Authors: Amy Sullivan, MS ⊕, Truveta, Inc, Bellevue, WA, Sunny Guin, PhD ⊕, Truveta, Inc, Bellevue, WA, Jared Kern ⊕, Truveta, Inc, Bellevue, WA, Joud Roufael, MPH ⊕, Truveta, Inc, Bellevue, WA, Sarah Eng, MPH ⊕, Truveta, Inc, Bellevue, WA, Puja Rao, MPH ⊕, Truveta, Inc, Bellevue, WA, Nadia Tabatabaeepour, MPH ⊕, Truveta, Inc, Bellevue, WA, Amy Wu ⊕, Truveta, Inc, Bellevue, WA, Vidya Venkataraman, PhD ⊕, Truveta, Inc, Bellevue, WA, Esther Kim, PhD ⊕Truveta, Inc, Bellevue, WA,
- Nearly all patients (97.8%) had documented electronic health record (EHR) activity in the year before their first observed metastatic non-small cell lung cancer (mNSCLC) diagnosis.
- Healthcare activity was common immediately before diagnosis: 88.9% of patients had documented activity within the preceding 30 days.
- Median overall survival following the first observed mNSCLC diagnosis was 8.4 months, highlighting the importance of understanding the healthcare journey leading up to recognition of metastatic disease.
This report summarizes our poster presented at ISPE 2026, titled Characterizing pre-diagnosis healthcare activity and survival in metastatic non-small cell lung cancer.
Advances in targeted therapies have increased attention to metastatic non-small cell lung cancer (mNSCLC), but patients can follow very different paths before metastatic disease is recognized (1). Real-world healthcare activity during this period remains poorly characterized, particularly outside academic centers and specialty oncology clinics (2, 3). Understanding what happens before an mNSCLC diagnosis may provide important context about patients’ diagnostic journeys and opportunities to better understand care leading up to recognition of metastatic disease.
Using longitudinal EHR data from multiple US health systems, we examined healthcare activity during the year before the first observed mNSCLC diagnosis and described overall survival following diagnosis.
Methods
Using a subset of Truveta Data, we identified 11,956 patients with at least one mNSCLC diagnosis based on SNOMED CT codes between January 1, 2018 and July 13, 2026. The index date was defined as the first observed mNSCLC diagnosis.
We examined healthcare activity during the 365 days before the index date, including prior cancer history, biomarker testing, and EHR clinical activity. EHR activity included non-annulled encounters, condition diagnoses, laboratory test results, procedures, medication requests, and clinical observations. Events were aggregated using a rolling two-day window to account for clustered documentation. Overall survival was estimated descriptively from the index date through the most recent available follow-up, with patients followed until death or censored at their last observed EHR activity.
Results
Nearly all patients had healthcare activity before mNSCLC diagnosis
The median age at diagnosis was 69 years (IQR 63–76), and 49.0% of patients were female. Nearly all patients—11,692 of 11,956 (97.8%)—had at least one documented EHR activity during the year before their first observed mNSCLC diagnosis.
As shown in Figure 1 of the poster, healthcare activity remained common throughout the year leading up to diagnosis. Overall, 86.0% of patients had activity between 91 and 365 days before diagnosis, 85.5% between 31 and 90 days before diagnosis, and 88.9% within the final 30 days. Patients had a median of 14 activity events during days 91–365, five during days 31–90, and three during the 30 days immediately preceding diagnosis.
Most patients had a prior history of cancer
During the year before the first observed mNSCLC diagnosis, 82.8% of patients had a documented history of cancer. Nearly three-quarters (74.4%) had evidence of primary lung cancer, 29.9% had a non-lung primary cancer, and 61.0% had evidence of metastatic cancer at any body site. Patients could contribute to multiple categories.
Biomarker testing was identified in structured EHR data for only 7.6% of patients during the year before diagnosis; however, this only represents information that is readily found in structured EHR data. The poster notes that future studies could incorporate information from clinical notes to better characterize biomarker and staging information that may not be captured in structured data.
Overall survival was limited following mNSCLC diagnosis
Median overall survival following the first observed mNSCLC diagnosis was 8.4 months (95% CI 7.9–8.9). This limited survival underscores the importance of understanding patients’ healthcare trajectories before metastatic disease is recognized.
Discussion
In this large, multi-health-system cohort, nearly all patients with mNSCLC had documented healthcare activity during the year before their first observed metastatic diagnosis, and almost 9 in 10 had activity within the preceding 30 days. These findings demonstrate that the period before recognition of metastatic disease contains substantial real-world clinical information that may help researchers better understand how patients move through the healthcare system before diagnosis.
Our findings also highlight the value of longitudinal EHR data for studying cancer journeys outside academic centers or specialty oncology settings. Previous work has noted limitations in the representativeness and completeness of some real-world oncology data sources (3). By capturing encounters, laboratory results, diagnoses, procedures, medication requests, and clinical observations across participating health systems, longitudinal EHR data can provide a broader view of the healthcare activity preceding metastatic disease recognition.
This analysis was descriptive and does not establish whether earlier diagnosis was possible based on the documented healthcare activity. We also did not include information from clinical notes, which may contain important details about symptoms, biomarker testing, staging, and clinicians’ diagnostic reasoning. Future research incorporating clinical notes may help characterize these pre-diagnosis trajectories in greater detail and identify patterns that could inform studies of earlier detection.
Together, these findings show that patients with mNSCLC have substantial and heterogeneous healthcare activity before metastatic disease is recognized. Understanding that activity in the context of limited survival after diagnosis may provide a foundation for future research into real-world diagnostic pathways and opportunities to better contextualize care before metastatic diagnosis.
Data are constantly changing and updating. These findings are consistent with data analyzed for the ISPE 2026 study.
Citations
- M. Reck, K. F. Rabe, Precision Diagnosis and Treatment for Advanced Non-Small-Cell Lung Cancer. N Engl J Med 377, 849–861 (2017).
- L. A. Coombs, et al., Exploring Clinical Pathways to a Metastatic Lung Cancer Diagnosis: A Latent Class Analysis Using North Carolina Multipayer Claims Data. Cancer Epidemiol Biomarkers Prev 35, 1045–1052 (2026).
- L. T. Penberthy, et al., An overview of real-world data sources for oncology and considerations for research. CA Cancer J Clin 72, 287–300 (2022).
