6-Mercaptopurine (6MP) Intake during Maintenance for Childhood Acute Lymphoblastic Leukemia (ALL) - a Comparison of Self-Report and Electronic Monitoring: A Report from the Children's Oncology Group (COG) Study AALL03N1

Background: Adequate exposure to oral 6MP during maintenance is critical to sustain durable remissions. We have previously shown that poor systemic exposure to 6MP (as measured by poor adherence to oral 6MP) is associated with increased risk of relapse (JCO 30[17]:2094-101; Blood 124[15]:2345-53). Accurate assessment of 6MP intake is therefore critical to ensure timely interventions. Self-report is a convenient and inexpensive method to monitor 6MP intake in the clinical setting; however, literature in the non-oncology setting indicates that self-report is subject to over-reporting, and the extent of over-reporting is directly related to the number of missed doses. The accuracy of self/parent-report of 6MP intake during ALL maintenance is not known. We address this issue by comparing self-reported 6MP intake to electronic monitoring, and identify predictors of over-reporting of prescribed 6MP intake.

Methods: Participants included 416 children with ALL in first remission receiving oral 6MP (75 mg/m²/d) during maintenance. 6MP intake was measured electronically using the Medication Event Management System (MEMS – objective record), which recorded the dates/times of each 6MP bottle opening over 4 study months per patient (28 days/month). The patient (or parent if <12y-old) self-reported the number of days that 6MP had been taken over the past month at the end of each of the 4 study months (subjective record). A total of 1344 patient-months of self-report and MEMS data were evaluated. Objective and subjective records of 6MP intake were compared for each patient by study month, and patients were categorized as “perfect reporters” (self-report = MEMS report), “over-reporters” (self-report > MEMS report: ≥5 days/month for >50% of study months) and “others.” Logistic regression examined characteristics (age at study, sex, race/ ethnicity, 6MP non-adherence [MEMs-based adherence rate <95%], TPMT genotype, red cell TGN levels and 6MP dose-intensity) associated with over-reporting.

Results: Median age at study entry was 6y (2-20); 36% were non-Hispanic white; 67% were male; 38.4% had high-risk disease; 40.4% were non-adherent to oral 6MP. Subjective vs objective 6MP intake: Mean (±SD) days of 6MP ingestion per month ranged from 25.8±5.3 to 26.1±4.5 by subjective reporting vs 22.8±6.4 to 25.4±4.5 by objective measurement (adjusted mean values by generalized estimating equations analysis, Figure 1A). The correlation between self-report and MEMS by study month ranged from 0.36 to 0.58 (Table). The difference between subjective and objective reporting was 0 for 39.6% patient-months; the difference was <0 for 7.7% patient-months, and was >0 for 52.7% patient-months (Figure 1B). A total of 50 (12%) patients were identified to be perfect reporters, 98 (23.6%) as over-reporters, and 64.4% as others.  Predictors of over-reporting: Logistic regression modeling (adjusted for TPMT genotype, 6MP dose intensity and red cell TGN levels) identified the following independent predictors of over-reporting: i) Older age (Odds Ratio [OR]=1.07/y increase in patient age, 95% Confidence Interval [CI]=1.0-1.1, p=0.04); ii) Non-white race: Hispanic, OR=2.4, 95%CI=1.2-5.0, p=0.02); Asian, OR=3.1, 95%CI 1.2-8.3, p=0.02; African-American, OR=5.3, 95%CI 2.2-12.5, p<0.001 iii) Paternal education <college (OR=2.1, 95%CI=1.1-4.1, p=0.02); and iv) 6MP non-adherence (OR=8.6, 95%CI=4.7-15.9, p<0.0001) (Figure 1C).  While 77/98 (78.6%) of over-reporters were non-adherent, only 1 of the 50 (2%) of the perfect reporters was non-adherent. 

Conclusions: Over-reporting of 6MP intake is common during maintenance therapy for childhood ALL. Over-reporting was more likely in patients who were non-adherent, older, of non-white race, and came from households with lower paternal education. Subjective reporting of 6MP ingestion during maintenance therapy for childhood ALL may be unreliable, particularly in non-adherent patients, and should be used with caution.

Table. 6MP ingestion by self-report vs. MEMS

Study month	Patients (N)	Self-report (days) (mean, SD)	MEMS (days) (mean, SD)	Correlation	p-value
1	275	26.1 (4.5)	25.4 (4.5)	0.58	<0.0433
2	370	26.1 (4.3)	22.8 (6.4)	0.44	<0.0001
3	355	25.9 (4.9)	23.7 (6.3)	0.36	<0.0001
4	344	25.8 (5.3)	23.2 (6.7)	0.43	<0.0001