If you're designing a post-approval study or exploring real-world evidence for treatment efficacy, you've likely encountered a deceptively simple question: "Which line of therapy are we studying?" The answer, it turns out, is far more complicated than most researchers anticipate.
At ISPOR Europe this November, Flatiron Health had the opportunity to discuss our systematic approach to operationalize Lines of Therapy (LOT)—the sequence of anticancer treatments a patient receives over time—from real-world data. For biopharma researchers and sponsors tasked with generating evidence that answers questions about how treatments compare to current standard of care across different treatment sequences, understanding LOT is essential.
The challenge: Defining what’s routine
In clinical practice, Lines of Therapy are a routine part of the cancer treatment journey. A clinician may commonly use the terms"first-line," "second-line," or "third-line" therapy to describe a patient's treatment sequence and risk/benefit discussions of advancing to a new line of therapy between a patient and physician are common.
However, thinking about how to define, extract, and validate LOT in real-world data becomes exponentially more complex. Patient journeys aren't cleanly documented—treatment data lives across multiple clinical systems, with some information captured in structured fields like drug names, start dates, and end dates, while other critical context about anticancer therapies is buried in clinical notes. There are gaps in therapy, dosage changes, and maintenance regimens that follow different patterns depending on the disease context, all of which complicate the task of deriving meaningful treatment sequences.
Consider prostate cancer: A patient might receive surgery or radiotherapy to treat the non-metastatic cancer and then receive a treatment to reduce the risk of disease recurring or spreading—despite this approach a patient may still progress to metastatic disease and move through subsequent treatment lines in this new setting. On top of this, we also know that many available medical anti-cancer treatments are shifting to earlier in the cancer journey. But without being in the patient and physician's conversation, how do we define when one line ends and another begins? When is a treatment change appropriate within a line, and when does it signal a true progression to a new treatment approach? These aren't arbitrary distinctions—they fundamentally shape research findings and the conclusions researchers draw about treatment effectiveness.
Flatiron's approach: Systematic rules, local clinical knowledge
At Flatiron, we've developed an approach that applies several categories of disease-specific rules to standardize how LOT is derived from RWD. These rules define:
- Eligible therapies for inclusion in lines
- Maintenance therapy status
- Gaps in therapy
- Acceptable and forbidden treatment changes within a line
- Index dates that establish the starting and ending point for treatment sequencing
- Line settings that capture disease context (neoadjuvant, adjuvant, metastatic, etc.)
The approach’s strength lies in its capacity for global application while maintaining local clinical relevance. We recognize that clinical practice varies globally, with treatment guidelines differing by country, drug availability varying across regions, and healthcare infrastructure differences shaping how patients access care. What constitutes "current standard of care" in the UK may look fundamentally different in Germany or Japan, which is why our team of clinical experts embedded in each geography is critical—they ensure that while we maintain a consistent framework, we also account for local clinical nuances.
Quality assurance: The foundation of trustworthy data
The quality of the LOT output is directly dependent on the underlying source data, which is often dispersed across multiple systems and contains imperfections and gaps naturally present in real-world data sources. Structured data may contain incomplete or inaccurate records and unstructured clinical notes require careful interpretation. Our quality assurance process evaluates LOT sequences against local clinical expectations, flagging patterns that don't align with how oncologists actually practice medicine. These unexpected patterns are identified not only during our routine quality control of the whole patient cohort, but also during rigorous, patient-level review which ensures that all known information is accurately reflected in the derived LOT. This step is crucial—it transforms raw data into clinically validated and reliable evidence for research.
Additionally, generating rigorous and unbiased LOT is a critical challenge in RWD research. Flatiron’s approach is unique in its consistent, transparent methodology. Our focus on neutrality, consistency, and broad validation, support robust and high quality resulting evidence that is more aligned with the expectations of regulators and HTA bodies.
Why this matters for your research
For researchers designing post-approval studies or comparative effectiveness research, LOT precision directly impacts the validity of your findings. Misclassified lines of therapy can bias your results and at times undermine the credibility of your conclusions.
More broadly, LOT represents one critical dimension of the patient journey. When you combine LOT data with other clinical variables—biomarkers, comorbidities, performance status, treatment outcomes—you unlock the ability to answer complex questions about how treatments perform in specific patient populations across different treatment sequences as well as better understand an unmet need in specific lines or sub-cohorts.
If you're exploring real-world evidence strategies or designing research that requires rigorous LOT definition, Flatiron's team has the clinical expertise, data infrastructure, and proven track record to be your thought partner.
