Real-World Treatment Patterns Among Patients With EGFR-Positive Metastatic Non–Small Cell Lung Cancer

This poster was originally presented at ISPOR on May 19, 2026, in Philadelphia, PA.

Authors: Rohan Vashi, Dave Iwanyckyj, Fernando Otalora, Melanie Jardim

Affiliations: Amplity Inc, Langhorne, PA, USA

Introduction

Non–small cell lung cancer (NSCLC) is the most common lung cancer subtype, and accounts for approximately 80-85% of all lung cancers. Approximately 60% of patients are diagnosed at an advanced stage of disease, which remains particularly challenging to treat and is associated with poor outcomes. Treatment of metastatic NSCLC (mNSCLC) utilizes targeted therapy or chemotherapy and is often guided by molecular profiling.

Mutations in the epidermal growth factor receptor (EGFR) tyrosine kinase domain are a major oncogenic driver of NSCLC, promoting tumor growth and survival.  Optimal treatment sequencing remains unclear in EGFR-mutated mNSCLC, presenting challenges in clinical decision-making.

This research characterizes real-world treatment patterns among community and academic physicians and identifies reasons for treatment transitions, particularly among an EGFR-positive patient subgroup.

Methods

• This retrospective observational study utilized Amplity AnswerY™, our proprietary real-world database and platform built from HIPAA-compliant transcriptions of US prescriber–patient visits.
  • Using AI and NLP, it extracts, visualizes, and summarizes treatment discussions and clinical decisions and covers inpatient and outpatient care across more than 70 specialties since 2017.
  • Patients who had mentioned diagnosis of mNSCLC and EGFR-positive disease from 2020-2025 were included and followed from 1L through 3L treatments.
  • Unstructured clinical narratives were processed using a hybrid NLP approach (rule-based and machine learning), supplemented by a large language model configured to ensure data privacy with no storage or reuse of inputs for training. Extracted variables were standardized, and a subset underwent manual review for validation.
  • Trends in prescribing habits and switching reasons were documented between lines of therapy.

Results

• AnswerY identified 17,587 patients from 3812 providers with mNSCLC, of which 927 patients from 387 providers were EGFR-positive.

• All patients with mNSCLC, chemotherapy alone (42.2%) and a chemotherapy + immunotherapy combination (27.4%) were the most common 1L choice.
• The most common 1L-3L treatment sequence among patients reaching each line was chemotherapy alone, immunotherapy alone, then other therapies (42.2%/34.8%/40.1%). Treatment patterns as depicted in Figure 1 show numerous and highly diffused pathways that patients may follow when receiving treatment.

• In the same subgroup, the most common 1L-3L treatment sequence was an EGFR-targeted therapy excluding osimertinib, osimertinib, then other therapies (30.0%/29.1%/32.8%). Treatment patterns as depicted in Figure 2 reveal that the presence of an EGFR-mutation does not ease the complexity of treatment selection in mNSCLC.

• Overall, all patients with mNSCLC and those in the EGFR-positive subgroup demonstrated short times on 1L treatments. Furthermore, time on 1L chemotherapy and chemotherapy and immunotherapy is particularly lower than EGFR-targeted options, perhaps suggesting a treatment switch upon the return of biomarker results. (see Figure 3)
• Among all patients with mNSCLC, those on TKI monotherapies demonstrated the highest mean time on 1L treatment followed by osimertinib and other EGFR-targeted therapies.

• Among all patients with mNSCLC and EGFR-positive for the break, treatment switches from 1L to 2L included disease progression (41.2%/30.9%), AEs (7.3%/9.3%), and planned transition (9.3%/ 2.0%.) (See Figure 4)
• Most patients in both cohorts transitioning from 2L to 3L treatment was because of disease progression (23.1%/31.6%).

Conclusions

• AnswerY identified that among all patients with mNSCLC, chemotherapy is the most common 1L treatment option.

• Among the EGFR-positive subgroup a significant proportion of patients did not receive EGFR-targeted 1L treatment, despite being recommended by current treatment guidelines.

• Empiric therapy prior to biomarker testing is favored in clinical practice, even when outcomes may be magnified by waiting for biomarker testing data.

• AEs, although not the primary driver, account for a substantial proportion of treatment switches, highlighting the need for better tolerated therapies, combination strategies, and prophylactic measures.

• These data illustrate a need for increased provider education on guideline-directed care in patients with EGFR-positive mNSCLC, the impact of biomarker testing on outcomes, and continued investigations into understanding barriers to real-world treatment protocol adherence.

DOWNLOAD POSTER

AnswerY helps explain HCP behavior. See other NSCLC research.