Uncommon/rare oncogenic drivers in non‐small cell lung cancer: Consensus and contention

Abstract

The importance of uncommon/rare oncogenic drivers in non‐small cell lung cancer (NSCLC) was underscored during the 20th China Lung Cancer Summit. These drivers, while present in a significant proportion of NSCLC patients, remain a challenge for diagnosis and therapeutic targeting. In the never‐smokers/low smokers category with mutations such as EGFR and HER2, the efficacy of immune checkpoint inhibitors (ICIs) remains suboptimal, attributed to lower PD‐L1 expression and tumor mutation burden (TMB). However, heavy smokers, often with mutations like KRAS, may derive benefits from ICIs, as supported by trials like CheckMate‐057. With the complex landscape of these drivers and their clinical implications, the summit culminated in six pivotal consensus points, aiming to guide future research and clinical decisions. Despite the advancements, the detection, interpretation, and therapeutic strategies involving these drivers necessitate further exploration and standardization.

Abbreviations

• AMP
• Association for molecular pathology
• ASCO
• American society of clinical oncology
• CAP
• College of American pathologists
• EGFR
• Epidermal growth factor receptor
• FDA
• Food and drug administration
• FFPE
• Formalin fixed paraffin embedded
• ICIs
• Immune checkpoint inhibitors
• MTB
• Molecular tumor board
• NCCN
• National comprehensive cancer network
• NGS
• Next-generation sequencing
• NSCLC
• Non-small cell lung cancer
• ORR
• Objective response rate
• OS
• Overall survival
• PCT
• Patient-centric trials
• PD-L1
• Programmed death-ligand 1
• PFS
• Progression-free survival
• TMB
• Tumor mutation burden
• TP53
• Tumor protein P53
• WCLC
• World conference on lung cancer

The 20th China Lung Cancer Summit, hosted by the Chinese Association of Lung Cancer and the Guangdong Association of Clinical Trials/Chinese Thoracic Oncology Group (GACT/CTONG), was successfully held on March 4, 2023, in Guangzhou. The summit focused on uncommon/rare oncogenic drivers in non-small cell lung cancer (NSCLC) and engaged in academic discussions from the perspectives of precise detection of uncommon/rare oncogenic drivers, novel clinical trial designs, and perioperative treatment in lung cancer patients with uncommon/rare oncogenic drivers. Ultimately, the expert consensus was made. Approximately 40% of non-small cell lung cancer (NSCLC) patients carry common epidermal growth factor receptor (EGFR) gene sensitizing mutations [], while another 20%–30% of patients carry uncommon/rare oncogenic drivers that can be specifically targeted for treatment []. Uncommon/rare oncogenic drivers have significant clinical value in the diagnosis and treatment of NSCLC, but many clinical questions still remain unanswered. For instance, how to accurately detect uncommon/rare oncogenic drivers? Next-generation sequencing (NGS) technology serves as an important tool for detecting uncommon/rare oncogenic drivers, but how should the test report of NGS be interpreted? What oncogenic drivers can be used as tissue-agnostic targets? What is the benefit of first-line immune therapy for rare oncogenic drivers? How to effectively select drugs for uncommon/rare oncogenic drivers? How to conduct the perioperative treatment trials for uncommon/rare oncogenic drivers? To address these issues, based on relevant domestic and international guidelines, expert consensus, and recently released results of clinical trials, the summit conducted in-depth discussions and opinions exchange, ultimately reached the following six points of expert consensus.

The level of evidence and strength of recommendations for this consensus are shown in Table .

Clinical question: How to accurately detect uncommon/rare oncogenic drivers?

Consensus 1: For the detection of uncommon/rare oncogenic drivers, it is recommended to use a large gene panel using DNA-based next-generation sequencing (NGS) as the preferred method. RNA sequencing can be used as a supplementary approach. The sequencing gene panel should include a minimum of 50 oncogenic drivers.

TABLE 1 Level of evidence and strength of recommendations.

Level or strength	Evidence	Recommendations
Strong	Evidence from large-scale phase III randomized controlled studies or meta-analysis on IPD	The expert panel achieves a substantial agreement based on the strong evidence (90%)
Moderate	Evidence from phase II clinical trials	The majority of the expert panel achieves is in agreement based on the moderate evidence (>60%)
Weak	No randomized controlled trials	The expert panel doesn't achieve an agreement based on the weak evidence

Level of evidence: Moderate

Strength of recommendation: Moderate

In NSCLC, oncogenic drivers with a detection rate of less than 5% are typically defined as uncommon oncogenic drivers [], while there is no clear definition of the criteria for “rare oncogenic drivers”. Uncommon/rare oncogenic drivers include classic gene fusions of ALK, ROS1, RET, NTRK, and NRG1, EGFR exon 20 insertions, HER2 mutations and amplification, MET primary amplification and exon 14 skipping mutations, BRAF V600E as well as KRAS G12C mutations. Despite the low frequency for each uncommon/rare oncogenic drivers, the overall proportion is not insignificant. Moreover, given the large population of NSCLC patients in China, it is of great significance to establish standardized methodologies for the detection of uncommon/rare oncogenic drivers. Currently, several domestic and international expert consensuses have provided recommendations for the detection of certain uncommon/rare oncogenic drivers []. It should be noted that while these consensuses respectively focus on methodological recommendations for the detection of a specific gene, there are commonalities among them. For example, based on tissue samples, consensuses recommend RT-PCR, FISH, DNA/RNA-NGS, and IHC for detection of gene fusion; FISH and NGS are recommended for the detection of gene amplification; RT-PCR or DNA/RNA-NGS is recommended for MET exon 14 skipping mutations; NGS is recommended for the detection of other point mutations as well as small insertion and deletion mutations. Notably, if the detection of all the aforementioned types of mutations is required, NGS is the preferred method. Tumor tissue samples are always preferred for testing; however, if tissue samples are unavailable or unqualified, liquid biopsy-based NGS can also be considered as an alternative.

Although there is limited high-quality evidence directly comparing the detection performance of different methodologies for uncommon/rare oncogenic drivers several publications with large sample size have demonstrated the feasibility of using NGS to detect uncommon/rare oncogenic drivers. Solomon et al. [] compared the detection performance on MET amplification between algorithm optimized NGS and traditional FISH in more than 50,000 tissue samples across 23 cancer types. The data showed that when using FISH as the reference, DNA-based NGS achieved a sensitivity of 97%, a specificity of 89%, and an overall concordance of 91%. In another study involving 33,997 patients with solid tumors, different methodologies for NTRK fusion detection were compared. When RNA-NGS results were used as the reference, DNA-NGS exhibited a sensitivity of 81.1%, a specificity of 99.9%, and an overall concordance of 99.8%. The sensitivity of IHC was 87.9%, specificity was 81.1%, and overall concordance was 82.2% []. Davies et al. [] conducted MET exon 14 skipping mutation detection using RNA-NGS and DNA-NGS on 856 NSCLC samples. The results showed that the positive detection rate of RNA-NGS (17/404, 4.2%) was significantly higher than that of DNA-NGS (11/856, 1.3%). In 286 samples that were tested by both methods, RNA-NGS detected 10 samples with MET exon 14 skipping mutations, while only 4 were detected by DNA-NGS, suggesting the possibility of false negatives with DNA-NGS methodology. However, on the other hand, two samples with positive results by DNA-NGS, failed to have RNA-NGS results due to insufficient RNA quality, indicating the higher sample quality requirements of RNA-NGS. Another study conducted head-to-head comparison of DNA-NGS and RNA-NGS in clinical formalin fixed paraffin embedded (FFPE) samples of 1253 NSCLC patients and found that 6% (10/168) of patients with negative results of DNA-NGS oncogenic drivers test were rediscovered by RNA-NGS detection (including 7 samples of fusion mutations, 3 samples of MET exon 14 skipping mutations), and these patients were offered targeted therapy [].

Considering the numerous types and mutually exclusive nature of rare oncogenic drivers in NSCLC [], NGS offers several advantages compared to traditional molecular pathology approach of sequential testing for individual mutations. These advantages include parallel testing of multiple genes, improved sample utilization efficiency, shorter turnaround time from testing to clinical decision-making, exploration of unknown mutations, lower average testing cost per gene, and comprehensive investigation of drug resistance mechanisms []. The National Comprehensive Cancer Network (NCCN) NSCLC guidelines (version 1.2023) has given an important update that NGS is the most commonly used comprehensive genomic profiling method that allows efficient utilization of limited biopsy tissue samples while maximizing diagnostic genomic information. Using an NGS panel composed of a specific number of genes, typically covering at least 50 genes, is a feasible strategy. If no oncogenic drivers are detected by a DNA-based NGS panel, RNA-based NGS should be considered []. In addition to considering the number of genes, it is also important to consider the tiers of somatic variants on the covered genes. The standards and guidelines for the interpretation and reporting of sequence variants in cancer, which is developed by the Association for Molecular Pathology (AMP), American Society of Clinical Oncology (ASCO), and College of American Pathologists (CAP) in 2017, proposed a four-tiered system (tier I, II, III, IV) to categorize somatic sequence variations based on their clinical significances. Tier I represents variants with strong and significant clinical implications on treatment, prognosis and diagnosis []. Therefore, when designing an NGS panel, all tier I variants of relevant genes should be included.

Clinical question: How should the test report of NGS be interpreted?

Consensus 2: The lung cancer MDT (multidisciplinary team) should include a Molecular Tumor Board (MTB), consisting of molecular pathologists, clinical molecular biologists, geneticists, and bioinformatics experts.

Level of evidence: strong

Strength of recommendation: strong

In the mid-1990s, with the development of imaging and radiology the multidisciplinary team (MDT) model was established and started to flourish due to its patient-centered core as well as the consequent improved clinical treatment decision. Thus, the concept model has gradually been accepted and promoted by clinicians []. A MDT consists of a small team of experts from different clinical departments to collaborate and make comprehensive treatment plans for patients []. The lung cancer MDT team should include Pulmonologist (Respir Onco), oncologists, thoracic surgeons, pathologists, palliative care specialists, and other doctors from various clinical departments [].

With the development of molecular medicine and precision medicine, there is still a significant gap between clinicians and NGS testing, which contains various gene mutation information and can help make clinical treatment decisions (clinicians are faced with the enormous problem of how to interpret the various gene mutations found in NGS tests, including their function and impact on clinical treatment decisions.). In recent years, many countries have started to establish multidisciplinary Molecular Tumor Boards (MTBs) to address this challenge by providing an interpretation of such reports. Except clinicians, the MTB team also includes molecular pathologists, clinical molecular biologists, geneticists, pharmacists, bioinformaticians, basic translational scientists, etc. The group of people pool their expertise from multiple disciplines to make the most accurate clinical treatment decision for patients. Among all these, molecular pathologists and clinical molecular biologists play a particularly crucial role as they provide advanced and evidence-based interpretation of molecular test results to guide clinical practice [].

In a study involving 168 patients, the targeted therapy population who received treatment recommendations from an MTB showed a significant improvement in median progression-free survival (PFS), increasing from 49 to 86 days []. Kato et al. evaluated treatments in 429 patients, of whom 265 (62%) were matched with at least one drug. Patients who received treatment recommendations from the MTB had significantly improved PFS (HR = 0.68, p = 0.008) and overall survival (OS) (HR = 0.69, p = 0.036) compared to those who solely followed clinical doctor recommendations. Additionally, the treatment plans they received were more aligned with their actual needs []. Larson KL et al. conducted a clinical benefit assessment of the MTB model by analyzing 14 studies. Among 3328 cancer patients without standard treatment options after multiple lines of therapy, the response rate varied from 0% to 67% in those who received treatment recommendations from the MTB. In the only trial that included a control group and was based on clinical outcomes, patients who received MTB-recommended treatment also had a significantly improved progression-free survival compared with those receiving conventional treatment []. Due to variations in study populations, disease types, treatment criteria, and recommended therapies, no definitive conclusions can be made from these studies. What we can make sure is that MTBs can provide clinical benefits. The authors of all these studies highly recommend the establishment of MTBs in healthcare institutions to assist in clinical decision-making and enhance the understanding of disease at the genetic level for both patients and clinicians. This suggests the emergence of a “new type” of physician in the future, namely clinical molecular biologists, who will focus on the diagnostic aspects of clinical molecular biology and provide practical advice and guidance to clinicians in this field.

Currently, the main focus of international MTBs is to provide clinical treatment decision guidance at the molecular tumor level and the group varies in numbers of participants ranging from 6 to 40 individuals []. Different MTBs have different tasks and use different tools which can help translate increasingly complex genetic information into patient-centered clinical decisions, thereby shepherding precision medicine into daily practice. Even though standards and guidelines are currently absent []. Therefore, it is crucial to continue exploring various possibilities through clinical practice and nurturing talents in clinical molecular biology to further improve the survival time and quality of life of patients.

Clinical question: What oncogenic drivers can be used as tissue-agnostic targets?

Consensus 3: Focus on driver gene research in cancer types other than lung cancer, exploring the potential of pan-tumor therapies, with particular attention to fusion and rearrangement events involving uncommon/rare oncogenic drivers.

Level of evidence: weak

Strength of recommendation: weak

The use of NGS technology in tumor research has generated a vast amount of publicly available genomic data, enabling the possibility of identifying comprehensive lists of driver genes across multiple types of cancer. Martínez-Jiménez et al. [] used the OncoGenomics (IntOGen) platform for tumor driver gene analysis to examine somatic variations in over 28,000 tumors across 66 different cancer types. They identified 568 tumor driver genes, a significant increase compared with previous findings. The majority of these driver genes (360) showed high specificity and were only trigger one or two tumor types. However, there is a small subset of canonical driver genes (TP53, PIK3CA, KMT2C, ARID1A, KMT2D, LRP1B, PTEN, RB1, FAT4, KRAS) that play crucial roles. Each of these ten genes can contribute to the tumorigenesis of more than 20 different types of tumors, with TP53 being particularly noteworthy as it is associated with over 50 different types of cancer. In general, fusion and rearranged genes tend to have more pronounced pan-tumor targeting tendencies. For example, [], NTRK1-3 gene fusions can occur in mammary analogue secretory carcinoma (MASC), secretory breast carcinoma, gastrointestinal stromal tumor, thyroid cancer, NSCLC, colorectal cancer, and glioblastoma. Drugs like entrectinib and larotrectinib have been approved for the treatment of solid tumor patients with NTRK fusion genes. RET gene rearrangements can occur in thyroid cancer, NSCLC, and colorectal cancer. The RET inhibitor selpercatinib has received FDA approval for the treatment of adult patients with locally advanced or metastatic solid tumors with RET gene fusion that have progressed on or following prior systemic treatment. ALK gene rearrangements can occur in NSCLC, anaplastic large-cell lymphoma, inflammatory myofibroblastic tumor, neuroblastoma, renal cell carcinoma, and esophageal squamous cell carcinoma. ROS1 gene rearrangements can occur in NSCLC, gastric cancer, colorectal cancer, Cholangiocarcinoma, angiosarcoma, and glioblastoma. Currently, there are no approved pan-tumor indications for targeted therapies against ALK and ROS1 gene rearrangements.

Are all deleterious mutations pathogenic? NGS data have shown that deleterious genomic variants can be seen in non-malignant disease, which poses challenges in assessing the genomic variations targeted in cancer treatments. Another key issue is the presence of certain genomic variations are present in and impact the growth of specific histologies. The effectiveness of pan-tumor therapies for this type of genomic variation is still being explored []. The prospects of applying these relatively specific genes in lung cancer require further investigation. Therefore, more work needs to be done regarding questions such as extrapolating the use of drugs targeting rare oncogenic drivers. Hence, it is recommended to focus on driver gene research in cancer types other than lung cancer, exploring the potential of pan-tumor therapies, with particular attention to fusion and rearrangement events involving uncommon/rare oncogenic drivers.

Clinical question: What is the benefit of first-line immune therapy for rare oncogenic drivers?

Consensus 4: Except for NSCLC with BRAF mutations or co-occurring KRAS and TP53 mutations, other driver gene-positive NSCLC cases are difficult to benefit from first-line immunotherapy. NSCLC patients with BRAF mutations or co-occurring KRAS and TP53 mutations experience the greatest benefits from immune checkpoint inhibitors. On the other hand, EGFR mutations or ALK or ROS1 rearrangements are typically associated with lower PD-L1 levels and tumor mutation burden (TMB).

Level of evidence: moderate

Strength of recommendation: moderate

For NSCLC patients with rare oncogenic drivers, the majority exhibit objective response rates (ORR) of over 50%, while a small portion have ORR below 40% []. Currently, there is limited prospective research data on the combination treatment of immune checkpoint inhibitors (ICIs) for NSCLC with rare oncogenic drivers. Most clinical data available comes from subgroup analyses of NSCLC patients treated with ICIs.

Tobacco exposure is possibly associated with higher PD-L1 expression and TMB []. NSCLC patients can be divided into two groups based on the level of tobacco exposure: never-smokers/low smokers and heavy smokers. Among never-smokers/low smokers with EGFR mutations/HER2 mutations/ALK fusions/ROS1 fusions, the majority have ORR of less than 15% when treated with ICIs, resulting in limited or no progression-free survival (PFS) benefits. EGFR mutations and HER2 mutations are typically associated with low PD-L1 expression (TPS <1%, approximately 50%) and low TMB (median <3 mutations/megabase [mut/Mb]). Although ALK fusions and ROS1 fusions exhibit higher PD-L1 expression rates (1% ≤ TPS ≤ 50%), they belong to the lowest TMB subtype (median <3 mut/Mb). Patients with RET rearrangements, similar to those with ALK fusions, also belong to the low TMB expression group. Different studies report varying PD-L1 expression rates, which may contribute to the limited benefit of ICIs treatment in patients with driver gene-positive mutations [].

Heavy smokers often have accompanying KRAS mutations, KRAS co-occurring with STK11 or KEAP1 mutations, BRAF mutations, or MET exon 14 skipping mutations. Literature reports suggest that patients with these mutations can benefit from ICIs. In the phase III clinical trial CheckMate 057, comparing second-line anti-PD-1 monotherapy to chemotherapy, the subgroup with KRAS mutations achieved significant overall survival benefits from anti-PD-1 monotherapy (HR 0.52, 95% CI 0.29–0.95) []. Studies indicate that patients with KRAS co-occurring with TP53 mutations can significantly benefit from single-agent ICIs []. Changes in the immune microenvironment induced by STK11 and/or KEAP1 alterations, along with the resistance to ICIs, depend on the KRAS mutation status and are associated with poorer treatment outcomes []. Therefore, when selecting first-line ICI monotherapy, it is necessary to comprehensively assess the status of KRAS, STK11, and KEAP1 mutations. Limited efficacy data are available for patients with BRAF and MET exon 14 mutations, especially for MET exon 14 mutations. Patients with BRAF mutations often exhibit high PD-L1 expression rates and high TMB []. Retrospective studies suggest moderate clinical benefits of single-agent ICIs in non-selective PD-L1 patients with BRAF-mutated NSCLC []. Emerging evidence suggests that different oncogenic drivers have different effects on tumor immune microenvironment, which may lead to the different clinical benefits of ICIs. Patients with BRAF mutations or co-occurring KRAS and TP53 mutations derive the greatest benefit from ICIs, while EGFR mutations or ALK fusions or ROS1 rearrangements are typically associated with lower tumor PD-L1 levels and TMB, limited immune cell infiltration in the tumor microenvironment, and resistance to ICIs [].

Clinical question: How to effectively select drugs for uncommon/rare oncogenic drivers?

Consensus 5: Encourage innovative clinical trial methods to accommodate clinical research on rare oncogenic drivers. Patient-centered clinical trial methods are recommended.

Level of evidence: moderate

Strength of recommendation: moderate

Tissue-agnostic studies present numerous challenges in their design, implementation, and analysis processes []. Both the US Food and Drug Administration (FDA) and China's National Medical Products Administration Center for Medical Device Evaluation (CDE) have issued relevant guidance principles to address the challenges faced in drug development, market approval, and clinical trial design. These guidelines include “Guidance on Enhancing the Diversity of Clinical Trial Populations - Eligibility Criteria, Enrollment Practices, and Trial Design Considerations for Use in Industry” [], “Technical Guidelines on the Applicability of Single-Arm Clinical Trials to Support Antineoplastic Drug Marketing Applications” [], “Guidelines on Multiplicity Issues in Clinical Trials (Draft for Comments)” [], and “Guidelines on Statistical Design of Clinical Trials for Antineoplastic Drugs (Draft for Comments)” []. These guidance principles focus on enhancing innovative clinical trial design, expanding the diversity of clinical trial participants, broadening inclusion criteria, avoiding unnecessary exclusion of patients from clinical trials, and applying them to clinical trials for the treatment of rare diseases or targeted therapies to accelerate clinical drug development. Based on these guidance principles, the TAPISTRY (Tumor-Agnostic Precision Immuno-Oncology and Somatic Targeting Rational for You Platform Study, NCT04589845) study was initiated. It is a phase II global, multicenter, open-label, multi-cohort platform study aimed at evaluating the safety and efficacy of targeted therapy or immunotherapy in patients with unresectable, locally advanced, or metastatic solid tumors. All enrolled patients were determined to have specific oncogenic genomic alterations or high tumor mutational burden (TMB) using validated next-generation sequencing (NGS) analysis []. The study began in January 2021 and is expected to be completed by September 2023. Another representative large-scale innovative study, The National Lung Matrix Trial (NLMT), sponsored by Cancer Research UK and supported by the Stratified Medicine Programme Phase 2 (SMP2) screening platform, employs an adaptive approach to ensure that patients can immediately change treatment options or exit the study once a particular treatment regimen proves ineffective []. Preliminary results of this study have been published in Nature, indicating objective response rates exceeding 60% across multiple biomarker-driven cohorts using 19 drug-biomarker modules []. The study is still ongoing.

In recent years, the concept of patient-centric trials (PCT) has been proposed as another important idea to improve the efficiency of drug development and market approval. PCT refers to a design approach that prioritizes patient needs throughout all stages of clinical trials. This requires collaborative efforts from various stakeholders, forming international alliances to further accelerate the advancement and implementation of precision oncology treatments []. CTONG1702/1705 is an open-label, multicenter, phase II adaptive umbrella trial and real-world observational study []. Patients who meet the strict inclusion criteria of CTONG1702 based on large next-generation sequencing (NGS) panel and PD-L1 expression results are enrolled in parallel sub-studies with treatment regimens matched to their profiles. Patients who do not meet the CTONG1702 inclusion criteria are included in the compassionate use group, receiving treatment according to standard protocols. Patients who are unwilling to participate in the clinical trial and receive treatment according to routine standards are entered into the observational real-world study, CTONG1705. This patient-centric clinical research design, by establishing a compassionate use group, increases the efficacy and safety data of investigational drug treatments in a broader population of patients. At the same time, it also observes the clinical outcomes of patients in the real-world group receiving treatment according to standard clinical protocols. Such designs enable comprehensive collection of clinical data for patients with rare oncogenic drivers and explore the most rational treatment approaches, thus creating efficient clinical research design strategies. These novel trial designs must also meet the requirements of ethical review in medical research, imposing higher demands for the clinical justification and prospective design of such studies.

Clinical question: How to conduct the perioperative treatment trials for uncommon/rare oncogenic drivers?

Consensus 6: Platform trials are recommended as the basis for advancing research on perioperative treatment studies of uncommon/rare oncogenic drivers.

Level of evidence: moderate

Strength of recommendation: strong

Currently, there is an unmet need for perioperative treatment of NSCLC with uncommon/rare oncogenic drivers []. Results from several retrospective clinical studies and case reports [] support the investigation of neo-adjuvant/adjuvant targeted therapies during the perioperative period for NSCLC with uncommon/rare oncogenic drivers.

Unlike studies in advanced NSCLC, conducting clinical research on uncommon/rare oncogenic drivers in early-stage non-small cell lung cancer (NSCLC) face lots of challenges such as difficulties in patient enrollment and lengthy evidence generation. ALINA (NCT03456076) is a phase III, multicenter, randomized controlled, open-label study comparing alectinib versus platinum-based chemotherapy as adjuvant treatment for ALK-positive NSCLC in IB (≥4cm)-IIIA (AJCC 7th) postoperative setting []. Over a span of more than 3 years (from August 2018 to December 2021), a total of 257 patients were enrolled across 141 global study centers, with fewer than two patients were enrolled in each center. Another ongoing phase III, multicenter, randomized controlled, double-blind study is LIBRETTO-432 (NCT04819100), which compares selpercatinib versus placebo as adjuvant treatment for RET-positive staged IB-IIIA NSCLC []; this study plans to enroll 170 patients in 211 global centers over a period of 6 years (December 2021 to December 2027). The involvement of a large number of study centers and the extended enrollment duration highlights the challenges faced in conducting traditional prospective randomized controlled trials for uncommon/rare oncogenic drivers.

To address these issues, FDA and CDE have issued several relevant guidelines and guidance [] to accelerate the recruitment of underrepresented populations in clinical trials. In order to efficiently select eligible patients for enrollment, umbrella studies have been commonly conducted in recent years for perioperative research on NSCLC with uncommon/rare oncogenic drivers, such as NAUTIKA1 study (NCT04302025), PURPOSE study (ChiCTR2100053021), and so on. NAUTIKA1 is an ongoing Phase II umbrella study [] to investigate the safety and efficacy of targeted agents as neo-adjuvant and adjuvant therapy (adjuvant setting was followed the standard chemotherapy after surgery) in early staged resectable NSCLC patients with molecular alterations including ALK, ROS1, NTRK, BRAF V600, and RET. The preliminary data of 8 patients in the ALK fusion cohort was reported at the 2022 World Conference on Lung Cancer (WCLC) (data cutoff on May 9, 2022). All of the 8 patients completed at least 8 weeks of alectinib neo-adjuvant treatment with a 100% R0 resection rate without surgery delay or significant complications. The most common drug-related adverse event during treatment was nausea. These preliminary findings demonstrate the safety and feasibility of alectinib perioperative-treatment in early-stage ALK fusion-positive NSCLC. The PURPOSE study is another ongoing open-label phase II prospective umbrella study [] that aims to investigate the neo-adjuvant treatment based on driver gene and immune biomarkers in resectable stage II-IIIB NSCLC. It includes multiple arms targeting ALK, MET, RET, HER2, and other alterations. The study is currently recruiting patients and will bring more evidence for perioperative treatment of NSCLC with uncommon/rare oncogenic drivers.

This expert consensus focuses on the unsolved clinical questions of NSCLC with uncommon/rare oncogenic drivers, combined with the clinical experience and opinions of experts, the following six points of consensus are made. Due to the rapid development of this field, this consensus represents the most fundamental requirements in the current state of the field, and we welcome experts to provide criticism and suggestions for improvement.

• Consensus 1: For the detection of uncommon/rare oncogenic drivers, it is recommended to use a large gene panel using DNA-based next-generation sequencing (NGS) as the preferred method. RNA sequencing can be used as a supplementary approach. The sequencing gene panel should include a minimum of 50 oncogenic drivers.

• Consensus 2: The lung cancer MDT (multidisciplinary team) should include a Molecular Tumor Board (MTB), consisting of molecular pathologists, clinical molecular biologists, geneticists, and bioinformatics experts.

• Consensus 3: Focus on driver gene research in cancer types other than lung cancer, exploring the potential of pan-tumor therapies, with particular attention to fusion and rearrangement events involving uncommon/rare oncogenic drivers.

• Consensus 4: Except for NSCLC with BRAF mutations or co-occurring KRAS and TP53 mutations, other driver gene-positive NSCLC cases are difficult to benefit from first-line immunotherapy. NSCLC patients with BRAF mutations or co-occurring KRAS and TP53 mutations experience the greatest benefits from immune checkpoint inhibitors. On the other hand, EGFR mutations or ALK or ROS1 rearrangements are typically associated with lower PD-L1 levels and tumor mutation burden (TMB).

• Consensus 5: Encourage innovative clinical trial methods to accommodate clinical research on rare oncogenic drivers. Patient-centered clinical trial methods are recommended.

• Consensus 6: Platform trials are recommended as the basis for advancing research on perioperative treatment studies of uncommon/rare oncogenic drivers.

AUTHOR CONTRIBUTIONS

Conception and design: Yi-Long Wu. Manuscript writing: All authors. Final approval of manuscript: All authors. Accountable for all aspects of the work: All authors.

ACKNOWLEDGMENTS

Guangdong Provincial Key Lab of Translational Medicine in Lung Cancer (2017B030314120, Yi-Long Wu), Guangdong Provincial Peoples Hospital Scientific Research Funds for Leading Medical Talents in Guangdong Province (KJ012019426, Yi-Long Wu), the National Natural Science Foundation of China (Grant No.82072562, Qing Zhou).

CONFLICT OF INTEREST STATEMENT

Yi-Long Wu reports grants and personal fees from AstraZeneca and Boehringer Ingelheim, grants from Bristol Myers Squibb, and personal fees from Beigen, Eli Lilly, MSD, Hengrui, Pfizer, Roche, and Sanofi outside the submitted work. Wenzhao Zhong reports other support from AstraZeneca, Bristol Myers Squibb, MSD, Roche, and Innovent outside the submitted work. Qing Zhou reports lecture and presentation fees from Astra- Zeneca, Boehringer Ingelheim, BMS, Eli Lilly, MSD, Pfizer, Roche, and Sanofi outside the submitted work.

DATA AVAILABILITY STATEMENT

Data sharing is not applicable to this article as no new data were created or analyzed in this study.

ETHICS STATEMENT

Not applicable.

CONSENT TO PARTICIPATE

Not applicable.

References

Fang W, Huang Y, Hong S, Zhang Z, Wang M, Gan J, et al. EGFR exon 20 insertion mutations and response to osimertinib in non‐small‐cell lung cancer. BMC Cancer. 2019;19(1):595. [DOI: https://dx.doi.org/10.1186/s12885-019-5820-0]