O-216 Phase III trial comparing cisplatin (CDDP) plus weekly vinorelbine (VNR) and CDDP plus day 1+8 VNR schedules in advanced non small cell lung cancer (NSCLC)

Orul Sessions/Chemotherapy NSCLC: Randomized Trials: Doublet WYSUSDoublet

61

O-21 3

Operable stage IllA NSCLC: prospectively randomized multicenter German trial of surgery and postoperative radiotherapy versus “trimodality treatment” - analysis of surgical results and toxicity

Wilfried Ernst Erich Eberhardt ‘, Sdnke Korfee’, Christoph Pbttgen’, Horst Wagne+, Georgios StamatiS 4, Bernward Passlick3, Volker Petersens, Johannes Hiising6, Stephan Bildat’, Martin Stuschke2. ‘Department of /“tarnal Medicine (Cancer Research), Essen, Germany; 2 Department of Radiation Oncology, Essen, Germany; 3 Asklepios Klinik, Gaoting, Germany; 4 Ruhrlandklinik, Essen, Germany; 5 Klinikum Rechfs der Isar, Munich, Germany; 6 Department of Statistics and Biometrics, Essen, Germany; 7Deparfmenf of Infernal Medicine, Essen, Germany

Rationale: Selected patients (pts) with early stage IIIA-NSCLC (operable, l2 LN at mediastinoscopy, no clinical N2, no bulky or extranodal disease) are still considered for upfront surgery (S) at many centers in Europe and North America. We have tried to compare this frequent approach to an aggressive trimodality treatment - induction chemotherapy (CTx), preoperative CTx/RTx and S -within a multicenter randomized GERMAN KREBSHILFE trial*. Patients and methods: Based on initial staging investigations including mediastinoscopy, patients with operable IIIA (“operable N2”, central T3NO/l (1994!); WHO 0,i) were stratified (according TN-group, treatment center) and randomized: arm A = S and postoperative RTx (50-60 Gy); arm B = inductionCTx (3 x cisplatin (P) 60 mg/m2 d 1 + 7/etoposide (E) 150 mg/m2 d 3,4,5) + cc CTxIRTx (45 Gy; 1.5 Gy bid: 1 x P 50 mg/m’ d 2 + 9, E 100 mg/m* d 4,5,6) + S (+PCI-l5x2Gy=30Gyq3wks). Results: From 1 l/94 till 7/2001 overall 112 pts were randomized. 2 pts had not completed all planned study treatment at the time of analysis, 6 pts were not evaluable (mis-staging, pts refusal, early progression prior any study treatment); 104 pts were evaluable for complete study analysis: Pts characteristics: M 89 F 15; age median 59 (37-71); histopathology: SCC 48 adeno-ca 34 LCC 16 other NSCLC 6; arm A: n = 50 pts T3NO-1 3 Tl-2N2 37 T3N2 10; arm B: n = 54 pts: T3NO-1 3 Tl-2N2 44 T3N2 7. Arm A: S 50/50 probatory thoracotomy (PT) 5/50 RI/2 lo/50 complete resection (RO) 35/50; arm B: S 39/54 PT l/54 RI/2 6/54 RO 32/54; peri- and postoperative morbidity and mortality were not significantly different between both arms (parameters: rate of deaths, infections, stump insufficiencies, bleeding, embolisms, postop duration of hospitalization, time on respirator, time on intensity care unit). Conclusions: The chosen trimodality protocol was feasible and safe in this randomized multicenter setting. S following this complex bimodality induction therapy does not show increased morbidity or mortality rates. A detailed analysis of long-term survival data will be available at the time of the meeting. Although this trial had initially been planned as a phase-Ill design, the small number of randomized pts in this carefully selected stage (“operable IIIA”) makes this trial “a randomized phase-11 design” with a relatively unique treatment comparison (local only vs trimodality + PCI). *This trial was sponsored by a full trials grant from GERMAN KREBSHILFE BONN 0 214

El

Resectability after results of the first initially resectable conducted by the

two different chemotherapy regimens: step of a phase II randomised trial in stage I - llla non-small cell lung cancer European Lung Cancer working Party

Philippe Lothaire’ , Th Berghmans’ , J.J. Lafitte’, V. Giner3, M.C. Berchier4, A.P. Meert’, M. Paesmans’, P. Mommen’, V. Ninanes, J.P. Sculier’. ‘Institut Jules Bordef, Bruxelles, Belgium; ’ CHU Calmeffe, Lille, France; 3 Hospital Peset, Valencia, Spain; 4 CH, Hayange, France; 5 CHU Saint-Pierre, Bruxelles, Belgium Our Group is conducting a phase II randomised study in order to determine if a combination of cisplatin (50 mg/m’ dl) with standard drugs (mitomycin C 6 mg/m*, ifosfamide 3 g/m” dl) (MIP) or with newer drugs (gemcitabine IQ/m’ & vinorelbine 15 mg/m2 dlc8) (GNP) can be performed in patients with initially resectable I-llla non-small cell lung cancer, without reducing the resectability rate. The first step of the study was to assess the feasibility of each of the neoadjuvant chemotherapy regimens in the first 36 patients, with an early stopping rule if the rate of complete resection would be inferior to 65%. Results are summar&d in the following table: MIP GNP 18 pts 18 ptsResponse to chemotherapy OR 12 (67%) 12 (67%)Stable disease 3 2Progression 1 1 Stop for toxicity 1 3Not operated 3 3Results of surgery (30 pts):(Bi)lobectomy 13 IOPneumectomy 2 4Exploratory surgery 0 1 Complete resection (Ro) 15 (83%) 13 (72%) Pathological complete resection 3 Oln conclusion, very high objective response rates were documented; as the resectability rate was superior to 65% in each arm, according to the statistical design, the trial is pursued until a total sample size of 67 patients per arm.

0 215 =

The study of peri-operative NSCLC

S63

chemotherapy

in stage I-llla

Mei Lin Liao, Y.Z. Zhou, J.A. Ding, G.X. Ni, J.M. Zhao, W.H. Chen, B.H. Heng, J. Chen, H. Be. Shanghai Chest Hospital, Shanghai, China, Shanghai First Pulmonary Hospital, Shanghai, China, Shanghai Hua Dong Hospital, Shanghai, China

Objective: A number of studies had evaluated

the benefit of neoadjuvant chemotherapycombined surgery of stage Illa-lllb NSCLC, survival benefit was found in several papers. We attempt to evaluate the survival and prognosis of cisplatinum-based schedule as peri-operative CT for resectable stage I-llla NSCLC. Methods: a prospective, randomized, multicenter study was conducted by Shanghai Lung Cancer Team(supported by Shanghai Branch of Discipline Foundation) since 1995-1997 on 212 cases of stage I-illa NSCLC with curative resection(99 stage I, 47 stage II, 65 stage Ill), age of ~75, KPS>80, staged by 1997 AJC TNM Criteria. They were randomized to be 103 cases with I-2 cycles of pre-operative CT and 108 cases with no pre-operative CT, 2-4 cycles of post-operative CT were used for stage I/ and stage llla NSCLC, it was totally 4 cycles of MVP or MOP CT schedules each case. Follow-up team had been trained, the follow-up rate should be >95%, last follow-up date was March of 2002. Lobectomy was performed for most patients. Accumulated survival, log rank, MST, Cox uni-variance and multi-variance analyses were used as statistics for evaluation. Results: The two arms were well balanced for baseline demographic and clinical characteristics(p>0.05 for all). Stage I NSCLC had the best yr-survival. No statistical survival difference was found between the group with pre-op CT and with no pre-op CT, p=O.O74, 0.087 and 0.097, respectively, 5-yr survival rates were of 31.98%: 36.68%. A statistical survival difference was only shown in stage II NSCLC, p=0.042,5-yr survival rates and MST were worse in the group with pre-op CT, 20%:65.2% and 24 months:48months, respectively. But no difference was seen in stage I and stage llla NSCLC. Stage and post-op CT were the only two meaningful parameters with statistical survival difference by multi-variance analyses, p=O.OOO all, but no difference was found in others 4 parameters(age, sex, type and pre-op CT). The response rate of pre-op CT was of 50%, though there was no statistical difference, the responders were with slightly better yr-survival rates than MRcNR patients, 38.9% and 33.3%, respectively. In the cases with patological “T” down stage and “T” unchanged after pre-op CT have a better yr-survival rates than “T” up-stage, p=O.O3. It showed us 5-yr survival rates were of 41.67%,40.51%:11.76%, respectively, thus, effective chemotherapy are beneficial to survival. Besides, in the cases with >3 cycles of post-op CT have better survival rates than less cycles, Conclusion: a prospective, randomized, multicenter peri-op CT study for stage I-llla NSCLC conducted in Shanghai, China., it showed there had no benefit in survival between with pre-op CT arm and with no pre-op CT arm. In stage II NSCLC, pre-op CT cases had a worse yr-survival than with no pre-opCT, p=O.O42, but no difference was seen in stage I and stage llla NSCLC. The responder of CT and “T” down stage, “T” unchanged had better survival rates than those of not response and “T” up-stage. From multivariance analyses, stage and post-op CT were the two meaningful parameters to yr-survival, 3-4 cycles of post-op CT had a better statistical higher yr-survival than less cycles, nutrition, supportive treatment, immunity status and prevention of toxicity might be the next study worthy to conduct.

THURSDAY, 14 AUGUST 2003

Chemotherapy Trials: Doublet pzGq

NSCLC: Randomized versus Doublet

Phase Ill trial comparing cisplatin (CDDP) plus weekly vinorelbine (VNR) and CDDP plus day 1+8 VNR schedules in advanced non small cell lung cancer (NSCLC)

Vittorio Gebbia’, Domenico Galetta*, Michele Caruso3, Vito Lorusso’, Ernest0 Durini4, Roberto Valenza’, Maria Ft. Valerio’, Giuseppe Colucci’. ’ University of Palermo, Palermo, Italy; a Oncological Institute of Sari, Bari, /fa/K 3 Oncological Center of Cafania, Catania, Italy; 4 Civil Hospifale of Tricase, Trcase Lecce, Italy; 5 VE Hospital of Gela, Gela, Italy The VNR/DDP combination regimen stands among the most active for advanced NSCLC being able to yield a response rate (ORR) of 28-40% and a 1 -year survival rate of 36%. The VNR/CDDP regimen is however not unique since several schedules have been developed by different investigators on the basis of clinical and toxicological considerations. The main criticism to

S64

Oral Sessions/Chemotherapy

NSCLC: Randomized Trials: Doublet IWSUSDoublet

the CDDPAveekly VNR regimen is represented by the incidence of leukopenia, vomiting and therapy discontinuations which could lead to a lower doseintensity. On the other hand the q21 schedule has been reported to be associated to lesser toxicity while retaining clinical efficacy. A phase ill trial was started in 2001 to identify the best tolerated and active schedule of the CDDPNNR regimen. Patients with poor prognosis stage IIIB or IV NSCLC were randomized to schedule A of CDDP iOOmg/ms q 28 days plus VNR 25 m/m’ weekly, or schedule B of CDDP 80 mg/m” q 21 days plus VNR 30 mg/m” on day 1+8. Comparison of toxicity, ORR, TTP and overall survival were all endpoints of the study. To date 186 pts have been enrolled. Interim analysis of mature toxicity data has shown an excess of severe IeukopeniaIneutropenia (p = 0.032) and therapy omissions in the weekly VNR arm (0.041). To day the ORR according to an intent-to-treat fashion is 38% for arm B and 35% for arm A (p = NS). These data are in accord to retrospective analysis showing that grade 4 leukopenia, grade 4 neutropenia, and therapy discontinuation were less frequently reported with schedule B than A (4-28% vs 4550%; 40-47% vs 65-78%; ~5% vs 1223% respectively), and that dose-intensity > 90% was reported for schedule B as compared to 65-70% for schedule A. The schedule with CDDP 80 mg/ms and VNR 30 mg/ms on day 1+8 is equiactive but better tolerated than the CDDP plus weekly VNR 25 mg/ma schedule with a lower rate of treatment omissions, therapy discontinuations, and severe grade 4 hematological side-effects. Final data concerning TTP and survival are awaited.

I

0 217

A randomized phase Ill trial of gemcitabine (G) in combination with carboplatin (C) or paclitaxel (P) versus paclitaxel plus carboplatin in advanced (Stage 1118,IV) Non-Small Cell Lung Cancer (NSCLC)

Joe Treat’, Belani P. Chandras, Martin Edelmar?, Mark Socinski“, R. Goni$, R. Ansari”, Heidi Gillenwater’, J. Roger@, Coleman Obasajug, R. Comis’. ’ Coalition of National Cancer Cooperative Groups, Philadelphia, USA; 2 Univ of Pittsburgh Cancer Ins, Pittsburgh, USA; 3 Univ of Maryland Greenebaum Cancer Ctr, Baltimore, USA; 4 Univ of N Carolina at Chapel Hi//, Chapel Hi//, USA; 5 Westat, Rockville, USA; 6 Northern Indiana Ca Research Consortium, South Band, US; 7 Univ of Va. Cancer Ctr, Charlottesville, USA; 8 West Virginia Univ, Morgantown, USA; g Lilly Oncology Indianapolis, USA Two-drug regimens are the “standard of care” for treatment of advanced NSCLC. The role of non-platinum containing doublets continues to evolve. We are conducting a multicenter, randomized trial designed to compare the efficacy of gemcitabine-containing platinum and non-platinum doublet regimens with the most commonly used regimen of P+C to identify the combination associated with the best therapeutic index. Between July 2000 and Jun 2002, 474 chemonaive patients (pts) with advanced stage (IIIB/IV) NSCLC, and ECOG PS < 2 were randomized to: (GC) G 1,000 mg/ms IV on days 1,8 plus C AUC 5.5, day 1 vs. (GP) G 1,000 mg/ms IV on days 1, 8 plus paclitaxel200 mg/m*, day 1 vs. the control arm (PC) paclitaxel 225 mg/ms plus carboplatin AUC 6.0, day 1. All three regimens were administered as 21 day treatment cycles. The primary end point was overall survival (OS). Secondary endpoints were TTP, RR, and quality of life (QoL) analysis using FACT-L questionnaire. Results: At the time of this preliminary analysis, 352/474 pts have sufficient follow-up for response evaluation. Toxicity data are available for 461 pts. Pts characteristics include: (M/F 292/l 82) median age 63.9 years (range 32.891 .O), ECOG PS (O-142/1-263), stage (IIIB-46/lV-428). Response and toxicity data are summarized in the table below. Best Overall response

G C - ~102

GP - n=105

PC - rk98

CFVPR (%) SD (%) Toxicities - grade 314 Anemia Neutropenia with sepsis Thrmbocytopenia Alopecia (gr 2) Neuro-Sensory

i/32 (32) 45 (44) G C (~151) 30 24 1 62 15 2

2/39 (39) 33(31) GP (~155) 7 15 1 9 65 7

o/34 (35) 26(27) PC (=155) 5 17 7 16 63 9

GC resulted in a higher incidence of uncomplicated grade 3/4 myelosuppression but was associated with less alopecia and neurosensory toxicity. Accrual to the study continues and survival data will be presented.

I.0 218

Smgle-agent Gemcitabine (G) and Docetaxel (D) Given Sequentially Every 3 Weeks in Advanced NSCLC: Final Results and Quality of Life Assessment of a Randomised Phase II Study show Effective Treatment

Joerq Mezqer’, Lothar R. Pilz’, Gisela Koschels, Katrin Scho@, D. Hruskas, Christian Manegold4. ‘St. Vincentius Kliniken, Karlsruhe, Germany; ’ Deutsches Krebsforschungszent, Heidelberg, Germany; 3 Allgemeines Krankenhaus, Hamburg-Harburg, Germany; 4 Thoraxklinik Heidelberg, Heidelberg, Germany; 5 Klinik Schillerhdhe, Gerlingen, Germany Gemcitabine (G) and Docetaxel (D) have been shown to be effective in chemotherapy-na’ive and pretreated patients (pts) and are not cross-resistant. We have tested G/D in doses of 1250 mg/ms/i 00 mg/ms and a q3w schedule G: day 1,8/D: day 1 (Proc ASCO 2001; 20: 337a [Abstract 13461) giving G or D initially up to 6 cycles and in case of tumor progression switching to D and G, respectively, (up to 6 cycles). 330 patients with median age of 64 (range 29-85) entered the study (stage at study entry IIIB/IV 11%/89%; WHO performance status 0 or l/21 : 81 %/I 9%; histology adenomatous/squamous/miscellanous 48%/26%/26%), and all 321 evaluable patients have been analysed. In median survival, no statistical significant difference can be seen between the treatment arms A: G followed by D, 6.3 mos; B: D followed by G, 8.6 mos (Kaplan-Meier: log-rank p=O.206). The corresponding confidence intervals (Cl) are AJB [5.2;7.2]/[7.1;10.3]. The l-Year-Survival-Rate of A/B is 28%/31% with 12(7.5%)/17(10.5%) censored observations. So far the treatment arms A and B can be statistically considered as equally effective. The quality of life (QoL) evaluation using the EORTC QLQ-C30 questionnaire showed no difference between A and B. 88%/96% (A/B) of pts participated in the QoL evaluation. 57%/50% pts were evaluable per protocol for baseline data (1st QLF) and after the 2nd cycle (3rd QoL evaluation). The individual difference of the first 13 questions was not statistically significant (p-value is 0.3913 in the Wilcoxon rank sum test with an estimate of -0.077 and a Cl of [-0.154;0.077]). The use of G and D as carried out in the two arms is effective. There was no significant difference in QoL assessment results between the two drugs.

I0 219

Preliminary Results of Four Arm Cooperative Study (FACS) for Advanced Non-Small Cell Lung Cancer (NSCLC) in Japan

Koii Takeda’, Shunichi Negoro’ , Yasuo Ohashis, Nagahiro Saijos, Yutaka Nishiwaki4, Tomohide Tamura3, Kazuhiko Nakagawas, Kaoru Kubota4, Yutaka Ariyoshis, Masahiro Fukuoka5. ’ Osaka City Genera/ Hospital, Osaka, Japan; ’ University of Tokyo, Tokyo, Japan; 3 National Cancer Center Hospital, Tokyo, Japan; 4 National Cancer Center Hospital East, Kashiwa, Japan; 5 Kinki University School of Medicine, Osakasayama, Japan; 6 Aichi Prefectural Hospital Okazaki, Japan FACS was designed to compare 3 platinum-based combination regimens to cisplatin (80 mg/ms, day 1) plus irinotecan (60 mg/m*, days 1, 8, 15) (IP, q4wks) as a reference arm. The experimental regimens were: carboplatin (AUC 6, day 1) plus paclitaxel (200 mg/ms, day 1) (TC, q3wks); cisplatin (80 mg/ms, day 1) plus gemcitabine (1,000 mg/ms, days 1, 8) (GP, q3wks); cisplatin (80 mg/m*, day 1) plus vinorelbine (25 mg/m *, days 1,8) (NP, q3wks). Each patient was required to meet the following criteria: histologically and/or cytologically documented NSCLC, clinical stage IV or IIIB (malignant pleural effusion and/or metastasis in the same lobe), PS (ECOG scale) 0 or 1, age >= 20, ~75 years old, no prior chemotherapy, measurable lesion, adequate organ functions. Between October 2000 and June 2002, 602 patients were registered from 44 hospitals in Japan. Age, gender, PS, stage, LDH and albumin were well balanced across arms. Objective tumor response rates in intention-to-treat (ITT) basis were almost equivalent across arms. As of January 2003, toxicity data of 182 pts were assessable. Frequency of grade 3 or greater neutropenia was lower in GP, however, thrombocytopenia was higher in GP. Frequency of grade 2 or greater nausea was lower in TC, however, sensory neuropathy was higher in TC. Frequency of grade 2 or greater diarrhea was lower in TC, GP and NP.

Results:

No. of randomized pts. RR in ITT No. of toxicity assessable G3-4 neutropenia G3-4 thrombocytopenia G2-4 nausea G2-4 sensory neuropathy G2-4 diarrhea

pts.

IP

TC

GP

NP

151 30% 45 80% 0% 62% 2% 49%

150 31% 46 87% 9% 22% 20% 4%

151 28% 52 82% 31% 54% 0% 4%

150 31% 39 87% 3% 39% 0% 5%

Response and toxicity data of all patients in each arm and combined survival will be presented in the meeting. Final analysis of the survival by arm will be performed on November 2003.