各位好！众所周知NADIM研究是第一个NSCLC领域的新辅助化免II期研究，该研究的临床和转化部分多次发表在顶刊上。继新辅助单免报道了5年生存情况后，NADIM研究也迎来收官，为我们呈现了5年生存和围绕该研究的所有结果。同时各地的专家学者也针对该领域发表了相应的专家共识，那么到底在这个话题上，还有哪些值得探索的领域？一起来看看。
Perioperative chemotherapy and nivolumab in non-small-cell lung cancer (NADIM): 5-year clinical outcomes from a multicentre, single-arm, phase 2 trial
Mariano Provencio, Ernest Nadal, Amelia Insa, Rosario García Campelo, Joaquín Casal, Manuel Dómine, Bartomeu Massuti, Margarita Majem, Delvys Rodríguez-Abreu, Alex Martínez-Martí, Javier de Castro, David Gómez de Antonio, Iván Macia, Santiago Figueroa, Luís Fernández Vago, Virginia Calvo, Ramón Palmero, Belén Sierra-Rodero, Cristina Martínez-Toledo, Marta Molina-Alejandre, Roberto Serna-Blasco, Atocha Romero, Alberto Cruz-Bermúdez
The Lancet Oncology October 14, 2024
Background Perioperative immunotherapy improves short-term outcomes in resectable non-small-cell lung cancer (NSCLC). We now report 5-year survival from the NADIM trial to assess its long-term benefit.
背景：围手术期免疫治疗可改善可切除非小细胞肺癌（NSCLC）的短期结局。本研究旨在报告NADIM试验的5年生存率，以评估其长期益处。
Methods: NADIM was a multicentre, single-arm, phase 2 trial conducted across 18 hospitals in Spain. Patients were aged 18 years or older, had an Eastern Cooperative Oncology Group performance status of 0 or 1, and had histologically or cytologically confirmed, treatment-naive, resectable stage IIIA NSCLC (American Joint Committee on Cancer, 7th edition criteria). The neoadjuvant treatment consisted of three cycles of intravenous paclitaxel (200 mg/m²) and carboplatin (area under the curve 6 mg/mL per min) with nivolumab (360 mg). After surgery, 1 year of adjuvant treatment with intravenous nivolumab monotherapy was administered (240 mg every 2 weeks for 4 months, followed by 480 mg every 4 weeks for 8 months). The primary endpoint was 24-month progression-free survival, with 5-year progression-free survival and overall survival as secondary endpoints, assessed in the intention-to-treat population (ie, all patients who received neoadjuvant treatment). Toxicity profile was also assessed as a secondary endpoint. This trial is registered at ClinicalTrials.gov (NCT03081689) and is complete; this is the final report of the trial.
方法：NADIM是一项在西班牙18家医院进行的多中心、单组、2期试验。患者年龄为18岁或以上，东部肿瘤协作组体能状态为0或1，并且患有组织学或细胞学证实的、未经治疗的、可切除的Ⅲ A期非小细胞肺癌（美国癌症联合委员会，第7版标准）。新辅助治疗包括三个周期的静脉注射紫杉醇（200毫克/平方米）和卡铂（曲线下面积6毫克/毫升/分钟）以及纳武利单抗（360毫克）。手术后，给予静脉注射纳武单药治疗1年（每2周240毫克，持续4个月，随后每4周480毫克，持续8个月）。主要终点是24个月无进展生存期，5年无进展生存期和总生存期作为次要终点，在意向治疗人群（即所有接受新辅助治疗的患者）中进行评估。毒性特征也作为次要终点进行了评估。本试验已在ClinicalTrials.gov（NCT 03081689）上注册并已完成;这是试验的最终报告。
Findings: Between April 26, 2017, and Aug 25, 2018, 51 patients were assessed for eligibility, of whom 46 comprised the intention-to-treat population (34 [74%] male and 12 [26%] female, median age 63 years [IQR 58–70]). Follow-up was concluded at 60 months (data cutoff July 11, 2023; median follow-up 60·0 months [IQR 60·0–60·0]). 5-year progression-free survival in the intention-to-treat population was 65·0% (95% CI 49·4–76·9), and overall survival was 69·3% (53·7–80·6). Disease progression occurred in 11 (24%) patients; 14 (30%) patients died, including nine (20%) from disease relapse and five (11%) from non-tumour-related causes. Treatment-related adverse events (TRAEs) of grade 3 or worse occurred in 14 (30%) of 46 patients during neoadjuvant treatment and in seven (19%) of 37 during adjuvant treatment. The most common grade 3 or worse TRAEs were increased lipase and febrile neutropenia (three [7%] each) during neoadjuvant treatment, and elevated serum lipase (four [7%]) and elevated serum amylase (three [8%]) during adjuvant treatment. Serious TRAEs included elevated serum lipase and neutropenia (one [2%] each) during neoadjuvant treatment, and elevated serum lipase (one [3%]) during adjuvant treatment. No treatment-related surgery delays, deaths, or unexpected long-term toxicities were reported.
结果：2017年4月26日至2018年8月25日期间，51名患者接受了资格评估，其中46名患者为意向治疗人群（34名[74%]男性，12名[26%]女性，中位年龄63岁[IQR 58-70]）。随访在60个月时结束（数据截止日期为2023年7月11日;中位随访时间60·0个月[IQR 60·0-60·0]）。5-意向治疗人群的年无进展生存期为65.0%（95%CI 49.4 - 76.9），总生存期为69.3%（53.7 - 80.6）。11例（24%）患者出现疾病进展; 14例（30%）患者死亡，其中9例（20%）死于疾病复发，5例（11%）死于非肿瘤相关原因。46名患者中有14名（30%）在新辅助治疗期间发生了3级或更严重的治疗相关不良事件（TRAE），37名患者中有7名（19%）在辅助治疗期间发生了3级或更严重的治疗相关不良事件（TRAE）。最常见的3级或更差TRAE是新辅助治疗期间脂肪酶升高和发热性中性粒细胞减少（各3例[7%]），以及辅助治疗期间血清脂肪酶升高（4例[7%]）和血清淀粉酶升高（3例[8%]）。严重的TRAE包括新辅助治疗期间血清脂肪酶和中性粒细胞减少症升高（各1例[2%]），以及辅助治疗期间血清脂肪酶升高（1例[3%]）。没有报告与治疗相关的手术延迟、死亡或意外的长期毒性。
Interpretation: Perioperative chemoimmunotherapy showed a promising long-term benefit with no concerning safety data, reinforcing its use in resectable stage IIIA NSCLC.
结论：围手术期化学免疫疗法显示出有希望的长期益处，并且没有令人担忧的安全性数据，强化了其在可切除的Ⅲ A期非小细胞肺癌中的应用。
1.首先来看下目前非小细胞肺癌新辅助化免的几大III期RCT研究，目前的结果均表明对于局部晚期NSCLC新辅助化免的疗效显著优于新辅助化疗。但整个围术期治疗模式中，仍然存在许多争议。
2.上细节：
首先NADIM研究是第一个进行新辅助化疗联合免疫治疗的II期研究，45%的MPR率为今后的多项III期RCT研究提供了坚实的数据基础。本文报道了NADIM研究的5年生存结果、探索性分析及转化研究成果。此外NADIM研究近期还在JITC杂志上发表了关于HLA I类缺陷与新辅助免疫治疗疗效的文章。
其次，从结果上看5-year PFS 65·0% (95% CI 49·4–76·9), and 5-year OS 69·3%，再一次印证了化疗联合免疫的疗效。副反应上因小样本研究，在大III期RCT相继报道的当下，显得略无出彩之处。但是，毕竟作为第一个化免治疗的II期研究，文章结果部分为后来人呈现了标杆式的分析方式——复发模式、与PFS/OS相关的临床因素、多角度的亚组分析、组织学疗效预测标志物和ctDNA的交织。这样的话，是不是就算III期RCT按照类似的思路分析》也有NADIM先登城楼之功？还得是NADIM啊~！会玩~！
第三，新辅助治疗模式今后的发展方向上讨论部分提了几个新的概念。A.治疗后3年内无复发或可作为新辅助治疗的一个探索性终点，研究观察到3年内无复发患者很少出现晚期复发或晚期副反应。B.pCR可以作为长期生存的替代终点，围绕pCR/non pCR的临床试验设计将成为下一个研究热点。C. ctDNA可能作为新辅助治疗中的辅助决策工具，III期RCT研究将为我们提供初步数据。D.辅助治疗怎么做仍然是有待于进一步研究。E. PD-L1和TMB都不是稳健地疗效预测标记物，迫切需要新的免疫治疗疗效预测标记物。
3. 关于NSCLC的新辅助治疗目前国际上有两大共识，IASLC的多学科共识已经广为熟知。这里介绍下STS（The Society of Thoracic Surgeons）协会的共识。
10.1016/j.athoracsur.2024.09.041
目录
1. INTRODUCTION
2. Material and Methods
2.1 Study design and participants (appendix 1)
2.2 Procedures [appendix 1 (pp 2–5)]
2.3 Outcomes
2.4 Statistical analysis (appendix 1)
3. Results [appendix 1 (pp 8, 9)][(appendix 1 p 12)][appendix 1 (p 10)][Figure 1A-B][appendix 1 p 13][appendix 1 p 14][appendix 1 p 15](table 1)(table 2) (appendix 1 p 16)(appendix 1 p 17)(figure 2, appendix 1 p 11)(appendix 1 p 18)(figure 3A)(appendix 1 p 19)(appendix 1 p 20)(appendix 1 p 21)(figure 3B)(appendix 1 p 22-5)(figure 4C)(figure 4D)(appendix 1 p 26)(appendix 1 pp 7, 27–29)
4. Discussion
— 图表汇总—
2.1 Study design and participants
appendix 1
3. Results
[appendix 1 (pp 8, 9)] Table S1. Baseline Characteristics of Patients in the Intention-to-Treat Population †
appendix 1 p 12 Figure S1. CONSORT flow diagram.
Flow diagram depicts the disposition of patients throughout the phases of the study, including, neoadjuvant treatment, surgical resection and adjuvant treatment. Percentages shown are relative to intention-to-treat population (n=46).
[appendix 1 (p 10)] Table S3. Disease relapses and salvage therapy in NADIM patients.
Figure 1. Long-term survival analysis of the intention-to-treat population of the NADIM trial
(A) Swimmer plot summarising follow-up and relevant events (end of treatment, disease progression, cancer-related death, and non-cancer-related death) for patients (n=46). Each bar represents one patient. The left column shows clinical and molecular characteristics (from left to right: sex, histology, lymph node status, pathological response, adherence to adjuvant treatment, and ctDNA clearance.
(B) Progression-free survival and overall survival Kaplan–Meier curves of the intention-to-treat population (n=46) considering all events. Shaded areas represent 95% CIs.
[appendix 1 p 13] Figure S2. Cancer-related survival analyses.
Progression-free survival and Overall survival Kaplan-Meier curves of the Intention-to-treat population (n=46) considering only cancer related events. Shaded area represents 95% CI (confidence interval) of the survival function.
[appendix 1 p 14] Figure S3. Survival analyses in the Per-protocol population
Progression-free survival and Overall survival Kaplan-Meier curves of the Per-protocol population (i.e. patients receiving at least 1 cycle of adjuvant treatment, n=37). Shaded area represents 95% CI (confidence interval) of the survival function.
[appendix 1 p 15] Figure S4. Survival analyses comparison between Per-protocol and Non-Per-protocol populations.
Progression-free survival and Overall survival Kaplan-Meier curves of the Per-protocol population (i.e. patients receiving at least 1 cycle of adjuvant treatment, n=37) versus Non-per-protocol population (n=9) are shown. Curve comparison was carried out using Log-rank test. P-values are two-sided. P-values of less than 0.05 were considered statistically significant. Shaded area represents 95% CI (confidence interval) of the survival function.
Abbreviations: PP, Per-protocol; NPP, Non-Per-protocol; HR, Hazard Ratio.
appendix 1 p 16 Figure S5. Survival analyses comparison by relapse site.
Progression-free survival and Overall survival Kaplan-Meier curves of patients who disease relapses exclusively locally (n=3) versus other types of relapses (n=8) are shown. Curve comparison was carried out using Log-rank test. P-values are two-sided. P-values of less than 0.05 were considered statistically significant. Shaded area represents 95% CI (confidence interval) of the survival function.
Abbreviations HR, Hazard Ratio.
appendix 1 p 17 Figure S6. Survival analyses comparison by Central Nervous System (CNS) relapses.
Progression-free survival and Overall survival Kaplan-Meier curves of patients who disease relapses at CNS (n=4) versus non-CNS relapses (n=7) are shown. Curve comparison was carried out using Log-rank test. P-values are two-sided. P-values of less than 0.05 were considered statistically significant. Shaded area represents 95% CI (confidence interval) of the survival function. Abbreviations HR, Hazard Ratio.
figure 2. Baseline clinicopathological characteristics and long-term survival
Forest plot of univariate Cox proportional HR for progression-free survival and overall survival by clinicopathological features at diagnosis in the intention-to-treat population (n=46) considering cancer-specific events. Number of events (disease progressions or deaths) and total number of patients for each group are shown. The reference categories for each variable are: sex, male (n=34); ECOG performance status, 0 (n=25); smoking, former smoker (n=25); histology, squamous-cell carcinoma (n=16); nodal stage, N0 (n=9); affected nodes, single (n=12). Age (n=46), packs per year (n=44), and tumour lesion size (n=46) are continuous variables.
ECOG=Eastern Cooperative Oncology Group. HR=hazard ratio. *N1 not included as there were only 3 patients with N1 disease, meaning too much error to calculate the Cox model.
appendix 1 p 11 Table S4. Five-year PFS and OS rates based on baseline clinical characteristics.
Table S5. Tissue pre-treatment and posttreatment genomic alterations. (See Appendix 2)
Table S6. ctDNA longitudinal analysis of genomic alterations. (See Appendix 2)
Table S7. Emerging Mutations in tissue and blood. (See Appendix 2)
appendix 1 p 18 Figure S7. Survival analyses considering patients’ resection and adjuvant status.
A) Progression-free survival and Overall survival Kaplan-Meier curves of patients who received surgery (n=41) versus those who do not receive surgery (n=4).
B) Progression-free survival and Overall survival Kaplan-Meier curves of patients who do not receive surgery (n=5) versus those who received surgery but did not initiate adjuvant treatment (n=4).
Curve comparison was carried out using Log-rank test. P-values are two-sided. P-values of less than 0.05 were considered statistically significant. Shaded area represents 95% CI (confidence interval) of the survival function.
Abbreviations HR, Hazard Ratio.
figure 3. Post-hoc survival analyses by pathological response and adjuvant treatment adherence
(A) Progression-free survival and overall survival Kaplan–Meier curves of patients who underwent tumour resection (n=41) who had a complete pathological response (n=26) versus those with a major or incomplete pathological response (non-complete pathological response, n=15).
(B) Progression-free survival and overall survival Kaplan–Meier curves of patients who had a non-complete pathological response (n=15) and completed ≥50% of adjuvant treatment cycles (n=11) versus those who completed <50% of adjuvant treatment cycles (n=4).
Curve comparison was carried out using the log-rank test. p values are two-sided; p<0·05 was considered statistically significant. Shaded areas represent 95% CIs of the survival function. HR=hazard ratio.
appendix 1 p 19 Figure S8. Survival analyses by clinical response.
Progression-free survival and Overall survival Kaplan-Meier curves of patients who achieve stable disease (SD, n=11) versus those who achieved Complete or Partial Response (CR or PR, n=35). Curve comparison was carried out using Log-rank test. P-values are two-sided. P-values of less than 0.05 were considered statistically significant. Shaded area represents 95% CI (confidence interval) of the survival function.
Abbreviations HR, Hazard Ratio.
appendix 1 p 20 Figure S9. Adjuvant adherence and Pathological Responses.
A) Number of patients who received ≥50% or <50% of adjuvant cycles considering pathologic response status.
B) Clinical characteristics of patients with low number of adjuvant treatment.
Abbreviations: CPR, Complete pathological Response; NCPR, Non-Complete Pathological Response; pR pathologic Response; PFS, progression-free survival; OS, Overall survival.
appendix 1 p 21 Figure S10. Survival analyses of CPR patients with ≥50% of adjuvant cycles.
Progression-free survival and Overall survival Kaplan-Meier curves of the patients who achieved complete pathological response and completed ≥50% of adjuvant cycles (n=22). Shaded area represents 95% CI (confidence interval) of the survival function.
Figure 4. Post-hoc survival analyses by molecular parameters
Progression-free survival and overall survival Kaplan–Meier curves of patients with available data for different endpoints.
(A) PD-L1 TPS (n=28), stratified into two categories: positive (≥1%, n=18) and negative (<1%, n=10).
(B) TMB (n=29), categorised as high (n=14) or low (n=15) based on cohort median (5·89 mutations per megabase).
(C) ctDNA status (n=43), stratified according to mutant allelic fraction from all detected mutations in blood with a threshold of 1% (sumMAF ≥1%, n=12; sumMAF <1%, n=31).
(D) ctDNA clearance after neoadjuvant treatment (n=40), stratified according to detection of ctDNA at the end of neoadjuvant treatment (ctDNA clearer, n=27; ctDNA non-clearer, n=13).
Curve comparison was carried out using the log-rank test. p values are two-sided; p<0·05 was considered statistically significant. Shaded areas represent 95% CIs of the survival function.
ctDNA=circulating tumour DNA. HR=hazard ratio. SumMAF=summatory mutant allelic fraction. TMB=tumour mutational burden. TPS=tumour proportion score.
appendix 1 p 22 Figure S11. PD-L1 and TMB association with survival.
Forest plot of univariate Cox proportional-hazards ratio (HR) for progression-free survival (PFS) and overall survival (OS) by PD-L1 TPS and Tumor mutational Burden (TMB) levels at diagnosis. The reference category for each comparison is indicated in the graph. No statistically significant differences were observed. Abbreviations: HR, hazard ratio; CI, confidence interval; Neg, Negative PD-L1.
appendix 1 p 23 Figure S12. Specific mutations and survival.
A) Progression-free survival and Overall survival Kaplan-Meier curves of the patients with tumors harboring mutations in STK11, KEAP1, EGFR, or Rb (n=9) vs no specific mutations (n=20).
B) Progression-free survival and Overall survival Kaplan-Meier curves of the patients with tumors harboring mutations in KEAP1 (n=4) vs no specific mutations (n=20).
Curve comparison was carried out using Log-rank test. P-values are two-sided. P-values of less than 0.05 were considered statistically significant. Shaded area represents 95% CI (confidence interval) of the survival function. Abbreviations HR, Hazard Ratio.
appendix 1 p 24 Figure S13. TCR repertoire clonality and survival.
Progression-free survival and Overall survival Kaplan-Meier curves of patients with tumors with high (n=11) or low (n=11) top1% clonal space. Top 1% clonal space is defined as the aggregate frequencies of the top 1% most frequent clones. Patients were stratified using the top1% clonal space cohort median as threshold (≥0.166115). Curve comparison was carried out using Log-rank test. P-values are two-sided. P-values of less than 0.05 were considered statistically significant. Shaded area represents 95% CI (confidence interval) of the survival function. Abbreviations HR, Hazard Ratio.
appendix 1 p 25 Figure S14. Tumor-Immune Prognostic Score and survival.
The Tumor-immune Prognostic Score (TIP score) was determined in 20 patients, 4 patients (20%) had a TIPs of 1 (Poor prognosis), 9 patients (45%) had a TIP score of 2 (Intermediate prognosis), and 7 patients (35%) had a TIP score ≥3 (Good prognosis). None of the patients had a TIP score of 0, and two patients had a TIP score of 4. A) Left panel: progression-free survival and Overall survival Kaplan-Meier curves of patients with tumors with TIPS 1, TIPS 2, and TIPS ≥3. The 5-year PFS rate augments with increasing TIP score categories, with 25% (95%CI 0.9-66.5), 55,6% (95%CI 20.4-80.5) and 100% (95%CI undefined) in poor (TIP score 1; n=4), intermediate (TIP score 2 n=9), and good (TIP score ≥3; n=7) TIPS, respectively (log-rank p=0.016, Chi-square p=0.0445). Similar results were observed for 5-year OS rates, with 25% (95%CI 0.9-66.5), 66.7% (95%CI 28.2-87.8), and 100% (95%CI undefined) in TIPS 1, TIPS 2, and TIPS ≥3, respectively (log-rank p=0.0333, Chi-square p=0.0434). Curve comparison was carried out using Log-rank test. P-values are two-sided. P-values of less than 0.05 were considered statistically significant. Shaded area represents 95% CI (confidence interval) of the survival function.
Abbreviations HR, Hazard Ratio. The right panel displays a bar graph and chi-square analysis depicting the number of patients with disease progression or death based on TIP score.
B) Progression-free survival and Overall survival Kaplan-Meier curves of patients with tumors with TIPS <2 and TIPS ≥3. Patients with TIPS ≤2, (n=13) showed a worse prognosis compared to those with TIPS ≥3, with a 5-year PFS and OS rates of 46.2% (95%CI 19.2-69.6, log-rank p=0.024) and 53.8% (95%CI 24.8-76.0, log-rank p=0.0462) Curve comparison was carried out using Log-rank test. P-values are two-sided. P-values of less than 0.05 were considered statistically significant. Shaded area represents 95% CI (confidence interval) of the survival function.
appendix 1 p 26 Figure S15. Survival analyses by ctDNA clearance stratified by pathologic response status.
A) Association between pathological response and ctDNA clearance status (n=36).
B) Progression-free survival and Overall survival Kaplan-Meier curves of CPR patients with undetectable ctDNA after neoadjuvant treatment (n=16) versus those with ctDNA detection (n=7).
C) Progression-free survival and Overall survival Kaplan-Meier curves of Non-resected and NCPR patients with undetectable ctDNA after neoadjuvant treatment (n=11) versus those with ctDNA detection (n=6).
ctDNA clearance data may be especially valuable in the non-resected patients, given that of the 4 non-resected cases with available ctDNA, the two patients who had undetectable ctDNA levels after neoadjuvant treatment were alive and progression-free at 5 years. In contrast, the two patients who maintained ctDNA detection died within 5 months of follow-up. Curve comparison was carried out using Log-rank test. P-values are two-sided. P-values of less than 0.05 were considered statistically significant. Shaded area represents 95% CI (confidence interval) of the survival function. Abbreviations HR, Hazard Ratio.
appendix 1 pp 7, 27 Figure S16. Mutational landscape in tissue and blood through perioperative treatment.
A) Oncoprint of genomic alterations detected at baseline in all tumors (n=29) and B) after neoadjuvant treatment in tumors with incomplete pathological response (n=7).
C) Oncoprint of genomic alterations detected at baseline (n=43), D) after neoadjuvant treatment at surgery (n=40), and E) at six months of adjuvant treatment with nivolumab (n=32). All mutations are described in detail in Table S2 and S3.
Abbreviations PR, pathological Response; DP, disease progression; CPR, complete pathological response; MPR, major pathological response; IPR, incomplete pathological response; SNPs, single nucleotide polymorphism; InDels, insertions or deletions.
appendix 1 pp 7, 28 Figure S17. Emerging mutations in tissue and blood.
Oncoprint of new genomic alterations detected in tissue at surgery (A), and in blood at surgery (B) and at 6 months of adjuvant treatment (C). Newly acquired mutations that are not present in the previous time point are defined as emerging mutations and depicted. Emerging mutations are listed in detail in Table S4.
appendix 1 pp 7, 29 Figure S18. Emerging, but not baseline, mutations in KRAS and PIK3CA genes are associated to worse outcomes to perioperative chemoimmunotherapy.
A) Progression-free survival and Overall survival Kaplan-Meier curves of patients with emerging mutations in KRAS or PIK3CA in tissue or blood (n=5) vs those with no KRAS and PIK3CA emerging mutations in tissue and blood (n=25).
B) Progression-free survival and Overall survival Kaplan-Meier curves of patients with baseline mutations in KRAS or PIK3CA in tissue or blood (n=13) vs those with no KRAS and PIK3CA baseline mutations in tissue and blood (n=17).
Curve comparison was carried out using Log-rank test. P-values are two-sided. P-values of less than 0.05 were considered statistically significant. Shaded area represents 95% CI (confidence interval) of the survival function.
Abbreviations: HR, Hazard Ratio.
