Nonadherence to Oral 6-Mercaptopurine (6MP) in a Multi-Ethnic Cohort of Children with Acute Lymphoblastic Leukemia (ALL) and Its Impact On Relapse – a Children's Oncology Group (COG) Study (AALL03N1)

Nearly 20% of children with ALL relapse within 5y from diagnosis. Second-line therapies are toxic and salvage is poor. Systemic exposure to 6MP is critical for durable remissions; low systemic exposure due to nonadherence to oral 6MP could potentially increase relapse risk. We have previously reported on adherence to oral 6MP in non-Hispanic whites and Hispanics, (J Clin Oncol 2012;30:2094-101); this report extends follow-up for assessment of disease status by 68,250 person-days for non-Hispanic whites and Hispanics; it also includes adherence data for a previously unreported cohort of African American and Asian children. The goal of this report was to: i) describe adherence to oral 6MP in a multi-ethnic cohort of children with ALL; ii) identify determinants of adherence; iii) describe impact of adherence on relapse; and iv) define a clinically-relevant level of adherence needed to minimize relapse risk.

Microprocessor chips in Medication Event Monitoring System (MEMS) caps recorded date/time of 6MP bottle openings for 6 mos/ patient. Adherence rate was defined as days of 6MP bottle opening, divided by days of prescribed 6MP (removing days when 6MP was withheld for toxicity/illness from denominator). Monthly red cell thioguanine nucleotide (TGN) levels were used to demonstrate that MEMS bottle openings were accompanied by 6MP ingestion. Analyses used Generalized Estimating Equations.

462 patients (168 Hispanics; 157 non-Hispanic whites; 69 Asians; 68 African Americans) yielded 76,055 person-days of adherence data. Median age at participation was 6y (2-20); 67% were males; 40% had high-risk disease per NCI criteria; 61% reported income <$50k/y; 14% reported single-caregiver households. Among patients with normal TPMT activity, each 1% increase in MEMS-based adherence was accompanied by a 14 unit (pmol/8·10⁸ red cells) increase in TGN (p=0.01). Adherence declined from mo 1 (94.4%) to mo 6 (89.2%, p<0.0001). Multivariate longitudinal analysis revealed adherence to be significantly lower in adolescents (≥12y: 84.5% vs. <12y: 92.6%, p=0.0003, Fig A); patients from single-caregiver households (87.2% vs. 92.0%, p=0.03, Fig B); patients with low income (<$50k/y: 89.4% vs. ≥$50k/y: 93.8%, p=0.02, Fig C); and Hispanics (90.5±1.6%), Asians (85.3±3.7%) and African Americans (85.3±2.9%) compared with non-Hispanic whites (95.3±1.2%, p<0.0001, Fig D). Adherence for the adolescents (≥12y) from low-income (<$50k/y) families with single-caregivers was significantly lower when compared with that for <12-year-olds from high-income families with multiple caregivers (79.9% vs. 96.7%, p=0.0002); this difference was observed across all racial/ ethnic backgrounds Reasons for missing 6MP included forgetfulness (79%), logistical barriers (19%), and active refusal (2%). After a median follow-up of 5.4y, multivariate analysis (adjusting for clinical/sociodemographic factors) revealed that adherence <95% was associated with an increase in relapse risk (reference: adherence ≥95%; 94.9%-90%: Hazard Ratio [HR]=3.3, 95% Confidence Interval [CI], 1.0–11.6, p=0.06; 89.9%-85%: HR=3.4, 95%CI, 0.9–13.0, p=0.07; <85%: HR=4.5, 95%CI, 1.3–15.1, p=0.02), leading us to use <95% as the cut-point for adherence with a clinically unacceptable increase in relapse. Using this definition, 45% of the patients were non-adherers. The cumulative incidence of relapse was significantly higher among non-adherers (18.8% vs. 4.9%, p=0.0003, Fig E). Furthermore, non-adherers were at a 3.7-fold increased risk of relapse (95%CI, 1.4–10.2, p=0.01), after adjusting for sociodemographic/clinical variables. The adjusted risk of relapse attributable to non-adherence was 47% for this cohort that had entered maintenance in 1^st CR.

Non-adherence to 6MP is prevalent in children with ALL; 45% consume <95% of prescribed 6MP. Adolescents, Hispanic, African American and Asian children, those with low annual household income, and those from single-caregiver households are more likely to be non-adherent. Forgetfulness is the most common reason for non-adherence. Adherence rates <95% significantly increase relapse risk; 47% of relapses after entry into maintenance are attributable to non-adherence to oral 6MP. Results of this study have led to a COG-wide intervention using cell phone reminders and directly supervised therapy to enhance adherence to oral 6MP.

Relling:St. Jude Children's Research Hospital: Dr. Mary Relling receives a portion of the income St. Jude receives from licensing patent rights related to TPMT polymorphisms and GGH polymorphisms. Dr. Mary Relling receives a portion of the income St. Jude receives from licensing patent rights related to TPMT polymorphisms and GGH polymorphisms. Patents & Royalties; Sigma-Tau Pharmaceuticals: Research Funding.