Eosinophilia As a Biomarker Predicting the Occurrence of Chronic Graft Versus Host Disease,

Chronic graft-versus-host disease (cGVHD) is a major complication after allogeneic hematopoietic stem cell transplantation (HSCT) known to impair quality of life and functional status and adversely affects long-term survival. Given its impact on outcomes post HSCT, biological markers predicting cGVHD appearance could be useful. Eosinophilia has been reported to occur in the setting of both acute and cGVHD. Even though many physicians refer to eosinophilia as a predictor of GVHD, it is unclear whether eosinophilia can serve as a biomarker predicting occurrence of cGVHD in the general HSCT population.

In order to evaluate whether eosinophilia can predict the development of cGVHD, we have chosen to study a cohort of HSCT patients who developed eosinophilia while being tapered off their immunosuppressive drugs.

From 01/2000 to 12/2009, a matched series of 112 patients who underwent allogeneic HSCT at City of Hope was identified. Of those, 56 patients presented with eosinophilia (>0.5×10^9 /L on two consecutive visits within 4 weeks, not related to relapse or documented allergy, drug reaction or parasitic infection) after day 120, while being tapered from their immunosuppressive drugs (case group). A control group (with no eosinophilia present on two consecutive visits within 4 weeks after day 120) was matched for the following criteria: age (±10 years), transplant year (± 2 years), HSC source (peripheral blood versus bone marrow), donor type (related vs. unrelated), female donor to male recipient, occurrence of acute GVHD and type of GVHD prophylaxis(tacrolimus/sirolimus vs. other).

A total of 112 patients were included in the analysis (Table no. 1). The most prevalent indication for HSCT was acute leukemia. Fifty seven percent of patients underwent myeloablative HSCT. There was no statistically significant difference between the case and control group for the matching criteria: age (p=0.88), HSC source, donor type, female donor to male recipient, occurrence of acute GVHD and type of GVHD prophylaxis (p=1). There was also no statistically significant difference between the groups for the different disease risk category (i.e. low risk, intermediate risk, high risk and non malignant) (p=0.4581). In the group of patients presenting with eosinophilia, 89% (50 out of 56) developed cGVHD, as opposed to 59% (33 out of 56) in the control group (p=0.0002, Chi square test). In this setting, eosinophilia was found to have a positive predictive value of 89%. In a multivariate logistic regression model, looking at the mentioned above known cGVHD risk factors, patients with eosinophilia were nearly 7 times more likely to develop cGVHD compared to those without eosinophila (p=0.0003, 95% CI 2.41–20.13).

The median time from the development of eosinophilia to first GVHD sign was 20 days. Eighty four percent of patients both in the eosinophilia and the control groups presented with extensive GVHD grading, with skin, mouth and liver being the most prevalent sites for cGVHD occurrence. With a median follow up of 51 month for the entire cohort (range 5.5–132m) 80% of patients were alive in the eosinophilia group vs. 87% in the control group.

The development of eosinophilia in the setting of tapering or discontinuation of immunosuppression may serve as a good predictor of cGVHD. Eosinophilia in this setting may be an indicator for early intervention.

No relevant conflicts of interest to declare.