Carboplatin plus Paclitaxel versus Carboplatin plus Stealth Liposomal Doxorubicin in patients with advanced ovarian cancer: activity and safety results of the MITO-2 randomized multicenter trial S. Pignata 1, G. Scambia 2, A. Savarese 3, R. Sorio 4, E. Breda 5, G. Ferrandina 2, V. Gebbia 6, P. Musso 7, C. Gallo 8, F. Perrone 9 1 Istituto Nazionale Tumori, Napoli; 2 Università Cattolica del Sacro Cuore, Roma; 3 Istituto Regina Elena, Roma; 4 CRO, Aviano; 5 Ospedale Fatebenefratelli, Isola Tiberina, Roma; 6 Casa di Cura La Maddalena, Università di Palermo; 7 Ospedale Oncologico M.Ascoli A.R.N.A.S., Palermo; 8 Seconda Università di Napoli; 9 Istituto Nazionale Tumori di Napoli,Italy.

ASCO conflict of interest statement MITO-2 is an independent, academic study. Sponsor of the study is NCI Naples, that is responsible for trial design, study coordination, data analysis and has the property of the database. The study was partially supported by funds from Schering- Plough. Schering-Plough Italy supplied pegylated liposomal doxorubicin (PLD). Sandro Pignata received honoraria from Schering-Plough.

Introduction (1) Carboplatin plus paclitaxel is standard first-line chemotherapy for patients with advanced ovarian cancer 1-3 Single-agent pegylated liposomal doxorubicin (PLD) is a standard option for platinum resistant relapsed ovarian cancer 4 1 Ozols RF et al, J Clin Oncol 2003, 21: Neijt JP et al, J Clin Oncol 2000, 18: du Bois A et al, J Natl Cancer Inst 2003, 95: Gordon AN et al, J Clin Oncol 2001, 19:

Introduction (2) Combination of carboplatin and PLD is highly active as second-line chemotherapy in patients with advanced ovarian cancer in late relapse Ferrero JM et al, Proc Am Soc Clin Oncol Ferrero JM et al, Ann Oncol 2007, 18:

Study objective MITO-2 is a randomized phase III study testing whether carboplatin plus PLD is more effective than carboplatin plus paclitaxel as first-line treatment of patients with advanced ovarian cancer

RandomRandom Strata: Center PS (0-1, 2) Stage (IC, II, III, IV) Residual disease after surgery (absent,  1 cm,  1 cm, no surgery) Control arm Carboplatin AUC 5, day 1 Paclitaxel 175 mg/m 2, day 1 Treatment repeated every 21 days, for 6 cycles Experimental arm 1:1 Study design Carboplatin AUC 5, day 1 PLD 30 mg/m 2, day 1 Treatment repeated every 21 days, for 6 cycles

Study population Inclusion criteria Cyto/histological diagnosis of ovarian cancer FIGO Stage IC – II – III – IV Age  75 ECOG Performance Status 0-2 No previous chemotherapy Main exclusion criteria ANC  2000/  L, platelets  /  L Creatinine  1.25 x UNL, SGOT and SGPT  1.25 x UNL Life expectancy of less than 3 months

Study endpoints Primary endpoint Progression-free survival (PFS) Secondary endpoints Overall survival (OS) Objective response rate (RECIST) Toxicity (NCI – CTC v2.0) Quality of Life (EORTC QLQ C30)

Sample size 2-tailed  : 0.05 Power: 80% Hazard Ratio: 0.80 Median PFS in control arm: 18 months Median PFS in experimental arm: 22.5 months  632 events (progressions) needed  820 patients planned

Study conduction First patient enrolled: January 17, 2003 Last patient enrolled: November 9, active Institutions (41 Italy, 1 Portugal) 820 randomized pts (809 Italy, 11 Portugal) Preplanned early safety analysis: –first 50 pts receiving carboplatin + PLD 1 Preplanned interim activity analysis: –first 50 pts eligible for RECIST assigned to carboplatin + PLD 2 1 Pignata S, BMC Cancer 2006; 6: Pignata S, Oncology 2009; 76: 49-54

Baseline characteristics Carbo + PaclitaxelCarbo + PLD (n = 410) Age median (range) 57(21-77)57(25-77) ECOG Performance Status (97%)397(97%) 212(3%)13(3%) FIGO Stage IC38(9%)38(9%) II40(10%)37(9%) III243(59%)247(60%) IV89(22%)88(22%) Residual disease after surgery Absent152(37%)150(37%)  1 cm 68(17%)69(19%)  1 cm 117(28%)114(28%) No surgery73(18%)67(16%)

Treatment compliance Carbo + PaclitaxelCarbo + PLD (n = 410) Pending information29 (7%)27 (7%) Did not start treatment4 (1%)6 (1%) Number of cycles received* 110 (3%)11 (3%) 213 (3%)19 (5%) 38 (2%)13 (4%) 46 (2%)11 (3%) 510 (3%)14 (4%) 6330 (88%)309 (82%) *p=0.39

Treatment compliance: delays Cycle 2Cycle 3Cycle 4Cycle 5Cycle 6 Delays due to hematologic toxicity Delays due to non hematologic toxicity Carboplatin + Paclitaxel Carboplatin + PLD Number of patients

Toxicity (1) Any grade Severe (G  3) C+PC+PLDp*C+PC+PLDp* Toxic deaths0.8%0.5%1 Anemia59%68%0.0074%10%  RBC transfusions2%6%0.002 Neutropenia73%80%0.0449%43%0.09 Febrile neutropenia2%1%0.21 Thrombocytopenia19%48%  %16%  Platelet transfusions0.3%2%0.06 Bleeding0.3%1%0.37-1%0.24 C+P: carboplatin + paclitaxel, 399 patients; C+PLD: carboplatin + PLD, 386 patients *Chi square or Fisher exact test as appropriate

Toxicity (2) Any grade Severe (G  3) C+PC+PLDp*C+PC+PLDp* Allergy6%5%0.602% 0.86 Heart2%4% %2%0.06 Fatigue44%43%0.863% 0.94 Constipation32% 0.991% 0.73 Nausea47%51%0.212% 0.95 Vomiting29%30%0.832%3%0.42 Diarrhoea13%6%  %-0.25 Hair loss63%14%  Skin toxicity6%20%  %0.01 Stomatitis9%20%  %0.5%0.62 Neurotoxicity47%15%  %0.2%0.004 *Chi square or Fisher exact test as appropriate C+P: carboplatin + paclitaxel, 399 patients; C+PLD: carboplatin + PLD, 386 patients

Activity analysis: flow of patients Analysis performed according to “intention to treat” principle Pending information Eligible for RECIST Not eligible for RECIST Non-target lesions only Elevated CA-125 only No lesions, normal CA-125 Carbo + Paclitaxel (n=410) 10 pts 83 pts 88 pts 73 pts 156 (38%) 18 pts 99 pts 80 pts 79 pts 134 (33%) Carbo + PLD (n=410)

Objective response – RECIST Women with target lesions Carbo + Paclitaxel (n=156) Carbo + PLD (n=134) p (  2 )* Objective response 92 (59%)76 (57%) 0.70 Complete response 24 (15%)22 (16%) Partial response 68 (44%)54 (40%) No response 64 (41%)58 (43%) Stable disease 45 (29%)41 (31%) Progressive disease 9 (6%)7 (5%) Not evaluated 10 (6%)10 (7%) *Objective response vs no response

Activity Women not eligible for RECIST Carbo + Paclitaxel Carbo + PLD p (  2 ) Non-target lesions only Complete response (CR) 27 / 83 (33%)29 / 99 (29%)0.64* No CR / No PD 46 / 83 (55%)48 / 99 (48%) Progressive disease 2 / 83 (2%)4 / 99 (4%) Not evaluated 8 / 83 (10%)18 / 99 (18%) Elevated Ca125 only Ca125 normalized 73 / 88 (83%)69 / 80 (86%)0.56** *Complete response vs not **Ca125 normalized vs not

Primary endpoint Number of events required for final analysis (632) has not been reached yet As of May 4, 2009, with a median follow-up of 35 months, 531 progressions have been recorded Only overall curves are shown

PatientsEvents Median PFS (months) 1-yr PFS 2-yr PFS (95%CI ) 65.0%41.9% Patients at risk Progression-free survival* Months *May 2009

PatientsEvents Median OS (months) 1-yr OS 2-yr OS (95%CI 48.3-n.a.) 88.8%73.8% Patients at risk Overall survival* Months *May 2009

Preliminary conclusions (1) Toxicity profile of carboplatin plus PLD as first-line treatment of advanced ovarian cancer is markedly different from carboplatin plus paclitaxel Carboplatin plus PLD is associated with: –Higher incidence of anemia and thrombocytopenia (rarely requiring transfusions) –Higher incidence of stomatitis and cutaneous toxicity (that are rarely severe) –Lower incidence of hair loss and neurotoxicity

Preliminary conclusions (2) There was no statistically significant difference in response rate between carboplatin plus PLD and carboplatin plus paclitaxel Final analysis for the primary endpoint (PFS) will be performed as soon as the required number of events will be reached

All the patients and their families The Investigators and the staff at each participating center: S.Pignata, S. Greggi, C.Pisano – NapoliG.Scambia, D.Lorusso – RomaF.Cognetti, A.Savarese – Roma A.Veronesi, R.Sorio – AvianoV.Zagonel, E.Breda – RomaG.Ferrandina – Campobasso V. Gebbia, R.Agueli – PalermoC.Malzoni, A.Vernaglia – AvellinoL.Frigerio, L.Carlini – Bergamo M.Nardi, P.Del Medico – Reggio C.P.Musso – PalermoA.Febbraro, MC Merola – Benevento P.Scollo, G.Scibilia – CannizzaroE.Galligioni, V.Murgia – TrentoA.Gambi, S.Tamberi – Faenza A.Brandes, S.Rimondini – BolognaA.Ravaioli, E.Pasquini – RiminiN.Gebbia, MR.Valerio – Palermo E.Aitini, G.Cavazzini – MantovaD.Natale, C.Chiapperino – PenneF.Artioli, L.Scaltriti – Carpi V. Lorusso, A. Latorre – Bari / LecceAM.D’Arco, A. Fabbrocini – Nocera InfC.Gridelli, F.DelGaizo – Avellino B.Massidda, V.Pusceddu – CagliariS. De Placido, R. Lauria – NapoliG.Lelli, M.Marzola - Ferrara V.Fosser, R.De Vivo – VicenzaS.Tumolo, M.Boccalon - PordenoneG.Giardina, S.Danese – Torino G.Colucci, E.Naglieri – BariD. Amadori, N. Riva – ForlìA. De Matteis, E.Rossi – Napoli G.Lucarelli, G.Nettis – Acquaviva d.F.T.Gamucci,M.Giampaolo - FrosinoneS.Palazzo,R.Biamonte – Cosenza V.Montesarchio – NapoliA.Cardone, G.Balbi – NapoliG.Fasola, C.Sacco – Udine ML.Geminiani, V.Arigliano – BudrioO.Campos, I.Henriques – Coimbra, Portugal Coordinating center:F.Perrone, M.Di Maio, A.Morabito, E.De Maio, J.Bryce, G.Canzanella – Napoli Statistician center:C.Gallo, G. Signoriello, P.Chiodini, N.Lama - Napoli Acknowledgements