FDA Surrogate Endpoint Table: How to Use It Without Assuming Qualification

For clinical development teams designing confirmatory trials, selecting an efficacy endpoint is a high-stakes design decision. When considering a surrogate measure rather than a direct clinical outcome, sponsors frequently consult the Table of Surrogate Endpoints That Were the Basis of Drug Approval or Licensure, published jointly by the Center for Drug Evaluation and Research (CDER) and the Center for Biologics Evaluation and Research (CBER). However, teams routinely fall into an expensive regulatory trap: treating an entry on this table as an approved 'catalog,' an evidentiary guarantee, or a portable validation that can be lifted into a new investigational program without dialogue with the agency.

The central reader decision is straightforward: how should a sponsor use FDA's surrogate endpoint table as a structured discussion map across disease, population, mechanism, and approval type, without mistaking a historical row match for formal qualification or universal scientific validation?

To answer that question, development teams must draw clear methodological boundaries. Evaluating a surrogate table row is not a therapeutic-area-specific case review, nor does it replicate the protocol-level risk analysis required by ICH E8(R1) Critical-to-Quality Endpoints. It does not address statistical multiplicity management across secondary endpoints, as examined in Multiple Endpoints in Clinical Trials, nor does it define the estimand strategies or sensitivity models detailed in ICH E9(R1) Estimands and Statistical Analysis Plans for Endpoints. Furthermore, surrogate biomarkers must never be confused with Clinical Outcome Assessments (COAs), which capture how a patient feels, functions, or survives directly through patient, clinician, observer, or performance reports—a distinction formalized in COA Type Selection and Fit-for-Purpose COAs under PFDD Guidance 3.

Instead, using the surrogate table requires understanding its statutory origin, deciphering its five standardized columns, respecting its qualifying footnotes, separating traditional from accelerated approval pathways, and knowing precisely when to request a formal PDUFA Type C novel-surrogate consultation.

The Table of Surrogate Endpoints is not an informal agency summary; it is a statutory instrument established under Section 3011 of the 21st Century Cures Act of 2016, which codified Section 507 of the Federal Food, Drug, and Cosmetic Act (FD&C Act; 21 U.S.C. § 357). Section 507 established a formal regulatory framework for the Qualification of Drug Development Tools (DDTs).

Crucially, Section 507(c)(1)(E) mandates that the Secretary make publicly available, and update on at least a biannual basis, two distinctly numbered transparency lists:

• List (i) — Qualified Drug Development Tools: A comprehensive list of all drug development tools (including biomarkers and COAs) formally qualified under Section 507(a) for a specified Context of Use (COU). Qualification is for that COU and may be used for that COU by any person in drug or biologics development. Qualification is not required to use a DDT in an individual IND, NDA, or BLA; non-qualified tools are reviewed in those applications.

• List (ii) — Approval-Basis Surrogate Endpoints: A comprehensive list of all surrogate endpoints which were the basis of approval or licensure of a drug or biological product under Section 505 of the FD&C Act or Section 351 of the Public Health Service Act (PHS Act), explicitly including accelerated approvals granted under Section 506(c) (21 U.S.C. § 356(c)).

The Table of Surrogate Endpoints is List (ii), not List (i). Listing on the surrogate table does not constitute a Section 507(a) qualification determination. While qualification grants portable regulatory standing for a defined context of use, the surrogate table is an informational transparency register of what review divisions have accepted in historical marketing applications or indicated acceptance of in published guidance.

Section 507(e)(9) of the FD&C Act establishes the binding statutory definition of a surrogate endpoint: a marker, such as a laboratory measurement, radiographic image, physical sign, or other measure, that is not itself a direct measurement of clinical benefit, and:

1. (A) Traditional Approval: is known to predict clinical benefit and could be used to support traditional approval; or

2. (B) Accelerated Approval: is reasonably likely to predict clinical benefit and could be used to support accelerated approval under Section 506(c).

This statutory binary governs the structure of every row in the public register, dictating both the evidentiary burden required to support marketing approval and the postmarketing legal obligations imposed on the sponsor.

The surrogate endpoint register is organized into four distinct sub-tables on FDA's portal and in its downloadable spreadsheet:

• Adult Non-Cancer Related Endpoints

• Adult Cancer Related Endpoints

• Pediatric Non-Cancer Related Endpoints

• Pediatric Cancer Related Endpoints

Across all four tables, CDER and CBER apply five standardized column headings, with pediatric tables incorporating a dedicated Age Range column to capture developmental and maturational windows:

In a 2022 methodological analysis of the surrogate table's construction (Amur et al., Frontiers in Pharmacology), FDA authors explained that verbatim label indications were deliberately generalized to prevent the table from becoming an unmanageable compendium. Highly specific population descriptors, visit schedules, response rates, and analytical protocols were intentionally omitted. Oncology rows represent the most heavily generalized entries in the fleet.

Consequently, a development team that finds a verbatim string match for its candidate drug in the Disease or Use and Surrogate Endpoint columns has not found a locked protocol definition. The table does not specify whether a viral load endpoint requires undetectable levels at 24 weeks or 48 weeks, nor does it define whether tumor shrinkage must be durable for 6 months. Operationalizing a table row requires pulling the original NDA/BLA review documents, the product's Summary Basis of Approval, and applicable disease-specific FDA guidance documents.

The footnotes appended to table entries establish strict operational conditions that govern row interpretation:

• Footnote * (Mechanism-Agnostic): FDA's mechanism-agnostic marker: many mechanisms of action are associated with the surrogate, so it is not directly related to a particular causal pathway.

• Footnote § (Oncology Tumor Burden Context): Endpoints based on changes in tumor burden may be used for traditional or accelerated approval depending on context of use, including disease, effect size, effect duration, residual uncertainty, and benefits of other available therapy. Dual labeling is not automatic dual eligibility.

• Footnote ˟ (Anticipated / Not Yet Approved): Highlights surrogates that FDA anticipates could be appropriate as primary efficacy endpoints, even though they have not yet supported an approved NDA or BLA. Relying on a footnote-˟ row represents higher regulatory risk than relying on an approval-tested row, requiring thorough pre-trial alignment.

• Footnote ¤ (Bone Mineral Density Restrictions): Bone mineral density is an acceptable primary endpoint for establishing efficacy for the treatment of male or glucocorticoid-induced osteoporosis after efficacy based on new morphometric vertebral fractures has been established in postmenopausal women.

• Footnote † (Clinical Endpoints Required): Clinical endpoints were required for the approvals. The table does not further specify whether those clinical measures were co-primary or supportive.

• Footnote # (Biomarker Composites): Identifies composite endpoints composed of multiple biomarkers. Mixed composites that combine biomarkers with clinical outcome assessments or functional scores are excluded from the table entirely.

The entry in the Type of Approval Appropriate for column dictates the statutory standard under which the marketing application will be evaluated:

When a row indicates Traditional, FDA is labeling that listed use as a surrogate known to predict clinical benefit under Section 507(e)(9)(A), the statutory counterpart of a BEST Validated Surrogate Endpoint in a defined context. BEST describes a validated surrogate as supported by a clear mechanistic rationale and clinical data providing strong evidence that an effect on the surrogate predicts a specific clinical benefit, and as able to support marketing approval in that defined context without additional studies to demonstrate the clinical benefit directly. The Type of Approval column is that row-level split. It is not a BEST qualification letter, and it is not proof that the same measure is validated for a new disease, population, or mechanism.

A Traditional row is therefore not the 21 CFR 314.510 / 601.41 confirmatory-study track used for reasonably likely surrogates. It still does not lock visit windows, assays, or analysis rules. Illustrative April 29, 2026 table readings—not disease-specific development advice—include Hypertension / Patients with hypertension / Blood pressure / Traditional / Mechanism agnostic; Hypercholesterolemia / Serum LDL cholesterol / Traditional / Lipid-lowering; Hepatitis C / Sustained viral response (HCV-RNA) / Traditional / Antiviral; and HIV-1 / Patients with HIV-1 / Undetectable plasma HIV RNA / Traditional / Antiviral.

When a row indicates Accelerated, the surrogate is governed by Section 506(c) of the FD&C Act (21 U.S.C. § 356(c)) and 21 CFR 314.510 (for drugs) or 21 CFR 601.41 (for biologics). It is classified in the BEST glossary as a Reasonably Likely Surrogate Endpoint—supported by robust pathophysiological, epidemiologic, or therapeutic rationale, but lacking definitive clinical trial evidence that effects on the marker translate directly to prolonged life or improved functioning.

Accelerated approval is conditional. Under 21 CFR 314.510, approval is subject to the requirement that the applicant study the drug further, in adequate and well-controlled trials carried out with due diligence, to verify and describe clinical benefit where there is uncertainty as to the relation of the surrogate to clinical benefit. If confirmatory trials fail to verify clinical benefit, FDA can withdraw the indication. Expedited-withdrawal procedures described in the December 2024 draft remain draft until that guidance is finalized.

Rows marked Accelerated/Traditional (predominantly in oncology and accompanied by footnote §) do not grant sponsors automatic discretion to choose their preferred pathway. Rather, the approval pathway is determined by the totality of clinical evidence: a substantial, durable complete response rate in a treatment-refractory setting may support traditional approval, whereas a modest, transient response rate in an earlier line of therapy may only support accelerated approval.

Furthermore, Section 506(c) distinguishes surrogate endpoints from intermediate clinical endpoints—clinical measurements that can be assessed earlier than irreversible morbidity or mortality (IMM) and are reasonably likely to predict an effect on IMM or other clinical benefit. An intermediate clinical endpoint directly measures a clinical event (such as a non-fatal stroke or progression-free interval), whereas a surrogate biomarker measures a biological or physical proxy. Development teams must not conflate the two.

The fundamental barrier preventing a sponsor from 'borrowing' an entry from the surrogate table is the principle of Context of Use (COU). In regulatory science, a biomarker is never validated in the abstract; it is validated solely for a specific drug mechanism, disease staging, patient population, and clinical setting.

There are three primary scientific reasons why a listed surrogate cannot be assumed transferable to a new clinical program:

1. 1. Mechanism-Specific Dissociation: A drug may alter a biomarker through an off-target or non-canonical biological pathway that fails to affect the downstream disease process, or it may induce toxicities that negate biomarker gains. Historical putative-surrogate failures—not current table rows—illustrate the point: in CAST, suppression of ventricular arrhythmias was followed by increased mortality, and torcetrapib raised HDL cholesterol while increasing cardiovascular deaths.

2. 2. Disease Stage and Treatment Landscape: A surrogate accepted in late-stage, refractory disease where no standard of care exists may be rejected in early-stage disease where available therapies offer prolonged survival. As available therapy changes, the risk-benefit tolerance shifts, and previously accepted surrogates may no longer be considered sufficiently robust.

3. 3. Absence of Published Evidentiary Dossiers: Independent analysis by the Federation of American Scientists notes that FDA's public table does not include the strength of evidence for each marker's association with a clinical outcome or a justification for surrogate selection. The table reveals that a surrogate supported an approval, not how much evidence was required. Sponsors cannot replicate an evidence package that is not published.

When a development team proposes using a biomarker that has not previously served as the primary basis of approval in the proposed context of use—whether it is a completely novel biomarker, a listed biomarker in a new indication, or an established biomarker evaluated with a novel mechanism of action—FDA has established a formal early-consultation pathway: the PDUFA Type C meeting for a biomarker not previously used as the primary basis of approval in the proposed context of use.

This specialized Type C track is designed to evaluate the feasibility of the surrogate as a primary efficacy endpoint, identify evidentiary gaps, and align on the preclinical and clinical studies required before Phase 3 protocol lock.

Unlike ordinary Type C meetings, a Type C requested as an early consultation on a new surrogate has additional package rules:

• Exception 1: Complete Package Due at Request: Ordinary Type C packages follow the August 2026 Formal Meetings tables (typically due before the meeting, not with the request). For a Type C requested as an early consultation on a new surrogate as the primary basis of approval in a proposed context of use, the complete meeting package is due at the time of the meeting request. A request that omits that package does not qualify for this consultation.

• Exception 2: Preliminary Human Data Required: To qualify for this consultation, the package must include preliminary human data showing that the drug affects the biomarker at a dose that appears generally tolerable. Preclinical in vitro or animal data alone do not meet that package rule.

• Exception 3: Explicit Cover Letter Identity: The meeting request cover letter must explicitly declare that the purpose of the meeting is to discuss a new surrogate endpoint not previously used for accelerated or traditional approval in the proposed context of use.

In its guidance document Considerations for Discussion of a New Surrogate Endpoint(s) at a Type C PDUFA Meeting Request (a 2018 discussion guide, not a form), FDA lists topics sponsors should incorporate into a comprehensive background package:

1. Clinical Outcome Definition: Precise description of the targeted clinical outcome (feeling, functioning, or survival) that the surrogate is intended to substitute for.

2. Surrogacy Rationale: The therapeutic or clinical rationale for proposing a surrogate rather than measuring the primary clinical outcome directly (e.g., extreme trial duration, rare disease event frequency, or irreversible disease progression).

3. Evidentiary Relationship: Comprehensive preclinical, epidemiologic, natural history, and clinical data linking changes in the biomarker to changes in the target clinical outcome.

4. Biological and Causal Pathway: The biological plausibility and causal mechanism connecting drug target engagement, biomarker modulation, and disease pathophysiology, including analysis of potential off-target pathways.

5. Threshold and Duration: The magnitude, cut-off threshold, and duration of biomarker response required to reliably produce meaningful clinical benefit in the target population.

6. Analytical Assay Validation: Comprehensive assay performance data, including analytical precision, accuracy, reproducibility, limits of detection and quantitation, dynamic range, sample stability, and standardized collection procedures across trial sites.

7. Subgroup Consistency: Evidence demonstrating that the biomarker response and its relationship to clinical outcomes remain consistent across patient demographics, disease severities, and clinical subgroups.

To prevent regulatory misalignment during clinical protocol design, the following matrix compares the four primary regulatory statuses an endpoint can hold under FDA review:

The operational decision sequence for evaluating a candidate surrogate endpoint is summarized in the following workflow diagram:

flowchart TD
    Start["Candidate primary endpoint identified"] --> CheckCOA{"Is the measure a direct clinical outcome (PRO, ClinRO, ObsRO, PerfO)?"}
    CheckCOA -- Yes --> RouteCOA["Evaluate under PFDD Guidance 3; not a surrogate-table row"]
    CheckCOA -- No --> MatchTable{"Does the biomarker match a current table row across columns?"}
    MatchTable -- Match found --> EvalCols["Read Disease or Use, Patient Population, Surrogate, Type of Approval, and Mechanism"]
    EvalCols --> CheckFootnote["Read footnotes, including anticipated-not-yet-used and tumor-burden context"]
    CheckFootnote --> CheckContext{"Do disease, population, mechanism, and available therapy match that row's setting?"}
    CheckContext -- Match --> DivisionReview["Discuss the row with the review division before locking it as primary"]
    CheckContext -- Different setting --> NovelSurrogate["Treat as a new context of use"]
    MatchTable -- No match --> NovelSurrogate
    NovelSurrogate --> CheckData{"Do preliminary human data show biomarker response at a generally tolerable dose?"}
    CheckData -- No --> GenerateData["Generate preliminary human biomarker-response data before this Type C track"]
    CheckData -- Yes --> TypeCPrep["Prepare the complete background package using the Type C discussion guide"]
    TypeCPrep --> SubmitTypeC["Submit the Type C request with the full package at the same time"]
    SubmitTypeC --> FormalMeeting["Discuss feasibility and gaps; the meeting does not list, qualify, or accept the endpoint"]

Once a development team establishes formal alignment with the review division—whether confirming applicability of a listed row or establishing feasibility through a Type C novel-surrogate consultation—the surrogate must be rigorously translated into trial operations.

The trial protocol must specify the precise biological specimen handling, chain of custody, and analytical assay validation parameters under ICH E6(R3) Endpoint Governance. Furthermore, as emphasized in Statistical Analysis Plans for Endpoints, the Statistical Analysis Plan (SAP) must lock all mathematical transformation rules, handling algorithms for values below the lower limit of quantitation (LLOQ), visit window definitions, and intercurrent event handling before unblinding.

FDA's surrogate endpoint table is a living statutory list and a discussion map. Used that way—read every column and footnote, keep Traditional distinct from Accelerated, and treat listing as neither qualification nor portable validation—it supports an honest conversation with the relevant CDER or CBER review division before a surrogate is locked as the trial's primary endpoint.