Prognostic Features of Anti-Cancer Drugs Response in Resected/Unresected Primary Non-Small Cell Lung Cancer: A Retrospective Cohort Study

Technical Summary: Prognostic Features of Anti-Cancer Drugs in Resected/Unresected Primary Non-Small Cell Lung Cancer

Problem Statement
Low response rates to chemotherapy represent a critical challenge in oncology, contributing significantly to early mortality in cancer patients. A major difficulty lies in identifying "chemotherapy non-responders" early in the treatment course. While tumor size changes (RECIST criteria) are standard for assessing response, they may not correlate directly with overall survival (OS), and baseline variables like cancer stage or tumor grade often fail to predict longitudinal chemotherapy response accurately. Furthermore, treatment discontinuation due to toxicity or rapid disease progression in non-responders leads to poor outcomes, yet these patients are poorly characterized in existing medical literature. There is a need to identify prognostic baseline variables, specifically pre-chemotherapy white blood cell (WBC) counts, and to distinguish effective drug classes and formulations for different patient subgroups.

Methodology
This study employed a multi-center retrospective cohort design analyzing electronic medical records (EMR) from three branches of Severance Hospital in Seoul, South Korea.

• Cohort: The study included 2,459 deceased patients with primary non-small cell lung cancer (NSCLC) diagnosed between January 1, 2010, and January 1, 2020. Patients with small-cell lung cancer, mesothelioma, secondary lung cancer, or those lost to follow-up were excluded.
• Stratification: To identify chemotherapy response patterns, the population was stratified not by tumor response or dosage, but by chemotherapy prescription frequency:
  • Group A: 1–15 prescriptions (n=1,289).
  • Group B: 16–30 prescriptions (n=648).
  • Group C: ≥31 prescriptions (n=522).
    This stratification was chosen to reflect a patient's inherent ability to withstand chemotherapy over time rather than just the total dose received.
• Data Collection: Variables included demographics, ECOG performance status, cancer stage, surgical history, radiotherapy, comorbidities, and laboratory values (RBC, WBC, platelets, albumin). Specifically, WBC counts from 7 days prior to chemotherapy initiation were analyzed for 2,129 patients.
• Statistical Analysis: Multivariate regression was used to identify risk factors for 1-year mortality. Kaplan-Meier analysis and log-rank tests were applied to estimate Overall Survival (OS) and Progression-Free Survival (PFS). Comparisons were made between resected and unresected groups, and across different WBC count ranges (normal vs. leukocytosis vs. leukopenia).

Key Results

• Stratification and Survival: Group A patients (low prescription frequency) were identified as chemotherapy non-responders. Despite having significantly fewer new metastases post-diagnosis (23.4%) compared to Group B (19.3%) and Group C (19.2%), Group A had the poorest survival outcomes.
  • Post-treatment OS: Group A: 9 months; Group B: 19 months; Group C: 36.6 months (p < 0.0001).
  • Metastasis during Chemotherapy: Group A had significantly lower rates of metastasis during treatment (20%) compared to Group B (33%) and Group C (43%), yet they died earlier, suggesting intolerance or resistance rather than lack of disease progression.
  • Early Mortality: Group A had a 9.9% mortality rate within 1 month of treatment initiation, compared to 0.3% in Group B and 0% in Group C.
• Prognostic Significance of WBC:
  • Leukocytosis: Pre-chemotherapy leukocytosis (WBC ≥ 10.8 x 10³) was a strong negative prognostic factor. In unresected Group A patients, those with leukocytosis survived only 4.7 months compared to 9.6 months for those with normal WBC counts (p < 0.0001).
  • Resected Patients: In resected Group A, leukocytosis (≥ 10.8 x 10³) resulted in a survival of 5.3 months versus 19.5 months for normal WBC counts (p = 0.004).
• Risk Factors: Males and patients with Stage III/IV disease were at higher risk of 1-year mortality. Surgery and radiotherapy were associated with lower mortality risks in Groups A and B.
• Drug Formulations:
  • Oral RTKI: Oral Receptor Tyrosine Kinase Inhibitors (RTKI) were found to be effective in non-responders (Group A), with users surviving 13.5 months compared to 7.5 months for injection/systemic users (p = 0.001).
  • Immunotherapy: Biologics (e.g., trastuzumab, bevacizumab) were not found to be beneficial compared to RTKIs in this cohort.
  • Standard Chemotherapy: Group A patients were non-responsive to first-line doublet regimens (e.g., platinum-based combinations), with a 79.2% 1-year mortality rate for doublet users in this group.

Significance and Claims
The paper claims to be the first to demonstrate that stratifying patients by chemotherapy prescription frequency is a viable method for identifying chemotherapy response in retrospective analyses. The study posits that Group A represents a population of "chemotherapy non-responders" who are intolerant to standard regimens and prone to early mortality due to toxicity or resistance, rather than rapid tumor progression.

Key clinical implications drawn by the authors include:

1. Pre-treatment Monitoring: Pre-chemotherapy leukocytosis is a major risk factor for early mortality and should trigger close monitoring and potentially different dosing strategies, particularly in unresected patients.
2. Formulation Selection: Oral RTKIs may offer a survival benefit for non-responders who cannot tolerate standard injectable chemotherapy.
3. Dosing Safety: The authors suggest that standard manufacturer-recommended doses may be unsafe for non-responders and propose that a "Minimum Safe Dose" (MSD) or Low-Dose Metronomic (LDM) chemotherapy should be considered for these intolerant or resistant populations.
4. Future Directions: The study suggests that pre-treating patients with leukocytosis with anti-inflammatory drugs (e.g., dexamethasone) prior to chemotherapy initiation could reduce early mortality, though this is presented as a suggestion for future prospective study rather than a confirmed finding.

The study concludes that while cumulative 5-year survival remains low (10.3%), identifying non-responders early through prescription patterns and WBC counts can guide safer treatment selection and improve clinical management for NSCLC patients.