Hasan, Riad Al

Abstract Background: Until 2022, the United Arab Emirates (UAE) lacked in-country allogenic hematopoietic stem cell transplant (allo-HSCT) services, forcing patients abroad for curative treatment while local hematologists managed pre- and post-transplant care. This fragmented model created unique challenges in graft-versus-host disease (GVHD) monitoring and survival tracking. We present the first analysis of outcomes in this underserved population. Aims: To evaluate overall survival (OS) and GVHD rates in adults receiving allo-HSCT abroad with UAE-based follow-up, establishing benchmarks for emerging local transplant programs. Methods: Retrospective multicenter study of adults undergoing allo-HSCT outside the UAE (2009–2023) with follow-up at three tertiary centers (SSMC, SKMC, Tawam). Of 454 screened patients, we excluded pediatric cases (n=245), autologous HSCT (n=76; excluded due to distinct indications/outcomes), and rare allo-HSCT indications (n=33), focusing on the most common diagnoses: acute myeloid leukemia (AML, n=54), B/T-acute lymphoblastic leukemia (B/T-ALL, n=38), and sickle cell disease (SCD, n=8). Kaplan-Meier (KM) analysis estimated 1/3/5-year OS. Observed median survival was calculated for deceased patients. GVHD grading followed NIH consensus criteria. Log-rank tests compared subgroups (donor type). Results: Survival: AML 1/3/5-year OS: 96%/79.3%/55.5% (95% CI 48–63). KM-estimated median not reached; observed median in deceased patients: 20 months. Matched unrelated donor (MUD) recipients had inferior survival vs. matched related donor (MRD)/mismatched related donor (MMRD) (p=0.03). B-ALL: 1/3/5-year OS: 92%/76.4%/69.2% (95% CI 61–77). T-ALL (n=7): 1/3-year OS: 80%/50% (limited by small sample) SCD (n=8): 100% survival at all timepoints; no GVHD observed. GVHD: Any-grade GVHD: 68.5% (AML), 73.3% (B-ALL), 71.4% (T-ALL)—exceeding CIBMTR benchmarks (50–60%). Acute GVHD 35.2% (AML), 43.3% (B-ALL), 28.6% (T-ALL). Chronic GVHD 46.3% (AML), 40% (B-ALL), 57.1% (T-ALL). Conclusion: Despite high GVHD rates, survival outcomes support selective overseas referral for AML/B-ALL patients, particularly younger recipients. SCD results highlight the safety of HSCT in non-malignant disease. These findings underscore the need for local HSCT capacity to mitigate care fragmentation and establish a critical benchmark for UAE programs. Prospective studies should evaluate socioeconomic drivers of referral patterns and GVHD disparities.

Abstract Background: Graft-versus-host disease (GVHD) remains the leading cause of non-relapse morbidity and mortality in allogeneic hematopoietic-stem-cell transplantation (allo-HSCT) recipients. Clinical scoring systems (e.g., EBMT, HCT-CI, DRI, etc.) can stratify risk but show modest predictive performance (C-statistics 0.49–0.67) when it comes to GVHD. To address this gap, we developed GVHD-Intel 1.0, a real-time, scalable, and interpretable machine learning (ML) framework from an international study to predict both acute (aGVHD) and chronic GVHD (cGVHD), providing individual and bulk predictions using common input variables and HLA alleles. Methods: This retrospective global study, approved by respective institutional review boards, included electronic health records from three tertiary care centers in Abu Dhabi, UAE, and King Hussein Cancer Center (KHCC), Jordan, from 2009 to 2023. The model was developed in phases. The UAE399 cohort (399 HSCT recipients) served as the baseline dataset. To enhance model generalizability, 55 adult allo-HSCT recipients from KHCC were added to the training set. Two distinct test cohorts were used: UAE46 (46 allogeneic HSCT recipients from UAE) and KHCC150 (150 adult allogeneic HSCT recipients from KHCC), collected later and independently of training data. Variables included donor type, sex mismatch, disease status, time from diagnosis to transplant, and agents used in conditioning and GVHD prophylaxis (e.g., TBI, ATG, alemtuzumab use etc.). HLA alleles (A, B, C, DQ, DR) were symbolically encoded and transformed via one-hot encoding. GVHD events were classified as aGVHD (<100 days) or cGVHD (≥100 days). Missing values were excluded, and class imbalance was handled via class weighting. GVHD-Intel 1.0, co-developed with MBZ University of Artificial Intelligence, features modular ensemble architecture, drug-level feature engineering, novel HLA encoding for low-resolution compatibility scoring, and built-in interpretability using SHAP and LIME. The framework is cloud-deployable for both real-time and batch predictions. Results: The GVHD-Intel 1.0 was externally tested on two independent, temporally and institutionally distinct cohorts (UAE46 and KHCC150), demonstrating strong and consistent predictive performance for both acute and chronic GVHD. For cGVHD prediction, the model achieved an AUC (area under the curve or C-statistic for binary outcomes) of 0.832, F1 score of 0.896, accuracy of 82.7%, PPV of 0.930, sensitivity of 0.864, Brier score of 0.178, and log loss of 0.543. For aGVHD, performance metrics included an AUC of 0.802, F1 score of 0.860, accuracy of 77.6%, PPV of 0.930, sensitivity of 0.799, Brier score of 0.196, and log loss of 0.581. Model calibration remained strong across both tasks, with Brier scores <0.20. Conclusions: GVHD-Intel 1.0 is a robust, interpretable machine learning framework developed from real-world, multicenter data. By integrating widely available clinical variables and HLA allele data, it outperforms conventional clinical risk scores in predicting both acute and chronic GVHD in allo-HSCT recipients. Slightly reduced performance in aGVHD prediction is likely due to the lack of relevant variables. Next steps include prospective validation across multiple transplant centers globally and continued refinement through expanded data integration. GVHD-Intel 1.0 holds strong potential as a clinically valuable, scalable decision-support tool that can be enacted widely and easily by practicing transplant physicians worldwide in all types of settings.

e13102 Background: Hereditary mutations of the homologous recombination (HR), the key DNA double-strand break repair pathway, are one of the well-known causes of breast and ovarian cancers. Few reports exist on these mutations and the utilization of PARP inhibitors in the Middle Eastern region. Methods: This is a single-center retrospective study of breast and ovarian cancer patients who underwent genetic testing for g BRCA and HR defects (HRD) respectively between August 2021 and May 2023. Data was summarized using descriptive statistics and compared using counts and percentages. Response to treatment was assessed using RECIST criteria. Results: The prevalence of pathogenic mutations was 22.8%, with 26.3% of breast cancers harboring g BRCA mutations and 24% of ovarian cancers showing HRD positivity. Middle Eastern patients comprised 40% of the g BRCA-positive breast cancer group, while Asians accounted for 66% of the HRD-positive ovarian cancer patients. In mutated breast cancer, 53.3% had a positive family history of cancer. 33.3% of mutated breast cancer group and 66.6% of the ovarian cancer group received PARPis. Objective response to PARPi was 40% and 100% in breast and ovarian cancers, respectively. Conclusions: Our study revealed a high prevalence of g BRCA mutations (26.3%) in breast cancer and HRD positivity (24%) in ovarian cancer). Ethnic profiles differed between these cancers, with Middle Eastern and Southeast Asian patients dominating the breast and ovarian groups, respectively. Interestingly, the TNBC rate in breast cancer remained around 20% regardless of g BRCA status. While disparities between HRD groups were more pronounced in ovarian cancer, as expected, unexpected trends emerged in breast cancer. Further research with larger cohorts is needed to confirm these findings. [Table: see text]