Longitudinal data sets can impart valuable insight from physicians back to drug researchers.
In complex clinical settings like chemotherapy, knowing what’s truly happening in physician practices is difficult. Longitudinal data research—which is based on observations of the same patients over time—gives drug developers direct access to investigate real patient history instead of relying on physician assumptions, recollections, and opinions collected through traditional primary market research. This source of valuable insight allows research teams to standardize data interpretation across multiple parameters, customize market definitions, and assess real market opportunity for new and existing drugs.
Longitudinal data can reveal market behavior that may otherwise be overlooked, setting the framework for proper clinical trial design by assessing the real available patient pool. This contextual data can also be analyzed to better predict a drug’s most likely market launch analog including usage in off-label markets, physician compliance, and potential lost opportunity.
But most importantly, longitudinal data-based analysis allows market research teams to accurately peg the expected impact a drug might have in the real world. This market positioning naturally supports the drug discovery process enabling quicker, more accurate, and more cost-effective insight into the potential drug behavior for unimpeachable action plans.
Effective market positioning is one of the biggest challenges in the discovery process. Toxicity, lack of clinical efficacy, and lack of commercial value are some of the major impediments that can hamper the process.
Because it takes eight to nine years from the time of drug synthesis to approval, primary and secondary market research is crucial to design appropriate clinical trials to maximize a drug’s success. Correct positioning may even result in expedited review for drugs that are active in life-threatening, underserved diseases, bringing them to market significantly faster than other compounds.
Longitudinal data-mining is a research method that can help speed market positioning. This contextual research identifies potential market size and toxicity issues not present in highly-selected clinical trial patients, calculates total drug consumption and cost of competitive therapies for proper pricing of the new product, and provides insight into real-world therapy duration and time-to-progression.
While longitudinal data sets, for the most part, impart valuable insight that sometimes results in actionable information, one must be wary of longitudinal data supplied using disparate, mismatched systems, especially those designed specifically as software and not data-collection tools. Proper examination into the collection methodology (including breadth and completion of data sets), the ability to standardize and customize definitions across physician and patient groups, and consistency of quantitative analysis across multiple parameters support successful utilization of longitudinal data.
The key indicator of the success of any new drug is defined by its impact in a real-world population. However, it doesn’t show when and why the dose was dialed down.
Longitudinal data provides this insight, presenting an opportunity to investigate real patient history, instead of physician assumptions and opinions alone. An examination of administration-level data and patient record audits yields granular data, that helps researchers quantify lost opportunities for existing drugs and predict market behavior for new compounds.
Data-capture at the drug administration level is usually quicker, more accurate, and more cost-effective than undergoing a patient record audit. Laboratory test data complements this administration-level data, revealing the complete picture of the treatment history. In addition, administration-level longitudinal data based on a stable geographically representative sample may more accurately characterize drug behavior on the national level in a timely manner.
Based on the actual number of drug administrations per patient, longitudinal data gives researchers the ability to calculate true duration of therapy—with a shared definition of any particular line—resulting in more accurate costs of therapy per patient. For example, longitudinal data shows that the real-world number of Avastin (Genetech) administrations is about two times lower than the number of Avastin administrations reported in clinical trials (Figure 2), therefore, the true duration of therapy and cost of therapy per patient is much lower than the one projected using clinical trial data. (Based on IntrinsiQ data.)
Determining toxicities of the current agents based on longitudinal data highlights the most common side effects and their frequency in real-world population. Evaluating the affect these toxicities also have on dose delays, dose reductions, and compliance enables one to better understand the true drug burden on an individual patient-level or defined patient cohort.
In addition, being able to view off-label use signals how positioning a drug for entry into a smaller tumor market may have larger indications later on, whereas a riskier, ineffective trial in a larger tumor market could destroy the drug. Physician adoption of the drug in off-label markets might also provide near-term perceptions of a product’s usage as well as input into clinical publication and indication strategies.
Carefully-projected longitudinal data can be successfully used to estimate the size of the population eligible for clinical trials, potential market size, and number of patients with certain duration of response. In addition, longitudinal data can be used to define study criteria such as primary endpoints, power of the study, sufficient sample size, and expected endpoint results. Longitudinal research helps identify high-risk drug candidates (ones with inadequate side-effect profile and small target market) early in the drug discovery process, bringing more effective compounds to the market faster with a higher likelihood of clinical success.
About the Author
Ed Kissel has 20 years of experience in the healthcare industry, with 15 years in primary and secondary market research and strategic forecasting focusing on oncology markets. Previously, he led the oncology market research function at GlaxoSmithKline.
Filed Under: Drug Discovery