| Setting | Treatment |
|---|---|
| ANAL MARGIN, T1N0 well/moderately differentiated, sphincter NOT involved | WIDE LOCAL EXCISION ALONE
1 cm margin, histological clearance >1 mm without damaging the sphincter · registry data show no 5-yr OS difference vs chemoRT (n=2,243; 85.3% vs 86.8%) If margin ≤1 mm → chemoradiation
or re-excise if feasible ⚠ Does NOT apply to anal CANAL tumours |
| ANAL CANAL, T1N0 | Concurrent chemoradiation — 5-FU + MITOMYCIN C + RT
~45–50.4 Gy to the primary Elective inguinal irradiation may be OMITTED (TROG 99.02)
inguinal failure without it is only 1.9% in T1N0 vs 12.5% in T2N0 — which is why elective nodal RT starts at T2 · ⚠ precise T1N0 dose band extrapolated: RTOG 0529 had no T1N0 stratum; confirm against current NCCN |
| T2N0 | 5-FU + mitomycin C + RT — dose-painted IMRT (RTOG 0529)
50.4 Gy/28 fractions to the tumour, 42 Gy/28 fractions to the elective nodal volume — INCLUDING elective inguinal nodes · (ACT II) independently validated 50.4 Gy/28 fx as the standard backbone |
| T3–T4 and/or NODE-POSITIVE | 5-FU + mitomycin C + RT at ESCALATED dose and EXPANDED volume (RTOG 0529)
this is what materially differs from T2N0 — 54 Gy/30 fractions to the tumour; 45 Gy/30 fractions to the elective nodal volume; involved nodes ≤3 cm → 50.4 Gy, >3 cm → 54 Gy Tumours >5 cm or node-positive carry the greatest locoregional relapse risk |
| WHY MITOMYCIN — the evidence, not the eponym | ⚠ "Nigro protocol" is an EPONYM, not a citation Mitomycin is NOT replaceable by cisplatin (RTOG 98-11)
mitomycin arm: 5-yr DFS 68% vs 58% for cisplatin, with lower colostomy failure Mitomycin/5-FU + 50.4 Gy/28 fx confirmed as standard (ACT II)
n=940, 2×2 factorial · maintenance chemotherapy conferred no benefit and is not justified Capecitabine + mitomycin + RT
an accepted substitution for infusional 5-FU |
| 🔴 HIV-POSITIVE modifier, not a separate pathway | Treat with the SAME curative-intent chemoradiation, expect the SAME outcomes
full-dose RT is feasible in the ART era CONTINUE ART concurrently — do not interrupt CD4 <200 is the actionable threshold
expect more acute / late toxicity · consider dose reduction or omission of mitomycin (some series cut chemo 25% while keeping full RT dose) · countervailing data show no toxicity difference above/below CD4 200 — treat <200 as a trigger for vigilance and MDT discussion, not automatic de-escalation ⚠ Median CD4 FALLS BY HALF during the first 3 months of chemoradiation
and stays below baseline through year 1 — maintain OI prophylaxis, don't misattribute the fall to HIV progression |
| 🔴 PERSISTENT / RECURRENT after chemoradiation | ⚠⚠ DO NOT BIOPSY OR OPERATE EARLY — regression is slow
this is the commonest management error Re-evaluate at 8–12 weeks with exam + DRE
a regressing lesion is OBSERVED, not resected 6 months from treatment start is when you declare a complete clinical response (ACT II)
persistent disease = residual disease at 26 weeks · act earlier only for progression during therapy or biopsy-proven residual disease Salvage = ABDOMINOPERINEAL RESECTION
with myocutaneous flap reconstruction + permanent colostomy — re-irradiation is not an option, so surgery IS the salvage · counsel realistically: poor disease-specific and recurrence-free survival |
| METASTATIC / inoperable locally recurrent | 1st line → carboplatin (AUC 5) + weekly paclitaxel (InterAACT) plus retifanlimab (POD1UM-303 / InterAACT-2)
OS 20 vs 12.3 months vs cisplatin/5-FU, with better toxicity · FDA approved 15 May 2025 — the first and only FDA-approved first-line IO regimen for advanced anal cancer; NCCN v1.2025 lists it category 2B THE COURSE — carboplatin AUC 5 d1 + paclitaxel 80 mg/m² d1, 8, 15 + retifanlimab 500 mg d1, on a 28-day cycle
chemo STOPS at 6 cycles; the PD-1 CONTINUES ALONE to 13 cycles total (≈1 year), then stops — finite, not "until progression" ⚠ Substituting another PD-1 — use EACH DRUG'S OWN approved dose
retifanlimab 500 mg q4wk (also has a 375 mg q3wk schedule); pembrolizumab 200 mg q3wk or 400 mg q6wk — never 500 mg · mg do NOT transfer between antibodies: potency, affinity and clearance differ, and each is dosed to SATURATE the PD-1 receptor, which is the real target ⚠ Equivalent at TARGET ENGAGEMENT ≠ proven clinical equivalence
both saturate PD-1 fully — pembrolizumab has a flat exposure–response across ~5–7.5× and saturates well below 200 mg (why 400 mg q6wk was approved on PK modelling alone); retifanlimab shows full sustained occupancy at every dose tested, half-life ~18 days. But only retifanlimab has the phase-3 in this disease — swapping PD-1s is a mechanistically sound class extrapolation, not an evidence-based equivalence. Say which you mean Retifanlimab monotherapy after platinum failure (POD1UM-202)
later lines: nivolumab (ORR 24%), pembrolizumab (ORR 17%, 42% stable disease) ⚠ Ulcerating anal primary + myelosuppressive chemo = perianal neutropenic-infection risk
sitz baths + bowel regulation · NO rectal instrumentation during the nadir (suppositories / enemas / DRE seed infection) · low threshold for review (abscess / Fournier's) · consider primary G-CSF to shorten the nadir · routine antibacterial prophylaxis is NOT indicated |
| RADIOTHERAPY — technique the delivery standard for every curative setting above | IMRT is standard (RTOG 0529)
cuts acute grade ≥2 GI/GU toxicity; comparable efficacy, no isolated nodal failures in elective volumes · 3D-conformal is no longer acceptable practice ⚠ Treatment breaks WORSEN outcome
manage toxicity aggressively — don't interrupt |