High Intra-Individual Variability In Systemic Exposure To 6 Mercaptopurine (6MP) In Children With Acute Lymphoblastic Leukemia (ALL) Contributes To ALL Relapse: Results From a Children’s Oncology Group (COG) Study (AALL03N1)

Prolonged (∼2y) exposure to 6MP is associated with durable remissions in children with ALL. 6MP exerts cytotoxicity through conversion to thioguanine nucleotide (TGN) that is incorporated into DNA. Erythrocyte TGN concentration correlates with cytotoxicity. Systemic exposure to 6MP is influenced by: i) prescribed dose (dose intensity [DI]), ii) patient (pt) adherence, and iii) inherited capacity to metabolize 6MP by thiopurine methyltransferase (TPMT). Previous studies have shown the individual impact of these variables, but the contribution of each of these variable measured simultaneously in determining relapse risk is unknown – and was explored in the current study. Methods: We enrolled 744 pts in 1^st remission and receiving oral 6MP (75 mg/m²/d) during maintenance. Monthly TGNs (∼6/ pt) yielded 3,953 measurements. 6MP DI (dose prescribed/ protocol dose) was ascertained for 113,682 days. TPMT genotyping classified pts as TPMT wild-type homozygotes (WT, 93.2%), heterozygotes (het, 6.7%) and homozygotes for deficient variants (def, 0.1%). Coefficient of variation (CV) was used to determine intra-individual variability in TGN (IQR=17.6-40.3), DI (3.2-28.2) and ANC (29.0-59.7). Using TGN CV% <30 vs. ≥30, DI CV% <20 vs. ≥20, and ANC CV% <50 vs. ≥50, pts were classified as having stable TGN (s₃₀TGN) vs. variable TGN (v₃₀TGN), stable DI (s₂₀DI) vs. variable DI (v₂₀DI), and stable ANC (s₅₀ANC) vs. variable ANC (v₅₀ANC), respectively. 459 of these pts contributed 75,720 pt-days of electronically-monitored (MEMS) adherence data. Adherence rate was defined as days 6MP MEMS bottle was opened/days 6MP was prescribed. Results: Median age at diagnosis was 5y (1-19); 68% were males; 43% had high-risk disease. TGN levels, 6MP DI, and cumulative incidence (CI) of relapse are summarized in Table . 59 relapses occurred after a median follow-up of 6y (CI of relapse: 8.8% at 5y). TGN levels were higher (300 vs. 140 pmol/8*10^8 erythrocytes, p<0.0001), 6MP DI was lower (0.65 vs. 0.86, p=0.0002) and risk of relapse tended to be lower (5y CI: 2.5% vs. 9.2%, p=0.1) among TPMT het/def vs. WT. Subsequent analyses were restricted to TPMT WT pts. Determinants of ALL relapse: Multivariable Cox regression analysis (adjusted for clinical factors) identified only adherence rate <95% to be associated with relapse (HR=2.5, p=0.03); neither low TGN (<140 [median]: HR=2.0, p=0.1), nor low DI (<0.86 [median]; HR=1.3, p=0.5) increased relapse risk. However, pts with v₃₀TGN had lower TGN levels (126.3 vs. 149.3, p<0.0001), lower 6MP DI (0.86 vs. 0.90, p=0.01), lower adherence rates (85.4% vs. 92.7%, p=0.0002) (Table) and higher 5y CI of relapse (12.4% vs. 5.3%, p=0.02) than pts with s₃₀TGN (Fig 1A ). Multivariable Cox regression confirmed the increased relapse risk associated with v₃₀TGN (HR=2.2, p=0.04). Similarly, pts with v₂₀DI were at a higher relapse risk than pts with s₂₀DI (5y CI: 13.4% vs. 6.7%, p=0.008, Fig 1B ); multivariable analysis confirmed this association with v₂₀DI (HR=1.8, p=0.04). Determinants of high intra-individual variability in TGN: Multivariable logistic regression analysis demonstrated that non-adherence (OR=2.1, p=0.0005) and high intra-individual variability in 6MP DI (v₂₀DI: OR=5.5, p<0.0001) were independently associated with v₃₀TGN. The variability in 6MP DI, in turn, was due to high intra-individual variability in ANC (v₅₀ANC: OR=2.7, p<0.0001; using multivariable logistic regression analysis). Conclusions: Absolute TGN levels do not help prognosticate relapse risk; instead, high intra-individual variability in TGN levels, determined by both intra-individual variability in 6MP DI and lack of adherence to 6MP contribute to relapse risk in this population. These findings reinforce the need to minimize fluctuations in 6MP prescribed dose, as well as use measures to enhance adherence.

View largeDownload slide

View largeDownload slide

Close modal

Evans: St. Jude: In accordance with institutional policy, St. Jude allocates a portion of the income it receives from licensing inventions and tangible research materials to those researchers responsible for creating this intellectual property. Patents & Royalties, Under this policy, I and/or my spouse have in the past received a portion of the income St. Jude receives from licensing patent rights related to TPMT polymorphisms as clinical diagnostics., Under this policy, I and/or my spouse have in the past received a portion of the income St. Jude receives from licensing patent rights related to TPMT polymorphisms as clinical diagnostics. Other. Relling:St. Jude: In accordance with institutional policy, St. Jude allocates a portion of the income it receives from licensing inventions and tangible research materials to those researchers responsible for creating this intellectual property. Patents & Royalties, Under this policy, I and/or my spouse have in the past received a portion of the income St. Jude receives from licensing patent rights related to TPMT polymorphisms as clinical diagnostics., Under this policy, I and/or my spouse have in the past received a portion of the income St. Jude receives from licensing patent rights related to TPMT polymorphisms as clinical diagnostics. Other.