Anal Squamous Cell Carcinoma: Choosing the Chemoradiation Backbone
A woman with cirrhosis-related cytopenias needs curative chemoradiation for anal cancer, and the trial that established the current standard may not be testing what it looks like it's testing.
R.M., a 61-year-old woman, spent thirty years cataloguing government documents at the state library before retiring, and now spends most of her time restoring her late husband's rose garden — work she has had to do in shorter stretches since a routine colonoscopy for rectal bleeding turned up a 3.5cm anal mass, biopsy-confirmed as squamous cell carcinoma, with a single enlarged inguinal node on staging PET (T2N1). Fifteen years ago she was treated and cured of hepatitis C with direct-acting antivirals, but the cirrhosis it left behind has not gone away: her spleen is enlarged on imaging, and her baseline counts — platelets 84,000, ANC 1,650 — reflect real hypersplenic sequestration, not incidental lab noise.
Standard curative treatment for anal SCC is concurrent chemoradiation, and RTOG 98-11 is the trial the current mitomycin-based standard rests on — though not in the form it is usually quoted. The primary report (Ajani et al., JAMA 2008) found mitomycin reduced colostomy failure but did not significantly improve disease-free or overall survival; the five-year disease-free and overall survival advantage everyone now cites emerged only in the long-term update (Gunderson et al., JCO 2012), after longer follow-up. But that comparator arm didn't test concurrent cisplatin against concurrent mitomycin — it gave induction cisplatin and 5-FU for two cycles, roughly six weeks, before radiation even started, then continued cisplatin concurrently. The delay to radiation, not the drug itself, is a live candidate for why that arm underperformed. ACT II tested the cleaner comparison — concurrent mitomycin versus concurrent cisplatin, no induction on either arm — and found no significant difference in complete response rate or long-term progression-free survival between them, with a materially different toxicity signature: more grade 3–4 hematologic toxicity on the mitomycin arm. R.M.'s platelets of 84,000 and ANC of 1,650 are not a contraindication to either drug; they are the reason the two trials stop being interchangeable, because only one of them measured what concurrent mitomycin does to a marrow that is already sequestering. Which trial actually answers her question is the whole argument.
Tumor board, before the radiation plan is finalized
I want to start from what actually cures anal cancer: dose-intense concurrent chemoradiation. RTOG 98-11 is still the trial our guidelines point to, and on long-term follow-up (Gunderson et al., JCO 2012) its mitomycin arm beat the comparator on five-year disease-free and overall survival — real, durable outcomes, not surrogates. I'll say plainly what that citation costs me: the primary publication (Ajani et al., JAMA 2008) found no significant survival difference at all, and the separation only appeared with more follow-up. Her counts are low, not absent. I'd rather manage transfusions and growth factor support through a proven regimen than switch drugs on a toxicity worry before we've even started.
You're right that 98-11 is the trial everyone cites — but her spleen is already doing to her counts what mitomycin's cumulative marrow suppression would do on top of it, and mitomycin's myelosuppression is notoriously delayed and cyclical, harder to manage prospectively than cisplatin's more predictable pattern. If a treatment break happens mid-radiation because her counts collapse, we've traded a toxicity risk for the exact local-control risk dose-intense RT was supposed to protect against.
I'd also flag, separately from efficacy, that mitomycin carries a rare but real hemolytic-uremic-syndrome risk — not something I'd want layered onto a cirrhotic patient's already-reduced physiologic reserve if an equally effective alternative exists.
Both of you are arguing from 98-11 as if it directly compared concurrent mitomycin to concurrent cisplatin. It didn't — the comparator arm added two cycles of induction cisplatin/5-FU that delayed radiation start before concurrent treatment ever began, and delayed RT initiation is itself a known adverse prognostic factor in anal cancer. ACT II (James et al., Lancet Oncology 2013) is the trial that actually isolates the drug: concurrent mitomycin versus concurrent cisplatin, no induction on either arm, and it found no significant difference in complete response or long-term outcomes — with meaningfully less grade 3–4 hematologic toxicity on the cisplatin arm.
So I don't think this is "proven regimen versus toxicity worry." It's two trials answering two different questions, and the one that isolates the actual variable in front of us says cisplatin is a reasonable, evidence-supported substitution — not a downgrade — for a patient whose marrow reserve makes mitomycin the higher-risk choice.
Agreed: cisplatin/5-FU with concurrent radiation, with weekly counts through treatment and a low threshold for transfusion support rather than dose reduction if platelets drift further.
Not agreed: the radiation oncologist's own reservation was stated plainly rather than smoothed over — cisplatin's efficacy equivalence rests on one trial, and if guideline committees revise their mitomycin preference downward on stronger future evidence, this case's choice will look better justified in hindsight than it currently feels in the room. For today, radiation dose-intensity was treated as the variable nobody was willing to risk, and the drug choice was the one both other voices agreed could flex around it.