Fact-checked by Grok 2 weeks ago

Surrogate endpoint

A surrogate endpoint is a biomarker, laboratory measurement, radiographic image, physical sign, or other intermediate clinical measure used as a substitute for a direct assessment of how a patient feels, functions, or survives, with the expectation that changes in the surrogate reliably predict effects on true clinical outcomes.^[1]^[2] These endpoints are employed in clinical trials to enhance efficiency, reducing required sample sizes, trial durations, and costs compared to traditional endpoints like overall survival or disease-specific morbidity, thereby supporting accelerated regulatory pathways such as the FDA's accelerated approval process for drugs addressing serious conditions with unmet needs.^[3]00403-6/fulltext) Validation of a surrogate demands robust epidemiologic, pathophysiologic, and trial-level evidence establishing its causal linkage to the clinical outcome, yet empirical data reveal frequent discrepancies where surrogate improvements fail to correlate with patient benefits.^[4]^[5] Prominent applications include oncology, where metrics like progression-free survival or objective response rates serve as surrogates for overall survival, facilitating rapid evaluation of novel therapies amid high unmet needs, though systematic reviews indicate only partial or inconsistent predictive validity across cancer types.00403-6/fulltext)^[6] In cardiovascular medicine, blood pressure reduction has proven a reliable surrogate for stroke and myocardial infarction risk reduction in many interventions, exemplifying success through established causal pathways.^[7] Conversely, failures underscore risks, as seen in antiarrhythmic drugs that suppressed ventricular ectopy—a surrogate for sudden death—but instead elevated overall mortality due to unpredicted proarrhythmic effects, highlighting how surrogates can overlook off-target harms or incomplete mechanistic capture.^[5] Such discordances have prompted calls for stricter post-approval confirmation trials and meta-awareness of potential overreliance, particularly in fields like oncology where surrogate-based approvals outnumber those tied to definitive outcomes.^[8]^[9]

Definition and Fundamentals

Core Definition

A surrogate endpoint is a clinical trial outcome measure employed as a proxy for a direct assessment of patient benefit, such as how a patient feels, functions, or survives.^[1] These endpoints typically consist of biomarkers, laboratory values, imaging findings, or physiological indicators that are anticipated to correlate with or predict the effects on true clinical endpoints, like mortality or disease progression, based on biological plausibility or empirical associations.^[2] Unlike direct clinical endpoints, surrogates aim to provide earlier indicators of therapeutic efficacy, facilitating faster drug development and regulatory review, particularly for serious conditions where prolonged trials measuring survival would delay treatment availability.^[10] The concept relies on the premise that changes in the surrogate reliably reflect causal impacts on the ultimate health outcome, though this linkage requires substantiation through mechanistic understanding or trial data.^[11] For instance, reductions in low-density lipoprotein cholesterol levels have been used as surrogates for cardiovascular event reduction in certain statin trials, predicated on established pathophysiologic roles in atherosclerosis.^[12] Regulatory bodies like the FDA authorize their use under frameworks such as accelerated approval, where surrogates deemed "reasonably likely" to predict clinical benefit support provisional licensure, contingent on confirmatory studies verifying patient-level effects.^[9] Surrogates encompass a range of measurable entities, including viral load reductions in HIV treatments as proxies for progression to AIDS or tumor response rates in oncology as stand-ins for survival gains.^[13] Their adoption has grown, with FDA data indicating that approximately 45% of new drug approvals between 2010 and 2012 relied on surrogate endpoints.^[2] This approach balances expedited access against the risk of over-reliance on imperfect predictors, necessitating rigorous evaluation to avoid misleading inferences about net clinical value.^[14]

Distinction from True Clinical Endpoints

True clinical endpoints represent direct measures of patient benefit, encompassing outcomes such as overall survival, progression-free survival without symptoms, or improvements in how a patient feels, functions, or survives, as established by regulatory bodies like the U.S. Food and Drug Administration (FDA).^[1]^[15] These endpoints capture the ultimate clinical impact of an intervention, serving as the gold standard for assessing therapeutic efficacy because they reflect tangible health improvements or harms without intermediary assumptions.^[15] In contrast, surrogate endpoints are indirect biomarkers or physiological measures—such as changes in blood pressure, tumor size, or laboratory values—employed as proxies for true clinical endpoints when the latter are impractical due to prolonged timelines, high costs, or ethical constraints in trials.^[1]^[16] The rationale is that favorable alterations in the surrogate reliably predict corresponding effects on the true endpoint, allowing accelerated drug development and approval; however, this substitution hinges on empirical validation demonstrating causal linkage, which is absent in unproven surrogates.^[16]^[17] The primary distinctions lie in their biological positioning and evidentiary requirements: true endpoints occupy the end of the causal chain from intervention to patient outcome, directly verifiable through long-term observation, whereas surrogates reside intermediately on that pathway, necessitating rigorous prospective validation to confirm that intervention-induced changes fully mediate and predict the clinical effect without extraneous influences.^[18]^[11] Practically, surrogates enable shorter, smaller trials by yielding quicker, less invasive data, but this expediency introduces risks of discordance, where surrogate improvements fail to translate to clinical benefit—or even correlate with harm—as seen in cases like antiarrhythmic drugs suppressing ventricular ectopy (a surrogate) yet increasing mortality in the Cardiac Arrhythmia Suppression Trial (CAST) of 1989.^[19] Such failures underscore that surrogates do not inherently equate to clinical relevance, demanding separate confirmation trials for true endpoints post-approval to mitigate overreliance on incomplete proxies.^[5]^[20] Validation criteria further delineate the two: a surrogate must exhibit trial-level and individual-level associations with the true endpoint, capturing the net intervention effect via frameworks like Prentice's criteria, which require the surrogate to fully predict clinical outcomes both within and across studies.^[11] Absent this, surrogates risk misleading approvals, as evidenced by oncology trials where progression-free survival (a surrogate) has inconsistently predicted overall survival, prompting regulators to mandate confirmatory data on direct patient outcomes.^[21]^[19] This distinction emphasizes causal realism, prioritizing endpoints grounded in observable patient-centered results over convenient but potentially decoupled markers.

Validation Criteria

Validation of surrogate endpoints requires demonstrating that the surrogate reliably predicts the effect of an intervention on the true clinical outcome, encompassing both biological plausibility and empirical evidence from clinical data.^[22] This process typically involves assessing correlation at the individual patient level (within-trial surrogacy) and consistency of treatment effects across trials (trial-level surrogacy), often through meta-analyses of randomized controlled trials.^[23] Failure to validate can lead to misleading conclusions, as seen in cases where surrogates like blood pressure reduction predicted cardiovascular events in some contexts but not others due to off-target effects.^[24] The foundational framework for validation, proposed by Prentice in 1989, outlines four necessary and sufficient conditions for a surrogate to fully substitute for the true endpoint in binary outcome settings.^[23] These criteria emphasize causal linkage:

The treatment must significantly alter the surrogate endpoint compared to control.
The surrogate must correlate with the true clinical endpoint in the control arm, serving as a prognostic marker.^[25]
The treatment must affect the true endpoint.^[25]
The net effect of treatment on the true endpoint must be entirely attributable to its effect on the surrogate, with no residual direct or indirect effects bypassing the surrogate.^[25]

Empirical testing of Prentice's criteria often uses statistical models, such as copula-based approaches for continuous surrogates or proportion-mediated effects for binary ones, but these conditions are rarely fully met, limiting validation to specific contexts like disease mechanisms or drug classes.^[26]^[23] Regulatory bodies like the FDA evaluate surrogates case-by-case, requiring a mechanistic rationale (e.g., pathophysiological linkage) supplemented by data from multiple trials showing consistent prediction of clinical benefit, such as survival or symptom relief.^[22]^[27] For accelerated approvals, surrogates are provisionally accepted if supported by historical data, but post-approval confirmatory trials are mandated to verify translation to true outcomes; between 2005 and 2022, only a subset of oncology surrogates like progression-free survival met rigorous validation thresholds across indications.^[28] Validation challenges persist due to heterogeneity in patient populations and interventions, underscoring the need for context-specific assessment rather than universal acceptance.^[29]

Historical Development

Origins in Clinical Research

The concept of surrogate endpoints emerged in clinical research as a pragmatic response to the protracted timelines required for observing definitive clinical outcomes, such as mortality or major morbidity, in randomized controlled trials for chronic conditions. Early informal applications leveraged intermediate physiological measures to infer treatment benefits, particularly in cardiovascular disease, where blood pressure reductions in antihypertensive interventions during the 1960s and 1970s were linked to lower rates of stroke and coronary events in trials like the Veterans Administration Cooperative Study.^[30] Similarly, serum cholesterol lowering was used as a proxy for reduced atherosclerotic events in lipid-modifying studies predating formal surrogate validation frameworks.^[7] These approaches reflected an intuitive causal chain—treatment alters a measurable biomarker, which in turn influences the ultimate health outcome—but lacked rigorous statistical criteria to ensure reliability.^[31] The explicit terminology and methodological foundations crystallized in the late 1980s, driven by urgent needs in infectious disease research amid the HIV/AIDS crisis and expanding trial designs. The term "surrogate marker" debuted in 1988, applied to biomarkers like gamma-glutamyltransferase for identifying carriers of non-A, non-B hepatitis, highlighting their role in expediting diagnostic and therapeutic assessments.^[32] This was rapidly advanced by R.L. Prentice's 1989 publication in Statistics in Medicine, which defined a surrogate endpoint as a substitute that fully mediates the treatment's effect on the true endpoint, proposing operational criteria: the treatment must significantly affect both the surrogate and true endpoint, and all treatment effect on the true endpoint must be explained by the surrogate.^[33] Prentice's criteria, rooted in causal inference, underscored that mere correlation was insufficient; the surrogate had to lie on the causal pathway to avoid misleading conclusions.^[23] These origins underscored surrogate endpoints' potential to accelerate evidence generation while exposing validation challenges, as subsequent trials revealed discrepancies between surrogates and clinical reality. For example, suppression of premature ventricular contractions as a surrogate for mortality reduction in post-myocardial infarction patients guided antiarrhythmic drug use in the 1980s but proved invalid in the Cardiac Arrhythmia Suppression Trial (initiated 1987), where such suppression correlated with increased mortality.^[16] This early tension informed ongoing refinements, emphasizing empirical testing over assumption in clinical research.^[24]

Regulatory Formalization

The regulatory formalization of surrogate endpoints emerged prominently through the U.S. Food and Drug Administration (FDA) in response to the HIV/AIDS epidemic of the 1980s, where urgent needs for rapid drug access prompted early reliance on biomarkers like CD4 cell counts and viral load reductions as proxies for survival benefits.^[34] This culminated in the FDA's 1992 establishment of the accelerated approval pathway via regulation, enabling provisional marketing authorization for drugs addressing serious or life-threatening conditions based on surrogate endpoints deemed reasonably likely to predict clinical benefit, contingent on confirmatory post-approval trials demonstrating direct outcomes.^[35] The pathway's use expanded from initial HIV applications, such as the 1987 approval of zidovudine (AZT) informed by surrogate data on mortality reduction, to broader therapeutic areas.^[36] The Food and Drug Administration Modernization Act (FDAMA) of 1997 codified and reinforced this framework by statutorily authorizing accelerated approvals and mandating FDA programs to foster surrogate endpoint development where they reliably forecast clinical outcomes, thereby integrating surrogates into statutory drug review criteria.^[37] This legislation emphasized validation requirements, aligning with statistical criteria like those proposed by Prentice in 1989 for surrogates to fully capture treatment effects on true endpoints. Subsequent FDA initiatives, including the 2018 public release of a table documenting over 100 surrogate endpoints previously accepted as primary efficacy measures for approvals, aimed to enhance transparency, guide developers, and standardize discussions on endpoint suitability.^[9] Prescription Drug User Fee Act (PDUFA) reauthorizations have further institutionalized surrogate use through commitments like those in PDUFA VI (2017–2022), which obligated the FDA to offer structured consultations on novel surrogates, including evidence frameworks linking them to clinical outcomes via Type C meetings and background packages.^[38] These measures prioritize empirical validation—such as trial-level correlations or meta-analyses—while requiring post-approval verification to mitigate risks of surrogate failure, as seen in select historical cases where initial predictions did not fully align with long-term benefits. Internationally, bodies like the European Medicines Agency (EMA) adopted parallel approaches, such as conditional marketing authorizations under Regulation (EC) No 507/2006, harmonizing surrogate reliance with confirmatory obligations.^[1]

Applications in Medicine

Cardiovascular Disease

In cardiovascular disease, surrogate endpoints are frequently utilized in clinical trials to assess interventions targeting dyslipidemia, hypertension, atherosclerosis progression, and heart failure, allowing for shorter study durations compared to waiting for hard clinical outcomes like myocardial infarction or cardiovascular mortality. Common examples include reductions in low-density lipoprotein cholesterol (LDL-C) levels for lipid-lowering agents, blood pressure measurements for antihypertensive therapies, and changes in left ventricular ejection fraction or biomarkers such as B-type natriuretic peptide (BNP) for heart failure treatments.^[7]^[39]^[40] LDL-C lowering has served as a primary surrogate endpoint in numerous trials of statins and other dyslipidemic agents, predicated on the causal role of elevated LDL-C in atherogenesis and plaque formation leading to events. For instance, randomized controlled trials such as the 1994 Scandinavian Simvastatin Survival Study (4S), involving 4,444 patients with coronary heart disease, showed that simvastatin-induced LDL-C reductions of approximately 35% correlated with a 34% decrease in major coronary events, establishing early empirical support for this surrogate. Subsequent large-scale studies, including the 2005 Treating to New Targets (TNT) trial with atorvastatin, reinforced this by demonstrating dose-dependent LDL-C reductions predicting proportional declines in cardiovascular events across diverse populations. The U.S. Food and Drug Administration (FDA) has relied on LDL-C as a basis for approvals of agents like PCSK9 inhibitors (e.g., evolocumab in 2015) and bempedoic acid in 2020, with post-approval outcome trials such as FOURIER (2017) confirming the surrogate's predictive validity by showing LDL-C reductions translating to 20% relative risk reductions in major adverse cardiovascular events.^[41]^[42]^[9] Blood pressure reduction functions as a surrogate for stroke, heart failure, and overall cardiovascular events in hypertension trials, supported by pathophysiological links to endothelial damage and vascular remodeling. Meta-analyses of trials like the 1997 Systolic Hypertension in the Elderly Program (SHEP), which enrolled over 2,300 participants, indicated that achieving systolic blood pressure below 150 mmHg reduced stroke risk by 36%, with consistent trial-level correlations across classes of antihypertensives including ACE inhibitors and calcium channel blockers. Regulatory bodies, including the FDA and European Medicines Agency, accept blood pressure changes for expedited approvals in this domain, as evidenced by endorsements in guidelines from the American College of Cardiology.^[30]^[43] In heart failure, surrogates such as improvements in left ventricular ejection fraction (LVEF) or reductions in BNP/NT-proBNP levels have been applied to evaluate therapies like beta-blockers and SGLT2 inhibitors, aiming to predict hospitalizations and mortality. The 2001 MERIT-HF trial with metoprolol succinate, for example, used LVEF increases alongside symptom scores to support efficacy in 3,991 patients, correlating with a 34% reduction in all-cause mortality in subsequent analyses. However, validation remains mechanism-specific, as not all LVEF changes fully capture long-term outcomes across drug classes.^[40]^[44] Imaging-based surrogates, including coronary plaque volume assessed via intravascular ultrasound or computed tomography angiography, have gained traction for anti-atherosclerotic agents, offering direct visualization of disease modification. The 2009 ASTEROID trial with rosuvastatin demonstrated plaque regression in 507 patients, with LDL-C reductions exceeding 50% linked to volumetric decreases, informing applications in serial monitoring for high-risk cohorts. These endpoints facilitate early-phase decisions but require integration with biological plausibility for broader acceptance.^[7]

Oncology

In oncology, surrogate endpoints such as progression-free survival (PFS), overall response rate (ORR), and disease-free survival (DFS) are frequently employed to assess treatment efficacy in clinical trials, particularly for advanced malignancies where overall survival (OS) requires extended follow-up due to confounding factors like subsequent therapies.^[45] These endpoints enable accelerated drug approvals by the U.S. Food and Drug Administration (FDA), with PFS serving as a primary surrogate for OS in settings like metastatic colorectal cancer or non-small cell lung cancer, based on observed delays in disease progression that are presumed to translate to prolonged life.^[46] ORR, measured via criteria like RECIST, quantifies tumor shrinkage and has supported approvals for immunotherapies and targeted agents in hematologic and solid tumors.^[47] In curative intent trials, such as neoadjuvant breast cancer therapy, pathological complete response (pCR) predicts long-term outcomes like event-free survival.00403-6/fulltext) Empirical validation of these surrogates remains inconsistent, with FDA analyses from 2005 to 2022 reviewing 25 surrogate endpoint studies across various cancers finding strong correlations with OS in only a minority of cases, often limited to specific tumor types or treatment classes like cytotoxics rather than modern targeted therapies or immunotherapies.^[28] For instance, PFS has demonstrated moderate trial-level surrogacy for OS in colorectal cancer (correlation coefficient R² ≈ 0.6-0.8 in meta-analyses), supporting its use for agents like cetuximab, but weaker associations appear in breast or lung cancers influenced by crossover effects and post-progression treatments.^[48] The FDA has relied on surrogates for approximately 194 oncology approvals since 1992, including 57% of accelerated approvals, yet confirmatory trials frequently fail to verify OS benefits, highlighting dissociation risks.^[49] Notable failures underscore causal disconnects: bevacizumab (Avastin) received accelerated FDA approval in 2008 for metastatic HER2-negative breast cancer based on PFS improvements (hazard ratio 0.60-0.67) in the E2100, AVADO, and RIBBON-1 trials, but phase III confirmatory studies (e.g., AVADO OS data from 2010) showed no OS extension (hazard ratio 0.97, p=0.76) alongside increased toxicity like hypertension and bleeding, prompting revocation of the indication in 2011.^[50] Similarly, in glioblastoma, bevacizumab extended PFS by 4.4 months in the EF-5 trial (2009) leading to provisional approval, yet failed to improve OS (median 14.6 vs. 14.3 months) or quality of life in subsequent analyses, resulting in non-conversion to full approval in 2011.^[51] These cases illustrate how surrogates can overestimate benefits when biological mechanisms—such as vascular normalization without cytotoxicity—do not reliably impact mortality, exacerbated by evolving standard-of-care post-progression.^[52] Regulatory frameworks acknowledge these limitations; the FDA's 2010 guidance permits PFS as a surrogate for traditional or accelerated approval only if mature OS data are infeasible and the endpoint captures meaningful clinical impact, yet post-approval requirements for OS confirmation are inconsistently met, with over 50% of accelerated oncology approvals lacking verified clinical benefit after five years of follow-up as of 2024.^[8] In hematologic malignancies, surrogates like complete response rates have higher predictive validity for OS in acute leukemias due to standardized remission definitions, but even here, minimal residual disease negativity as a surrogate requires further validation against relapse rates.^[53] Overall, while surrogates accelerate access to therapies in a field where OS trials can span years, their deployment demands rigorous, context-specific validation to avoid approving interventions that prolong treatment exposure without survival gains.^[54]

Infectious Diseases

In infectious diseases, surrogate endpoints commonly include reductions in pathogen burden, such as viral load measurements or microbiological eradication, which are used to predict clinical outcomes like disease progression, hospitalization, or mortality. These markers enable shorter trial durations compared to direct endpoints like survival, as pathogen control often causally precedes symptom resolution or prevention of complications. For instance, the U.S. Food and Drug Administration (FDA) has relied on such surrogates for accelerated approvals of antimicrobials, provided they demonstrate a reasonable likelihood of clinical benefit based on prior validation studies.^[9]^[55] A prominent example is in human immunodeficiency virus (HIV) treatment, where plasma HIV RNA levels (viral load) serve as a validated surrogate for clinical progression and survival. FDA guidance specifies that sustained viral suppression to undetectable levels, often assessed over 24-48 weeks alongside CD4+ cell count increases, supports full approval of antiretroviral drugs, as these changes correlate strongly with delayed AIDS onset and reduced mortality in long-term observational data from trials like ACTG 175 and community cohorts. This approach revolutionized HIV drug development in the 1990s, allowing approvals without waiting years for survival data, though confirmatory trials with clinical endpoints are required post-approval to verify durability. Similarly, for cytomegalovirus (CMV) in immunocompromised patients, reductions in CMV viremia (DNAemia) have been deemed a validated surrogate by the FDA for traditional approval in composite endpoints, based on correlations with prevention of end-organ disease in transplant settings.^[56]^[55]^[57] For antibacterial agents, microbiological eradication—defined as clearance of the target pathogen from infection sites, such as blood or sputum—functions as a primary surrogate endpoint in non-inferiority trials for conditions like pneumonia or bacteremia. This measure, often evaluated 5-7 days post-treatment, predicts clinical cure rates exceeding 80-90% in validated scenarios, with biomarkers like procalcitonin decline providing additional correlation to bacterial clearance and reduced antibiotic duration needs. The FDA has accepted such endpoints for approvals when linked to historical data showing pathogen elimination reduces morbidity, though challenges arise in polymicrobial infections or emerging resistance, where eradication may not fully capture host immune responses or relapse risks.^[58]^[59] In vaccine development for infectious diseases, immune correlates like neutralizing antibody titers are employed as surrogates for protective efficacy against infection or severe disease. For example, in enterovirus 71 (EV71) vaccines, a neutralizing titer threshold of 1:32 has been validated as predictive of protection based on challenge studies and seroepidemiology, while for influenza, hemagglutination inhibition titers correlate with reduced infection rates in randomized trials. During the COVID-19 pandemic, neutralizing antibody levels against SARS-CoV-2 were explored as provisional surrogates, with titers of 100-250 associated with lower infection risk in observational data, though FDA approvals emphasized clinical efficacy over pure surrogacy due to variability in T-cell contributions and waning immunity. Validation remains context-specific, requiring trial-level analyses to confirm surrogates capture population-level protection without overestimating durability.^[60]^[61]^[62] Despite successes, limitations persist; surrogates may dissociate from clinical outcomes in heterogeneous populations, such as when viral load reductions in HIV fail to prevent non-AIDS events like cardiovascular complications driven by persistent inflammation. In antibacterials, microbiological success does not always translate to symptom resolution if underlying comorbidities dominate, necessitating hybrid endpoints combining surrogates with clinical assessments to mitigate risks of approving ineffective agents. Regulatory bodies like the FDA mandate post-approval studies to address these gaps, emphasizing that surrogates, while accelerating access, require rigorous epidemiologic and mechanistic substantiation to avoid overreliance.^[16]^[63]

Neurological Disorders

In multiple sclerosis (MS), magnetic resonance imaging (MRI)-derived measures, such as the number and volume of gadolinium-enhancing lesions or T2 hyperintense lesions, serve as surrogate endpoints for relapse rates and disability progression assessed by the Expanded Disability Status Scale (EDSS).^[64] Meta-analyses of randomized controlled trials have demonstrated that treatment effects on MRI lesion activity predict reductions in relapses with high accuracy, enabling MRI metrics to support regulatory approvals for disease-modifying therapies like interferons and monoclonal antibodies.^[65] However, correlations weaken for long-term EDSS worsening, with trial-level R² values around 0.57, indicating MRI surrogates capture short-term inflammatory activity but less reliably predict irreversible neurodegeneration.^[66] For Alzheimer's disease (AD), cerebrospinal fluid (CSF) levels of amyloid-beta (Aβ) and phosphorylated tau (p-tau), or amyloid positron emission tomography (PET) imaging, are proposed surrogates for cognitive decline measured by scales like the Alzheimer's Disease Assessment Scale-Cognitive Subscale (ADAS-Cog).^[67] Anti-amyloid monoclonal antibodies, such as lecanemab approved by the FDA in 2023, relied on amyloid plaque reduction via PET as a surrogate under accelerated approval, correlating with modest slowing of cognitive decline in phase 3 trials (e.g., 27% reduction in decline on CDR-SB).^[9] Yet, validation remains incomplete; historical failures like solanezumab reduced Aβ but yielded no clinical benefit, and prospective analyses question whether amyloid clearance alone predicts outcomes across AD stages due to multifactorial pathology including tau and neuroinflammation. Blood-based biomarkers like plasma p-tau181 show promise for trial enrichment but lack established trial-level surrogacy for clinical endpoints.^[68] In Parkinson's disease (PD), potential surrogates include dopamine transporter (DaT) imaging via single-photon emission computed tomography (SPECT) to assess nigrostriatal degeneration, or quantitative motor assessments like bradykinesia via wearable sensors.^[69] DaT-SPECT changes have been explored in early PD trials but fail Prentice criteria for validated surrogacy, as they correlate with diagnosis yet not consistently with symptom progression or levodopa response.^[70] Digital biomarkers from accelerometers, tracking gait or tremor, are emerging for prodromal PD but remain investigational without FDA endorsement as primary endpoints, reflecting challenges in capturing heterogeneous non-motor symptoms like cognition and mood.^[71] Across these disorders, surrogates accelerate trials for slowly progressive conditions—e.g., FDA's accelerated pathway for neurogenetic diseases requires surrogates "reasonably likely" to predict benefit—but risks arise from dissociation, as seen in AD where biomarker improvements precede clinical endpoints by years without guaranteeing net patient benefit.^[72] Empirical support is strongest in relapsing MS, moderate in amyloid-targeted AD therapies, and weakest in PD, underscoring the need for causal validation beyond correlation.^[73]

Advantages and Empirical Support

Acceleration of Drug Development

Surrogate endpoints accelerate drug development by substituting for time-intensive clinical outcomes, such as overall survival or disease-specific morbidity, which demand prolonged patient follow-up—often years—to observe statistically significant effects. In contrast, surrogates like tumor response rates or biomarker reductions manifest earlier, enabling trials to reach endpoints in months rather than years, thereby compressing development timelines from discovery to approval. This approach underpins regulatory frameworks like the FDA's Accelerated Approval pathway, which has relied on such endpoints to expedite access for therapies targeting unmet needs in serious conditions.^[1]^[74] Empirical analyses confirm measurable time savings; for instance, oncology trials employing surrogate endpoints, such as progression-free survival, exhibit drug development durations roughly 11 months shorter than those using overall survival.^[75] These reductions stem from shorter accrual and follow-up periods, as surrogates require fewer events to achieve power—e.g., detecting progression in hundreds of patients versus thousands for mortality endpoints. In cardiovascular applications, surrogates like LDL cholesterol reduction have similarly hastened approvals for lipid-lowering agents, avoiding multi-year morbidity tracking.^[76] The FDA has cataloged over 100 instances since 1992 where surrogates formed the basis for approvals across therapeutic areas, underscoring their role in halving median review times for eligible submissions compared to traditional pathways.^[9] Beyond temporal gains, surrogates lower developmental costs by necessitating smaller cohorts and simplified designs, with estimates indicating up to 30-50% reductions in trial expenses for surrogate-based studies in oncology and infectious diseases.^[77] This efficiency has proven pivotal in fields with rapid disease evolution, such as HIV, where viral load surrogates facilitated antiretroviral approvals within 1-2 years of pivotal trials, versus decades for earlier agents reliant on survival data.^[16] However, acceleration hinges on prior validation of the surrogate's predictive validity, as unverified proxies risk approving ineffective drugs, though successful cases demonstrate net benefits in hastening beneficial interventions to market.^[78]

Evidence of Successful Correlations

In cardiovascular disease, reductions in low-density lipoprotein (LDL) cholesterol levels serve as a validated surrogate endpoint, correlating strongly with decreased risks of major adverse cardiovascular events such as myocardial infarction and stroke, as evidenced by meta-analyses of statin trials showing that each 1 mmol/L reduction in LDL cholesterol predicts a proportional 20-25% relative risk reduction in these outcomes.^[79] ^[80] Similarly, blood pressure lowering has been empirically supported as a surrogate, with trials involving over 500,000 patients demonstrating that antihypertensive interventions reducing systolic blood pressure by 10 mmHg correlate with 20-30% reductions in stroke and coronary events across diverse drug classes.^[16] For infectious diseases, particularly HIV/AIDS, plasma HIV viral load suppression below detectable limits (typically <50 copies/mL) has shown robust correlation with delayed disease progression and improved survival, as confirmed in longitudinal studies where sustained virologic response predicted reduced AIDS-defining events and mortality with high specificity in antiretroviral therapy trials.^[81] The U.S. Food and Drug Administration (FDA) recognizes HIV viral load as sufficiently predictive for traditional drug approvals, based on evidence from pivotal trials like those establishing highly active antiretroviral therapy standards in the 1990s, where viral suppression rates aligned closely with clinical benefit metrics.^[2] In oncology, certain surrogate endpoints exhibit successful correlations in specific contexts; for instance, progression-free survival (PFS) has validated surrogacy for overall survival (OS) in advanced colorectal cancer following first-line chemotherapy, with meta-analyses reporting trial-level correlation coefficients exceeding 0.8 and predictive values where PFS improvements consistently translated to OS gains in regimens like FOLFOX or FOLFIRI combinations.^[82] Pathologic complete response (pCR) after neoadjuvant therapy in early-stage breast cancer also demonstrates strong linkage to event-free survival and OS, as supported by FDA analyses of approvals where pCR rates predicted long-term outcomes with hazard ratios indicating 50-70% risk reductions in recurrence.^[83] These examples highlight instances where surrogates capture causal pathways to clinical benefit, enabling reliable inferences; however, validation requires context-specific trial data, as aggregated evidence from FDA-reviewed cases underscores that only surrogates with demonstrated trial-level and individual-level associations warrant reliance.^[9]

Limitations and Empirical Failures

Theoretical Shortcomings

Surrogate endpoints presuppose that alterations in the biomarker or intermediate measure fully mediate the treatment's impact on the definitive clinical outcome, a stringent requirement encapsulated in Prentice's 1989 criteria. These stipulate that the surrogate must not only respond to the intervention and causally influence the true endpoint but also exhaustively account for the treatment's entire effect on that endpoint, precluding any independent pathways.^[33] Verifying this exhaustive mediation demands comprehensive knowledge of disease pathophysiology, which is theoretically unattainable in multifaceted biological processes where unmeasured confounders or parallel mechanisms persist.^[5] A core theoretical deficiency arises from the potential for dissociation, wherein the surrogate captures only a subset of causal influences while overlooking off-target or compensatory pathways. Treatments may modulate the surrogate favorably yet exacerbate the clinical outcome through unrelated channels, such as toxicity or unintended physiological disruptions, invalidating the surrogate's predictive fidelity.^[19] This violates causal realism, as historical associations between surrogate and outcome—often derived from observational data or prior trials—do not ensure that novel interventions propagate effects identically, given variations in dosing, patient heterogeneity, or disease stage.^[84] Furthermore, surrogates inherently risk confounding by extraneous factors that mimic treatment signals without advancing clinical benefit, undermining the assumption of unidirectionality. Prentice's framework, while operational, proves insufficient theoretically because it relies on statistical surrogacy within specific trial contexts, failing to generalize across interventions or populations where mechanistic divergence occurs.^[5] Consequently, surrogates prioritize expediency over causal completeness, potentially endorsing therapies that yield illusory progress absent verifiable linkage to patient-centered endpoints like survival or function.^[19]

Case Studies of Dissociation

One prominent example of surrogate endpoint dissociation occurred in cardiovascular disease trials of torcetrapib, a cholesteryl ester transfer protein (CETP) inhibitor developed to raise high-density lipoprotein (HDL) cholesterol levels as a surrogate for reduced cardiovascular risk. In the phase 3 ILLUMINATE trial involving 15,067 high-risk patients, torcetrapib combined with atorvastatin increased HDL cholesterol by 72.4% and decreased low-density lipoprotein (LDL) cholesterol by 24.9% compared to atorvastatin alone after 12 months.^[85] However, the trial was terminated early on December 2, 2006, after an interim analysis revealed a 60% relative increase in all-cause mortality (93 deaths versus 58 in the control group) and higher rates of cardiovascular events, including angina and heart failure, attributed to off-target effects such as elevated blood pressure and aldosterone levels rather than lipid modulation.^[85]^[86] This dissociation highlighted how favorable changes in lipid surrogates failed to predict clinical benefit and instead masked adverse outcomes.^[87] In oncology, bevacizumab (Avastin), a vascular endothelial growth factor inhibitor, exemplified dissociation between progression-free survival (PFS) as a surrogate and overall survival (OS) in metastatic breast cancer. The U.S. Food and Drug Administration (FDA) granted accelerated approval in February 2008 based on the E2100 trial, where bevacizumab plus paclitaxel extended median PFS from 5.8 months to 11.3 months compared to paclitaxel alone, interpreted as evidence of clinical benefit via tumor progression delay.^[50] Subsequent confirmatory phase 3 trials, including AVADO (2009) and RIBBON-1 (2009), confirmed PFS improvements (e.g., hazard ratio of 0.67 in AVADO) but showed no OS benefit, with median OS ranging from 25.3 to 31.9 months across arms without significant differences.^[88] In July 2010, the FDA initiated withdrawal proceedings, finalized in November 2011, after determining that PFS failed to reliably predict durable clinical outcomes like OS or improved quality of life, leading to the revocation of the breast cancer indication.^[50]^[88] This case underscored the risks of relying on intermediate endpoints in heterogeneous cancers where post-progression therapies can confound surrogacy.^[89] Neurological disorders provided another clear instance with anti-amyloid monoclonal antibodies for Alzheimer's disease, using amyloid-beta plaque reduction as a surrogate for cognitive preservation. Aducanumab (Aduhelm) received FDA accelerated approval on June 7, 2021, primarily based on dose- and time-dependent reductions in cerebral amyloid plaques observed via positron emission tomography in the phase 3 EMERGE and ENGAGE trials, despite inconsistent effects on the primary clinical endpoint of slowing cognitive decline as measured by the Clinical Dementia Rating-Sum of Boxes scale.^[90] In EMERGE, high-dose aducanumab reduced amyloid by 59% at 18 months but yielded only a marginal 0.39-point improvement on the 18-point scale (p=0.01, not robustly replicated), while ENGAGE showed no clinical benefit despite similar amyloid clearance.^[91]^[92] Confirmatory studies faltered, with Biogen discontinuing aducanumab development by January 31, 2024, after Medicare non-coverage and limited uptake revealed the surrogate's inadequacy in predicting meaningful functional or survival gains, amid broader failures of over 10 anti-amyloid agents in phase 3 trials.^[93]^[94] This pattern reinforced skepticism toward amyloid as a validated surrogate, given multifactorial Alzheimer's pathology.^[91]

Broader Impacts on Patient Outcomes

Reliance on surrogate endpoints has resulted in the approval and clinical use of interventions that improve the surrogate measure but fail to enhance, or even worsen, patient-centered outcomes such as overall survival or quality of life.^[5] In cases of dissociation, where the surrogate does not reliably predict clinical benefit, patients may receive treatments exposing them to significant toxicities, adverse events, and financial burdens without corresponding reductions in morbidity or mortality.^[52] For instance, in oncology trials, meta-analyses across 36 studies found that only 23% of surrogate endpoints exhibited high correlation (r ≥ 0.85) with overall survival, with 52% showing low correlation (r ≤ 0.7), leading to the adoption of agents like bevacizumab, which improved progression-free survival in metastatic breast cancer but demonstrated no survival benefit and prompted withdrawal of approval in 2011 due to lack of efficacy.^[52] Specific trial-level failures underscore these risks. The Cardiac Arrhythmia Suppression Trial (CAST) demonstrated that encainide and flecainide suppressed premature ventricular contractions—a surrogate for arrhythmia risk—in post-myocardial infarction patients, yet these drugs increased mortality by 2.7-fold compared to placebo, with a 7.7% absolute increase in arrhythmic death or cardiac arrest within two years.^[5] Similarly, torcetrapib raised high-density lipoprotein cholesterol levels—a purported surrogate for cardiovascular protection—but accelerated mortality and cardiovascular events in coronary artery disease patients, halting development after phase III trials revealed excess harm.^[5] Such dissociations not only inflict direct patient harm through unanticipated adverse effects but also contribute to broader inefficiencies, including prolonged exposure to ineffective therapies that delay access to superior alternatives and inflate healthcare expenditures on non-beneficial interventions.^[19] These patterns erode confidence in therapeutic decisions and highlight the potential for surrogate-driven approvals to prioritize short-term metrics over long-term patient welfare. In aggregate, weak surrogate validation has fostered a landscape where up to half of oncology drugs approved on surrogates like progression-free survival lack confirmatory survival gains, subjecting patients to cumulative toxicities—such as neuropathy, hypertension, or immune-related events—without offsetting clinical advantages.^[52] This underscores the necessity for rigorous, context-specific validation to mitigate risks of harm outweighing negligible or absent benefits in real-world application.^[5]

Regulatory and Ethical Controversies

FDA Accelerated Approval Framework

The FDA's Accelerated Approval pathway, established by regulation in 1992 amid the HIV/AIDS crisis, enables provisional approval of drugs for serious or life-threatening conditions based on evidence of an effect on a surrogate endpoint reasonably likely to predict clinical benefit, rather than direct demonstration of improved survival or irreversible morbidity.^[95]^[35] This framework codified surrogate endpoints—such as laboratory measures, radiographic images, or physical signs—as acceptable for expediting access when traditional endpoints would delay availability for unmet needs.^[95] The program was statutorily reinforced in 2012 under the Food and Drug Administration Safety and Innovation Act, expanding its application beyond infectious diseases to areas like oncology.^[96] Under the framework, sponsors must conduct post-approval confirmatory trials to verify the surrogate's prediction of actual clinical benefit, with FDA retaining authority to require these studies as a condition of approval and to expedite or mandate their initiation.^[74]^[97] If trials confirm benefit, the approval converts to traditional; failure to confirm or substantial evidence of inefficacy can lead to withdrawal, though enforcement has varied, prompting scrutiny over incomplete validations.^[74] Sponsors submit progress reports every 180 days, and recent FDA guidance emphasizes that confirmatory trials must be "underway" prior to or shortly after accelerated approval, defined by factors like protocol finalization, institutional review board approval, and patient enrollment initiation.^[98] The framework prioritizes surrogates with historical validation from prior studies linking them to clinical outcomes, but lacks a formal prospective qualification process, relying instead on case-by-case assessment of plausibility and evidence.^[1] From 1992 to 2021, approximately 300 drugs received accelerated approval, predominantly using surrogates in oncology and rare diseases, though confirmatory success rates have been estimated at around 50-60% in some analyses, highlighting risks of dissociation.^[96] In December 2024, FDA issued draft guidance reinforcing stricter timelines and evidentiary standards for surrogates and confirmatories, aiming to address delays and withdrawals amid congressional mandates from the 2022 Consolidated Appropriations Act.^[99]

Debates on Validation Standards

Validation of surrogate endpoints requires demonstrating both statistical correlation with the true clinical endpoint at the trial level and that the surrogate fully captures the net effect of the intervention on the clinical outcome, as outlined in the Prentice criteria established in 1989.^[14] These criteria demand that the surrogate lies on the causal pathway between treatment and outcome, fully mediates the treatment effect, and is measured earlier than the clinical endpoint. However, debates persist over their stringency, with critics arguing they are rarely met in practice due to incomplete mediation in complex diseases, leading to calls for supplementary statistical methods like meta-regression analyses that assess trial-level associations via metrics such as the coefficient of determination (R²).^[14] ^[100] Regulatory agencies like the FDA classify surrogates into categories such as "candidate," "reasonably likely," or "validated," with accelerated approvals often relying on the "reasonably likely" threshold, which requires plausible biological rationale and supportive evidence rather than exhaustive prospective validation across multiple trials.^[1] This approach contrasts with more rigorous academic standards, sparking contention that regulatory leniency prioritizes speed over certainty, as evidenced by oncology trials where progression-free survival (PFS) surrogates correlate moderately with overall survival (OS) at the trial level (R² ≈ 0.6-0.8 in meta-analyses) but fail patient-level validation and occasionally dissociate in individual studies.^[21] ^[52] The EMA similarly emphasizes validation through consistent predictive value but lacks harmonized global thresholds, exacerbating debates on cross-agency comparability.^[101] A core controversy involves the scope of validation data: retrospective meta-analyses using subsets of trials often overestimate surrogacy strength, while prospective designs are resource-intensive and rarely funded, resulting in only a fraction of purported surrogates undergoing comprehensive testing.^[21] ^[52] Proponents of relaxed standards, often aligned with industry perspectives, highlight efficiencies in fields like cardiology where LDL cholesterol reductions robustly predict cardiovascular events (high R² >0.9), justifying context-specific flexibility.^[7] Detractors, including methodologists, counter that disease heterogeneity—particularly in oncology—demands universal biological plausibility alongside statistical metrics, warning that unvalidated surrogates risk approving ineffective therapies, as seen in cases where surrogate improvements did not translate to survival benefits.^[13] ^[102] Industry incentives further fuel the debate, as validation hurdles like high costs and long timelines discourage investment, leading to reliance on historical correlations without mechanistic confirmation, despite evidence that such shortcuts contribute to post-approval withdrawals.^[103] Emerging proposals advocate for adaptive frameworks incorporating machine learning for dynamic validation and mandatory post-approval clinical outcome confirmation, yet consensus remains elusive due to varying stakeholder priorities between rapid access and evidentiary rigor.^[104] ^[101]

Influence of Industry Incentives

The pharmaceutical industry has strong financial incentives to prioritize surrogate endpoints in drug development, as these measures enable accelerated regulatory approvals and earlier market entry, thereby shortening the time to revenue generation amid high research and development costs averaging $1.3 billion to $2.6 billion per approved drug.^[105] With effective patent exclusivity often limited to 10-15 years after approval due to development timelines, companies seek to minimize clinical trial durations, which surrogate endpoints facilitate by substituting for lengthy hard outcomes like overall survival, potentially reducing trial costs by 20-50% in fields like oncology.^[106] This economic pressure is evident in the preference for surrogate-based pathways, where post-approval confirmatory trials—required to verify clinical benefit—are often delayed or under-resourced, as firms can already capture market share and pricing power during the interim, with oncology drugs approved on surrogates generating billions in annual sales despite incomplete validation.^[107] Industry influence extends through funding mechanisms like the Prescription Drug User Fee Act (PDUFA), under which pharmaceutical fees constitute over 65% of the FDA's human drugs program budget as of PDUFA VII in 2023, fostering a performance-based review system that expedites approvals but raises concerns of regulatory capture.^[108] This structure aligns FDA priorities with industry timelines, including greater acceptance of surrogate endpoints in accelerated approvals, which rose from 10% of new molecular entities in the 1990s to over 45% in oncology by 2020, partly incentivized by user fee commitments to faster reviews.^[109] Critics, including analyses from the Institute for Clinical and Economic Review, argue that such dependencies may dilute scrutiny of surrogate validity, as evidenced by cases where industry-sponsored trials selectively emphasize favorable surrogates while downplaying dissociation risks, leading to approvals of therapies with marginal or unconfirmed patient benefits.^[107]^[110] These incentives can distort validation efforts, with industry-funded studies 1.5-2 times more likely to report positive surrogate results compared to independent trials, potentially prioritizing marketable drugs over rigorous causal linkages to clinical outcomes.^[110] For instance, in rare diseases, orphan drug incentives combined with surrogate reliance allow premium pricing—often exceeding $100,000 per patient annually—without full endpoint confirmation, amplifying profits but contributing to failures in 15-20% of accelerated oncology approvals where confirmatory data fail to substantiate benefits.^[111]^[89] While proponents highlight patient access gains, empirical reviews indicate that economic pressures may undervalue long-term confirmatory investments, as firms face limited penalties for delays beyond initial sales accrual.^[107]

Recent Developments and Future Directions

Post-2023 Scrutiny and Studies

In 2024, a JAMA analysis of 99 surrogate-clinical outcome pairs from 86 randomized clinical trials supporting 93 FDA accelerated approvals for nononcologic chronic disease treatments found that surrogates were statistically significantly associated with clinical outcomes in only 67% of pairs, with just 17% showing high-strength evidence (correlation coefficients ≥0.85 or R² ≥0.72).^[73] Most surrogates (59%) lacked supporting meta-analyses, and only three—forced expiratory volume in 1 second for chronic obstructive pulmonary disease, hemoglobin A1c for type 2 diabetes, and proteinuria for primary glomerular disease—exhibited consistent high-strength links to outcomes like mortality or morbidity.^[73] A separate 2024 JAMA study examining 129 oncology drug indications accelerated between 2013 and 2023 reported that only 63% with more than five years of follow-up converted to traditional approval, with 37% of recent (2021-2023) conversions relying on response rate plus duration of response rather than progression-free survival or overall survival, signaling growing dependence on less rigorously validated surrogates.^[112] This low conversion rate highlights empirical failures where initial surrogate improvements did not translate to durable clinical benefits, prompting critiques of FDA's accelerated pathway for potentially exposing patients to unproven therapies.^[112] November 2024 research in cancer trials demonstrated inconsistent surrogate reliability, with recurrence-free survival correlating strongly with overall survival in pancreatic ductal adenocarcinoma (ρ=0.88) and biliary tract cancer (ρ=0.87) but weakly in hepatocellular carcinoma (ρ=0.67) and colorectal liver metastases (ρ=0.53), largely due to effective salvage treatments post-recurrence that decoupled surrogates from long-term outcomes.^[6] Such variability underscores the context-specific nature of validation, where surrogates like disease-free survival are FDA-accepted for certain cancers (e.g., breast, colorectal) but fail in others without trial-level and individual-level causal evidence.^[6] In response to these reliability gaps, the CONSORT-Surrogate extension, published in July 2024, established reporting guidelines for randomized controlled trials using surrogates as primary outcomes, emphasizing mandatory disclosure of validation evidence, uncertainty assessments, and causal assumptions to mitigate risks of misinterpretation.^[113] Validation remains resource-intensive, often requiring long-term data unavailable at approval, as noted in contemporaneous BMJ commentary, fueling ongoing debates over surrogate thresholds for regulatory decisions.^[114] By August 2025, the FDA's updated table listed numerous surrogate-based approvals across indications, yet persistent scrutiny from these studies calls for stricter pre-approval validation to ensure surrogates predict clinical benefit without undue patient risk.^[9]

Emerging Validation Frameworks

Recent efforts to enhance the validation of surrogate endpoints emphasize rigorous empirical and statistical criteria to establish causal links between surrogates and clinical outcomes, addressing historical dissociations where surrogates failed to predict true benefits. Frameworks now integrate multi-stakeholder consensus processes, advanced causal inference, and standardized reporting to mitigate risks of over-reliance on unvalidated proxies.^[115]^[114] A 2023 framework developed through linked empirical methods—comprising a scoping review of definitions, an e-Delphi consensus among 219 experts, a stakeholder meeting with 28 participants, and a web survey of 80 respondents—aims to clarify surrogate endpoint definitions and interpretation in interventional trials. This approach endorses prior criteria like the Biomarker, Endpoint, and other Surrogate Taskforce (BEST) guidelines while highlighting variability in stakeholder views, such as clinicians prioritizing biological plausibility over patient-centered outcomes like quality of life. By distinguishing biomarkers and intermediate clinical outcomes as predictive substitutes for target endpoints, the framework facilitates better assessment of validation strength, requiring evidence of association at both individual and trial levels to interpret results reliably.^[115] The SPIRIT-Surrogate extension, published in 2024, extends the Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) checklist to mandate detailed reporting of surrogate endpoint rationale, validation evidence, and sensitivity analyses in trial protocols. Key additions include specifications on surrogate selection criteria, planned analyses for trial- and individual-level surrogacy (e.g., using Prentice's criteria), and contingencies for confirming clinical benefit post-approval. This promotes transparency and preemptive scrutiny, reducing approval of inadequately validated surrogates by ensuring protocols address potential causal disconnects.^[114] Emerging statistical methods incorporate causal frameworks like mediation analysis combined with joint modeling for longitudinal surrogates, as proposed in a 2025 approach applied to prostate-specific antigen (PSA) as a surrogate for disease-free survival. This method quantifies the proportion of treatment effect (PTE) mediated through the surrogate via joint models with random effects, accommodating meta-analytic data across trials and providing time-dependent estimates under assumptions like sequential ignorability and consistency. Advantages include handling unobserved heterogeneity and trial-level variability, outperforming traditional correlation-based metrics by directly estimating causal mediation, though it requires large datasets for robustness. Principal stratification techniques, extended in recent tutorials, further evaluate surrogates by partitioning populations based on potential outcomes, offering a counterfactual basis to assess whether surrogates capture intervention effects across subgroups.^[116]^[117]

References

[1]
Surrogate Endpoint Resources for Drug and Biologic Development
Jul 24, 2018 · A surrogate endpoint is a clinical trial endpoint used as a substitute for a direct measure of how a patient feels, functions, or survives.
[2]
FDA Facts: Biomarkers and Surrogate Endpoints
Dec 21, 2017 · Surrogate endpoints are used instead of clinical outcomes in some clinical trials. Surrogate endpoints are used when the clinical outcomes might ...
[3]
Surrogate endpoints in trials: a call for better reporting
Dec 12, 2022 · Using a surrogate endpoint as a substitute for a patient-relevant final outcome enables randomised controlled trials (RCTs) to be conducted more efficiently.
[4]
https://www.jclinepi.com/article/S0895-4356(23)00160-9/fulltext
[5]
The perils of surrogate endpoints - PMC - NIH
May 14, 2015 · Studies with surrogates will also generally have a smaller number of patients and a shorter time span. This limits such studies for evaluation ...
[6]
Surrogate endpoints in clinical trials: when is good...good enough?
Nov 21, 2024 · Validating surrogate endpoints in cancer clinical trials is critically important as they likely increase study feasibility, decrease cost, and ...
[7]
Two Decades of Cardiovascular Trials With Primary Surrogate ...
Mar 21, 2017 · We identified only 3 examples in which the surrogate trial was negative but a subsequent outcomes trial was conducted and showed benefit.
[8]
Has the FDA lost the plot on surrogate endpoints? - AgencyIQ
May 28, 2024 · According to the FDA, a surrogate or intermediate clinical endpoint is “a marker – a laboratory measurement, radiographic image, physical sign ...
[9]
Table of Surrogate Endpoints That Were the Basis of Drug Approval ...
Aug 29, 2025 · This surrogate endpoint table includes surrogate endpoints that sponsors have used as primary efficacy clinical trial endpoints for approval of new drug ...
[10]
Definition of surrogate endpoint - NCI Dictionary of Cancer Terms
Surrogate endpoints include a shrinking tumor or lower biomarker levels. They may be used instead of stronger indicators, such as longer survival or improved ...
[11]
Five criteria for using a surrogate endpoint to predict treatment effect ...
A surrogate endpoint in a randomized clinical trial is an endpoint that occurs after randomization and before the true, clinically meaningful, endpoint that ...
[12]
Surrogate endpoints: a key concept in clinical epidemiology
Dec 21, 2023 · Surrogate endpoints are biomarkers or intermediate outcomes that are used as substitutes for clinical outcomes of interest, often to expedite research or ...
[13]
Frequently asked questions on surrogate endpoints in oncology ...
Sep 13, 2024 · In this article, we provide a comprehensive review of surrogate endpoints currently used in oncology trials both in curative and advanced disease settings.
[14]
Biomarkers and Surrogate Endpoints: Lessons Learned From ... - NIH
Valid surrogate endpoints must be able to predict a clinically relevant endpoint, such as loss of vision or decrease in quality of life.
[15]
Evidence from Biomarkers and Surrogate Endpoints - PMC - NIH
A “clinical endpoint” is “a characteristic or variable that reflects how a patient feels, functions or survives.” Clinical endpoints are considered the gold ...
[16]
Biomarkers and Surrogate Endpoints In Clinical Trials - PMC - NIH
A surrogate endpoint is an outcome measure “used as a substitute for a clinically meaningful endpoint…changes induced by a therapy on a surrogate endpoint are ...Missing: distinction | Show results with:distinction
[17]
The Role of Surrogate Endpoints in the Evaluation of Efficacy ... - NIH
In contrast to patient-important endpoints, surrogates do not describe direct clinical benefit to the patient. However, a proven association between a surrogate ...
[18]
Evaluating surrogate endpoints, prognostic markers, and predictive ...
A surrogate endpoint is an endpoint observed earlier than the true endpoint (a health outcome) that is used to draw conclusions about the effect of ...
[19]
Surrogate End Points in Clinical Trials: Are We Being Misled?
Oct 1, 1996 · Unfortunately, the failure of surrogate end points to predict true outcome is not an isolated problem. Table 1 shows various examples from ...
[20]
Studies on surrogate end points - NCBI - NIH
The surrogate end point is expected to be based on research evidence other than survival or irreversible morbidity. Important points to note with regard to the ...Missing: distinction | Show results with:distinction<|separator|>
[21]
Surrogate endpoints in oncology: when are they acceptable for ...
Jul 21, 2017 · Surrogates may fall short in their ability to predict outcomes in hard endpoints for two broad categories of reasons: (1) technical factors in ...
[22]
[PDF] Guidance for Industry #267 - Biomarkers and Surrogate Endpoints in ...
The guidance suggests that the approach used for drug assays should be the starting point for validation of biomarker assays, although the FDA realizes that ...
[23]
Criteria for the validation of surrogate endpoints in randomized ...
The validation of surrogate endpoints has been studied by Prentice (1989, Statistics in Medicine 8, 431-440) and Freedman, Graubard, and Schatzkin (1992 ...
[24]
Surrogate Endpoints: Wishful Thinking or Reality? - Oxford Academic
Apr 19, 2006 · One of the earliest and best known validation measures is the set of criteria proposed by Prentice ( 4 , 5 ) to ensure that rejection of the ...
[25]
https://www.fnih.org/sites/default/files/final/pdf/6-McShane-Case%2520Study%2520Overview.pdf
[26]
Statistical Validation of Surrogate Endpoints: Another Look at the ...
For the statistical validation of surrogate endpoints, an alternative formulation is proposed for testing Prentice's fourth criterion, under a bivariate ...
[27]
Validating Surrogate Endpoints to Support FDA Drug Approval
Jul 8, 2025 · Surrogate endpoints are considered to be “validated” if they have undergone extensive testing that confirms their ability to predict a clinical outcome.
[28]
FDA validation of surrogate endpoints in oncology: 2005-2022
Sep 22, 2022 · Until surrogate outcomes have been thoroughly and robustly validated, they should be used with caution in drug approval decisions. Keywords: ...
[29]
Establishing a Public Resource for Acceptable Surrogate Endpoints ...
Feb 18, 2022 · Validated surrogate endpoints (SE's) are endpoints supported by a clear mechanistic rationale and clinical data providing strong evidence to ...
[30]
Surrogate endpoints in clinical trials: cardiovascular diseases
A surrogate endpoint in a cardiovascular clinical trial is defined as endpoint measured in lieu of some other so-called 'true' endpoint.
[31]
Biomarkers and surrogate endpoints - PMC - NIH
The use of the term 'surrogate marker' in medicine dates from the late 1980s [1], but it had been preceded by some years by the term 'biomarker'[2] and was ...
[32]
https://doi.org/10.1111/j.1423-0410.1988.tb03889.x
[33]
Surrogate endpoints in clinical trials: Definition and operational criteria
I discuss the idea of using surrogate endpoints in the context of clinical trials to compare two or more treatments or interventions in respect to some 'true' ...<|separator|>
[34]
Understanding the History and Use of the Accelerated Approval ...
Jan 4, 2022 · The Accelerated Approval Program, created in 1992, allows expedited approval of drugs for serious conditions using surrogate endpoints, with ...
[35]
Analysis of FDA's Accelerated Approval Program Performance ... - NIH
Jul 28, 2022 · Accelerated approval was established in 1992 by US Food and Drug Administration (“FDA”) regulation in response to the AIDS epidemic to address ...
[36]
FDA Accelerated Approval Gives Hope to Those in Greatest Need
AZT, approved in 1987, was the first drug approval based on a surrogate endpoint—and that ultimately led to the highly effective HAART drug that has allowed HIV ...Missing: history | Show results with:history
[37]
[PDF] FOOD AND DRUG ADMINISTRATION MODERIZATION ACT OF 1997
Nov 21, 1997 · ''(2) establish a program to encourage the development of surrogate endpoints that are reasonably likely to predict clinical benefit for ...
[38]
[PDF] Considerations for Discussion of a New Surrogate Endpoint(s ... - FDA
This is not a form but should be used as a guide and incorporated into a comprehensive background package. 1 PDUFA VI Commitment Letter, Page 22. 2 PDUFA VI ...
[39]
5.3 - Surrogate Endpoints | STAT 509 - STAT ONLINE
Examples of surrogate endpoints include CD4 counts in AIDS patients, tumor size reduction in cancer patients, blood pressure in cardiovascular disease, and ...
[40]
Surrogate end points in heart failure - ScienceDirect.com
A surrogate end point is defined as a measurement that can substitute for a true end point for the purpose of comparing specific interventions or treatments in ...
[41]
[PDF] LDL-C and HDL-C as Surrogate Endpoints - FNIH
Subsequently randomized clinical trials confirmed that the modification of levels of. LDL-C could reduce the occurrence rate of major cardiovascular events. COU ...
[42]
perils of surrogate endpoints | European Heart Journal
A potential surrogate is often considered as an intermediate endpoint in a clinical trial because it is found to predict outcome in observational studies. Thus, ...
[43]
[PDF] Endpoints in Cardiovascular Trials - FDA
“A surrogate endpoint, or 'marker,' is a laboratory measurement or physical sign that is used in therapeutic trials as a substitute for a clinically meaningful ...
[44]
[PDF] Clinical Trial Endpoints for the Approval of Cancer Drugs and ... - FDA
Treatment effect measured by PFS can be a surrogate endpoint to support accelerated approval, a surrogate endpoint to support traditional approval, or it ...
[45]
Surrogate Endpoints in Oncology & Use for Cancer Drug Approvals
Aug 25, 2025 · Surrogate endpoints, such as overall response rate and progression-free survival (PFS), are increasingly being used to support approvals of ...
[46]
Clinical endpoints in oncology - a primer - PMC - NIH
This review explores the roles of primary and surrogate clinical endpoints as well as the benefits and drawbacks of each specific endpoint.
[47]
Progression-free survival as a surrogate for overall survival in ...
Mar 26, 2020 · We aimed to identify all studies evaluating the validity of PFS as a surrogate for OS in oncology, and to describe their methodological characteristics.Data Extraction · General And Clinical... · Discussion
[48]
FDA Acceptance of Surrogate End Points for Cancer Drug Approval
Apr 27, 2020 · The FDA has used surrogate end points approximately 194 times to approve cancer drugs since 1992, and about 1 in 3 times, a surrogate was used ...
[49]
[PDF] Regulatory Decision to Withdraw Avastin (bevacizumab) First - FDA
Dec 15, 2010 · FDA has considered PFS as a surrogate endpoint of clinical benefit rather than a direct measure of clinical benefit. In granting accelerated ...
[50]
Regulatory and clinical consequences of negative confirmatory trials ...
Sep 9, 2021 · One trial (bevacizumab in glioblastoma) showed that the drug failed to improve both overall survival and quality of life, and one trial ( ...
[51]
The Strength of Association Between Surrogate End Points and ...
The use of surrogate end points in oncology has led to use of toxic drugs that do not improve survival. For example, bevacizumab (Avastin; Roche/Genentech) ...
[52]
What clinicians should know about surrogate end points in ...
Surrogate end points are markers that can be measured early which correlate with long-term outcomes to allow more rapid assessment of efficacy in clinical.
[53]
Bedside implications of the use of surrogate endpoints in solid and ...
Oct 2, 2024 · Conclusion. Surrogate endpoints have gained massive popularity in oncology, with PFS surpassing overall survival as the most common primary ...
[54]
CMV viral load kinetics as surrogate endpoints after allogeneic ... - NIH
FDA approval of VL surrogate endpoints revolutionized antiviral drug development for HIV and hepatitis C (8, 10). Clinical trials for HIV and hepatitis C ...
[55]
[PDF] Developing Antiretroviral Drugs for Treatment Guidance for Industry
in surrogate endpoints, primarily plasma HIV-RNA levels but also CD4. + cell ... HIV-RNA viral load, baseline CD4. + cell count, clade, and baseline ...
[56]
Jeffrey Murray - Forum for Collaborative Research
“FDA considers CMV viremia (DNAemia) as a validated surrogate endpoint to be used as a part of a composite endpoint to support traditional approval. Therefore, ...
[57]
Surrogate markers and microbiologic end points - PubMed
Aug 1, 2010 · A decreasing or low procalcitonin level after initiation of antibiotic treatment correlates well with bacterial eradication.
[58]
Surrogate Markers and Microbiologic End Points - Oxford Academic
A decreasing or low PCT level after initiation of antibiotic treatment correlates well with bacterial eradication.
[59]
Considerations for the Feasibility of Neutralizing Antibodies as a ...
A neutralizing antibody titer of 1:32 for EV71 vaccine has been recommended as the immunological surrogate endpoint because of its association with protection ...
[60]
Immunological surrogate endpoints of COVID-2019 vaccines - Nature
Feb 2, 2021 · These data suggest that NAb may be a potential CoP and NAb titers of 100-250 may be a protective antibody level against SARS-CoV-2 infection in ...
[61]
Validation of neutralizing antibody titers for estimating vaccine ...
Aug 31, 2023 · These analyses demonstrate that models using neutralizing antibody titers can provide rapid VE estimates which can inform vaccine design and selection.
[62]
Definitions, acceptability, limitations, and guidance in the use and ...
Jun 26, 2023 · A surrogate end point is expected to predict clinical benefit (or harm or lack of benefit or harm) based on epidemiologic, therapeutic, ...
[63]
Magnetic resonance imaging as a surrogate outcome for multiple ...
Magnetic resonance imaging (MRI) of lesions in the brain may be the best current candidate for a surrogate biological marker of clinical outcomes in relapsing ...
[64]
MRI lesions as a surrogate for relapses in multiple sclerosis - PubMed
Jun 3, 2013 · Our findings indicate that the effect of a treatment on relapses can be accurately predicted by the effect of that therapy on MRI lesions.
[65]
Surrogate endpoints for EDSS worsening in multiple sclerosis
The correlation between the treatment effect on MRI lesions and EDSS worsening was slightly weaker (R2 = 0.57) but significant. Conclusions: These findings ...
[66]
The validity of biomarkers as surrogate endpoints in Alzheimer's ...
A surrogate endpoint is expected to predict clinical benefit based on epidemiologic, therapeutic, pathophysiologic, or other scientific evidence”. The use of ...<|separator|>
[67]
Blood biomarkers as surrogate endpoints in Alzheimer's disease ...
May 8, 2025 · The objective of this study is to assess the association between short-term blood biomarker data and long-term cognitive and brain structure ...
[68]
Surrogate endpoints in Parkinson's disease research - PubMed
This paper outlines the definition, role, validity process, and risks associated with surrogate endpoints. The use of biomarkers in recent PD clinical trials is ...
[69]
Evidence from Biomarkers and Surrogate Endpoints - NeuroRX
4. Morrish, PK. How valid is dopamine transporter imaging as a surrogate marker in research trials in Parkinson's disease? Mov Disord. 2003; 18:S63-S70.
[70]
Digital Outcomes as Biomarkers of Disease Progression in Early ...
Nov 29, 2024 · Outcomes derived from digital health technologies (DHTs) are promising candidate markers for monitoring Parkinson's disease (PD) progression ...<|control11|><|separator|>
[71]
Trials for Slowly Progressive Neurogenetic Diseases Need ... - NIH
The FDA has an accelerated approval pathway that could be granted if surrogate endpoints are reasonably likely to predict a clinical benefit. Meaningful ...
[72]
Surrogate Markers and Clinical Outcomes for Nononcologic Chronic ...
Apr 22, 2024 · This study examines the strength of association between surrogate markers used as primary end points in clinical trials to support FDA ...
[73]
Accelerated Approval Program - FDA
Aug 21, 2025 · The use of a surrogate endpoint can considerably shorten the time required prior to receiving FDA approval. Drug companies are still required ...Ongoing · Verified Clinical Benefit · Withdrawn · Ongoing | Infectious Disease...
[74]
Estimation of Study Time Reduction Using Surrogate End Points ...
Apr 1, 2019 · Using surrogate end points for oncology drug approval was associated with reduced drug development time by approximately 11 months compared with an overall ...
[75]
Time to Review the Role of Surrogate End Points in Health Policy
Results from surrogate end points generally accrue more quickly than from final end points, thus allowing for clinical trials with shorter follow-up periods and ...
[76]
Frequently asked questions on surrogate endpoints in oncology ...
In this article, we provide a comprehensive review of surrogate endpoints currently used in oncology trials both in curative and advanced disease settings.
[77]
Surrogate Endpoints for Accelerated Approval
Apr 13, 2022 · The main purpose for using a surrogate endpoint is to shorten clinical development timelines or improve the feasibility of clinical studies in ...
[78]
Validated Surrogate Endpoint - NCBI - NIH
Sep 25, 2017 · Sustained virologic response 12 weeks after treatment (SVR12) has been considered a validated surrogate and is used as the primary endpoint in ...
[79]
Surrogate endpoints: a key concept in clinical epidemiology
Surrogate endpoints are biomarkers or intermediate outcomes that are used as substitutes for clinical outcomes of interest, often to expedite research or ...<|separator|>
[80]
https://www.sciencedirect.com/science/article/pii/S0895435623003402
[81]
Progression-free survival as surrogate and as true end point
In advanced colorectal cancer, in contrast, current evidence indicates that PFS is a valid surrogate for OS after first-line treatment with chemotherapy.
[82]
Surrogate endpoints for early-stage breast cancer - PubMed Central
Nov 29, 2021 · In the FDA's Surrogate Endpoint Table, the surrogate measures listed for breast cancer include (1) pCR, (2) EFS, (3) DFS, (4) objective response ...<|control11|><|separator|>
[83]
Debate: The slippery slope of surrogate outcomes - PMC - NIH
A concern is that changes in surrogate markers may not reflect changes in disease outcomes. Two recent clinical trials (Heart and Estrogen/Progestin Replacement ...
[84]
Effects of Torcetrapib in Patients at High Risk for Coronary Events
Torcetrapib therapy resulted in an increased risk of mortality and morbidity of unknown mechanism. Although there was evidence of an off-target effect of ...Missing: dissociation | Show results with:dissociation
[85]
The failure of torcetrapib: what have we learned? - PMC
Jun 9, 2008 · However, learning from the failure of torcetrapib, we must remain cautious in accepting promising results from surrogate end-point trials, and ...
[86]
Surrogate endpoints in randomized cardiovascular clinical trials
Jul 5, 2011 · An example is the use of LDL cholesterol lowering as a surrogate for reduction in CHD events in trials of statin therapy. In this case, a strong ...
[87]
The US FDAs withdrawal of the breast cancer indication for Avastin ...
... surrogate endpoint such as tumor shrinkage that suggests the drug has ... FDA news release: FDA begins process to remove breast cancer from Avastin label.
[88]
FDA validation of surrogate endpoints in oncology: 2005–2022
We sought to examine surrogate validation studies conducted by the US FDA published between 2005 and 2022.
[89]
Aducanumab (marketed as Aduhelm) Information - FDA
Jul 8, 2021 · Aduhelm is an amyloid beta-directed antibody indicated to treat Alzheimer's disease. Aduhelm is approved under the accelerated approval pathway.Missing: dissociation | Show results with:dissociation
[90]
Revisiting FDA Approval of Aducanumab
Jul 28, 2021 · Examples of failed surrogates, sometimes identified only after widespread harms have occurred, range from ventricular arrythmias (as a surrogate ...
[91]
Failure to demonstrate efficacy of aducanumab: An analysis of the ...
Nov 1, 2020 · After a pair of phase III trials failed to achieve their goals, review of the unblinded data suggested that mild, but not more advanced, ...
[92]
Lessons Learned from Approval of Aducanumab for Alzheimer's ...
Jan 29, 2024 · But 2 years later, aducanumab's failure to launch provides a cautionary tale about the complexities of dementia and the need for a thorough and ...
[93]
Aducanumab and the “post-amyloid” era of Alzheimer research?
Oct 6, 2021 · Overall, more than 10 different anti-amyloid agents have failed in large phase III trials, while dozens more have failed in smaller trials.
[94]
Accelerated Approval | FDA
Feb 24, 2023 · A surrogate endpoint used for accelerated approval is a marker - a laboratory measurement, radiographic image, physical sign or other measure ...Priority Review · Fast Track · Breakthrough TherapyMissing: framework | Show results with:framework
[95]
The Food and Drug Administration's Accelerated Approval Process ...
Accelerated approval was codified into statute in 2012. FDA has granted accelerated approval for about 300 drugs between 1992 and 2021. Hubbard called the ...
[96]
Project Confirm - FDA
A: As a condition of accelerated approval, the applicant should conduct confirmatory trials to verify the clinical benefit of the drug or demonstrate an effect ...
[97]
Accelerated Approval and Considerations for Determining Whether ...
Jan 7, 2025 · The CAA also requires sponsors to submit postmarketing reports to FDA on the progress of required confirmatory trials approximately every 180 ...<|control11|><|separator|>
[98]
A Quick Take on FDA's New Draft Accelerated Approval Guidance ...
Dec 16, 2024 · When operating under accelerated approval, FDA will approve a drug based on clinical trial data showing the drug has an effect on a surrogate ...
[99]
Surrogate Endpoints in Drug Development: A Review of Statistical ...
Mar 19, 2025 · A surrogate endpoint is an endpoint that is used as a substitute for a direct clinical outcome measure of how a patient feels, functions, or survives.Regulatory Considerations for... · Statistical Methodology · Applications · Discussion
[100]
Validity of Surrogate Endpoints and Their Impact on Coverage ... - NIH
... clinical endpoint. We screened the guidance ... Surrogate threshold effect: an alternative measure for meta-analytic surrogate endpoint validation.
[101]
Frequently asked questions on surrogate endpoints in oncology ...
Moreover, the surrogate endpoint usually lies somewhere in the middle of the chain between the start of treatment and death and can be reached sooner than OS, ...
[102]
POSB285 Five Identified Hurdles Associated with Surrogate ...
The main challenges in surrogate endpoint validation are limited knowledge of diseases, lack of incentives and harmonized criteria, costly and time-consuming.
[103]
[PDF] Surrogate Endpoints Under Attack – Is It Still Worth Performing ...
Dec 16, 2021 · Dr. Paddock noted that this uncertainty necessarily implies the risk of being wrong about surrogacy, and that failures on either side of the ...
[104]
Research and Development in the Pharmaceutical Industry
Apr 8, 2021 · In this report, the Congressional Budget Office assesses trends in spending for drug research and development (R&D) and the introduction of new drugs.
[105]
The role of imperfect surrogate endpoint information in drug ...
Since surrogates are imperfectly correlated with the final outcomes of interest, surrogate endpoint information provides weaker evidence of the benefit value ...Missing: shortcomings | Show results with:shortcomings
[106]
[PDF] strengthening the accelerated approval pathway: an analysis ... - ICER
Apr 26, 2021 · The FDA accepts surrogate endpoints for clinical studies of pharmaceutical products in several instances. Approval decisions through the ...
[107]
Speed, Safety, and Industry Funding — From PDUFA I to PDUFA VI
Dec 7, 2017 · We then consider the effect of user-fee legislation on FDA operations and the legislative process. Expanding Fees and FDA Commitments, Shrinking ...More User Fees For Other... · Fda Reauthorization Act Of... · Policy Implications
[108]
Prescription Drug User Fee Act Reauthorization (PDUFA VI ... - FDA
PDUFA authorized FDA to collect user fees to speed up drug reviews, improve public health, and provide faster access to new drugs.
[109]
The Impact of Industry Funding on Randomized Controlled Trials of ...
Feb 28, 2022 · Industry-funded trials investigating biologic therapies are more likely to yield statistically significant positive outcomes and use placebo comparators.
[110]
WVU research shows how much pharmaceutical companies are ...
Jun 12, 2023 · ... pharmaceutical companies are capitalizing on rare drug incentives ... “They get fast-track FDA approval and also get to use 'surrogate endpoints,' ...
[111]
https://wvutoday.wvu.edu/stories/2023/06/12/wvu-research-shows-how-much-pharmaceutical-companies-are-capitalizing-on-rare-drug-incentives
[112]
Reporting of surrogate endpoints in randomised controlled trial ...
Jul 9, 2024 · Surrogate endpoints are frequently used to improve trial efficiency (eg, to shorten duration of follow-up, reduce sample size, and, thus, lower ...
[113]
Reporting of surrogate endpoints in randomised controlled trial ...
Jul 9, 2024 · This report presents the SPIRIT-Surrogate, an extension of the Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) checklist.
[114]
A framework for the definition and interpretation of ... - PubMed Central
Interventional trials that evaluate treatment effects using surrogate endpoints have become increasingly common. This paper describes four linked empirical ...
[115]
Validation of a Longitudinal Marker as a Surrogate Using Mediation ...
Jun 27, 2025 · In this work, we propose a method for validating a longitudinal outcome as a surrogate of a time‐to‐event endpoint using a combination of joint ...
[116]
Surrogate Marker Evaluation: A Tutorial Using R - PMC - NIH
Sep 3, 2025 · Surrogate marker evaluation is one application of principal stratification, developed further by Gilbert and Hudgens (2008). The framework is ...