Human Challenge Study Design in Action - a Dengue Fever vaccine trial

A human challenge study, also known as a controlled human infection model (CHIM), is a type of clinical research study in which healthy volunteers are intentionally exposed to a specific pathogen (such as a virus, bacterium, or parasite) under controlled conditions. The primary goal of these studies is to better understand the pathogen's behavior, the human immune response to it, and to test the effectiveness of potential treatments, vaccines, or preventive measures. Human challenge studies can provide valuable insights into disease progression, immunity, and treatment efficacy in a controlled and ethical manner.

These studies are typically conducted under strict ethical and safety guidelines to minimize the risk to participants. Participants are closely monitored, and their informed consent is obtained. Human challenge studies have been used to study a variety of diseases, including influenza, malaria, Dengue fever, and COVID-19, among others. They play a crucial role in advancing medical and scientific knowledge and can accelerate the development of treatments and vaccines.

A human challenge study was mentioned as an alternative clinical trial design at the beginning of the COVID-19 pandemic when the world was desperate to find an effective and safe vaccine. I wrote an article about this: "Human Challenge Study Design for Covid-19 Vaccine Clinical Trials?"

Just this morning, Janssen Announces Promising Antiviral Activity Against Dengue in a Phase 2a Human Challenge Model. The results were from a phase 2a study titled "A Phase 2a, Randomized, Double-blind, Placebo Controlled Trial to Evaluate the Antiviral Activity, Safety, and Pharmacokinetics of Repeated Oral Doses of JNJ-64281802 Against Dengue Serotype 3 Infection in a Dengue Human Challenge Model in Healthy Adult Participants" that was posted on clinicaltrials.gov. Unfortunately, the clinical trial registration did not contain any description of the 'Challenge' part (i.e., how the healthy volunteers are exposed to the infectious agents (in this case, the Dengue virus). We will just need to wait for the formal publication of the study to know the details. 

In a paper by Porter et al "A human Phase I/IIa malaria challenge trial of a polyprotein malaria vaccine", the whole details about the human challenge study including the 'challenge' part were discussed. The 'sporozoite challenge' to the healthy volunteers was described below: 

 

Drugs Approved by FDA Despite Failed Trials or Minimal/Insufficient Data

I have been trying to collect the cases of that drugs were approved by the FDA despite the failed trials or minimal/insufficient data. For drugs treating rare diseases or diseases with unmet medical needs, the FDA may apply flexibility in approving the drug with loosened criteria. 

For diseases with clearly unmet medical needs such as ALS (Amyotrophic Lateral Sclerosis) and Alzheimer's disease, FDA officials have recently emphasized the urgent need for new treatments and pledged to use maximum "regulatory flexibility" when reviewing the NDA/BLA packages. By applying the maximum "regulatory flexibility", FDA has approved some drugs which do not meet the agency's traditional approval standards. Some of the approvals are really controversial and make me wonder if there is any boundary for the maximum "regulatory flexibility". 

The following paper on BioSpace.com listed six drugs that earned FDA approval without substantial evidence of effectiveness.

6 Drugs Approved Despite Failed Trials or Minimal Data

• Ipsen’s Sohonos (palovarotene) for the ultra-rare genetic disease fibrodysplasia ossificans progressive (FOP)
• Sarepta’s Elevidys as the first gene therapy for Duchenne muscular dystrophy (DMD)
• Biogen's Qalsody (tofersen) to treat patients with superoxide dismutase 1 (SOD1)-ALS, a rare subtype of the fatal neurodegenerative disease
• Biogen and Eisai got the nod for Aduhelm (aducanumab) for Alzheimer's diease,
• Jazz Pharmaceuticals and PharmaMar’s Zepzelca (lurbinectedin) for small cell lung cancer (SCLC) that had progressed on or after platinum-based chemotherapy
• Acadia Pharmaceuticals’ Nuplazid (pimavanserin) to treat hallucinations and delusions associated with psychosis in Parkinson’s disease.

Some of the approvals gave the sponsors the false hope that an innovative drug could be approved by the FDA even if the study failed to demonstrate the effectiveness as long as the drug was for the treatment of diseases with urgent unmet medical needs. A recent story about BrainStorm's ALS drug is exactly the case about this. 

MCID (Minimum Clinical Important Difference) for 6MWD - how low can we go?

I went back to watch the FDA CRDAC (Cardiovascular and Renal Drugs Advisory Committee) meeting to discuss Alnylam's drug Patisiran for the treatment of ATTR-CM (Transthyretin Amyloidosis) - a rare form of heart disease. The meeting discussion was centered on the clinical meaningfulness of the efficacy measures in the primary efficacy endpoint of 6MWD (how many meters patients can walk in 6 minutes) and the secondary endpoint of KCCQ - a patient-reported quality of life measure. 

The sponsor, Alnylam, conducted a phase III study called "APOLLO-B: A Study to Evaluate Patisiran in Participants With Transthyretin Amyloidosis With Cardiomyopathy (ATTR Amyloidosis With Cardiomyopathy)". The study results showed statistically significant differences in 6MWD and in KCCQ total score. However, the magnitude of the treatment differences was very small: 14.7 meters in 6MWD and 3.7 points in KCCQ at month 12.

To judge if the treatment difference is clinically meaningful, people will compare the magnitude of the treatment differences from the study with the MCID (minimal clinically important difference). MCID. MCID is the smallest change in a treatment outcome that individual patients would identify as important and which would indicate a change in the patients' management.  The MCID is a patient-centered concept that captures both the magnitude of the improvement and the value patients place on the change.  In other words, the MCID is the smallest amount of change in the score of a scale recognized by the patient without considering the side effects and cost. 

In FDA's briefing book for CRADAC meeting, FDA casted doubts about the Patisiram's efficacy: 

The 6MWT, a performance outcome (PerfO), is a practical simple test that measures the distance that a patient can quickly walk on a flat, hard surface in a period of 6 minutes (the 6MWD). It evaluates the global and integrated responses of all the systems involved during exercise. The results of the APOLLO-B trial showed a statistically significant but small treatment effect for the primary efficacy endpoint. Subjects treated with patisiran experienced an average decrease in their 6MWD of 13 m at Month 12 from an average 6MWD of 361 m at baseline, while subjects in the placebo arm experienced an average decrease in their 6MWD of 31 m at Month 12 from an average 6MWD of 375 m at baseline. The change from baseline at Month 12 in 6MWT (Hodges-Lehmann [HL] estimate of median difference) for patisiran vs. placebo was 14.7 m (95% confidence interval [CI] 0.7, 28.7; p-value 0.04). Literature has reported a range of meaningful differences (22 to 90 m) reflective of the heterogeneity in cardiomyopathy patients (Mathai et al. 2012; Shoemaker et al. 2012).

 The KCCQ, a patient-reported outcome (PRO) and a disease-specific measure for HF, is a 23-item self-administered questionnaire developed to measure the patient’s perception of their health status, which includes heart failure symptoms, impact on physical and social function, and how heart failure impacts their quality of life (QOL) within a 2-week recall period. The KCCQ-OSS has a 0-100 transformed score range where higher scores reflect better health status (based on the Physical Limitation, Symptom Frequency, Symptom Burden, Quality of Life and Social Limitations Domain Scores). In the APOLLO-B trial,the treatment effect for the first secondary efficacy endpoint, change from baseline at Month 12 in KCCQ-OSS was small (3.7 points on a 0 to 100 transformed score range; 95% CI 0.2, 7.2; p-value 0.04). On average, subjects treated with patisiran had an increase in KCCQ-OSS of 0.3 points at Month 12 from the average baseline score of 69.8 points, while subjects in the placebo arm had a decrease in KCCQ-OSS of 3.4 points at Month 12 from the average baseline score of 70.3 points.  

Sponsor, Alnyam's briefing book and presentation spent a lot of effort to defend that the small, but statistically significant treatment differences are clinically meaningful. 

Sponsor attempted to derive an MCID using KCCQ category as an anchor based on the data from the study itself (APOLLO-B study). 

Not surprisingly, the MCID they generated were much smaller (MCID in the range of 7 - 8 meters) than the MCIDs reported in the literature. If the MCID is indeed in the range of 7 - 8 meters, the 14.7 meters (treatment difference observed in Apollo-B study) would be clinically meaningful. 

During the FDA advisory committee meeting, most of the members were not convinced by sponsor's presentation to defend the clinical meaningfulness of  small treatment difference in 6MWD (about 14 meter). However, majority of them (9-3) still voted in favor of the Patisiran's efficacy and the benefit-risk profile. 

FDA panel endorses Alnylam's heart drug - after picking apart its supporting data 

Expert Perspective: FDA CRDAC Vote on Patisiran for ATTR-CM, with Ahmad Masri, MD

Pfizer's tafamidis is the only approved drug for the treatment of ATTR-CM. According to the product label, the treatment difference in 6MWD was much larger - 76 meters with 95% confidence interval 58, 94 meters at month 30. 

For the same 6MWD, the MCID may be different depending on the treating diseases, different patient population, whether patients receiving the background therapies,... However, a treatment difference of 14 meters is still not a convincing number to be clinical meaningful. Putting on the relative scale, the 14 meters in patients with baseline 6MWD 361 meter is less than 5%. It is difficult to convince people a treatment difference less than 5% is clinically meaningful. 

I am particularlly interested in the MCID of 6MWD in lung diseases (especially the pulmonary arterial hypertension). 

Anne E. Holland (2014) "An official European Respiratory Society/American Thoracic Society technical standard: field walking tests in chronic respiratory disease" stated

“Available evidence suggests a minimal important difference (MID) of 30 m for the 6MWD in adults with chronic respiratory disease.”

Jude Moutchia (2023) "Minimal Clinically Important Difference in the 6-minute-walk Distance for Patients with Pulmonary Arterial Hypertension" found:

The minimal clinically important difference in the derivation sample was 33 meters (95% confidence interval, 27–38), which was almost identical to that in the validation sample (36 m [95% confidence interval, 29–43]). The minimal clinically important difference did not differ by age, sex, race, pulmonary hypertension etiology, body mass index, use of background therapy, or World Health Organization functional class.

Here is a table containing some literatures with estimated MCID. The MCID was found to be in the range of 20 - 54 meters depending on the indication/disease. 

Study/Article	Indication/Disease	MCID Range	MCID Midpoint
Chan (2015)	ARF	20 – 30	25
du Bois (2011)	IPF	24 – 45	35
Gilbert (2009)	PAH	41	41
Granger (2015)	Lung Cancer	22 – 42	32
Holland (2009)	DPLD/IPF	29 – 34	32
Holland (2010)	COPD	25	25
Mathai (2012)	PAH	33	33
Nathan (2015)	IPF	22 – 37	30
Polkey (2013)	COPD	30	30
Puhan (2008)	COPD	35	35
Puhan (2011)	COPD	24 – 28	26
Redelmeier (1997)	CLD	54	54
Swigris (2010)	IPF	28	28

Latest update: 

In the end, the FDA did not approve Patisiran for the treatment of ATTR-CM  because the treatment difference in 6MWD was too small (way below the MCID)  and not clinically meaningful even though the FDA advisory committee voted in favor of the Patisiran's benefit and there was no issue with the safety and the manufacturing. 

Alnylam Announces Receipt of Complete Response Letter from U.S. FDA for Supplemental New Drug Application for Patisiran for the Treatment of the Cardiomyopathy of ATTR Amyloidosis

FDA blocks Alnylam’s bid to expand Onpattro label - PharmaLive

 "In its Complete Response Letter (CRL), the regulator said that Alnylam had not provided enough evidence of the therapy’s benefit in the proposed indication. At the same time, the FDA did not flag any problems with patisiran’s clinical safety, drug quality, manufacturing processes or study conduct.

“The CRL indicated that the clinical meaningfulness of patisiran’s treatment effects for the cardiomyopathy of ATTR amyloidosis had not been established,” according to the company’s announcement. In light of the rejection, Alnylam will no long work toward an expanded label for Onpattro in the U.S."

Platform trials in action beyond oncology trials

One of the complex innovative designs is the study with 'Master protocol' that has been getting popularity in recent years. Master protocol simply refers to the study of more than one drug with a single protocol. In clinical trials with Master protocol, the protocol document stands alone, without reference to specific drugs. Protocol amendments (or study-specific protocol) are used to provide details of each drug. Master protocols establish a trial network with infrastructure in place to streamline trial logistics and improve data quality, and facilitate data sharing and new data collection. Master protocols develop a common protocol for the network that incorporates innovative statistical approaches to study design and data analysis.

Platform trial is one type of Master protocol and refers to establishing the trial infrastructure and master protocol as a perpetuating effort, with drugs entering and leaving the platform.

Two prominent references for master protocols and platform trials are the following: 

 "Master Protocols to Study Multiple Therapies, Multiple Diseases, or Both" in 2017 by Woodcock and LaVange defines Master protocols include three type of trials: Umbrella, basket, and platform trials. 

FDA procedural guidance in 2022 "Master Protocols: Efficient Clinical Trial Design Strategies to Expedite Development of Oncology Drugs and Biologics Guidance for Industry" provided the definition for Master protocols:

a master protocol is defined as a protocol designed with multiple substudies, which may have different objectives and involve coordinated efforts to evaluate one or more investigational drugs in one or more disease subtypes within the overall trial structure.

A master protocol may be used to conduct the trial or trials for exploratory purposes or to support a marketing application and can be structured to evaluate, in parallel, different drugs compared with their respective controls or to a single common control. The sponsor can design the master protocol with a fixed or adaptive design with the intent to modify the protocol to incorporate or terminate individual substudies within the master protocol. 

Master protocols can include basket trial, umbrella trial, and platform trial and these trials can be schematically displayed as the following: 

Basket trial

Umbrella Trial

Platform Trial:

In  a paper by Berry et al, "The Platform Trial An Efficient Strategy for Evaluating Multiple Treatments", the platform trial was compared to the traditional trial: 

While platform trial has its challenges, we see the implementation trials in action, beyond the oncology trials. Here are examples of platform trials in areas other than the oncology trials. 

HEALEY ALS Platform Trial

The HEALEY ALS Platform Trial is a perpetual multi-center, multi-regimen clinical trial evaluating the safety and efficacy of investigational products for the treatment of ALS. This trial is designed as a perpetual platform trial. This means that there is a single Master Protocol dictating the conduct of the trial.

In this trial, multiple investigational products for ALS will be tested simultaneously or sequentially. Each investigational product will be tested in a regimen. Each regimen consists of a placebo-controlled trial, meaning that the active investigational product and matching placebo will be tested in each regimen.

The study website is here: https://www.massgeneral.org/neurology/als/research/platform-trial

The study master protocol can be accessed here

Several testing drugs have been completed or graduated from the platform, please see the news releases

PrecISE (Precision Interventions for Severe and/or Exacerbation-Prone Asthma) Network Study

PrecISE is an adaptive platform trial under master protocol with common biomarker screening where drugs enter when available and discontinue based on futility analysis. According to clinicaltrials.gov, 5 novel interventions are currently listed as the testing drug for study: Medium Chain Triglycerides (MCT), Clazakizumab, Broncho-Vaxom, Imatinib Mesylate, Cavosonstat; each active will be tested against its own control.

the study design was described in the paper by Israel et al "PrecISE: Precision Medicine in Severe Asthma: An adaptive platform trial with biomarker ascertainment"

REMAP-CAP: A Randomised, Embedded, Multi-factorial, Adaptive Platform Trial for Community-Acquired Pneumonia

REMAP-CAP platform trial was originally designed for identifying the treatments for community-acquired pneumonia. After the Covid pandemic in 2020, REMAP-CAP has quickly implemented the Pandemic Appendix to the Core Protocol so that the platform can respond rapidly in the event of widespread disease resulting from the novel 2019 coronavirus (COVID-19).

ACTIV (NIH) Accelerating Covid-19 Therapeutic Interventions and Vaccines

Working in an unprecedented time frame, the Accelerating COVID-19 Therapeutic Interventions and Vaccines (ACTIV) public–private partnership developed and launched 9 master protocols between 14 April 2020 and 31 May 2021 to allow for the coordinated and efficient evaluation of multiple investigational therapeutic agents for COVID-19. The ACTIV master protocols were designed with a portfolio approach to serve the following patient populations with COVID-19: mild to moderately ill outpatients, moderately ill inpatients, and critically ill inpatients.

The study design was described in the paper by LaVange et al "Accelerating COVID-19 Therapeutic Interventions and Vaccines (ACTIV): Designing Master Protocols for Evaluation of Candidate COVID-19 Therapeutics"

RECOVER Clinical Trials - for Long Covid

Long COVID is defined as "a multifaceted disease that can affect nearly every organ system" and can manifest as new or worsening chronic health problems, including but not limited to heart disease, diabetes, kidney disease, hematologic issues, and mental and neurologic conditions. The signs, symptoms, and conditions continue or arise anew 4 weeks or more after the initial symptomatic or asymptomatic infection and may be relapsing and remitting.

NIH RECOVER trial is also using platform protocols/clinical trials to investigate treatments for long covid and master protocol is posted on the study website: https://trials.recovercovid.org/design. There will be multiple master protocols: RECOVER-VITAL, RECOVER-NEURO, RECOVER-AUTONOMIC (coming soon), RECOVER-SLEEP (coming soon) to investigate different

It is ironic that the RECOVER trial is started after NIH spent 2.5 years and $1 billion on long Covid research and failed to test meaningful long Covid treatments. See the article "Underwhelming’: NIH trials fail to test meaningful long Covid treatments — after 2.5 years and $1 billion "

Clinicaltrialsarena.com has an featured article "Platform trials: an opportunity for rare dystrophies or gene therapies?". The article discussed the possibility of platform trials in rare diseases and gene therapies. PaVeGT platform trial was listed: 

PaVe-GT: Paving the Way for Rare Disease Gene Therapies

PaVe-GT will develop and test AAV-9, with a different gene for each indication, using a single master protocol. The NCATS-led Platform Vector Gene Therapy (PaVe-GT) pilot project seeks to increase the efficiency of clinical trial startup by using the same gene delivery system and manufacturing methods for multiple rare disease gene therapies. We will make program results and regulatory documents publicly available, with the intention of benefiting future gene therapy clinical trials for very rare diseases.

 In a paper by Collignon (2022) "An Economic Perspective on Platform Trials—The Gift and the Curse", the following conclusion was made about the platform trial: 

"...sharing a common control group in a platform trial can be viewed as a gift and a curse, and choosing whether to implement a platform trial or a more standard development program is complex since it is contingent on numerous factors that are disease and context dependent. To make such a decision, a clear framework needs to be implemented. Decision-making in the pharmaceutical industry is becoming increasingly more quantitative, and in practice, both development approaches would be compared according to a series of standard metrics, including

(1) duration of the clinical program;

(2) cost and expected reward of the clinical program;

(3) flexibility of the clinical program (eg, reporting of the analyses corresponding to the different treatments while maintaining data and trial integrity, ability to add sites, and so forth); (4) probability of success for each treatment, for all treatments, or for at least 1 treatment, assuming all or some of the treatments have a certain efficacy;

(5) probability of (multiple) false-positive findings (eg, achieving statistical significance) for each treatment or for at least 1 treatment, assuming all or some of the treatments are comparator-like; and

(6) the extent to which a positive readout informs the success of potential subsequent trials and the ability to de-risk the next phase of development.

Using these metrics, governance boards would then make their decisions on whether to progress developing a treatment according to a range of diverse considerations, such as portfolio opportunities and investment priorities."

Mediation Analysis vs. Landmark Analysis for Clinical Trial Data

Mediation analysis and Landmark analysis are two valuable statistical tools in different contexts, providing insights into different aspects of data analysis. Both methods can be utilized to investigate if a surrogate endpoint or short-term measure can predict the long-term clinical endpoint or to investigate if short-term measures can be mediators for the long-term clinical endpoint. Both mediation analysis and landmark analysis are useful tools for post-hoc, exploratory analyses, but not the primary analysis method for the primary efficacy endpoint. 

Mediation analysis is a statistical method used to explore and understand the mechanism or process through which an independent variable influences a dependent variable. It helps to determine whether the effect of an independent variable on a dependent variable is mediated (i.e., transmitted through) one or more intermediate variables, often referred to as mediators.

In mediation analysis, the main focus is on understanding the causal chain of relationships between variables. It involves three key components:

• Independent variable (X): The variable that is hypothesized to influence the dependent variable directly or indirectly through one or more mediators.
• Mediator (M): The variable(s) that mediates the relationship between the independent variable and the dependent variable. It explains how and why the independent variable affects the dependent variable.
• Dependent variable (Y): The variable that is influenced by the independent variable either directly or indirectly through the mediators.

The analysis aims to estimate the direct effect of the independent variable on the dependent variable and the indirect effect mediated through the mediator(s). Several statistical techniques can be used for mediation analysis, such as regression-based methods (e.g., ordinary least squares regression) or more advanced methods like structural equation modeling.

In our published paper "Contemporary risk scores predict clinicalworsening in pulmonary arterial hypertension - Ananalysis of FREEDOM-EV",  mediation analysis was used to test the hypothesis that improvements in risk score (a surrogate endpoint) contributed to reduced likelihood for clinical worsening (a long-term clinical endpoint). 

 

In a paper by Blette et al, "Is low-risk status a surrogate outcome in pulmonary arterialhypertension? An analysis of three randomised trials ", the mediation analysis was used to investigate surrogacy. The author stated:

"we performed a mediation analysis considering each candidate surrogate as an intermediate outcome. To avoid inconsistent mediation (ie, when direct and indirect effects cancel each other out, the direct effect is even larger than the total effect, or other situations that can result in a negative proportion mediated), we first empirically tested four criteria to justify the mediation analysis. Next, Cox proportional hazards models were fit for the clinical worsening and survival outcomes conditional on each candidate surrogate separately, as well as treatment, corresponding baseline risk score, and a fixed effect variable with three levels for trial membership (to allow for similarity within each trial). Results from these models were combined with parallel models that did not condition on the candidate surrogates, but otherwise conditioned on the same set of variables to perform the difference method for mediation, estimating the total effect and direct effects of treatment on clinical worsening and survival, as well as the indirect effects through each candidate surrogate. The proportion of the effect mediated through each surrogate risk score was estimated, along with 95% CIs via a bootstrap procedure."

In a previous post, the mediation analysis was discussed. "Mediation analysis and SAS CAUSALMED procedure"

Landmark analysis, also known as landmark survival analysis, is a statistical method commonly used in survival analysis to investigate the impact of time-dependent variables on the occurrence of an event of interest. It allows for the assessment of time-varying effects in longitudinal studies or clinical trials where the values of variables may change over time.

In landmark analysis, the follow-up time is divided into predefined intervals or "landmarks," and the analysis is performed separately for each landmark. The key steps in landmark analysis are as follows:

• Define landmarks: Choose specific time points of interest during the follow-up period.
• Create landmark cohorts: At each landmark, divide the study population into subgroups based on the status of the time-dependent variable(s) of interest.
• Analyze survival outcomes: Estimate survival probabilities or hazard rates for each subgroup defined by the landmark cohorts.
• Compare survival outcomes: Compare the survival outcomes between the different subgroups to assess the impact of the time-dependent variable(s) on the event occurrence.

Landmark analysis allows researchers to capture time-varying effects and observe how the relationship between variables changes over the course of a study. It is particularly useful when analyzing data with time-dependent covariates, treatment interventions, or changes in exposure levels over time.

In a paper by McLaughlin et al "Pulmonary Arterial Hypertension-Related Morbidity Is Prognostic for Mortality", the landmark analysis was used to assess the impact of morbidity events on the risk of subsequent mortality.

In a paper by Eisenstein et al "Clopidogrel use and long-term clinical outcomes after drug-eluting stent implantation" , landmark analyses were performed to explore the association of extended clopidogrel use and long-term clinical outcomes of patients receiving drug eluting stents (DES) and bare-metal stents (BMS) for treatment of coronary artery disease.

The tutorial paper "Landmark Analysis at the 25-Year Landmark Point" by Dr Dafni is a good reference for Landmark analysis. 

Mediation analysis and Landmark analysis differ in their goals, focus, and types of data they analyze:

Goals: Mediation analysis aims to understand the mechanism or process through which an independent variable influences a dependent variable, exploring direct and indirect effects. Landmark analysis, on the other hand, focuses on assessing time-varying effects and understanding how the occurrence of an event is influenced by time-dependent variables.

Focus: Mediation analysis emphasizes identifying mediators that explain the relationship between an independent variable and a dependent variable. Landmark analysis focuses on investigating the impact of time-dependent variables on survival outcomes or event occurrences.

Data: Mediation analysis typically requires cross-sectional or longitudinal data with variables measured at different time points. It is applicable to both continuous and categorical variables. Landmark analysis is commonly used in survival analysis, analyzing time-to-event data, and requires longitudinal data with time-dependent variables.

Both mediation analysis and landmark analysis are valuable statistical tools in different contexts, providing insights into different aspects of data analysis. Here is a table to compare the mediation analysis and the landmark analysis generated by ChatGPT, but I added the last row for implementing the mediation and landmark analyses. 

Aspect	Mediation Analysis	Landmark Analysis
Statistical Method	Investigates relationships and effects between variables	Analyzes time-varying effects and event occurrences
Time Dependency	Considers temporal aspect of data	Accounts for time-varying effects and changing values over time
Causal Inference	Aims to understand causal chain of relationships	Examines impact of time-dependent variables on event outcomes
Focus	Identifying mediators that explain relationships	Assessing time-dependent variables and event occurrences
Variables	Independent, mediator, and dependent variables	Time-dependent variables influencing event occurrences
Data Type	Cross-sectional or longitudinal with measured variables	Longitudinal data with time-dependent variables
Analytical Steps	Estimating direct and indirect effects using regression	Dividing follow-up into landmarks, comparing survival outcomes
Research Questions	Mechanisms and processes of variable influence	Impact of time-dependent variables on events or survival
Implementation	In SAS, Procedure CASUALMED allows you to estimate direct and indirect effects using different mediation models. It supports various regression-based mediation approaches, including Sobel, bootstrapping, and Bayesian estimation. In R, to conduct mediation analysis, the most commonly used package is "mediation." This package provides a comprehensive set of functions to estimate direct and indirect effects in mediation models. It supports various mediation methods, including the causal steps approach, bootstrapping, and structural equation modeling (SEM).	In SAS, Procedures LIFETEST, PHREG, and LIFEREG can all be used. You can divide the follow-up time into landmark intervals and analyze survival outcomes for different subgroups defined by the landmarks In R, you can utilize the "survival" package, which is widely used for survival analysis. The "survival" package provides functions to handle time-to-event data, perform survival analysis, and estimate survival probabilities. You can divide the follow-up time into landmarks and analyze survival outcomes for different landmark cohorts.

Personally, I prefer the mediation analysis to the landmark analysis. In the landmark analysis, the subjects who did not reach the landmark timepoint were excluded from the analysis and the analyses are performed on a subset of the overall population - sort of principle stratum. The endpoint measure at the landmark timepoint and whether or not the subjects reach the landmark timepoint itself is meaningful. Excluding it from the analysis is against the intention-to-treatment principle and may cause biases. 

Win Ratio and its application in clinical trials

The win ratio is a method for examining composite endpoints in clinical trials. It was introduced in 2012 by Dr Pocock and has since been widely adopted in cardiovascular trials. The win ratio accounts for relative priorities of the components and allows the components to be different types of outcomes. The win ratio is calculated by dividing the total number of winners by the total number of losers.

The win ratio was motivated by the Finkelstein–Schoenfeld (FS) test, with the aim of providing an estimate of the treatment effect (the win ratio) and confidence interval, in addition to a P-value. The general principle behind both the Finkelstein-Schoenfeld test and the win ratio is that they are both methods for examining composite endpoints in clinical trials. Win ratio can be considered as a popular name for the Finkelstein-Schoenfeld test. See my previous post "Finkelstein-Schoenfeld Method, Win Ratio, and Hodges-Lehman Estimates - Statistical Methods Based on All Paired Comparisons". In practice, we may see the terms "Win Ratio test" and "Finkelstein-Schoenfeld test" interchangeably. 

Win Ratio is defined differently than hazard ratio in survival analysis, however, both Win Ratio and Hazard Ratio are important statistical measures used in clinical trials to assess treatment outcomes and compare different treatments. However, they have distinct interpretations and applications, making them suitable for different types of clinical trial data and research questions. Here are comparison table for Win Ratio and Hazard Ratio:

Metric	Win Ratio	Hazard Ratio
Definition	Ratio of treatment success in the experimental group to the control group	Compares the risk of an event occurring in the treatment group to the control group over time
Clinical trial design	Usually fixed-duration studies	Both fixed-duration studies and event-driven studies
Endpoint	Composite endpoint, all events are considered	Composite endpoint, usually time to the first event
Interpretation	Great than 1: Experimental treatment has higher success rate than control group Equal to 1: Both groups have the same success rate<br> Less than 1: Experimental treatment has lower success rate than control group	Greater than 1: Higher risk of event in treatment group<br> Equal to 1: Equal risk of event in both groups<br> Less than 1: Lower risk of event in treatment group
Application	Non-inferiority or superiority trials to assess treatment efficacy	Survival analysis with time-to-event outcomes (e.g., overall survival, progression-free survival)
Focus	Treatment success rates	Risk of an event over time
Type of Data	Binary or categorical data	Time-to-event data

For win-ratio and hazard ratio comparison, please see the article by Ferreira et al "Use of the Win Ratio in Cardiovascular Trials".

Win Ratio approach has been used to re-analyze the clinical trials with a composite endpoint (either the primary efficacy or secondary efficacy endpoint). There is a table in Redfors et al 2020 paper that summarized the re-analysis results using the Win Ratio approach. We recently published a paper in Annals of American Thoracic Society "A Novel Approach to Clinical Change Endpoints: A Win Ratio Analysis of the INCREASE Trial" where the Win Ratio approach was used to analyze the secondary efficacy endpoint of clinical worsening events.

Win Ratio approach has also been prespecified as the primary analysis method to analyze the clinical trials with composite endpoints. A table in Redfors et al 2020 paper summarized seven different trials including ATTR-ACT, Chart-1, and TAVR-UNLOAD trials where the Win Ratio approach was listed as the primary analysis method.

This past week, Biotech company, Bridgebio Pharma announced their phase III study of acoramidis in transthyretin amyloid cardiomyopathy, or ATTR-CM. One of the primary efficacy endpoint is a composite endpoint - a hierarchical combination of 1) all-cause mortality, 2) cumulative frequency of cardiovascular-related hospitalization, 3) change from baseline in NT-proBNP, and 4) change from baseline in 6MWT over a 30-month fixed treatment duration. According to clinicaltrials.gov, the composite endpoint is analyzed using the following approach: 

Each subject will be compared to every other subject within a stratum over outcomes of all-cause mortality (death due to any cause), cumulative frequency of cardiovascular-related hospitalizations (number of times a subject is hospitalized for cardiovascular-related causes), change from baseline in NT-proBNP, and change from baseline in the total distance walked in 6 minutes (distance in meters).

The hierarchical approach with the Finkelstein-Schoenfeld test will be applied and the test recognizes the greater importance of the mortality endpoint. Scores are transformed to -1, 0, +1. The alternative hypothesis is a subject in the acoramidis treatment group will have a greater score than a subject in the placebo group.

In the company's presentation slides, results for the primary endpoint and secondary endpoints are listed. The win ratio is listed as 1.8, which indicates 80% more wins in acoramidis treatment group than in the placebo group.

The win ratio test requires a pairwise comparison (each subject in the treatment group is compared with each subject in the placebo (control) group). The number of pairs is the sample size (n) for the treatment group times the sample size (m) for the placebo group, i.e., n x m pairs. The win ratio test results are largely dependent on the rules or algorithms in determining the win/loss/tie for each individual pair. Therefore, the rules and algorithm need to be pre-specified. For the prospective clinical trials, these rules and algorithms need to be pre-specified in the statistical analysis plan (SAP) and the SAP needs to be submitted to FDA for review to avoid the potential biases or potential impression of biases. 

 

References:

• Pocock et al (2012) The win ratio: a new approach to the analysis of composite endpoints in clinical trials based on clinical priorities.
• Redfors et al, 2020 The win ratio approach for composite endpoints: practical guidance based on previous experience
  
• Ferreira et al (2020) Use of the Win Ratio in Cardiovascular Trials
• Maurer et al (2018) Tafamidis Treatment for Patients with Transthyretin Amyloid Cardiomyopathy