The Evolving Role of Adjuvant Therapies

Transcription

1 Therapy-predictive markers for adjuvant chemotherapy The Evolving Role of Adjuvant Therapies Micrometastasis Prognostic Markers (BRCA1) Cured No Further Tx Chemosensitive Rafael Rosell th European Perspectives in Lung Cancer Brussels 6-7 March 09 Targeted Therapy EGFR mutations HER2 mutations K-ras mutations Akt hyperactivity YES (%) Customized Chemo BRCA1 / RAP 80 Crosstalk between EGFR and integrin for invasion and metastasis Ricono et al. Cancer Res 09 1

2 Therapy-predictive markers in advanced NSCLC Chemosensitive LACE (Lung Adjuvant Cisplatin Evaluation) Adjuvant vrb/cis ANITA trial vrb/cis Targeted Therapy YES (%) HR=1.14 EGFR mutations Customized Chemo Del19, L858R RAP 80 / BRCA1 T790M HER2 mutations K-ras mutations K-ras mutations ti IL-6 / STAT3 / FoxP3 / Skp2 P< Antiangiogenics IGF Inhibitors PML / HAUSP / PTEN BMI-1 LKB1 inactivation HR=0.66 HR=0. Targeted therapy mtor inhibitors MEK inhibitors Antiangiogenics Pignon et al. JCO 08 Douillard et al. Lancet Oncol 06 2

3 1.00 ANITA: Survival in N2 patients How to predict relapse in stage I NSCLC TaqMan low-density array outcomes in early SCC of the lung S u r v i v a l Chemotherapy + PORT 0.25 Chemotherapy Observation Observation + PORT M o n t h s Douillard et al. Int J Rad Oncol Biol & Phys 08 Transcriptional profiling reveals sets of genes ( signatures ) for which expression in primary tumors correlates with metastatic relapse and poor survival The function of the gene signatures as a whole is poorly understood A selection of 29 genes analyzed by TaqMan Low Density Arrays found a 3-gene model (including the CSF-1/EGFR invasive signature) that predicts survival with 70% accuracy () Survival univariate analysis Analysis by quartiles Analysis by median M-CSF p=0.003 Fibronectin p=0.009 CA IX p=0.04 M-CSF p=0.002 Fibronectin p=0.002 CA IX p=0.009 PH4 p=0.01 PH4 p=0.01 CXCR4 p=0.03 KIAA0974 p=0.02 ANLN p=0.04 ANLN p=0.02 PGK1 p=0.02 Insulin R p=0.03 VEGFC p=0.03 VEGFC p=0.03 NTRK1 p=0.06 NTRK1 p=0.04 Ezrin p=0.003 EGFR p=0.05 3

4 Predicting survival in SCC by low-density arrays (29 genes examined) 3-gene model by TaqMan low density arrays Survival according to 3-gene model in resected SCC of the lung Hazard Ratio 95% CI p Tumor size 4 1 (ref.) > CSF (ref.) > EGFR (ref.) > CA IX (ref.) > < ß HR 95% C.I. p CSF-1 (>0.9) CA IX ( 0.56) EGFR (>0.76) Risk score = (0.93xCSF-1) + (1.4xCA IX) + (1.1xEGFR) low high N Median 95% C.I p Low 33 NR - < High

5 Survival according to 3-gene model in resected SCC of the lung, tumor size < 4 cm. High risk Low risk N Median 95% C.I p Low 15 NR High Survival according to 3-gene model in resected SCC of the lung, tumor size > 4 cm. Low risk High risk N Median 95% C.I p 15 NR Low High Characteristics Age - yr Median (range) Mean Sex N (%) Male Female Race N (%) Caucasian African-American Other Stage N (%) IA IB IIA IIB IIIA Tumor diameter cm Predictive accuracy of 3-gene signature Gdansk, Poland Training Cohort (N=66) 63 (37-76) (78.8) 14 (21.2) San Francisco, USA Validation Cohort (N=26) 67 (45-65) (65.4) 9 (34.6) Caucasian 66 (0) 21 (80.8) 8) - 3 (11.5) - 2 (7.6) (15.2) 32 (48.5) - 22 (33.3) 2 (3) (38.5) 12 (46.2) 1 (3.8) 3 (11.5) (1-9.5) 2.5 (1.5-8) 70 % 70 % 5

6 Survival according to our 3-gene model in validation set of stage Survival I NSCLC Several gene signatures occupy the same prognostic space Low risk 04 High risk Months N Median 95% CI p Low 13 NR High Massague. NEJM 07 6

7 BRCA-1 >12.28 <= >12.28 <= ERCC-1 0 >3.61 <= >3.61 <= RRM-1 0 >1.81 <= >1.81 <= Multivariate Cox model for survival Outcome according to BRCA1, ERCC1 & RRM1 (validation cohort) Survival Survival Survival Squamous BRCA1 Squamous ERCC1 Squamous RRM1 Customizing chemotherapy Cisplatin-based chemotherapy could be a mistake ERCC1 Stage I-III 1 Positive low risk? RRM1 Stage I 2 Negative high risk? P=0.04 P=0.70 Time to Relapse Time to Relapse Time to Relapse BRCA1 Squamous ERCC1 Squamous RRM1 Squamous P=0.38 Which patients really have a high risk of relapse? cisplatin-sensitive sensitive tumors cisplatin-based chemotherapy cisplatin-resistant tumors taxane-based chemotherapy BRCA-1 ERCC-1 RRM-1 P=0.08 P=0.78 P=0.69 ERCC1 RRM1 BRCA1 Stage I-III 3 Low levels High levels low risk high risk Rosell et al. PLoS ONE 07 1 Olaussen et al; 2 Zheng et al; 3 Rosell et al. PLoS ONE 07 7

8 1 HCCBR HCCEV Molarity of Cisplatin HCCBR116: IC = 4.1 x 6 M HCCEV1: IC = 2.3 x 7 M % Co ntrol % Control cisplatin bleomycin etoposide paclitaxel 0 HCCBR HCCEV M olarity of Taxol HCCBR116: IC = 7.73 x 9 M HCCEV1: IC = 6.21 x 6 M % Control 0 HCCBR HCCEV Molarity of Bleomycin HCCBR116: IC > 1.0 x 3 M HCCEV1: IC = 5.0 x 4 M 0 % Control % Con tr ol HCCBR116 0 HCCEV Molarity of Etoposide HCCBR116: IC > 1.0 x 4 M HCCEV1: IC = 9.12 x 7 M vinorelbine HCCBR116 HCCEV M olarity of Vinorelbine HCCBR116: IC 70 = 1.9 x 9 M HCCEV1: IC 70 = 1.7 x 5 M Survival to neoadjuvant gem/cis in NSCLC according to BRCA1 mrna expression Advanced NSCLC q4 q1 q2+q3 Survival (months) T1 BRCA1 T2 BRCA1 T4 BRCA1 N Median P q1=15 NR q2+q3= q4= Gem/Cis Docetaxel/Cis Docetaxel Taron et al. Hum Mol Genet 04 Pilot SCAT: Spanish Customized Adjuvant Therapy based on BRCA1 mrna levels in stage II-IIIA NSCLC R0 NSCLC pn1 / pn2 Q 1 BRCA1 Q2&3 BRCA1 Q 4 BRCA1 Gem/Cis Doc/Cis Doc Same survival with paclitaxel vs paclitaxel/carbo in stage IV NSCLC (Lilenbaum et al. JCO 05) Cobo et al. ASCO 08 8

9 Docetaxel Docetaxel/cisplatin Gemcitabine/cisplatin mrna expression RAP80 T1 RAP80 T2 RAP80 T3 P levels Median (min-max) 0.56 ( ) 1.26 ( ) 2.27 ( ) BRCA1 T1 8 () 7 (35) 5 (25) ( ) BRCA1 T2 4 (44.4) 3 (33.3) 2 (22.2) 7.67 ( ) BRCA1 T ( ) 1 (16.7) 2 (33.3) 3 () Median 95% CI p Docetaxel NR Docetaxel/Cisplatin Gemcitabine/Cisplatin Bennett & Harper. Nat Struct Molec Biol 08 9

10 Spanish Customized Adjuvant Therapy in completely resected N1 & N2 NSCLC S9900 study design S9900: progression-free survival by treatment arm 05/17/07, median F/U 53 mo 0% CONTROL Resected NSCLC pn1 / pn2 1:1 EXPERIMENTAL Docetaxel/Cis T1 RAP80 (T1-T3 BRCA1) T2-T3 RAP80 (T1-T2 BRCA1) T2-T3 RAP80 (T3 BRCA1) genotyping for EGFR & K-ras mutations Gem/Cis Docetaxel/Cis Docetaxel E L I G I B L E R A N D O M I Z E PACLITAXEL CARBOPLATIN x 3 cycles SURGERY SURGERY Clinical stage T2N0, T1-2N1, T3N0-1 Mediastinoscopy if LN > 1 cm on CT Stratification: IB/IIA vs IIB/IIIA 80% % % % 0% Median 3 yr 5 yr Preop 33 mos 47% 42% Control 21 mos 38% 32% HR=0.77 [ ], p= Months After Registration Courtesy K Pisters, ASCO 07 Courtesy K Pisters, ASCO 07

11 Stratification IB/IIA vs. IIB/IIIA Investigator s Centre Ch.E.S.T. Chemotherapy for Early Stages Trial R A N D O M I Z A T I O N Arm 1: Surgery Arm 2: Cisplatin/ Gemcitabine Surgery F O L L O W U P Cisplatin 75 mg/m2, d.1 & Gemcitabine 12 mg/m2 d.1&8, q3wks for a total of 3 cycles Surgery : within 14 days from randomization in arm 1 and 3-6 wks after the completion of CT in arm 2 Scagiotti et al. ASCO Progression-free survival Surgery only CT+Surgery (N=141) (N=129) Events 55% 49% Median PFS (yrs) Hazard ratio yr PFS 48% 53% 95% CI (0., 0.99) Log-rank p-value* 0.9 p-value** Pts at risk at years Surgery only: CT+Surgery: * unadjusted treatment p-value **global model p-value Scagiotti et al. ASCO Conclusions Customized chemotherapy based on RAP 80 / BRCA1 can improve PFS and overall survival both in the adjuvant and neoadjuvant settings Genotyping for EGFR mutations in early resected NSCLC could be useful to customize erlotinib or other EGFR TKIs Genotyping for K-ras mutations is recommended Further research required to identify high-risk stage I patients for appropriate neoadjuvant or adjuvant chemotherapy 11