Late and Classic aGVHD Differences Do Not Warrant Separate Approaches
Long-term outcomes did not differ among de novo late acute graft-vs-host disease and classic acute graft-vs-host disease, showing they should be treated the same way.
Although de novo late acute graft-vs-host disease (aGVHD) demonstrated more severe clinical and biomarker parameters and lower response to treatment than classic aGVHD, long-term outcomes did not differ among these 2 types of aGVHD and they should be treated the same way, according to data presented at the 2023 Transplantation & Cellular Therapy Meetings.
“Other than timing, there are no important differences between late and classic aGVHD,” Yu Akahoshi, MD, said in his presentation. “Late aGVHD should not be considered a distinct category and patients with late acute GVHD should be treated the same as classic acute GVHD patients including eligibility for clinical trials.”
Akahoshi, a postdoctoral fellow at Icahn School of Medicine at Mount Sinai Tisch Cancer Institute and transplant physician at Jichi Medical University Saitama Medical Center, Japan, explained that late aGVHD is classified as GVHD occurring over 100 days after an allogeneic hematopoietic cell transplant (HCT) whereas the classic type occurs between the day of HCT and 100 days after.
To assess the clinical outcomes, biomarker impact, and incidence of late and classic aGVHD, the investigators looked at 3721 adult patients who received their first HCT at 1 of 24 Mount Sinai Acute GVHD International Consortium (MAGIC) centers across North America, Europe, and Asia between January 2014 and August 2021. Of those patients, 3542 met the inclusion criteria, with 59 excluded for ex-vivo T-cell depletion and 120 excluded for umbilical cord blood transplantation.
In total, 1857 patients (52.8%) had aGVHD. Late aGVHD made up 7.5% of these patients and the other 45.3% had classic aGVHD. There were 193 patients (5.7%) who received systemic therapy for late aGVHD, 1245 (35.2%) for classic aGVHD, and 1438 (40.9%) total. The median onset of late aGVHD was day 151 (range, 102-330), and all late aGVHD occurred within a year of HCT, according to Akahoshi.
Compared with classic aGVHD, late aGVHD was more severe in patients with grade III or IV disease, at 17.1% with late vs 36.3% with classic (P < .001). By Minnesota aGVHD risk score, 15.0% of patients with classic disease had high risk compared with 30.1% of patients with late disease (P < .001). Similarly, 15.2% vs 32.6% had lower gastrointestinal stage 2 to 4 disease, respectively (P < .001).
At 28 days after initial systemic treatment, the overall response rates (ORRs) were significantly lower in those with late aGVHD than with classic aGVHD because there was more severe clinical presentation at the beginning of late aGVHD. There was an ORR of 72.0% in the classic group and 55.4% in the late group (P < .001). There were also fewer patients with classic than late aGVHD who received second-line treatment by day 28, at 14.8% vs 22.8%, respectively (P = .006).
To evaluate the impact of aGVHD on nonrelapse mortality (NRM), the investigators treated aGVHD as a time-dependent covariate. “When we analyzed late aGVHD, only patients without a prior history of a classic aGVHD who survived without relapse at day 100 were included. The development of both classic and late aGVHD increased significant NRM and resulted in poor overall survival,” Akahoshi stated.
One year after patients were treated with systemic therapy, there was no significant difference between rates of NRM for classic and late aGVHD (P = .551), although late aGVHD had more severe presentation at onset. In a multivariate analysis of the impact of GVHD on NRM over time using a Simon-makuch plot, classic aGVHD had a HR of 2.53 (95% CI, 2.13-3.02; P < .001) which was lower than that of late aGHVD (HR, 4.37; 95% CI, 3.03-6.30; P < .001). At 6 months, grades I to IV aGVHD were similar in NRM as well.
Previous data on MAGIC algorithm probability (MAP) used the GVHD biomarkers ST2 and Reg3α to determine 6-month NRM and could potentially risk stratify patients with classic aGVHD at treatment and 1 week after treatment. Investigators hypothesized that MAPs could also predict outcomes for late aGVHD. This assessment using MAPs showed patients with late aGVHD had greater cumulative incidence of NRM than classic aGHVD at time of treatment. After 1 week of treatment, the rate of high-risk MAP biomarker were higher with late aGHVD compared with classic aGVHD, and high MAP score was associated with increased NRM risk.
Statistically significant risk factors for late aGVHD were determined to be transplant recipient age of 55 years or older (HR, 1.84; 95% CI, 1.19-2.88; P = .007), transplant sex mismatch of female to male (HR, 1.53; 95% CI, 1.09-2.14; P = .014), use of reduced intensity conditioning regimen (HR, 1.61; 95% CI, 1.13-2.31; P = .008), and use of GVHD prophylaxis of triple post-transplant cyclophosphamide (HR, 0.28; 95% CI, 0.13-0.58; P < .001).
Though late aGHVD was more severe by clinical and biomarker parameters and associated with higher NRM, long-term data showed outcomes did not differ for late and classic aGVHD and these subcategories should not be treated differently.
