Clinical characteristics and whole-exome sequencing-based genomic profiling of post-transplant lymphoproliferative disorders following allogeneic hematopoietic stem cell transplantation

Abstract

Abstract Background: Post-transplant lymphoproliferative disorder (PTLD) is a lymphoid neoplasm that arises in the context of immunosuppression following solid organ transplantation or hematopoietic stem cell transplantation (HSCT). Given its poor prognosis, PTLD is a clinically significant complication; however, its rarity has limited the accumulation of clinical and genomic data. While several studies have reported the genomic profiles of PTLD following solid organ transplantation, little is known about PTLD occurring after allogeneic HSCT. In this study, we aimed to characterize the clinical features and genomic landscape of PTLD after allogeneic HSCT to inform future therapeutic strategies. Methods: We retrospectively analyzed patients who underwent allogeneic HSCT at our institution between 2006 and 2024. Clinical endpoints included the cumulative incidence of PTLD and overall survival (OS) following diagnosis. Only cases diagnosed by tissue biopsy were included. Competing risk analysis was used to estimate the cumulative incidence of PTLD, treating death from other causes as a competing event; Gray's test was applied for group comparisons. OS after PTLD onset was estimated using the Kaplan-Meier method and compared using the log-rank test. Whole-exome sequencing (WES) was performed using biopsy specimens obtained at PTLD diagnosis. Somatic variants were called against the hg19 reference genome and filtered by the following criteria: read depth <30, strand bias <0.05 or >0.95, and intronic/synonymous variants were excluded. Identified variants were compared with previously reported single nucleotide variants in PTLD after solid organ transplantation and DLBCL in immunocompetent individuals. Results: Among 2,430 allogeneic HSCT recipients during the study period, 31 developed PTLD. Histological subtypes included 15 monomorphic cases (11 DLBCL, 1 CLL/SLL, 3 PTCL), 3 polymorphic, 4 non-destructive, 2 classical Hodgkin-like, and 7 unclassified cases. EBER in situ hybridization positivity was observed in 30 of 31 cases (96.8%). Regarding donor type, 1 patient received a transplant from a related donor, 7 from unrelated donors, and 23 from cord blood (CB) donors, with no T-cell-depleted transplantation performed. The 10-year cumulative incidence of PTLD was 1.6% (95% CI: 1.1–2.3%), with a median time to onset of 7.4 months (range: 1.9–99.3 months). There was no significant difference in incidence between cord blood and other donor types (10-year incidence: 1.7% vs. 1.7%, p = 0.82). OS at 1 year after PTLD diagnosis was 39.0% (95% CI, 21.7–55.9%). First-line therapy included rituximab monotherapy (n=17), R-CHOP (n=3), and other regimens (Brentuximab Vedotin, ABVD, etoposide, gemcitabine, or Pola-R; n=1 each); immunosuppression reduction was used in 3 cases (1 was untreated due to disease relapse; 2 were undocumented). Patients who achieved an overall (CR or PR) response to first-line therapy had a 1-year survival rate of 75.5%, compared to 8.3% in non-responders (p<0.01). WES was performed on 21 biopsy specimens from 17 patients (monomorphic PTLD: n=8 [DLBCL n=5, PTCL n=3], classical Hodgkin-like PTLD: n=2, polymorphic PTLD: n=2, non-destructive PTLD: n=2, unclassified: n=3). The most frequently mutated gene was KMT2D (n=3, 18.8%), followed by SPEN, FAS, ZEB2, MGA, TLN2, NOTCH1, NOTCH2, HMCN1, FUT5, DNMT1, and DNAH5 (each n=2, 12.5%). Pathway classification revealed frequent alterations in epigenetic regulators and transport/metabolism-related genes (n=6 each), followed by transcription factors involved in lymphocyte differentiation (n=5), NOTCH signaling (n=3), and JAK-STAT and apoptosis-related pathways (n=2 each). Conclusion: PTLD following allogeneic HSCT is a rare complication with a 10-year cumulative incidence of 1.5%, and the 1-year overall survival of affected patients was 36%, indicating a poor prognosis. Lack of response to initial therapy was significantly associated with worse prognosis. The mutation profile observed in our cohort showed partial similarity to that reported in post-solid organ transplant PTLD or DLBCL among immunocompromised hosts. However, the number of detected mutations was low, and commonly reported mutations such as TP53 and MYC were not observed. This may be due to the fact that PTLD in the setting of profound immunosuppression following allogeneic HSCT is frequently EBV-driven, and a large proportion of our cohort consisted of CB recipients, who are known to experience delayed immune reconstitution.