onset of ocular GVHD. Recurrent acute (N=6, 27%), delayed acute (N=1, 5%), overlap (N=6, 27%), or classic chronic GVHD (N=3, 14%) preceded the development of ocular GVHD. Of those who developed ocular GVHD beyond day 100, in 6 pts (27%) the eye was the first organ to be affected by GVHD. Survival was measured from day 100 for pts never developing ocular GVHD and for those with ocular GVHD at day 100. Analyses of OS could be confounded by late onset of ocular GVHD. In order to attenuate for this time bias, OS of pts developing ocular GVHD beyond day 100 was calculated from onset of ocular GVHD. The 3-year overall survival (OS) of the cohort was 63.5%. Pts with ocular GVHD had a better survival outcome compared to no ocular GVHD (36/42 alive, 86% vs. 52/86 alive, 61%, P=0.004; mean survival-years 5.38 vs. 3.61, P=0.005, log-rank test). The 3-year OS of pts with ocular GVHD at day 100, new ocular GVHD beyond day 100 and the group of pts who never developed ocular GVHD were 80%, 85% and 47%. Relapse (N=19, 48%) and non-relapse (N=21, 52%) accounted for 40 deaths, and were not different in pts with and without ocular GVHD.

Conclusion: Ocular GVHD represents a major problem for long term survivors after allogeneic SCT and affects 40% of pts. A comprehensive ocular exam at day 100 by a dry-eye specialist should be considered, as 20% of pts have ocular GVHD at that time point. Diagnosis of ocular GVHD with Schirmer test alone is not accurate. In a third of the pts who develop ocular GVHD beyond day 100, the eye is the first organ to be affected. Ocular GVHD is a diagnostic criterion for chronic GVHD and confers a superior survival outcome. Even pts who develop ocular GVHD at day 100 seem to get this protective benefit.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author