When would you consider additional testing for NUDT15 and other polymorphisms related to 6-MP/6-TG metabolism, if not guided by a study protocol?

My practice is to send a pharmacogenetic panel that includes TPMT and NUDT15 (as well as some other useful tidbits such as UGT1A! for Gilbert's syndrome) when patients with ALL are diagnosed.

The cost-effectiveness of this strategy has been addressed, with different conclusions (van den Akker-van Marle et al., PMID 16886902; Donnan et al., PMID 21344614), but the cost in the negative study (which considered survival) is low and the chance to decrease morbidity is significant (Stocco and Crews, PMID 21796762). Clearly, the prevalence of TPMT and NUDT15 varies by ethnicity, so a targeted strategy can be more cost-effective (Wei et al., PMID 34709580; Zarca et al., PMID 32433339), but in the melting pot of America might be hard to implement.

At diagnosis is best, but not always feasible.