Single Centre, Retrospective Analysis of Extracorporeal Photopheresis (ECP) Therapy in the Patients Who Are Heavily Pre-Treated for Steroid Resistant Chronic Graft-Versus-Host Disease (GVHD)

*DB and DK contributed to the work equally.

Background

Extracorporeal photopheresis (ECP) is a recommended second or later line of therapy for chronic GVHD (cGVHD), and is beneficial not only because of increased response but also for its lack of adverse effects, specifically systemic immune suppression, mainly from steroids. In this single centre study, we attempted to evaluate not only therapeutic efficacy of ECP, but also its steroid-sparing effect by regularly analyzing steroid dose per body weight. We also attempted to identify any predictors of response or survival after ECP, which was not well defined before.

Patients and methods

A total of 75 cGVHD patients (pts) who received ECP for cGVHD from 2007 to 2021 at Princess Margaret Cancer Centre were included and evaluated retrospectively. Patients and disease characteristics are as follows: median age 48.5 years (range 17-70); male 42/75 (56%); organ involvement at the time of ECP: skin (93%), oral (53%), eye (51%), gastrointestinal (25%), liver (49%), lung (57%), and musculoskeletal (n=50, 67%). Sixty-eight (91%) and 7 pts (9%) had severe and moderate grade cGVHD, respectively. Sixty-eight pts (91%) received ECP as 4 ^th line or beyond. They were heavily pretreated with prednisone (98%), cyclosporine (57%), tacrolimus (24%), mycophenolate mofetil (64%), azathioprine (65%), rituximab (7%), imatinib (8%), ibrunitib (3%) and ruxolitinib (1%).

ECP was started twice weekly for the first 12 weeks, then twice every 2 weeks in 2012-2021, while the schedule was twice every 2 weeks from 2007-2012. If there was no response or clinical benefit noted in first 24 treatments, then ECP was discontinued. In general, we attempted to provide up to around 60 sessions based on the clinicians' discretion.

The overall response rate (ORR) and clinical benefit (CB) were assessed at months 3, 6 and 12 after staring ECP. As part of standard clinical practice, NIH consensus criteria were used for grading and response assessment. CB was assessed considering clinical response as well as steroid dose reduction. Treatment failure was defined as 1) resistance to ECP requiring treatment switch, 2) non-relapse mortality (NRM), 3) relapse, 4) intolerance to ECP. Failure free survival (FFS) and overall survival (OS)were calculated from the day of ECP initiation until the endpoints of failure or death, respectively.

Results

ECP was started a median of 28 months (range 1-125) after development of cGVHD. ECP was performed a median of 35 times (range 6-174) with a median duration of 11 months (range 1-53). Out of 75 pts, 48 completed planned ECP successfully and 27 stopped due to no response or benefit including, of whom 14 required additional therapy, 1 stopped due to line infection, and 1 stopped due to relapse of AML.

With a median 72 months of follow-up, ORR was attained in 21% (16/75), 57% (36/63) and 70% (32/46) at month 3, 6 and 12, respectively. At 6 months, ORR was observed in 47-64% across all organs assessed. No difference in ORR was noted according to the cGVHD grade; at 6 months, severe cGVHD showed similar ORR (57%) to those with moderate cGVHD (60%) (p=0.893). CB was noted in 23% (17/75), 62% (39/63), and 76% (35/46) at month 3, 6 and 12, respectively.

A total of 27/75 failures (36%) and 20/75 death (27%) occurred, due to the following: ECP resistance requiring switch to other therapy (n=14, 19%), NRM (n=11, 15%), relapse of primary disease (n=1, 1%) or ECP-related complication (n=1, 1%, line infection). In the overall cohort, FFS and OS at 12 months were 68.3% and 85.9%, respectively (Figure 1).

More than a half of pts stopped steroids completely within 12 months after starting ECP. The proportion of pts off steroids was 16%, 17%, 32%, and 64% at month 0, 3, 6 and 12 after starting ECP, respectively (Figure 2).

Risk factor analysis did not show any predictive markers for ORR at 6 months, while prognostic factor analysis suggested the development of musculoskeletal involvement as favorable prognostic factor for FFS (p=0.003, HR 0.315 [0.147, 0.673]) even with multivariate analysis.

Conclusion

Our study showed that: 1) ECP is a very effective treatment for heavily pre-treated cGVHD pts who have failed other therapies; 2) More than a half of pts can stop steroids completely within 12 months after starting ECP, thus avoiding long-term toxicity risk. Further study is warranted comparing ECP with other cGVHD treatment modalities.

Figure 1

View largeDownload slide

Figure 1

View largeDownload slide

Close modal