Details

Autor(en) / Beteiligte

• Myers, Regina M
• Getz, Kelly D
• Li, Yimei
• Huang, Yuan-Shung V.
• Citrin, Rebecca
• Elgarten, Caitlin W
• Laskin, Benjamin L
• Horowitz, Joseph P.
• Fisher, Brian T.
• Aplenc, Richard

Titel

Comparative Effectiveness of Rasburicase and Allopurinol in Children with Acute Lymphoblastic Leukemia: An Emulated Pragmatic Trial Using Observational Data

Ist Teil von

• Blood, 2018-11, Vol.132 (Supplement 1), p.830-830

Ort / Verlag

Elsevier Inc

Erscheinungsjahr

2018

Link zum Volltext

Quelle

Alma/SFX Local Collection

Beschreibungen/Notizen

• Introduction: Children with newly diagnosed acute lymphoblastic leukemia (ALL) are at risk for developing tumor lysis syndrome (TLS), an oncologic emergency with potentially severe consequences. Effective TLS management centers on prevention, achieved in large part through control of hyperuricemia with allopurinol or rasburicase. Though rasburicase is known to be very effective for children at high TLS risk, there are limited evidence-based data to guide its usage. Currently available guidelines rely on expert opinion and provide conflicting recommendations, especially for patients at intermediate TLS risk. To address this data gap, we emulated a randomized clinical trial (RCT) to test the efficacy of rasburicase in intermediate risk TLS. We hypothesized that rasburicase would not provide benefit for children with ALL who were classified as being at intermediate TLS risk based on white blood cell count (WBC), uric acid (UA), and creatinine (Cr). Study Design: We used observational data from the Pediatric Health Information Systems (PHIS) administrative database to emulate a RCT assessing the effectiveness of rasburicase initiation (Hernan & Robins Am J Epidemiol. 2016). The source population consisted of the subset of an established cohort of newly diagnosed ALL patients, who were treated at 6 centers that contributed laboratory values to PHIS from 2007-2012. We mimicked rasburicase initiation trials (“pseudo-trials”) over time, where each day of the initial diagnostic admission (up to day 4) represented the start of a new pseudo-trial. For each pseudo-trial, a child was considered eligible if he received at least one dose of allopurinol or rasburicase and had a maximum WBC <100/mL, UA <10mg/dL, and Cr <1.5 times the upper limit of normal (ULN) during the 2-day baseline evaluation window. Each child could thus contribute to a maximum of 4 pseudo-trials. For each trial, children were assigned to the non-initiator arm if they received allopurinol-only, or to the rasburicase-initiator arm if they started rasburicase. Rasburicase exposed patients were not eligible for subsequent trials. Study Outcome: The primary outcome was acute kidney injury (AKI), defined as a Cr >1.5 times the ULN. Secondary outcomes were receipt of dialysis, inpatient length of stay (LOS) and mortality. Follow-up for the outcomes started on day 3 of each pseudo-trial and ended 7 days later for AKI and dialysis or 35 days later for LOS and mortality. An intention-to-treat analysis was used. To adjust for potential confounding by baseline covariates, the propensity to receive rasburicase was estimated using a logistic regression model that informed calculation of inverse probability weights (IPW). IPW-adjusted frequency tables were constructed for AKI, dialysis, and mortality. For LOS, linear regression models with IPW and generalized estimating equations for repeated trials were used. Results: Of 945 eligible children, 44 initiated rasburicase (ever-initiators) and 901 never initiated rasburicase (never-initiators). Compared to the non-initiators, ever-initiators were more likely to be male (p=0.040), non-Hispanic white (p=0.018), have private insurance (p=0.028), and have higher WBC (p<.0001), UA (p<.0001) and creatinine (p=0.019) at baseline. The final analyses included 3,175 eligible trial patients. IPW-adjusted analyses suggested no differences in frequency of AKI (2.72% vs. 0.34%, p=0.545), receipt of dialysis (0.51% vs. 0%, p=1.000), mortality (0.06% vs. 0.04% p=1.000) or inpatient LOS (mean 14.7 days vs. 11.0 days, p=0.18) between the initiator and non-initiator arms. Conclusion: Our results suggest that administration of rasburicase may not decrease the frequency of AKI, receipt of dialysis, mortality or decrease LOS for children with newly diagnosed ALL at intermediate TLS risk. Though only 44 rasburicase initiators were identified, which limited the power, these results represent data from the largest pediatric ALL cohort with both laboratory values and resource utilization. If replicated in other cohorts, such findings have the potential to impact guidelines for rasburicase use and greater standardization of rasburicase utilization across centers. Similar to previous reports, we also found that males were more likely to initiate rasburicase, a finding that deserves further exploration, given that males are more likely to have G6PD deficiency, a contraindication to rasburicase use. Fisher:Pfizer: Research Funding; Merck: Research Funding.