Tracking a molecular tumor board's targeted therapy recommendation through prior authorization, appeal, compassionate use, and single-patient IND
Precision oncology routinely outruns the FDA label. A tumor's genomic profile may match a drug approved for a completely different cancer type, and the molecular tumor board (MTB) — a multidisciplinary panel of oncologists, pathologists, geneticists, and pharmacists — must decide whether to recommend that drug anyway, then build a documentation package strong enough to survive a payer's medical necessity review.
The FDA approves drugs for specific indications defined by the clinical trials that supported approval — usually a single tumor type, sometimes a single biomarker-defined subgroup. But cancer biology does not respect those boundaries: an activating BRAF V600E mutation behaves similarly whether it arises in melanoma, colorectal cancer, or a glioma, and a drug engineered to inhibit that specific mutant kinase can plausibly work across all three.
Surveys of oncology prescribing (IMS Health/IQVIA claims analyses, ASCO position statements) consistently find that 30–50% of systemic anticancer drug prescriptions occur off-label — for combinations, doses, patient populations, or disease settings not on the FDA label. In pediatric oncology, where few agents carry pediatric-specific approval, the rate climbs toward 60%. This is not fringe or experimental medicine; it is standard practice, built into the culture of tumor boards and supported by professional guidelines.
The legal foundation is straightforward: the FDA regulates drug manufacturers' marketing claims, not physician prescribing. Once a drug is approved and on the market, a licensed physician may prescribe it for any indication supported by their clinical judgment. The practical obstacle is not legality — it is payment. Insurers are not obligated to reimburse for uses their policies deem "experimental" or "not medically necessary," and that is where the tumor board's documentation becomes the whole ballgame.
Medicare's statutory framework explicitly recognizes off-label oncology use as reimbursable when supported by specific compendia — most importantly the NCCN Drugs & Biologics Compendium. A drug/indication pairing rated Category 1 or 2A in the NCCN Compendium is treated by CMS, and by most commercial payers who mirror Medicare policy, as "medically accepted" even without FDA label language to match.
A strong off-label request package typically includes:
• Comprehensive genomic profiling report — the specific alteration, variant allele frequency, and the biomarker's established or plausible functional link to the drug's mechanism • NCCN Compendium citation — the specific category rating (1, 2A, 2B, or 3) for that drug/indication/biomarker combination, since this single citation carries outsized weight with payers • Clinical evidence tier — ranked from strongest to weakest: randomized basket trials (e.g., NCI-MATCH, TAPUR) → single-arm basket trial cohorts → case series → individual case reports → mechanistic/preclinical rationale alone • Tumor board meeting minutes — a signed, dated record showing multidisciplinary discussion and consensus recommendation, which payers increasingly require as proof the request is not a single physician's unilateral decision • Prior therapy history — documentation that FDA-approved and guideline-preferred options have been exhausted or are contraindicated, since payers almost universally require "failed standard of care first" • Patient-specific clinical rationale — performance status, organ function, and why this particular targeted agent, at this point in the disease course, represents reasonable medical judgment
Evidence strength varies enormously by scenario. A BRAF-mutant biliary tract cancer treated with a BRAF/MEK combination that has already shown activity in a multi-basket trial cohort is a strong, well-supported request. A rare fusion event with only a single published case report is a weak one — still potentially justifiable, but far more likely to be denied on first pass and to require the full appeal and expanded-access pathway.
Prior authorization (PA) is the utilization management checkpoint nearly every payer places between an oncologist's prescription and dispensing of a high-cost specialty drug. For an off-label targeted therapy, PA is where the MTB's carefully built rationale meets a reviewer applying payer-specific medical policy — and where administrative friction, not clinical disagreement, often determines the outcome.
The PA packet lands with a payer's utilization management (UM) unit, typically staffed by nurses following payer-specific medical policy documents — internal criteria, often derived from but not identical to NCCN or ASCO guidelines, that define what counts as "medically necessary" for that plan. A first-level nurse reviewer checks the packet against a checklist: correct diagnosis code, documented prior therapy failures, presence of a compendium citation, and completeness of the required forms.
If the request does not cleanly match a pre-built approval pathway — which is common for off-label requests, since payer software is built primarily around on-label indications — the case escalates to a medical director, typically a physician (not necessarily an oncologist) employed by or contracted to the payer. This reviewer applies clinical judgment against the same medical policy but has discretion to approve exceptions.
Turnaround time varies by plan and urgency designation: standard requests often take 3–14 calendar days; expedited/urgent requests (when the physician certifies a delay would seriously jeopardize the patient) are supposed to resolve faster — 24–72 hours under many state laws and CMS rules for Medicare Advantage — though real-world practice frequently exceeds these windows.
The American Medical Association's 2022 prior authorization physician survey found 94% of physicians report PA delays access to necessary care, and 33% report a PA delay led to a serious adverse event for a patient. For oncology specifically — where "delay" often means disease progression — this administrative friction carries outsized clinical stakes.
Payer medical policies are written years in advance and updated on a lag, typically reviewed annually or semi-annually. A newly emerging off-label use — even one already supported by a basket trial readout or a fresh NCCN Compendium update — may not yet be reflected in the payer's written policy at the time of submission. The UM reviewer, bound to apply the policy as written, may deny a request that a medical director or external reviewer would readily approve on appeal once given the chance to apply clinical judgment rather than a checklist.
This creates a structural pattern: off-label targeted therapy requests are denied at first pass more frequently than on-label requests, not necessarily because payers object to the science, but because the administrative apparatus is calibrated to on-label defaults and the initial reviewer lacks discretion to deviate from written policy. This is precisely why appeal — the next stage — succeeds so much more often than the initial denial rate would suggest: it routes the same clinical facts to a reviewer empowered to exercise judgment.
A denial is rarely the end of the story. US payers are required by law and by contract to offer structured appeal mechanisms — internal peer-to-peer review, formal written appeal, and, if internal appeals are exhausted, an independent external review conducted by a party with no financial stake in the outcome. Each cycle takes time, but each cycle also substantially raises the odds of eventual approval.
1. Peer-to-peer review: the treating oncologist requests a direct phone conversation with the payer's medical director. This is often the fastest and highest-yield step — a clinical conversation can surface context (why standard-of-care agents failed, why the biomarker match is stronger than the written chart suggests) that a paper review missed. Many denials reverse at this stage without ever reaching a formal written appeal.
2. Formal written (internal) appeal: if the peer-to-peer call does not resolve the denial, the practice submits a structured written appeal — typically requiring new or reorganized evidence, an appeal letter directly rebutting the payer's stated denial rationale, and often a second physician attestation. Internal appeals are reviewed by a different medical director than the original denial, at least in principle providing a fresh set of eyes.
3. External (independent) review: once internal appeal rights are exhausted, patients in most US states have a statutory right to an Independent Review Organization (IRO) — a third party under contract with the state insurance regulator, with no financial relationship to the payer, that renders a binding decision. External review overturns a substantial share of oncology drug denials, particularly for off-label uses supported by NCCN Compendium citations, because IROs are explicitly instructed to weigh peer-reviewed literature and compendia rather than payer-internal policy alone.
Each additional appeal cycle meaningfully raises the cumulative probability of approval — internal audits at several large cancer centers report that combined peer-to-peer plus written appeal overturns roughly half of initial specialty-drug denials, and external review overturns a comparable share of what remains. Persistence through the full three-tier ladder is, empirically, one of the single highest-leverage actions a treating team can take.
Every appeal cycle buys time against the payer's written policy — and costs time against the patient's disease. A single peer-to-peer call might resolve in days; a full internal-then-external appeal ladder can consume 6–10 weeks. For an indolent tumor, that delay may be clinically tolerable. For an aggressively progressing cancer, it is not, which is why treating teams frequently pursue insurance appeal and the manufacturer expanded-access pathway in parallel rather than sequentially — hedging against the possibility that the appeal ladder simply takes too long to matter.
State "prompt pay" and expedited-appeal laws exist precisely because of this tension: when a physician certifies that a standard appeal timeline would seriously jeopardize the patient's life or health, payers are generally required to resolve expedited appeals within 72 hours (some states: 24 hours) rather than the standard multi-week timeline.
When insurance access stalls or an agent is not yet commercially marketed for any indication (still investigational), the treating physician can approach the drug's manufacturer directly and request the product through its expanded access — colloquially "compassionate use" — program. This route bypasses the insurance reimbursement system altogether: the manufacturer supplies drug directly, typically free of charge, outside normal commercial distribution.
Expanded access is a request made directly to the company that owns the drug, not to a payer. For an already-FDA-approved drug being used off-label, this typically means asking the manufacturer to donate product through a patient assistance or named-patient program — sidestepping the need for insurance approval entirely, since the drug is given rather than billed. For a drug that is not yet approved for any indication (a purely investigational agent the MTB wants to access based on early trial data), expanded access is the only route available outside of a clinical trial, and it requires the manufacturer's sponsorship of an expanded-access IND with the FDA.
Manufacturers evaluate expanded-access requests against their own internal criteria: available drug supply (companies must balance expanded access against enrollment needs for ongoing pivotal trials), whether the request could complicate the regulatory pathway to full approval, liability considerations, and — in the case of production-constrained biologics or cell therapies — genuine physical supply limits. The 21st Century Cures Act (2016) requires manufacturers of investigational drugs to publicly post their expanded-access policy, giving physicians a starting point for outreach, though it does not compel any company to grant a specific request.
Many large academic cancer centers maintain dedicated expanded-access or "special populations" pharmacy teams whose sole job is navigating these manufacturer relationships — because the process, while it bypasses insurance, still requires its own paperwork: an expanded-access application, an independent IRB (institutional review board) approval at the treating site, and FDA notification even when the manufacturer is supplying the drug free of charge.
The federal Right to Try Act, signed into law in 2018, created an alternative to FDA-mediated expanded access for a narrower population: patients with a life-threatening disease who have exhausted approved treatment options and are not eligible for a clinical trial. Right to Try allows a physician to request an eligible investigational drug (one that has completed FDA Phase 1 testing and remains in active development) directly from the manufacturer without an FDA application step.
In practice, Right to Try has seen limited real-world use compared with the traditional FDA expanded-access/single-patient-IND pathway (covered next) — manufacturers retain full discretion to decline, the "life-threatening disease, no other options" eligibility bar is strict, and the FDA's own single-patient IND process is widely viewed by oncologists as no more burdensome while offering the FDA's additional safety oversight. Right to Try nonetheless remains available as a parallel option, particularly useful when a manufacturer is reluctant to engage with FDA paperwork but willing to supply drug under the narrower state/federal Right to Try framework.
When insurance access has failed and manufacturer expanded access is unavailable or insufficient, the treating physician can file a single-patient Investigational New Drug (IND) application directly with the FDA — formally called an "individual patient expanded access" request. Despite its last-resort position in the pathway, this route has one of the highest approval rates of any FDA regulatory mechanism.
The FDA's review of a single-patient IND is deliberately narrow: the agency evaluates only whether the potential risks of the investigational treatment are reasonable given the patient's disease severity and lack of alternatives — it does not re-adjudicate efficacy, and it does not act as a payer weighing cost-effectiveness. Because the requesting physician has already, by the time this stage is reached, assembled substantial clinical rationale (surviving MTB review, PA denial, appeal), the FDA's safety-focused bar is comparatively easy to clear.
The FDA publishes its own expanded-access statistics and has repeatedly reported authorizing the overwhelming majority — commonly cited as over 99% — of individual patient expanded access requests it receives, with only a small number modified (e.g., a dosing adjustment requested) rather than denied outright. For requests marked emergency (immediate threat to life), the FDA maintains 24/7 on-call staff and can authorize by phone in a median of about four hours, with formal paperwork filed afterward. Non-emergency single-patient requests use FDA Form 3926, a form specifically simplified in 2016 to reduce physician burden, with a target review window of 30 days that the agency frequently beats in practice.
The FDA's own summary language is direct: expanded access requests are "rarely refused" — the emphasis of the review is safety, not efficacy, and the agency has structured the pathway explicitly so that regulatory review is not the bottleneck for a dying or severely ill patient's last therapeutic option. The real bottleneck, almost always, is upstream: assembling the physician time, manufacturer cooperation, and institutional IRB approval the application still requires.
FDA authorization of a single-patient IND is necessary but not sufficient to get drug to the patient. Three additional pieces must align:
• Manufacturer cooperation — the FDA can authorize the physician to use an investigational or off-label supply, but it cannot compel the manufacturer to provide it. In practice, by the time a case reaches single-patient IND, informal manufacturer discussions have usually already begun in parallel (Stage 4), so FDA authorization typically formalizes a supply arrangement rather than initiating one from scratch.
• Local IRB approval — the treating institution's institutional review board must approve the single-patient protocol, reviewing informed consent documentation and confirming the treatment plan is ethically sound for this specific patient. Many academic centers have expedited IRB tracks specifically for expanded-access cases to avoid this step becoming its own bottleneck.
• Ongoing safety reporting — once treatment begins under a single-patient IND, the physician takes on FDA sponsor-investigator responsibilities: adverse event reporting, periodic safety updates, and eventual closeout reporting to the agency, a meaningfully heavier administrative burden than standard off-label prescribing under an approved label.
Taken together, the single-patient IND pathway illustrates a broader truth about off-label access in oncology: the scientific and regulatory system is, in the end, quite permissive about letting a well-justified, safety-reviewed treatment reach a patient — the friction lives almost entirely in the administrative layers (documentation, insurance utilization management, and manufacturer logistics) that this simulation traces from end to end.