Pretreatment Levels of Circulating Tumor DNA May Predict Long-Term Survival in NSCLC

Get Permission

Pretreatment circulating tumor DNA (ctDNA) levels in patients with non–small cell lung cancer (NSCLC) appear to be a more accurate predictor of long-term survival than classic survival surrogates, according to research presented during the International Association for the Study of Lung Cancer (IASLC) 2021 World Conference on Lung Cancer.1

An exploratory analysis of the single-arm, phase II NADIM study showed that ctDNA levels before treatment initiation identified patients at high risk of disease progression and outperformed radiologic response in the prediction of survival. Conversely, neither tumor mutation burden (TMB) nor PD-L1 expression was found to be predictive of long-term survival, the study authors reported.

Atocha Romero, PharmD, PhD

Atocha Romero, PharmD, PhD

“The identification of patients with NSCLC who may obtain long-term benefit from chemoimmunotherapy is essential to optimize therapies,” said Atocha Romero, PharmD, PhD, of the Hospital Universitario Puerta de Hierro, Majadahonda, Spain. “Pretreatment ctDNA levels could thus be useful for the design of new clinical trials.”

Although neoadjuvant chemotherapy is recommended for patients with resectable NSCLC, 5-year overall survival remains poor. The results of the NADIM trial, in which patients with resectable, stage IIIA NSCLC were treated with neoadjuvant nivolumab plus chemotherapy, have shown encouraging results, said Dr. Romero. However, there are currently no available biomarkers to identify patients who may exhibit long-term benefit from chemoimmunotherapy.2

Using samples from the NADIM trial, Dr. Romero and colleagues evaluated the capacity of ctDNA levels before treatment initiation to predict overall surival and progression-free survival. The researchers compared its predictive value with classic survival surrogates, including pathologic response and clinical response, assessed according to Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1.

The ctDNA was analyzed by next-generation sequencing, using the Oncomine Pan-Cancer Cell-Free Assay. For each positive plasma sample, the sum of the mutant allele frequency for all detected mutations was calculated.

Pretreatment ctDNA Levels vs Radiologic Criteria and Pathologic Response

The median follow-up time on the study was 38 months, Dr. Romero reported. Of the 46 patients included in the trial, 43 pretreatment plasma samples were available for ctDNA analysis, whereas 29 and 28 samples were available for TMB assessment and PD-L1 staining, respectively. Analysis of the data showed no association between either PD-L1 expression or TMB and overall survival, according to Dr. Romero.

In addition, neither pathologic response nor clinical response based on RECIST criteria was predictive of overall or progression-free survival. When patients who died of COVID-19 (n = 2) were excluded, however, pathologic complete response (but not major pathologic response) was associated with improved progression-free and overall survival.


  • In patients with NSCLC, pretreatment levels of ctDNA more accurately predicted long-term survival than radiologic assessments.
  • Study investigators suggest that pretreatment levels of ctDNA may be useful for the design of new clinical trials.

Conversely, multivariate analysis showed that patients with low ctDNA levels (< 1% mutant allele frequency) in the baseline sample had significantly better progression-free survival (hazard ratio [HR] = 0.22) and overall survival (HR = 0.04) than patients for whom the opposite situation occurred.

The adjusted C-statistic (c) to predict progression-free survival for ctDNA was 0.68, which was superior to that of RECIST criteria (c = 0.61) and similar to that of pathologic response, Dr. Romero reported. Similarly, baseline ctDNA levels appeared to predict overall survival (c = 0.85) better than RECIST criteria (c = 0.68).

“Although the numbers are small, our data show that ctDNA clearance clearly identifies patients who respond to treatment even more accurately than radiologic response according to RECIST criteria,” said Dr. Romero. “I think ctDNA is informative and should be considered as a trial endpoint; it predicts long-term survival and may be useful to tailor subsequent therapies.”

DISCLOSURE: The NADIM trial was funded by Bristol Myers Squibb. Dr. Romero has received consulting fees from Boehringer Ingelheim.


1. Provencio M, Nada E, Blasco-Serna R, et al: Pre-treatment levels of ctDNA for long-term survival prediction in stage IIIA NSCLC treated with neoadjuvant chemo-immunotherapy. 2021 World Conference on Lung Cancer. Abstract OA20.02. Presented September 13, 2021.

2. Provencio M, Nadal E, Insa A, et al: Neoadjuvant chemotherapy and nivolu-mab in resectable non-small-cell lung cancer (NADIM): An open-label, multicentre, single-arm, phase 2 trial. Lancet Oncol 21:1413-1422, 2020.

Related Articles

Expert Point of View: Bruna Pellini, MD

Study discussant Bruna Pellini, MD, of the Department of Thoracic Oncology, Moffitt Cancer Center, Florida, noted that, regardless of PD-L1 or tumor mutation burden status, the NADIM trial demonstrated improved progression-free and overall survival for patients with resectable, stage IIIA non–small ...