Image from PLOS ONE
Micrograph showing GVHD
A prospective study did not reveal the mechanism driving response to extracorporeal photopheresis (ECP) in patients with chronic graft-versus-host disease (cGVHD).
However, researchers did find that responses occurred independent of risk factors, and results suggested that regulatory T cells (Tregs) are not the dominant mechanism of response to ECP.
Madan Jagasia, MBBS, of Vanderbilt University School of Medicine in Nashville, Tennessee, and his colleagues conducted this research and detailed the findings in Biology of Blood and Marrow Transplantation.
The study was funded by Therakos, Inc., a Mallinckrodt Pharmaceuticals Company.
The study included 77 patients with cGVHD. The median age at transplant was 49, 88% of patients were white, and 62% were male.
Patients had moderate (48%) or severe (52%) cGVHD. Sites of involvement included skin (86%), mouth (52%), gastrointestinal tract (29%), eye (62%), joint and fascia (51%), genital tract (11%), and lung (28%).
Patients had received a median of 2 (range, 0 to 7) prior lines of cGVHD therapy.
For this study, patients received 1689 ECP treatments, an average of 21.9 treatments per patient. The most common regimen was ECP twice a week for 4 weeks, then twice a week every 2 weeks for 8 weeks, then further tapering at the treating physician’s discretion.
Forty-eight patients (62.3%) completed all 6 months of ECP treatment. Reasons for early discontinuation included cGVHD progression (n=6), infection (n=4), cGVHD improvement (n=2), death from cGVHD-related cause (n=2), logistical issues (n=2), loss to follow-up (n=2), unknown reasons (n=4), and other various reasons (n=7) such as a finance issue and non-adherence.
Response
Provider-assessed response rates differed from response rates according to 2005 National Institutes of Health (NIH) consensus criteria.
According to providers, the response rate was 62.3% (48/77), with 14% of patients (n=11) achieving a complete response and 48% (n=37) attaining a partial response. Nineteen percent of patients (n=15) had stable disease, 14% (n=11) progressed, and 4% (n=3) did not have follow-up for cGVHD-related reasons.
Eight patients did not have enough data to assess NIH response. For the 69 evaluable patients, the NIH response rate was 43.5% (n=30), with 6% of patients (n=4) achieving a complete response and 38% (n=26) attaining a partial response. Fifteen percent of patients (n=10) had stable disease, 38% (n=26) progressed, and 4% (n=3) did not have follow-up for cGVHD-related reasons.
Risk factors
The researchers said there was no significant difference between responders and nonresponders when it came to age at treatment, platelet count, cGVHD severity, donor gender, and donor type.
For provider-assessed response, the median Karnofsky performance score at study entry was significantly higher in responders than nonresponders (90 vs 80; P=0.001). Responders also had a significantly shorter time from transplant to study entry compared to nonresponders (1.43 vs 2.06 years; P=0.051).
Similarly, according to NIH response, time from transplant to study entry was significantly shorter for responders than nonresponders (1.23 vs 1.92 years; P=0.04).
However, in a logistic regression model, time from transplant to study entry was not associated with provider-assessed or NIH response.
Tregs
The researchers found no significant difference in Treg percentages between responders and nonresponders.
For provider-assessed response, the baseline Treg frequency was 4.4% in responders and 4.8% in nonresponders (P=0.4). Treg percentages at the end of study were 4.2% and 5.5%, respectively (P=0.2). And the change in Treg frequency was 0.3% and 1.3%, respectively (P=0.3).
For NIH response, the baseline Treg frequency was 4.7% in responders and 4.4% in nonresponders (P=0.3). Treg percentages at the end of study were 4.4% and 4.7%, respectively (P=0.6). And the change in Treg percentages was 0.3% and 0.7%, respectively (P=0.4).
These findings run contrary to the researchers’ hypothesis that response to ECP would be associated with an increase in the percentage of Tregs.
The researchers did note that the number of Tregs varied between patients, and the team raised the possibility that the mechanism of Tregs in ECP is not visible by measuring cell abundance.
However, the researchers also said future studies should explore additional mechanisms of action for ECP and look particularly at other T-cell populations, dendritic cells, inhibitory cytokines, and proinflammatory cytokines.