Paper Details

ABSTRACT

Introduction:  Asparaginase, an essential component of acute lymphoblastic leukemia (ALL) treatment, can result in several toxicities that may lead to its discontinuation and consequently, increase relapse risk.  We aim to assess the relationship between serum asparaginase activity (SAA) levels and host genetic polymorphisms in predicting risk of developing asparaginase-associated toxicities (AAT), specifically hypersensitivity, thromboembolic event (TE) and acute pancreatitis.  

Methods: Eligible patients (pts) with newly-diagnosed ALL were enrolled on the phase 3 Dana-Farber Cancer Institute (DFCI) ALL 16-001 trial (NCT03020030) from 2017 to 2022.  All pts received the first dose of pegaspargase 2500 IU/m²/dose on Day 6 of Induction; subsequently, SAA levels were collected on Induction Days 18, 25 and 32.  Serum samples from pts who consented to optional studies were used for panel-based genotyping of top-ranking candidates. Cumulative incidence of toxicity in relation to SAA levels and genotypes was compared by log-rank test and Cox regression. 

Results: Of 560 pts enrolled on DFCI 16-001, 335 (59.8%), 324 (57.9%) and 419 (74.8%) pts had SAA levels measured on Days 18, 25 and 32 of Induction, respectively.  Median SAA levels of the cohort were 0.6670 IU/mL (IQR 0.4860 – 0.8360), 0.2990 IU/mL (IQR 0.2035 – 0.4035) and 0.0520 IU/mL (IQR 0.0180 – 0.0820) on Days 18, 25 and 32, respectively. Pts with Day 32 SAA levels in the lowest quartile were at significantly increased risk of developing grade ≥3 hypersensitivity (HR: 2.6, 95% CI 1.0-6.7, p=0.02) or switch to Erwinia-based asparaginase (HR: 3.8, 95% CI 2.4-5.9, p=1.4x 10^-8) at subsequent asparaginase exposure post-induction compared to those with Day 32 SAA levels in the higher quartiles. In contrast, pts with Day 32 SAA levels in the higher quartiles were not at significantly increased risk of developing grade ≥3 acute pancreatitis or TE compared to those with Day 32 SAA level in the lowest quartile.  However, when examining the decline between Days 18 and 32 SAA levels, pts in the highest quartile were at significantly increased risk of developing grade ≥3 TE (HR: 2.5, 95% CI 1.0-6.3, p=0.048) compared to those in the remainder quartiles.

DNA for genotyping was available for 78.4% (439/560) of pts.  Pts with ASNS rs3832526 3R3R genotype had increased risk of developing grade ≥3 hypersensitivity (HR 4.8, 95% CI 1.0–13.7, p=0.03).  Pts with HLA DRB10701 genotype had increased risk of developing hypersensitivity overall (HR 4.3, 95% CI 1.4–12.9, p=0.005), whereas those with HLA haplotypes, composed of DRB10701 and DQB10202 alleles, had a trend toward increased risk of developing grade ≥3 hypersensitivity (HR 3.0, 95% CI 0.9– 10.4, p=0.06). The combination of ASNS and HLA haplotypes also increased the risk of grade ≥3 hypersensitivity (HR 3.6, 95% CI 1.3–9.9, p=0.007). Individuals with IL16 rs11556218 had increased risk of developing TE overall (HR 6.6, 95% CI 1.5–29.6, p=0.004), whereas individuals with ADAMSTS17 rs72755233 had increased risk of developing grade ≥3 TE (HR 2.2, 95% CI 1.0–6.5, p=0.04). The ASNS rs3832526 significantly increased the risk of grade ≥3 acute pancreatitis (HR 2.6, 95% CI 1.2–5.9, p=0.02), while the ASNS rs3757676 variant was protective of developing acute pancreatitis (HR 0.3, 95% CI 0.1–0.7, p=0.006).  

Among pts with Day 32 SAA levels in the lowest quartile, two pts carried the ASNS rs3832526 3R3R genotype and both developed grade ≥3 hypersensitivity.  Among those with Day 32 SAA levels in the higher quartiles, pts who carried the IL16 rs11556218 variant were at significantly increased risk of developing TE than those without the variant (HR 10.4, 95% CI 1.9 – 57.1, p=0.0008).  Among pts with Day 32 SAA levels in the higher quartiles, homozygotes for the ASNS rs3757676 variant were less likely than those without the variant to develop acute pancreatitis (HR 0.2, 95% CI 0.07–0.7, p=0.005), and none developed grade ≥3 acute pancreatitis.

Conclusion: While prior reports have demonstrated an association with either SAA levels or variants in ASNS, HLA, IL16 and ADAMSTS17 and increased susceptibility to develop AAT, our results demonstrate that host genetic polymorphisms could modulate the risk of specific AAT among pts with different SAA levels.  To the best of our knowledge, this is the first report to illustrate a dynamic interplay between SAA levels and gene polymorphisms in developing AAT.