Association of metabolomics with PD-1 inhibitor plus chemotherapy outcomes in patients with advanced non-small-cell lung cancer

Liang Zheng, Fang Hu, Lin Huang, Jun Lü, Xiaohua Yang, Jianlin Xu, Shuyuan Wang, Yinchen Shen, Runbo Zhong, Tianqing Chu, Wei Zhang, Ying Li, Xiaoxuan Zheng, Baohui Han, Hua Zhong, Wei Nie, Xueyan ZhangJournal for ImmunoTherapy of Cancer 2024 · DOI

TL;DR

Researchers have identified a serum metabolite, N-(3-Indolylacetyl)-L-alanine, that serves as a potent independent predictor for how well advanced lung cancer patients will respond to immunotherapy-chemotherapy combinations. This blood-based biomarker offers a more convenient and dynamic alternative to traditional tissue biopsies, potentially allowing clinicians to better tailor treatments for patients lacking specific genetic mutations.

What Problem Does This Paper Address?

For patients with advanced non-small-cell lung cancer (NSCLC) who don't have "driver" mutations (like EGFR or ALK), the standard of care is a combination of immune checkpoint inhibitors (ICIs) and chemotherapy. However, only about 20% of patients see a sustained positive response.

Currently, the primary tool for predicting success is testing for PD-L1 expression in tumor tissue. While useful, PD-L1 is far from a perfect predictor; it requires invasive tissue biopsies, and its levels can vary wildly across different parts of a tumor or over time. There is an urgent need for "liquid biopsy" biomarkers—markers found in the blood—that are easier to collect and more reflective of the body's current metabolic and immune state.

Methods

The study employed untargeted metabolomics to scan the serum of patients for small-molecule signatures associated with treatment success.

  1. Sample Collection: Serum was collected from 250 patients before they started treatment.
  2. Analytical Techniques: The team used Liquid Chromatography-Tandem Mass Spectrometry (LC-MS) to identify thousands of metabolites.
  3. Statistical Filtering: They used Principal Component Analysis (PCA) and Orthogonal Partial Least Squares Discriminant Analysis (OPLS-DA) to separate "responders" (Progression-Free Survival $\ge$ 12 months) from "non-responders" ($<$ 12 months).
  4. Validation: To ensure the results weren't just statistical noise, they split the combination therapy group into a Discovery set (n=50) and a Validation set (n=150), while also including a chemotherapy-only control group (n=50) to confirm the markers were specific to the immunotherapy combination.

Datasets

The study utilized a primary cohort of 250 patients from Shanghai Chest Hospital (treated between 2019 and 2021).

  • Combination Group: 200 patients receiving PD-1 inhibitors (pembrolizumab, tislelizumab, or sintilimab) plus chemotherapy.
  • Control Group: 50 patients receiving only cytotoxic chemotherapy.
  • Clinical Data: All patients had stage IIIB-IV NSCLC and lacked sensitizing driver mutations.

Key Insights and Ideas

  • The "Aha" Metabolite: Out of thousands of candidates, N-(3-Indolylacetyl)-L-alanine emerged as the most significant predictor. High levels of this metabolite were consistently found in patients with poorer prognoses.
  • Independent Power: In multivariate analyses—which account for age, sex, smoking, and tumor stage—this metabolite remained a significant independent predictor of progression-free survival (PFS).
  • Synergy with PD-L1: The study found that the metabolite works best when used alongside PD-L1 testing. Patients with both low metabolite levels and high PD-L1 expression had the longest survival times, suggesting these two markers capture different, complementary aspects of the tumor-immune interaction.
  • Metabolic Mirroring: The authors suggest that because N-(3-Indolylacetyl)-L-alanine is an amino acid derivative, its levels may reflect the underlying "metabolic exhaustion" or immune suppression within the tumor microenvironment.

Key Results

  • Survival Gap: Patients with low levels of N-(3-Indolylacetyl)-L-alanine had a median PFS of 18.0 months, compared to just 11.0 months for those with high levels ($p=0.003$).
  • Hazard Ratio: The multivariate analysis showed a Hazard Ratio (HR) of 0.60, meaning patients with low levels of the metabolite had a 40% lower risk of disease progression or death.
  • Response Rates: The Objective Response Rate (ORR) was significantly higher in the low-metabolite group (43.0%) compared to the high-metabolite group (24.0%).
  • Specificity: Another metabolite, methomyl, also showed predictive value in univariate analysis, but it lost statistical significance in the more rigorous multivariate model, highlighting N-(3-Indolylacetyl)-L-alanine as the more robust biomarker.

Limitations and Future Work

The authors acknowledge several hurdles before this can reach the clinic:

  1. Retrospective Nature: The study looked back at existing data; a prospective, multi-center trial is needed to confirm the findings.
  2. Survival Data: The study focused on Progression-Free Survival (PFS) because the Overall Survival (OS) data was not yet "mature" (not enough time had passed to see final outcomes).
  3. Untargeted vs. Targeted: This was an untargeted study meant for discovery. Future work must develop a standardized, absolute quantitative kit for this specific metabolite to make it practical for hospital labs.
  4. Biological Mechanism: While the correlation is strong, the exact biological reason why this specific metabolite influences immunotherapy response remains a subject for future basic experimental research.

Summary updated

Abstract

The combination of ICIs and chemotherapy has established itself as the new standard of care for first-line or second-line treatment in patients with advanced NSCLC lacking oncogenic driver alterations. Therefore, identifying biomarkers that can predict the efficacy and prognosis of immunotherapy plus chemotherapy is of paramount importance. Currently, the only validated predictive biomarker is programmed cell death ligand-1 (PD-L1), but its predictive value is not absolute. Our study suggests that the detection of N-(3-Indolylacetyl)-L-alanine in patient serum with untargeted metabolomics prior to combined therapy may predict the efficacy of treatment. Compared with detecting PD-L1 expression, the advantage of our biomarker is that it is more convenient, more dynamic, and seems to work synergistically with PD-L1 expression.
BibTeX
@article{zheng2024association,
  author = {Liang Zheng and Fang Hu and Lin Huang and Jun Lü and Xiaohua Yang and Jianlin Xu and Shuyuan Wang and Yinchen Shen and Runbo Zhong and Tianqing Chu and Wei Zhang and Ying Li and Xiaoxuan Zheng and Baohui Han and Hua Zhong and Wei Nie and Xueyan Zhang},
  title = {{Association of metabolomics with PD-1 inhibitor plus chemotherapy outcomes in patients with advanced non-small-cell lung cancer}},
  journal = {Journal for ImmunoTherapy of Cancer},
  year = {2024},
  doi = {10.1136/jitc-2023-008190},
}
Figure