BACKGROUND: aGVHD occurs unpredictably following ASCT and is a significant cause of morbidity and mortality. Biological factors predicting organ involvement and severity remain to be elucidated. Tregs are a suppressive subset of CD4+ T lymphocytes that are implicated in the prevention of aGVHD. Tregs can express the gut-homing integrin α4β7 or the skin-homing marker cutaneous lymphocyte antigen (CLA), thus potentially conferring organ-specific immune regulation. We hypothesized that at engraftment, increased frequencies of Tregs with either a gut- or skin-homing phenotype are associated with a decrease in gut or skin aGVHD, respectively.

METHODS: Patients (pts) undergoing a T cell replete ASCT were enrolled. Overall aGVHD grade and organ stage was determined using the modified Glucksberg criteria. Initial, recurrent, and maximal stage/grade of aGVHD was analyzed. The frequency of CD4+ Tregs, defined as CD25+ Foxp3+CD127lo was quantified from recipient CD4+ T-lymphocytes at the time of neutrophil engraftment using polychromatic flow cytometry. Tregs with a gut- or skin-homing phenotype were identified by the presence of α4β7+ or CLA+, respectively. These subsets were expressed as the percentage of the total Treg population.

RESULTS: Forty-three pts underwent ASCT with a matched related donor [N= 26 (60%)] or an unrelated donor [N= 17 (40%)]. The median age at ASCT was 45 years (range, 21–70). Ablative conditioning was utilized in 26 (60%) pts and a reduced intensity regimen was performed in 17 (40%) pts. The stem cell source was peripheral blood (PB), bone marrow (BM), or umbilical cord blood (UCB) for 32 (74%), 9 (21%), and 2 (5%) pts, respectively. Pts received aGVHD prophylaxis with a calcineurin inhibitor and either methotrexate [N=25 (58%)] or mycophenolate mofetil [N=18 (42%)]. aGVHD grade II–IV or grade III–IV occurred in 34 (79%) and 9 (21%) pts, respectively at a median of 23 days post ASCT (range, 9–93). The gastrointestinal (GI) tract was involved in 29 (67%) pts followed by skin in 20 (47%) pts. Twenty (47%) pts previously diagnosed with aGVHD had recurrent grade II–IV aGVHD. Treg engraftment samples were obtained at a median of 19 days post ASCT (range, 10–31). The median percentage of Tregs and their CLA+ skin- or α4β7+ gut-homing subsets were: 4.9% (range, 0.80–35.4%), 1.9% (range, 0.12–9.79), and 16.2% (range, 2.7–47.5), respectively. The percentage of total Tregs at engraftment between pts with or without grade II–IV aGVHD was not different (4.4% vs. 5.0%; P=0.40). Pts with any stage skin aGVHD had a significantly lower frequency of CLA+ Tregs at engraftment (1.7% vs. 2.6%; P=0.04). A trend was noted with regards to the association between α4β7+ Tregs and the occurrence of gut aGVHD [15.2% vs. 20.1%; P=0.11 (gut aGVHD vs. no gut aGVHD)]. Increasing frequencies of CLA+ Tregs and α4β7+ Tregs were associated with lower stages of skin [odds ratio (OR), 0.68; 95% confidence interval (CI), 0.46–0.99; P=0.04] and less severe gut aGVHD (OR, 0.94; 95% CI, 0.88–0.99; P=0.04), respectively. Pts with recurrent grade II–IV aGVHD, any recurrent skin aGVHD, and any recurrent gut aGVHD tended to have lower percentages of total Tregs, CLA+ Tregs, and α4β7+ Tregs when compared to pts without recurrent aGVHD, however these values were not statistically significant (4.1% vs. 7.5%; P=0.07), (1.8% vs. 2.2%; P=0.08), and (10.4% vs. 18.7%; P=0.07), respectively. Pts with recurrent gut aGVHD had significantly lower Tregs (3.1% vs. 7.5%; P<0.01). Using multivariable logistic regression (adjusting for donor type) Tregs became a significant predictor for recurrent grade II–IV aGVHD (OR, 0.88; 95% CI, 0.78–0.99; P=0.04] and CLA+ Tregs conferred a protective effect with the prevention of any stage skin aGVHD (OR, 0.67; 95% CI, 0.45–0.98; P=0.04).

CONCLUSIONS: CLA+ Tregs at engraftment are associated with the prevention of skin aGVHD, while total Tregs predict for recurrence of both grade II–IV aGVHD and any stage gut aGVHD. Increased frequencies of CLA+ Tregs and α4β7+ Tregs are associated with less severe skin or gut aGVHD, respectively. Tregs and their subsets could help explain the clinical heterogeneity of aGVHD with respect to aGVHD incidence, severity, and patterns of organ involvement. In the future, Tregs at engraftment could be used to risk stratify pts prior to aGVHD onset for clinical trials examining novel immunosuppressive regimens aimed at increasing Tregs and preventing aGVHD.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author

Sign in via your Institution