Abstract
Background
Hepatic veno-occlusive disease (VOD) is a rare complication of hematopoietic cell transplantation (HCT) that is associated with high mortality. Developing a clinical risk score to identify patients at high risk for VOD would inform clinical decision-making and aid efforts to study preventive strategies.
Methods
We retrospectively analyzed data on 13,097 patients receiving allogeneic HCT as reported by 221 centers to the Center for International Blood & Marrow Transplant Research between 2008-2013. All ages, disease types (malignant, non-malignant), and conditioning regimen intensities were included. The cohort was randomly divided into training and validation sets. The primary outcome was development of VOD by day+100 post HCT with or without multi-organ failure as reported by the reporting institutions. A multivariate logistic regression model was built using the training set to identify independent prognostic factors associated with the primary outcome. Age was analyzed as a continuous variable within 5 strata (Table 1). A risk score was constructed in the training set using the significant factors from the regression model: the magnitude of the odds ratios (OR) determined the magnitude of the risk score term coefficients. Risk scores were calculated by addition of the risk score coefficients corresponding to each patients' characteristics. The score's predictive ability was confirmed in the validation set.
Results
Baseline characteristics are shown in Table 1. The training and validation sets were similar across all characteristics. VOD incidence at day+100 was 4.9% (n=637). VOD diagnosis was established by biopsy/autopsy (n=139), Seattle criteria (n=115) or Baltimore criteria (n=289) in 543 patients (85% of patients). Diagnostic information was missing on 3 patients. The remaining 91 patients were diagnosed using other clinical evidence (e.g. ascites, ultrasound). The multivariate model identified 6 independent prognostic variables: age, hepatitis B/C (HBV/HCV) serology, Karnofsky performance score (KPS), disease type/status, conditioning regimen and sirolimus use. See Table 1 for non-significant variables assessed. Age was inversely associated with development of VOD across all strata; only the <10 and 20-40 strata were statistically significant (Figure 1). For ages <10 years, 10-20, 20-40, 40-60, and 60-80 the ORs were 1.10 (95% confidence interval [CI] 1.03-1.17, p=0.004), 1.01 (95% CI: 0.95 - 1.07, p=0.81), 1.04 (95% CI: 1.01-1.08, p=0.01), 1.01 (95% CI: 0.98-1.04, p=0.61), and 1.05 (95% CI: 0.96-1.15, p=0.28), corresponding to a 10%, 1%, 4%, 1% and 5% increased risk of VOD for each year decrease in age within their respective strata. Myeloablative (MAC) regimens were associated with higher risk of VOD, and busulfan-based (BU) MAC regimens guided by pharmacokinetic (PK) monitoring were associated with higher risk than those without PK guidance. KPS <90% (OR 1.47, 95% CI 1.11-1.95, p=0.007), HBV+ serology with or without HCV+ serology (OR 2.19, 95% CI 1.35-3.56, p=0.002), and use of prophylactic sirolimus (OR 2.39, 95% CI 1.54-3.71, p = 0.0001) were associated with higher risk. Within the training set, each patient's risk score was calculated and the patients were stratified into 4 groups according to their risk score percentile: low (A, ≤40th percentile), intermediate (B, >40th, ≤80th percentile), high (C, >80th, ≤90th percentile) and very high risk (D, >90th percentile) with ORs for development of VOD in the training and validation sets as shown in Table 2 and Figures 2 and 3. The incidences of VOD within groups A, B, C and D in the training set were 1.15%, 4.34%, 8.70%, and 17.84%, respectively; the corresponding incidences in the validation set were 1.96%, 4.43%, 9.72%, and 14.33%, respectively. There was no significant difference in the c statistic by set (p = 0.07).
Conclusion
The risk score successfully stratified allogeneic HCT patients by risk of developing VOD and was validated in an independent subset of patients. The model demonstrates strong discriminatory ability to identify a high risk cohort to focus efforts for further study.
Villa:Jazz Pharmaceuticals: Employment, Equity Ownership.
Author notes
Asterisk with author names denotes non-ASH members.