- 著者: Le Treut J, Sault MC, Lena H, Souquet PJ, Vergnenegre A, Le Caer H, Berard H, Boffa S, Monnet I, Damotte D, Chouaid C
- Corresponding author: Chouaid C (Department of Pneumology, Centre Hospitalier de Créteil, Créteil, France)
- 雑誌: Annals of Oncology
- 発行年: 2013
- Epub日: 2013-02-13
- Article種別: Original Article (Phase II clinical trial)
- PMID: 23406729
背景
Large-cell neuroendocrine carcinoma (LCNEC) of the lung は全肺癌の約 1% を占める rare、aggressive high-grade neuroendocrine tumorで、 1991年 Travis らによって組織学的に独立した entity として記載された (large cells with abundant cytoplasm、high mitotic rate、extensive necrosis、neuroendocrine growth pattern)。WHO classification では LCNEC を large-cell carcinoma の subtype (i.e., non-small cell lung carcinoma [NSCLC] の subtype) として分類するが、 small cell lung carcinoma (SCLC) と genetic alterations を共有し (Jones et al. ClinCancerRes 2005)、clinical behavior (male preponderance、smoking history、aggressive course) も類似する。これに対し typical / atypical carcinoidは低悪性度であり、 mitotic rate / necrosis profile が大きく異なる。LCNEC patients の 5-year survival rate は 15-57% と heterogeneousで、 NSCLC より significantly worse だが SCLC とは同等の prognosisを示す (Travis et al. JThoracOncol 2006)。Treatment では、 LCNEC が cisplatin-based SCLC-style chemotherapy regimensに比較的良好な response を示すと retrospective studiesで報告されてきた (Rossi et al. LungCancer 2005) が、 small sample sizes (n=12-45 patients) + retrospective design という limitationを抱え、prospective evidenceは皆無であった。これまで未解明だった重要な課題として足りなかったのは、(1) LCNEC への cisplatin + etoposide doublet chemotherapy efficacy の prospective multicenter validation、(2) 中央病理レビューによる LCNEC 診断 accuracy の前向き検証 (実臨床では misclassification が頻発と推測されていた)、(3) Cisplatin-etoposide regimen 毒性プロファイルの prospective characterization、の3点であり、本研究はフランス Groupe Français de Pneumo Cancérologie (GFPC) 主導の n=42 advanced LCNEC patients を対象とする世界初の前向き phase II trialで初めて体系的に検証する。
目的
未治療・PS 0/1の stage IV / stage IIIB (pleural T4) advanced LCNEC patientsに対する cisplatin (80 mg/m^2 day 1) + etoposide (100 mg/m^2 day 1-3) 21-day cycle doublet chemotherapy の (a) primary objective: objective response rate (RECIST v1.0) / disease control rate (CR + PR + SD)、(b) secondary objectives: pathology diagnosis accuracy (中央 6-pathologist panel review、independent review with consensus resolution)、treatment toxicity (NCI Common Toxicity Criteria v3.0)、median progression-free survival (PFS) + overall survival (OS)、を体系的に prospectively評価する。
結果
Patient enrollment + baseline characteristics (Table 1)：18 study centers が 42 patients を enroll、mean age 59 ± 9 years (range 33-73)、69% males (n=29 patients)、PS 0 in 48% / PS 1 in 52%、stage IV metastatic 88% (n=37 cases) + stage IIIB with pleural T4 12% (n=5 cases)、mean time from diagnosis to study inclusion 0.9 ± 1.2 months、mean weight loss 5.9% ± 7% (range 0-29%)。Cisplatin-etoposide chemotherapy given to 41/42 patients (98%)、median 4 cycles/patient (range 1-6)、median cisplatin dose 520 mg、median etoposide dose 1,740 mg、cohort n=42 patients。
Centralized pathology review — 27.5% reclassified (Figure 2)：40/42 patients (95%) で centralized pathology review が可能。中央病理委員会判定で 11/40 patients (27.5%) が reclassified : SCLC 9 cases (22.5%)、undifferentiated NSCLC 1 case (2.5%)、atypical carcinoid 1 case (2.5%) (cohort n=40 patients)。SCLC への reclassification 約 9-fold dominant over other reclassifications。これは LCNEC misclassification の structural problem を prospectively で初めて定量化した重要 finding (cohort n=40 patients、 約 2.7-fold of patients are reclassified)。
Efficacy — DCR 64% primary endpoint achieved：Intention-to-treat analysis of overall population (n=42 cases): disease control rate (DCR) = 64% (primary endpoint achieved、threshold > 50% retain)、stabilization 26%、partial response 38%、disease progression 36% (Table 2、cohort n=42 patients)。Confirmed LCNEC subgroup (n=29 patients、excluding 11 reclassified): stabilization 31%、partial response 34%、progression 35% (no significant difference from overall population、p=0.18、chi-square test)。
Survival outcomes (Figure 1)：Median follow-up = 37.2 months、cohort n=42 patients、Kaplan-Meier analysis。Median PFS = 5.2 months (95% CI 3.1-6.6 months) vs median OS = 7.7 months (95% CI 6.0-9.6 months) (intention-to-treat、Figure 1)、1-year PFS rate = 14.3% (95% CI 6.7-27.8%)、1-year OS rate = 26.8% (95% CI 15.7-41.9%)。Confirmed LCNEC subgroup (n=29 patients): median PFS 5.0 (95% CI 4.0-7.9)、median OS 8.0 months (95% CI 3.7-7.9)、cohort n=29 patients。Reclassified subgroup (n=11 cases、SCLC/NSCLC/carcinoid): median PFS 3.1 months (95% CI 2.8-8.5)、median OS 7.0 months (95% CI 3.0-9.0)、cohort n=11 cases。No significant difference between confirmed LCNEC vs reclassified groups (p=0.55、Figure 2)、LCNEC OS 8.0 vs reclassified OS 7.0 = 約 1.14-fold difference (not statistically significant)。
Comparison with extensive SCLC：LCNEC outcome は extensive SCLC とほぼ同等 — phase III SCLC trial (cisplatin-etoposide vs cisplatin-irinotecan) の median PFS 4.8 + OS 9.4 months、phase III SCLC trial (cisplatin-etoposide vs cyclophosphamide-epirubicin-vincristine triplet) の median OS 8.4 months と比較し、本研究 OS 7.7 months は 約 0.82-0.92-fold (cohort n=42 LCNEC vs historical SCLC trials)。LCNEC が advanced SCLC と類似した chemotherapy responsiveness と poor prognosisを示すことが prospective evidence として確認された。
Toxicity profile (Table 2)：n=41 patients receiving ≥1 chemotherapy cycle (98%)、median 4 cycles/patient (cohort n=41 patients)。Grade 3/4 toxicity in 59% of patients (n=24): neutropenia 32% (n=13、Grade 3/4)、thrombocytopenia 17% (n=7)、anemia 12% (n=5)、vomiting 7% (n=3)、renal toxicity 5% (n=2)、decline in general health 12% (n=5)、asthenia 2% (n=1)、alopecia 2%。Any-grade toxicity in 95% of patients。Chemotherapy cycle postponement (leukopenia/neutropenia) in 23/41 patients (56%)、dose reduction required in 9/41 patients (22%)。Reasons for chemotherapy discontinuation: disease progression 38%、completion of 6 cycles 32%、toxicity 20%、death 10% (cohort n=41 patients、no treatment-related deaths attributed to chemotherapy)。
考察/結論
① 先行研究との違い: これまでの LCNEC chemotherapy efficacy reports は retrospective、small cohort (n=12-45 patients)、heterogeneous regimens (cisplatin + irinotecan / paclitaxel / vinorelbine / docetaxel 混在) という limitations を抱えていた (Rossi et al. LungCancer 2005; Niho et al. LungCancer 2009) のと対照的に、本研究は LCNEC への cisplatin-etoposide doublet chemotherapy efficacy を 世界初の prospective multicenter phase II trial (n=42 patients) で評価した点で先行研究と決定的に異なる。さらに既存報告は centralized pathology reviewなしの local diagnosis を信頼したのと相違し、本研究は 6-pathologist panel による rigorous central review を導入し misclassification rate 27.5% という structural diagnostic problemを定量化した点が methodologically distinct。Historical SCLC trials の median OS 8.4-9.4 months と本研究 LCNEC OS 7.7 months の同等性は、SCLC-style chemotherapy regimens の LCNEC への applicability を prospective level で初めて establishした。② 新規性: 本研究で初めて、 advanced LCNEC に対する cisplatin-etoposide doublet chemotherapy のprospective phase II efficacy + safety data を提供した novel evidence であり、DCR 64% という primary endpoint achievement と median OS 7.7 months という outcome が world-first proof-of-concept。これまで報告されていない知見として、(a) Centralized pathology review reclassification rate 27.5% (cohort n=40 patients)、(b) Cisplatin-etoposide regimen LCNEC = extensive SCLC equivalence in efficacy、(c) Grade 3/4 toxicity rate 59% + treatment discontinuation 20% という novel quantitative safety profile、(d) 1-year survival 26.8% という poor outcome の prospective characterization、を実証した。③ 臨床応用と意義: 本研究の bench-to-bedside translational significance として、(a) Cisplatin-etoposide doublet chemotherapyを advanced LCNEC patientsの first-line standard treatment として臨床応用支持、(b) Centralized pathology review の clinical trial 必須化 (misclassification 27.5% defenseで 適切 patient selection)、(c) LCNEC = SCLC-style treatment regimen の臨床現場での合理化、(d) Aggressive supportive care (growth factor、antiemetic、blood product transfusion) 必要性の確立、(e) WHO + IASLC LCNEC treatment guidelines に直接 informった、 臨床的有用性が極めて大きい (実臨床での LCNEC patients への cisplatin-etoposide treatment は本試験を根拠に標準化)。④ 残された課題と今後の検討: limitationとして、(1) 単群 phase II design (non-randomized) のため direct comparison に NSCLC-style regimens (carboplatin + paclitaxel、pemetrexed-based) を含めた randomized trial が今後の研究として必要、(2) Sample size n=42 patients は statistical power に限界、larger cohort + multicenter validation が future studyとして残された課題、(3) Molecular subtype-based treatment selection — 後続 research で RB1/TP53 mutated SCLC-like LCNEC vs NSCLC-like LCNEC の chemotherapy response差が示唆されており、 genotype-guided treatment が今後の検討、(4) Second-line + salvage therapy regimen optimization (irinotecan、taxanes、topotecan、lurbinectedin への移行) が今後の展望、(5) Immunotherapy (PD-1/PD-L1 inhibitor) era での LCNEC への nivolumab、pembrolizumab、atezolizumab + chemotherapy combination の clinical utility validation、(6) Targeted therapy potential (KIT、PDGFRα、PDGFRβ、MET、DLL3 等 receptor tyrosine kinasesの LCNEC expression profile reported)、(7) LCNEC-specific biomarker development (neuroendocrine differentiation markers、ASCL1、NEUROD1、POU2F3 等 SCLC subtype-equivalent profiling)、が今後の研究テーマ。これらlimitationにもかかわらず、本研究はadvanced LCNEC に対する prospective phase II evidence を世界で初めて提供した重要な clinical milestone evidence であり、 LCNEC oncology research の clinical-NEC standard treatment establishmentに寄与した。
方法
試験設計: Prospective、multicentre、single-arm、open-label、phase II trial (GFPC 0302)、フランス 18 study centers (2004年5月-2009年12月 enrollment、ethics committee approval Marseille 2 number 03/71、written informed consent)。対象: histologically documented LCNECで stage IV または stage IIIB with pleural involvement、performance status (PS) 0/1、age 18-75 years、no prior chemotherapy、measurable target lesion (non-irradiated)、no peripheral neuropathy grade ≥2、life expectancy > 3 months、biological status compatible (bilirubin < 1.25 × ULN [upper limit of normal]、transaminase < 3 × ULN、alkaline phosphatase < 2.5 × ULN、polymorphonuclear neutrophil > 1.5 G/L、platelet > 100 G/L)。Pathological definition of LCNEC: cohesive sheet of large tumor cells with endocrine pattern、one or several nucleoli、expression of ≥2 of 3 neuroendocrine markers (CD56、synaptophysin、chromogranin)。Mixed histological features や cytology-only diagnosis は除外。Excluded: SCLC、bronchioloalveolar carcinoma、prior chemotherapy、symptomatic brain metastasis、unstable heart disease、uncontrolled infection、grade > 2 neuropathy、metastatic malignancy history within 5 years。Treatment regimen: cisplatin 80 mg/m^2 IV over 30-min on day 1 + etoposide 100 mg/m^2 on days 1-2-3、21-day cycles、up to 6 cycles。Growth-factor therapy はinvestigator discretion。Clinical evaluation + laboratory tests at each cycle initiation。Response assessment: RECIST v1.0、CT または MRI for radiologically measurable tumors、after 3 + 6 cycles。Centralized review by panel of 6 pathologists (independent review with consensus resolution、reclassified patients continued participation)。統計: Open phase II design with 2 interim analyses (after n=20 and n=35 enrollments、protocol stop for futility if < 3 / < 6 disease control patients respectively)、planned n=42、disease control rate (RR) reject if < 30% / retain if > 50% (α=5%、power 80%)。Quantitative data = median + range、qualitative data = percentage、group comparison = chi-square test。Primary endpoint objective response rate = responders / total study patients (intention-to-treat、 treatment discontinuation due to toxicity classified as failure)。Treatment toxicity = NCI Common Toxicity Criteria v3.0。PFS + OS by Kaplan-Meier method (baseline to radiological progression or death for PFS、to death for OS)、median follow-up by inverse Kaplan-Meier method。Data cut-off date April 2012。No cell line / mouse strain (clinical trial)。