Approximately 20 to 25% of patients diagnosed with NSCLC are eligible for surgical resection (1), yet 30% to 55% of those who undergo curative surgery will encounter recurrence and ultimately die (2, 3). Despite the improvement in treatment of NSCLC patients, the 5-year survival rates remain relatively low, hovering around 36% (4). Over the past decade, neoadjuvant platinum-based chemotherapy, followed by surgery, has become the standard of care. Nevertheless, only 5 to 6 % improvements have been noted in 5-year recurrence-free survival and overall survival rates compared to surgery alone (5).
Currently, combination of chemotherapy with immunotherapy has demonstrated superior clinical efficacy and prolonged event-free survival for patients with resectable NSCLC. Trials such as CheckMate 816 (6), SAKK 16/14 (7), and NADIM (8), have showcased the enhanced efficacy and manageable safety of combination therapy compared to monotherapy. However, for most previous clinical trials, cytotoxic drugs with standard doses and anti-PD1 antibody were administrated on the same day (9), which may result in unsatisfactory eradication rates and relatively high incidence of severe treatment-related adverse events (TRAEs). To improve the cure rate and reduce TRAEs, the use of low-dose chemotherapy may represent a feasible alternative, given it favorable tolerability profile without compromising efficacy (10, 11). Although various regimens of low-dose chemotherapy have been utilized in cancer therapy (12, 13), the effects of dosage and administration sequence on the efficacy and safety of NSCLC treatment remain controversial. According to our previous research findings, the administration of immunotherapy 3-5 days after chemotherapy significantly reduces lymphocyte damage (14). Therefore, we hypothesize that by coordinating the dosage of chemotherapy and the timing of immunotherapy administration, we can enhance the synergistic effects between the two interventions. Ultimately, this could improve clinical efficacy while reducing toxic side effects.
To monitor the therapeutic efficacy during the process of chemo-immunotherapy, there is an urgent requirement for reliable indicators. The T cell receptor (TCR) repertoire and the presence of circulating tumor DNA (ctDNA) in blood have been proven to be efficient predictors of therapeutic responses in cancer treatment (15, 16). Chemotherapeutic agents have the capability to induce immunogenic death of tumor cells, reduce tumor burden, and promote the infiltration of CD8 T lymphocyte (17, 18). The recognition of tumor antigens by CD8 T lymphocytes relies primarily on the presence of a diverse and polymorphic TCR within an individual (19). Hence, it is imperative to explore the dynamic changes of TCR diversity and tumor burden in ctDNA throughout the course of therapy.
In this study, we report the results of our pilot clinical trial assessing the efficacy and safety of modified neoadjuvant chemo-immunotherapy in patients with resectable NSCLC. These results indicate that the combination of low-dose chemotherapy followed by delayed immunotherapy could potentially induce tumor cell damage, promote antigen exposure, and support T cell expansion, thereby enhancing therapeutic efficacy and reducing toxic side effects. Monitoring the dynamic changes of TCR repertoire and ctDNA does not only offer insights into the mechanism underlying our therapeutic approach to some extent but also furnishes valuable data for monitoring treatment efficacy.
Study design and participants
38 Patients with resectable stage IIA to IIIA NSCLC, as per the staging criteria outlined by the American Joint Committee on Cancer, 8th edition, and possessing adequate organ function, along with an Eastern Cooperative Oncology Group performance-status rating of 0 to 1, were included in this study. All patients had not received any prior cancer treatment, exhibited measurable disease according to the Response Evaluation Criteria in Solid Tumors (version 1.1) and had confirmed histological or cytological evidence of the disease. Exclusion criteria included the presence of identified EGFR mutations or ALK translocations; a diagnosis of malignancies beyond NSCLC within 5 years; previous exposure to specific therapies such as anti-PD-1, anti-PD-L1, or anti-PD-L2 agents or medications targeting T cell receptor modulation (e.g., CTLA-4, LAG-3, or TIM-3); a history of allogeneic organ or hematopoietic stem cell transplantation (except corneal transplantation); ongoing repercussions from prior interventions; pregnancy or breastfeeding; and HIV positivity. Moreover, the availability of pre- and post-adjuvant treatment tumor tissues and serial blood samples facilitated comprehensive analyses including whole transcriptome sequencing, TCR sequencing, and ctDNA sequencing.
Trial design and treatment
This study was a single-center, single-arm, pilot clinical trial (Registration Number: ChiCTR2000033092). Eligible patients aged from 45 to 72 received 2 cycles of neoadjuvant therapy of low-dose platinum-based chemotherapy: albumin-bound paclitaxel 180-220 mg/m2 plus cisplatin 60 mg/m2 for lung squamous cell carcinoma or pemetrexed disodium 500 mg/m2 plus cisplatin 60 mg/m2 for lung adenocarcinoma respectively on day 1 plus sintilimab 200 mg on day 5 of each cycle, followed by radiographic evaluation before undergoing radical surgery. Surgery was conducted within 6 weeks after the completion of neoadjuvant treatment. Then patients would receive 2 cycles of adjuvant therapy with the same regimen.
Study outcomes
The primary endpoints include objective response rate (ORR: the proportion of patients with complete response (CR) or partial response (PR) according to response evaluation criteria in solid tumors) and major pathologic response (MPR) (less than 10% residual viable tumor cells in the primary tumor and sampled lymph nodes) according to the pathologist review. Secondary endpoint included pathologic complete response (pCR) (0% residual viable tumor cells in the primary tumor and sampled lymph nodes), disease free survival (DFS), treatment-related adverse events (TRAEs), occurrence of surgical complications, and intraoperative and perioperative complications.
Plasma collection, cell-free DNA extraction, and sequencing
Blood samples were collected into cell-free DNA blood streck tubes. Isolated plasma and paired white blood cell (WBC) control were stored at -80[until extraction. Cell-free DNA (cfDNA) and WBC-derived DNA were extracted from 4 mL plasma and WBC using QIAamp Circulating Nucleic Acid Kit (Qiagen, Hilden, Germany) and Mag-Bind® Blood & Tissue DNA HDQ 96 kit (OMEGA) according to the manufacturer’s instructions, respectively. DNA extracts were sheared to 100 bp fragments using an S220 focused ultrasonicator (Covaris, Woodbrun, MA, USA). The adaptors (NadPrep®) were attached to the terminal of fragments and followed by PCR (Bio-Rad T100, USA) replication. The libraries were constructed using NadPrep Plasma Cell-Free DNA Dual-Ended Molecular Tag Library Construction Kit (for MGI) (Nanodigmbio). The target customized probes YuceOne Plus (ctDNA) synthesized by IDT were used for hybridization enrichment. Target enrichment was performed with NadPrep Hybrid Capture Reagents (Nanodigmbio). The enriched DNA fragments were sequenced with the MGISEQ platform. The mean depth of coverage was 3445× for cfDNA and 1419× for matched peripheral blood samples.
Circulating tumor DNA (ctDNA) analysis
SOAPnuke was implemented to cut the adapter and remove low-quality raw reads (20). Genecore was performed for cluster reads by unique molecule index (UMI) and remove the replicated reads derived from PCR (21). Clean reads were aligned against the human reference genome (hg19) with BWA (v0.7.12) (22) and duplicated reads were removed by samtools (v0.1.19) (23). Subsequently, generated BAM files were used for downstream analysis. We compared the ctDNA samples and matched control blood sequencing data to identify the somatic mutations, including single nucleotide variants (SNVs) and small insertions and deletions (Indels) using mutation caller Vardict (v1.7.0) with default parameters (24). Next, the caller was run with dbSNP (version 147), 1000G (phase3_release_v5), CLINVAR (version 151 ClinVar (nih.gov)), and COSMIC (version 81) data for known polymorphic sites. Substitutions and indels would be filtered out according to the following criteria: 1) low variant allelic fractions (VAF < 0.0005), 2) removed any variants with greater than 1 supporting read in the matched germline sample, 3) variants present in the above genome database. In addition, the filtered mutations were annotated by snpEff (v4.3) (25) with NCBIrefseq (https://www.ncbi.nlm.nih.gov/refseq/). The maximum VAF is considered as the tumor burden during the therapy.
Whole transcriptome sequencing
The total RNA of tumor samples was isolated using the RNeasy FFPE Kit (Qiagen, GER) according to the manufacturer’s instructions. The concentrations of RNA were quantified using Qubit™ RNA HS Assay Kit (ThermoFisher Scientific, USA). RNA purity and integrity were checked using the Take3 (BioTek, USA) and the RNA Cartridge kit of the Qseq100 Bio-Fragment Analyzer (Bioptic, CHN), respectively. Then, RNA sequencing (RNA-seq) libraries were constructed using the rRNA depletion module (H/M/R) and NadPrep DNA Library Preparation Module for MGI (MGI, CHN). The libraries were sequenced on the MGISEQ-T7 platform with 100 bp paired-end reads.
RNA-Seq raw data quality control, gene expression analysis, and immune cell abundance estimation
Raw sequencing data (raw reads) from MGISEQ-T7 was processed to filter out low-quality reads. Clean reads from each sample were aligned against the hg19 reference by STAR (v2.7.8) (26). The read counts and transcripts per million (TPM) values were calculated based on aligned RNA sequencing reads to reference transcripts downloaded from gencode (v27lift37) database, as implemented in rsem (v1.3.0) (27). Based on the gene expression matrix, R packages xCell (28) and web tool ImmunCellAI (29) were used to estimate the infiltration abundance of immune cells for each sample.
T cell receptor sequencing and analysis
For T cell receptor (TCR) sequencing, total RNA was isolated from peripheral blood mononuclear cells (PBMCs) by RNeasy Plus Mini Kit (Qiagen, USA) according to the manufacturer’s instructions. Qubit™ RNA HS Assay Kit (ThermoFisher Scientific, USA) was applied to determine the final concentration. Then TCR β-chain of CDR3 was amplified by Multi-IR Library Prep Kit (iGeneTech, CHN). Sequencing procedures were utilized by the MGISEQ-T7 platform with 100-bp paired-end reads. Fastq reads were trimmed based on their low-quality 3’ ends to base. Trimmed pair-end reads were integrated according to overlapping alignment with the modified Needleman-Wunsch algorithm. MiXCR (v2.1.10) (30) was performed to identify the CDR3 sequences of V-D-J gene segments with reference sequences from the IMGT (31). VDJtools (v1.2.1) (32) was utilized to assess the immune repertoire sequencing. A frequency-based correction was performed on samples. The diversity of the TCR repertoire was calculated with the Shannon-Wiener index formula 
Statistical analysis
All statistical analyses were implemented by R 4.0 software or SPSS 27.0. Medians and inter quartile range (IQRs) were provided for distributions of observation duration and time intervals. Wilcoxon’s Rank-Sum test was used to compare across independent groups. Wilcoxon’s Signed-Rank test was used for the pairwise comparison samples. All reported P values are two-sided, and P values less than 0.05 were considered statistically significant.
Overview of the patient cohort
From May 2020 to November 2023, 50 patients diagnosed with NSCLC at stage IIA-IIIA were screened for eligibility, of which 38 received two cycles of modified neoadjuvant chemo-immunotherapy (Fig. 1A). Each cycle featured two phases, corresponding to chemotherapy and immunotherapy (Fig. 1B). Following modified neoadjuvant chemo-immunotherapy, 31 patients underwent R0 resection (Figure 1A and Table S1). The postoperative tissues were subjected to pathological evaluation. Samples from a representative patient (P2) who achieved pCR after neoadjuvant treatment exhibited morphological alterations, characterized by extracellular mucin pools and fibrosis (Fig. 1C). Computed tomography scans demonstrated the obliteration of lung lesions following neoadjuvant therapy (Fig. 1D). 74.2% (27/31) participants received the modified adjuvant chemo-immunotherapy (from time-point T4 to T8) (Fig. 1A and B; Table S1). Detailed treatment regimens are provided in Table S2.
Most patients were men (78.7%) and had a smoking history (81.6%). The median age of patients was 64 years (range: 37-77). All patients had ECOG performance status of 0 or 1. Baseline patient characteristics are summarized in Table 1 and detailed in Table S1.
Clinical efficacy and safety
After neoadjuvant therapy, 39.5% (15/38) patients achieved radiological downstaging (Fig. 2A; Table 2). The ratio of patients with CR, PR, and stable disease (SD) was 5.3% (2/38), 68.4% (26/38), and 26.3% (10/38), respectively, yielding an ORR of 73.7% (28/38) (Fig. 2B). 91.3% (21 of 23) were recurrence-free (Fig. 2C). The median interval between the last dose of sintilimab and surgery was 33 days (interquartile ranges (IQRs) 5.55 days) (Table S1). 47.4% (18/38) patients achieved a MPR and 31.6% (12/38) patients achieved a pCR (Fig. 2D).
For patients receiving adjuvant therapy, the median interval between the surgery and the adjuvant therapy was 32.1 days (IQRs 7.2 days) (Table S1). Up to January 18, 2024, the median follow-up was 14.37 months (IQRs 24.22 months) (Table S1). The median OS and DFS are not reached (Fig. 2E). 44.7% (17/38) patients experienced at least one TRAE (Table 3). Grade 3 TRAEs occurred only in one patient. The most common grade 1 or 2 TRAEs were nausea (15.8%). Previously unreported toxic effects were not observed, and there were no instances of discontinuation of modified neoadjuvant chemo-immunotherapy due to toxicity.
Low-dose chemotherapy combined with delayed immunotherapy has a positive effect on T lymphocytes
To investigate whether dynamic changes in peripheral blood TCR repertoire were associated with treatment efficiency during modified chemo-immunotherapy, 49 peripheral blood samples were collected from serial time-points and evaluated using TCR sequencing (Fig. 3A). Since the predominant number of TCR clones is an important indicator of immune response and treatment efficacy (15, 33–35), the frequency of the top 20 TCR clones was analyzed at each treatment phase. Significant increase of top 20 TCR clones were observed in phase 1 to 3 (all P < 0.0001; Fig. 3B; Fig. S1). An overall increase of top 20 frequency was observed in the first cycle of neoadjuvant therapy (C1). In contrast, no significant changes were observed during adjuvant therapy (Fig. 3C; Table S3). These results suggested that the peripheral blood-derived top 20 TCR clones were continuously increased not only during low-dose chemotherapy, but also during immunotherapy within neoadjuvant therapy.
According to clonotype abundance, the TCR clones were divided into 3 clonotypes, including large (frequency > 0.001), median (0.001-0.0001), and small (< 0.0001) clones. A markedly increase of large clones was observed during neoadjuvant therapy (all P < 0.05) (Fig. 3D and E), consistent with the trend of top 20 TCR clones. No significant difference was observed in the diversity, evenness, and clonality of TCR clones across the whole treatment procedure (Fig. 3F; Fig. S2; Table S4). The changes of T lymphocytes in the tumor microenvironment across the neoadjuvant therapy were analyzed based on transcriptomic data. As shown in Figure S3, the CD4+ and CD8+ T lymphocytes tended to increase during neoadjuvant therapy. Collectively, the neoadjuvant treatment had a positive effect on T lymphocytes.
Reduced tumor burden in ctDNA was observed during neoadjuvant therapy
To monitor the dynamic changes of tumor burden in ctDNA during therapy, peripheral blood samples were collected at serial time-points (T0 to T7) and evaluated via ctDNA sequencing. The landscape of genomic alternations was shown in Figure 4A. Among the detected genes, TP53 (50%) was the most frequently mutated gene, followed by ARID1A (11%), DNMT3A (11%), and PIK3CA (11%) (Fig. 4A; Table S5). Throughout the entire treatment, the maximal somatic variant allelic frequency (maxVAF) of ctDNA had a declining trend, especially in PH1 (T0 to T1) with dramatically decreased (P < 0.05) (Fig. 4B; Fig. S4), indicating that low-dose chemotherapy would induce tumor cell death thus reducing tumor burden. Notably, we observed that the changes in TCR and ctDNA exhibited opposite trends during the neoadjuvant therapy (Fig.3B and 3D; Fig. 4B). Taken together, these results suggest that low-dose chemotherapy combined with delayed immunotherapy not only cause the increase of tumor antigens shed into blood but also stimulate the body’s immunity to produce more tumor antigen-specific T lymphocytes.
In recent clinical trials, the combination of neoadjuvant chemotherapy with immunotherapy by administering them concurrently in patients with resectable NSCLC has brought about longer event-free survival (6, 8, 36). Although chemotherapy promotes tumor cell death and T lymphocyte infiltration (17, 18), it may also have negative effects on T lymphocyte proliferation (37). Consequently, the combination of chemotherapy with immunotherapy has been a topic of controversy. Therefore, the optimal dose, timing of administration, and sequence of therapy need urgent investigation.
Our previous findings from both in vitro experiments and clinical trials have suggested that administering PD-1/PD-L1 inhibitors at 1-10 days (especially 3-5 days) after chemotherapy in lung cancer patients may be more effective than administering them before or concurrently (14). In this trial, we reported the safety and efficacy of low-dose platinum-based chemotherapy combined with delayed immunotherapy for patients with stage IIA to IIIA resectable NSCLC. After modified neoadjuvant therapy, ORR was achieved in 73.7% (28/38) of patients, which was higher than the report of CheckMate-816 (54%) (6) and comparable to those of NADIM (76%) (8) and LungMate 002 (76%) (38) trials. MPR was achieved in 47.4% (18/38) patients, with 31.6% (12/38) achieving pCR. During a median follow-up of 14.37 months, 91.3% (21/23) patients were recurrence-free. These results indicated that modified neoadjuvant chemo-immunotherapy was effective in patients with resectable NSCLC. Only 2.6% (1/38) patients experienced serious TRAEs after receiving modified neoadjuvant treatment, which is lower than the rates reported in CheckMate 816 (33.5%), NAMID (30.0%) (8), and LungMate 002 (6.0%) (38) trials. The reduction of chemotherapeutics may contribute to the low incidence of serious TRAEs. Additionally, we did not observe treatment discontinuation due to toxicity. These results provide evidences that low-dose platinum-based chemotherapy combined with delayed immunotherapy as a neoadjuvant therapy is safe and less toxic.
Previous research has illustrated an association between clinical benefit and TCR repertoire as well as ctDNA level at baseline (39, 40). However, the dynamic changes in the TCR repertoire and ctDNA level during neoadjuvant treatment have been less studied. In our study, we found that the predominant TCR clones were greatly increased after the treatment of chemotherapy from phase 1 to 3, indicating that low-dose platinum-based chemotherapy did not inhibit the proliferation of predominant T cell clones which create a beneficial microenvironment for subsequent immunotherapy. Moreover, ICI treatment significantly increased the number of predominant TCR clones during neoadjuvant therapy, suggesting that immunotherapy has a positive impact on TCR clones, consistent with the study by Forde et al (41). During adjuvant therapy, no significant fluctuation in the predominant TCR clones was observed, probably because the tumor lesions were cleared. Under such circumstances, no tumor antigen was released, thus no tumor antigen-specific T cells were proliferated. Similarly, no significant change was observed in evenness or clonality of TCR. MaxVAF of ctDNA was dramatically dropped during phase 1, which is accordance with previous studies (42–44). Notably, the maxVAF of ctDNA and predominant TCR clones showed opposite trends of change in phase 1. Moreover, the maxVAF of ctDNA was close to 0 during adjuvant therapy stage.
Our study had several limitations that need to be acknowledged. First, it was a pilot study with a relatively small cohort and lacked a control arm. Therefore, validation of our findings in larger cohorts is necessary. Second, the follow-up period was short, and longer-term results are needed to investigate whether modified therapy provides long-term survival benefits. Finally, innovative techniques should be incorporated to further explore the underlying mechanisms of superior clinical efficacy and reduced TRAEs.
In this study, the modified neoadjuvant low-dose platinum-based chemotherapy combined with delayed immunotherapy in patients with resectable NSCLC exhibited promising efficacy and feasibility. The multi-omics analyses provide the dynamic trajectories of the TCR repertoire and tumor burden, which can aim in better understanding the mechanisms of superior efficiency and the feasibility of modified chemo-immunotherapy. In the future, it is possible to conduct large-scale clinical research using this modified chemo-immunotherapy.
All data produced in the present study are available upon reasonable request to the authors
We would like to thank Innovent Biologics, Inc for providing sintilimab for this study.
Author’s contributions
C.Y. Liang: Conceptualization, Formal analysis, Methodology, Writing - original draft. Q. Song: Formal analysis, Methodology, Funding acquisition, Writing - original draft. W.H. Zhou and N. Li: Formal analysis, Methodology, Writing - original draft. C.H. Pan, Q. Xiong B. Yang, and X.L Zhang: Formal analysis. S.H Zhao, W.W. Shi, X. Yan, J.T. Guo, S.J. Sun, Z.H. Dong, and Y.P. Long: Validation, Writing - reviewing & editing. B. Yang, J. Li, Y. Hu, and H.T. Luo: Conceptualization, Investigation, Validation, Verification of underlying data, Supervision, Writing - reviewing & editing.
Funding
This study was supported by the National Natural Science Fund of the People’s Republic of China (NSFC) No.81902912.
Declarations
Ethics approval and consent to participate
Ethics committee/IRB of General Hospital of the People’s Liberation Army gave ethical approval (S2020-144-01) and written informed consent was obtained from all participants.
Consent for publication
Not applicable.
Competing interests
The authors declare no conflicts of interest.
