Gaining FDA Support for a Prognostic Biomarker, Giving Patients Hope

Gaining FDA Support for a Prognostic Biomarker, Giving Patients Hope

As a clinician scientist, I get really excited about helping sponsors develop new treatments for patients. For some diseases, the lack of a validated prognostic biomarker is an impediment to developing effective treatments. In this blog post, I’ll discuss how we worked with the Critical Path Institute (C-Path) Polycystic Kidney Disease Outcomes Consortium to gain FDA support for a prognostic biomarker for a rare disease, autosomal dominant polycystic kidney disease (ADPKD). Having a validated biomarker will help spur new research and clinical trials to find a treatment for ADPKD.

What is Autosomal Dominant Polycystic Kidney Disease?

Polycystic Kidney Disease (PKD) is a debilitating genetic disease that affects at least 1 in 500 people globally. It comes in two forms- autosomal dominant and autosomal recessive. Autosomal dominant PKD (ADPKD) is one of the most common life-threatening genetic diseases whereas autosomal recessive PKD is relatively rare.

Normal kidneys are roughly fist-sized and weigh about a third of a pound. Over the course of several decades, PKD patients develop renal cysts that cause the kidneys to increase in size and weight- often to many pounds. The sequelae of renal enlargement include severe pain, increasing abdominal girth, hypertension, hematuria, kidney stones, and kidney infection. About 40-50% of PKD patients develop end stage kidney disease (ESRD), which then requires either renal transplantation or dialysis. Since there are no medications that can slow or stop the progression of PKD, the goal of treatment is to manage symptoms.

Developing a disease progression model for ADPKD

One of the factors that has hindered the development of a treatment for ADPKD is the lack of understanding of the disease progression, which is highly variable between patients. The commonly used endpoints for renal function only show changes very late in the course of the disease. Thus, it is difficult to assess the effectiveness of investigational new drugs using these endpoints. Validating a biomarker that could predict disease progression at an earlier stage when patients are more likely to respond to medication is a key step in enabling the development of treatments.

In 2011, we won a competitive bid to work with the Critical Path Institute’s Polycystic Kidney Disease Outcomes Consortium to develop disease progression models of ADPKD. The Consortium is a collaboration between C-Path, the PKD Foundation, Clinical Data Interchange Standards Consortium (CDISC), four academic medical centers, and three pharmaceutical companies. We investigated the use of Total Kidney Volume (TKV)—measured using magnetic resonance imaging (MRI), computed tomography (CT) or ultrasound (US)— as a prognostic biomarker for worsening of kidney function and response to therapy in ADPKD patients.

Qualification of this novel biomarker involved a number of steps. First, we needed to harmonize data from a number of disparate sources including several longitudinal patient registries that had been maintained for decades and observational clinical trial data. To do this, C-Path worked with CDISC to create a data standard for ADPKD. Next, the data standard was used to remap the data from the various sources into a single database. We then worked with the C-Path data management team to curate the database to make it modeling ready. Finally, we discussed the data analysis and modeling strategies with the FDA pharmacometrics and biostats teams as well as PKD clinicians to find the best method to elucidate the role of the interrelated variables: age, glomerular filtration rate (eGFR) and the trajectory of TKV.

Model building and validation

Briefly, we needed to link a key clinical outcome such as ESRD to the trajectory of TKV. We chose two additional endpoints— 30% and 57% worsening of renal function— since they occur in a shorter timeframe and predict the longer term outcome, ESRD. To tease out potential confounding effects between the various predictors, we used Multivariate Cox models. Then, we leveraged a simultaneous joint modeling framework of TKV and the time-to-event outcome to account for the longitudinal time varying covariate – TKV. Our team employed a parametric survival submodel to simulate the probability of avoiding the clinical outcome according to a given baseline eGFR, baseline TKV, and baseline age. Now that the model was complete, we validated it using a standard 5‑fold cross validation strategy.

Our model will support ADPKD drug development in several ways. For example, it was deemed appropriate for simulating clinical trials to determine economically feasible trial durations. Likewise, the results of our analysis showed that longitudinal TKV, eGFR and age were significant predictors of the likelihood of developing the clinical outcome. We are working on several peer-reviewed scientific papers that will detail the results of this work.

Overcoming project challenges

This complex project came with numerous challenges. For example, the database was continually being updated. Thus, data analysis was an iterative process where we generated outputs as the dataset matured. Another challenge was finding a technology that facilitated working with busy clinicians. To ensure that we got their critical input, we provided them with results that were viewable on mobile devices. This meant that they could easily check project results while seeing patients in the clinic. The infrastructure that comes with working at Certara— a technology enabled company— was indispensable in enabling us to complete this project in record time.

Gaining regulatory support for a novel biomarker for ADPKD

This spring, all of our hard work finally paid off when we received a Biomarker Letter of Support from the FDA. This biomarker will help support ADPKD drug development programs in their study design considerations, such as enrichment for patient subgroups that are more likely to experience progression of their renal dysfunction over the course of the trial. Optimizing patient selection for clinical trials is not only an ethical imperative, it will also increase the likelihood of a drug development program’s success. In addition, sponsors will be able to use this biomarker to support regulatory decision making for a given Investigational New Drug (IND) development program.

For more information, please read the Letter of Support from the FDA’s Center for Drug Evaluation and Research (CDER).

All information presented derive from public source materials.

Learn more about how Certara’s rare disease solutions can help you!

I hope that you’ll read this case study on how evidence from a previous indication was used to build models that supported the approval of an orphan drug.

Samer Mouksassi

About the Author

Samer Mouksassi

More Posts by This Author

Dr. Mouksassi started his clinical pharmacology career serving as a clinical pharmacist in a research hospital where he led the development of a Therapeutic Drug Monitoring service using Bayesian Feedback. His research focused on developing population pharmacokinetics/pharmacodynamics models to help clinicians to select the best dose for pediatric patients. Since joining Certara in 2007, he’s helped sponsors apply model-based drug development for more than a hundred studies. These contributions generated more than 60 abstracts/manuscripts in the field of pharmacometrics spanning numerous therapeutic areas. Dr. Mouksassi earned his PhD in pharmaceutical sciences from the University of Montreal, Faculty of Pharmacy and is a fellow of the American College of Clinical Pharmacology. He earned his PharmD from the Lebanese University in 2003.