Perspectives on early‐stage lung cancer identification and challenges to thoracic surgery

Lung cancer is one of the leading causes of cancer‐related death worldwide with the long‐term survival rate for stage I cancers treated with surgery reaching over 80%, but rapidly reducing as disease develops with stage III cancer patients presenting with a long‐term survival rate of just 20%. Although surgery remains the most effective strategy for treating early stage lung cancers, screening and early diagnosis remain critical for improving prognosis in lung cancer. In addition to advancements in surgical techniques, the rapid development of artificial intelligence (AI), liquid biopsy, and other technologies in recent years has led to new breakthroughs in the diagnosis of this disease. Clinical application of neoadjuvant therapies using immune checkpoint inhibitors have also shown promise for improved efficacy in the treatment of early stage lung cancers. This editorial was designed to summarize the recent progress around the screening and diagnosis of early stage lung cancers and the challenges associated with various surgical strategies for treating these patients.

1. ENHANCEMENTS IN AI AND LIQUID BIOPSY TECHNOLOGIES BRING NEW BREAKTHROUGHS AND CHALLENGES

The National Lung Cancer Screening Study (NLST) ¹ in the United States and the NELSON study ² in Europe provide compelling evidence of the efficacy of lung cancer screening using low‐dose computed tomography (LDCT) to reduce lung cancer mortality. However, the application of this technology also increases false‐positive rates and presents with some inherent economic costs. A study published in Science demonstrated that blood tests could offer exciting possibilities for detecting many cancer types, including lung cancers, at a relatively early stage using a minimally invasive technique. ³ However, given the clinical significance of lung cancer screening, there are some challenges associated with liquid biopsy applications, including the need for clearer screening standards, more accurate selection of high‐risk patients, and more individualized screening strategies that still need to be addressed, before wide scale adoption.

It is worth noting that the widespread use of chest computed tomography (CT) has increased the detection rate of pulmonary nodules considerably, helping identify lung cancers earlier, however, the differentiation of cancer from benign pulmonary nodules remains a challenge in clinical practice. International guidelines recommend the application of mathematical models to assist in these evaluations, with the most broadly adopted including the Mayo, VA, and Peking University mathematical models. These models have all been shown to facilitate the greatest diagnostic accuracy, making them the most useful in clinical settings. ⁴ , ⁵ , ⁶

The application of AI and radiomics in imaging diagnostics of lung cancer increases the sensitivity of lung cancer screening and simultaneously reduces the burden on radiologists. This has meant that several computer‐aided diagnosis systems (CADx), developed by several teams, also show significant potential in improving early cancer detection. ⁷ , ⁸ , ⁹ However, nearly all AI‐and radiomics‐based diagnostic systems make diagnoses using only imaging characteristics without the support of any clinical information, such as a history of smoking or tumors, which may limit their utility. At present, CADx systems are only used to assist in the detection of pulmonary nodules in order to reduce the workload of radiologists, but these systems will need to be validated in broader populations before their widespread clinical application.

The application of liquid biopsy, especially circulating tumor DNA (ctDNA) for the early diagnosis of various cancer, holds significant promise for non‐invasive cancer detection. However, its detection rate poses significant challenges, especially in the early stages of lung cancer, when the tumor is small, increasing the need for extremely sensitive detection. Chen et al. first confirmed the feasibility of applying ctDNA for the diagnosis of early stage lung cancer in China ¹⁰ using ultrasensitive cancer personalized profiling by deep sequencing (CAPP‐Seq) technology. The team from Stanford University also detected ctDNA in 42% of stage I lung cancer patients ¹¹ while the circulating cell‐free genome atlas study (CCGA) has demonstrated a sensitivity of more than 90% for stage III lung cancer, but only about 20% for stage I lung cancer using this technology. Taken together this suggests that ctDNA detection of stage II lung cancers is feasible, but more work needs to be done to satisfy the sensitivity required for reliable stage I detection. Multi‐omics analysis of circulating tumor markers is one of the future directions for liquid biopsy and has been shown to exhibit higher accuracy in the diagnosis of early‐stage lung cancers. ¹²

2. ONGOING GENOMIC DISCOVERY AROUND EARLY‐STAGE LUNG CANCER FACILITATES UPDATES TO CLINICAL TREATMENT STRATEGIES FOR THIS DISEASE

Tremendous progress in sequencing technologies and the broad array of complex bioinformatic methods have provided scientists with an opportunity to uncover the details surrounding the evolutionary and biological factors influencing cancer development and prognosis, allowing for the identification of new drug targets, improving clinical decision‐making, and facilitating improved individualized treatment. TRACERx, a large‐scale genomic study carried out by the Cancer Research Institute in the UK, revealed the evolutionary trajectories of different cancers under various selection pressures using dynamic and multi‐region sampling, combined with genomic, transcriptomic, epigenetic, and other omics data, and explored the impact of evolution on cancer prognosis. ¹³ , ¹⁴ , ¹⁵ , ¹⁶ Another study from Singapore systematically evaluated the genomic landscape of lung adenocarcinoma in the East Asian population and provided an important data source for the accurate diagnosis and treatment of East Asian lung cancer patients. ¹⁷

However, due to the conservative treatment strategies used in Western countries for very early stage lung cancers, the molecular characteristics of lung adenocarcinoma presenting as ground‐glass nodules (GGNs) remain poorly understood. Several studies from our team have unraveled the genomic and immune microenvironment characteristics of lung cancer presenting as GGNs, ¹⁸ , ¹⁹ and have revealed that GGN‐associated lung cancers exhibit a lower tumor mutation burden, less active immune environment, decreased expression of immune activation markers, and less infiltration of most immune‐cell subsets than their solid tumor counterparts. HLA loss of heterozygosity was significantly less common in lung adenocarcinomas with GGN components than in those without. These characteristics may indicate some of the mechanisms underlying their indolent clinical course, but further evaluation is required to confirm this hypothesis. In addition, a series of studies on the molecular characteristics and immune microenvironment of pre‐invasive and early stage lung cancers from China have provided valuable insights into the surgical and comprehensive management of such cancers. ²⁰ , ²¹ , ²²

Surgery remains the most effective therapeutic strategy for the treatment of early stage lung cancer and lobectomy and systematic lymph node dissection have been standard surgical treatments since the 1990s. Given the increasing detection rates for small peripheral lung nodules and the characteristics of GGNs, limited resections in the form of sublobar resection may be beneficial in patients presenting with these pathologies. Two RCT studies, JCOG0802 and CALGB140503, comparing lobectomy and sub lobectomy in the treatment of early stage lung cancer, have reported perioperative results after a long period of enrollment. These evaluations have found no differences in the surgical complications and mortality rates between sublobar resection and lobectomy, ²³ , ²⁴ but the results of the long‐term follow‐up of JCOG0802, released in 2021, provide more evidence‐based support for the application of sub lobectomy. However, none of these studies specifically focused on elderly patients, despite the fact that the elderly population experiences a high incidence of lung cancer and presents with unique clinical characteristics and more complex complications. The choice of surgical resection for elderly patients may be quite different from that of younger patients. This lack of data is partially addressed by the STEPS study initiated by Yang et al. which was the first randomized controlled trial focusing on elderly patients with lung cancer, ²⁵ but there remains a need for more in‐depth evaluations in the future. However, the data from these studies suggest that there may be a need to revise the current surgical treatment guidelines used for early‐stage lung cancer.

3. DESPITE PROMISING SHORT‐TERM RESULTS, LONG‐TERM OVERALL SURVIVAL RATES ARE STILL LACKING FOR NEOADJUVANT IMMUNOTHERAPY

A meta‐analysis from the NSCLC Meta‐analysis Collaborative Group has confirmed the role of neoadjuvant chemotherapy in the treatment of lung cancer. This and other studies have suggested that neoadjuvant treatment with immune checkpoint inhibitors (ICIs) may be a promising new approach for managing bulky but resectable lung cancers. Immunotherapies enhance T‐cell activation priming antigen recognition and increasing the breadth and durability of tumor‐specific T‐cell responses. Published phase II‐III clinical trials have also confirmed that this strategy has a high pathological response rate. ²⁶ , ²⁷ In addition, translational research along with these clinical trials has explored predictive biomarkers for neoadjuvant treatment with ICIs. However, these results remain controversial. Studies, including Checkmate 159, NADIM, and Checkmate‐816, concluded that dynamic changes in ctDNA could predict the treatment response of patients, while studies such as LCMC3 and NEOSTAR found significant differences in T cell clonal expansion between patients with major pathologic response (MPR) and non‐MPR. Nevertheless, the feasibility of this treatment strategy is determined by its influence on surgical outcomes and most published studies conclude that the R0 resection rate was > 87%, and the operation delay rate was < 20%. There was no significant increase in intraoperative blood loss or operation time, suggesting that this strategy is safe and effective.

The tremendous efficacy of neoadjuvant ICIs has attracted attention from all fields; however, many unanswered questions remain, which pose significant challenges for both oncologists and thoracic surgeons. First, what role does surgery play and will other tumor ablation therapies work as effectively? Second, are specific immunotherapies more effective than others in the neoadjuvant context? And third, what are the determinants of response to neoadjuvant immunotherapy? Understanding why neoadjuvant immunotherapy is effective for some patients, but not for others, may allow the identification of specific biomarkers of response and expand our understanding of the various mechanisms of resistance to these therapies.

4. TARGETED ADJUVANT AND IMMUNOTHERAPIES ARE EXPECTED TO IMPROVE TREATMENT, BUT MORE ACCURATE SELECTION IS NEEDED

The benefit of adjuvant chemotherapy in patients with stage II NSCLC has been confirmed but remains limited, ²⁸ with the wide spread use of targeted therapy in advanced lung cancers showing some efficacy, further studies on adjuvant targeted therapy for early stage lung cancers have also been conducted. Two clinical trials carried out in China have confirmed prolonged disease‐free survival (DFS) in patients with resected stage II‐III lung cancer carrying EGFR mutations. ²⁹ , ³⁰ However, the OS results of the ADJUVANT study were negative even with significant clinical benefits. The most promising ADAURA study demonstrated a significant improvement in DFS in the targeted therapy group. However, it is worth noting that ADAURA prolonged the use of TKI for three years and adopted third‐generation EGFR‐TKIs, which means that these improved OS results were expected.

The unmet needs here include how to select high‐risk patients for adjuvant treatment and which treatment scheme is the best choice for different patients. Two studies from European and North American patients have shown a close relationship between postoperative minimal residual lesions (MRD) and tumor recurrence. ³¹ , ³² This suggests that monitoring of MRD using highly sensitive ctDNA detection methods may help to accurately distinguish between high ‐and low‐risk groups of postoperative recurrence. Chen et al. sponsored the first study evaluating the application of ctDNA to monitor postoperative MRD in Asia (DYNAMIC). ³³ In addition, a prospective study using specific methylation sites to monitor MRD (MEDAL) is still enrolling patients, and is expected to be a more effective strategy for the postoperative surveillance of lung cancer. ³⁴ Interventional clinical trials based on MRD have been launched for the treatment of lung cancer and these studies will provide more choices for postoperative monitoring of lung cancer.

IMpower010, a randomized, multicenter, open‐label, phase 3 trial, demonstrated a DFS benefit for atezolizumab versus best supportive care following adjuvant chemotherapy in patients with resected stage II‐IIIA NSCLC with no safety concerns. Atezolizumab after adjuvant chemotherapy is a promising treatment option for patients with resected early stage NSCLC. Although only atezolizumab has been shown to be effective so far, there is hope for the future use of other ICIs in the adjuvant setting for lung cancer.

In conclusion, the diagnosis, treatment, and postoperative follow‐up of patients with early‐stage lung cancer requires further refinement, but the development of more accurate screening strategies and individualized treatment options are promising developments for the future of these patients. We will have to work hard to ensure these technologies continue to mature and overcome their current and future challenges.

AUTHOR CONTRIBUTION

Li X and Chen KZ: Conceptualization; Literature collection; Writing‐Original Draft. Yang F: Writing‐Review & Editing; Supervision. Wang J: Conceptualization; Resources; Supervision; Writing‐Review & Editing.

CONFLICTS OF INTEREST

The authors declare no conflicts of interest. Professor Xiao Li and Jun Wang are members of Chronic Diseases and Translational Medicine editorial board and is not involved in the peer review process of this article.

ETHICS STATEMENT

No ethical issues involved.

DATA AVAILABILITY STATEMENT

All data were from the cited references.