Locally Advanced Rectal Cancer: A Genotype That Complicates the Stronger Regimen
The trial that made induction FOLFIRINOX attractive for threatened-margin rectal cancer never had to account for a patient whose own genes slow irinotecan clearance to a crawl.
M.D., a 49-year-old man, drives a long-haul route through three states most weeks and has been putting off a colonoscopy since his first invitation letter arrived at 45, until blood in his stool three months ago finally moved him to call. MRI staging showed a mid-rectal tumor with the mesorectal fascia involved on one side — a threatened circumferential resection margin, the finding that pushes a case toward total neoadjuvant therapy rather than surgery-first management, since a positive margin at resection is one of the strongest predictors of local recurrence in rectal cancer. Pre-treatment pharmacogenomic testing, increasingly routine before an irinotecan-containing regimen, returned UGT1A1*28/*28 — homozygous for the polymorphism that slows glucuronidation of irinotecan's active metabolite, SN-38, and substantially raises the risk of severe neutropenia and diarrhea at standard dosing.
PRODIGE 23 established induction FOLFIRINOX, given before chemoradiation and surgery, as superior to standard chemoradiation-first management for locally advanced rectal cancer — better disease-free survival and a lower rate of the kind of positive margins M.D.'s imaging is already flagging as a real risk. That trial, though, enrolled a general population and did not stratify or adjust for UGT1A1 status, which was not routine testing at the time it was designed. M.D.'s genotype is not a modest risk factor; *28/*28 homozygosity is specifically associated with markedly increased severe toxicity at standard irinotecan doses, well beyond the heterozygous *1/*28 pattern, for which the Dutch Pharmacogenetics Working Group recommends no dose reduction at all — the 70% starting dose applies only to homozygous poor metabolizers like him. Whether the trial's aggregate benefit still applies to a patient whose predicted toxicity profile sits meaningfully outside the range that benefit was measured in is the actual question the team has to answer before writing the first prescription. There is no pre-existing deficit available to blame if this goes badly — an absolute neutrophil count of 4,200, platelets of 265,000, and a bilirubin of 0.6 that rules out the Gilbert's-pattern hyperbilirubinemia which might otherwise have raised the question before genotyping did — so any toxicity that follows will be the genotype's doing and nothing else's.
Colorectal tumor board, before induction chemotherapy starts
His MRI shows a threatened margin — exactly the presentation PRODIGE 23 (Conroy et al., Lancet Oncology 2021) showed benefits most from induction FOLFIRINOX, with a real reduction in positive- margin rates at resection. From where I sit, that's the outcome that determines whether I can get him a clean operation. I'd want a real, specific reason to move away from the regimen shown to improve exactly that outcome.
The margin benefit is real, and I'm not disputing it. But PRODIGE 23's population wasn't genotyped for UGT1A1, and *28/*28 homozygosity — not the more common heterozygous pattern — is specifically associated with severe, sometimes life-threatening neutropenia and diarrhea at standard irinotecan dosing, because his own glucuronidation capacity for SN-38 is markedly reduced. Giving him the standard dose isn't the same treatment the trial tested in an average patient; pharmacologically, it's a meaningfully higher relative exposure.
This isn't a theoretical genotype-based caution — there's specific published dosing guidance for exactly this homozygous pattern, which tells me the risk is well-characterized enough to act on directly rather than simply avoid.
That's the piece I'd add: we don't have to choose between the surgeon's margin argument and the pharmacologist's toxicity argument. Pharmacogenomic guidance for *28/*28 patients — the Dutch Pharmacogenetics Working Group's 70% starting dose, and the irinotecan label's own instruction to reduce the starting dose by at least one level — specifies a reduced dose rather than omission — we can keep the three-drug regimen PRODIGE 23 actually tested, adjusted for his real, documented metabolic risk, rather than either accepting standard-dose toxicity or giving up the drug's contribution to the induction benefit entirely.
Agreed: induction FOLFIRINOX with irinotecan started at a reduced dose per published *28/*28 dosing guidance, weekly counts through the first two cycles, and a pre-specified plan to escalate the irinotecan dose at cycle three only if the first two are tolerated without grade 3–4 neutropenia or diarrhea.
Not fully agreed: the surgeon's preference for standard-dose therapy from the outset, on the reasoning that under-dosing risks under-treating the margin threat PRODIGE 23 was designed to address, was heard and recorded rather than adopted — the team's working plan escalates toward full dose only if tolerated, rather than starting there and reducing after a toxicity event occurs.