ICH S9 Waives the Nonclinical Package for Advanced Cancer — Until the Cure Rate Moves

ICH S9 is the guideline most oncology nonclinical teams reach for first, and for good reason: for patients with advanced cancer, it lets a program skip studies that would otherwise sit on the critical path. It is also one of the more frequently over-applied guidelines in the oncology toolkit, because the waiver is read as a blanket exemption for "cancer drugs" rather than the narrowly scoped, population-specific allowance it actually is.

What S9 actually covers

ICH S9 reached Step 4 on October 29, 2009, and it was written for a specific population: patients with cancer whose disease is advanced, life-threatening, or severely debilitating, and whose therapeutic options are limited. The intent behind the waivers is explicit in the guideline itself — for patients in that situation, a conventional nonclinical program adds development time without adding information that changes the risk-benefit calculus for the patient, and the guideline was also written to reduce unnecessary animal use. That rationale does not travel automatically to every oncology program; it travels to the population S9 defines.

The waivers, and what they don't touch

Inside its scope, S9 is specific about which studies move and which don't:

• Carcinogenicity. Not warranted to support marketing for therapeutics treating advanced-cancer patients; S9 defers to ICH S1A for the general appropriateness question.
• Safety pharmacology. A stand-alone study is not needed to support trials in late-stage or advanced disease; where there is a specific concern, the S7A core battery or follow-up and supplemental studies should still be considered.
• Reproductive and developmental toxicology. A fertility and early embryonic development study is not warranted for trials or marketing in this population, and a pre- and postnatal toxicology study is generally not warranted either.
• Genotoxicity — the exception that proves the rule. Not considered essential to support clinical trials in this population, but genotoxicity studies are still expected to support marketing, per ICH S2. S9 trims the program; it does not zero it out.

S9 doesn't say cancer drugs skip nonclinical testing. It says a specific population's program can skip specific studies — and it is explicit about where that line sits. Why the scope boundary is the part worth reading closely

Where the waiver stops

Two boundaries matter more than the waiver list itself. First, scope: S9 does not cover healthy-volunteer studies, cancer prevention or chemoprevention programs, supportive care products, or cell and gene therapy products — an oncology CGT candidate needs its nonclinical package built against ICH S6 and the relevant CGT-specific guidance, not S9's advanced-disease framework, and often alongside the kind of biodistribution work covered in ICH S12. Second, durability: FDA's S9 Questions and Answers document addresses this directly, noting that once a high cure rate and low recurrence rate are established for a given use, carcinogenicity studies or a full reproductive and developmental toxicology program may still be expected before marketing authorization. A drug that earned S9's waivers in a late-line, short-survival setting does not necessarily keep them if the development program moves the same molecule into an earlier line with durable responses.

None of this argues against using S9 — it remains the right tool for the population it was written for, and a correctly scoped S9 program genuinely saves time a conventional nonclinical package would spend. The failure mode is treating "this is an oncology drug" as the test for applying it, instead of checking the population, the study list, and the point in development where the waiver's own logic stops holding.