Reports

This report is based on medical evidence presented at sanctioned medical congress, from peer reviewed literature or opinion provided by a qualified healthcare practitioner. The consumption of the information contained within this report is intended for qualified Canadian healthcare practitioners only.

ABSTRACTS IN PERSPECTIVE based on presentations from the 44th Annual Meeting of the American Society of Clinical Oncology (ASCO)

Chicago, Illinois / May 30-June 3, 2008

EDITORIAL OVERVIEW:

Yoo-Joung Ko, MD, MMSc, SM, FRCPC

Division of Hematology/Medical Oncology, Sunnybrook Health Sciences, Centre, Toronto, Ontario

The landmark SHARP (Sorafenib HCC Assessment Randomized Protocol) study presented at ASCO 2007 demonstrated that a systemic therapy prolonged both overall survival (OS) and time to progression (TTP) in patients with advanced hepatocellular carcinoma (HCC) compared with placebo. Subsequent to these findings, sorafenib was approved both in Canada and the US for the treatment of advanced HCC. However, SHARP investigators cautioned that further research was needed in different ethnic groups and in patients with liver cancer with different etiologies.

Advanced HCC Subgroup Analyses

To study the influence of prognostic factors on treatment efficacy, a subset analysis of SHARP in patients with macrovascular invasion (MVI) and/or extrahepatic spread (EHS) was carried out by Dr. Morris Sherman, Associate Professor of Medicine, University of Toronto, Ontario, and colleagues. An overall response rate of 2.4% in MVI/EHS patients on sorafenib was observed vs. 0.9% on placebo, while disease stabilization was seen in 65.6% compared with 62.7%, respectively, for a disease control rate of 41.2% in the sorafenib group vs. 27.8% for placebo patients. Median TTP and OS for patients with MVI/EHS were 4.1 and 8.9 months and for the placebo cohort 2.7 and 6.7 months, respectively. Findings confirmed that the multikinase inhibitor prolonged OS and TTP regardless of the presence of MVI and/or EHS and that results in these patients are consistent with the overall SHARP cohort.

Further data on patients from the Asia-Pacific region were reported by Dr. Ann-Lii Cheng, National Taiwan University Hospital, Taipei. In this phase III study, over 70% of Asian patients with advanced HCC had hepatitis B infection as the cause of their liver cancer compared to 18% of the original SHARP cohort. They were also younger and had more tumour sites and lung metastases. Clinical benefit was of the same order of magnitude as in the SHARP trial, i.e. virtually identical with a prolonged OS of 47% (6.2 months vs. 4.1 months) and a median TTP of 74% (2.8 months vs. 1.4 months) vs. placebo (SHARP results: 44% and 73%, respectively). Despite the advanced stage of disease in these Asian-Pacific patients, sorafenib was equally well tolerated as in SHARP, confirming that it is a safe and effective treatment in different ethnic groups and in HCC arising from hepatitis B infection.

ECOG performance status (PS) is another important measure of the efficacy of treatment. Dr. Jean-Luc Raoul, Centre Eugène Marquis, Centre de Lutte Contre le Cancer, Rennes, France, provided evidence from an analysis of SHARP that patients with an ECOG PS of 1-2 respond as well to sorafenib as those with a PS of 0. In sorafenib-treated patients with a PS of 0, the disease control rate was 46.6% vs. 39.9% in those with a PS of 1-2, while median TTP and OS rates for patients with a PS of 0 were 5.5 months and 13.3 months and in those with a PS of 1-2, they were 5.3 months and 8.9 months, respectively. No notable differences in the distribution or frequency of drug-related adverse events such as diarrhea, fatigue, hand-foot reaction and hypertension between groups were observed and those that did occur were generally mild to moderate in severity.

In a cost-effectiveness model, Ms. Noemi Muszbek, Senior Research Associate, United BioSource Corporation, London, UK, determined that sorafenib is more expensive than best supportive care (BSC) in HCC because patients live longer and costs of providing BSC after progression are higher. Her Markov model determined that sorafenib is cost-effective as it falls within previously established thresholds deemed acceptable for society to pay ($50,000-$100,000 per life years gained).

Treatment Safety in Elderly Patients with Renal Cell Carcinoma

Randomized clinical trials usually include only highly selected patient groups and do not always reflect patients in real-world practice who may have multiple comorbidities. As noted by Dr. Ronald M. Bukowski, Cleveland Clinic Foundation, Ohio, the ARCCS (Advanced Renal Cell Carcinoma Sorafenib) expanded-access program in North America is a valuable study, as it included a broad range of patients who had not been studied in previous sorafenib trials and provides a realworld assessment of patients with advanced RCC.

In elderly patients over the age of 65, disease control rates at 84% were virtually identical to those under the age of 65 at 83%. Similar median progression-free survival (PFS) and OS were also observed between groups at 38.1 weeks and 46.6 weeks for patients >65 years old vs. 34.9 weeks and 50.3 weeks for those 65 years of age, respectively. Rates of grade 3 adverse events were low and comparable between the two age cohorts, the most common being hand-foot skin reactions and rash. Nor were there any differences in cardiacrelated events between the two age groups, although patients were not eligible for study entry if they had cardiac arrhythmias in excess of grade 1, active coronary artery disease, ischemia or uncontrolled hypertension.

Alternative to Resistant GIST

As an inhibitor of both angiogenesis and multiple kinases, sorafenib has several potent anti-cancer pathways through which to control tumour growth and may therefore be able to overcome resistance to other targeted agents. In a study led by Dr. Lauren Wiebe, University of Chicago, Illinois, patients with either imatinib-resistant or imatinib/sunitinib-resistant gastrointestinal stromal tumours (GIST) were treated with sorafenib for a median of four cycles. In updated data presented, of 29 patients, four (14%) achieved a partial response (PR) and disease stabilized in 18 patients (62%) for an overall disease control rate of 76%. Median PFS and OS in the imatinib-alone resistant group were 3.4 months and 13.6 months, respectively, while in the dual-resistance group, they were 5.7 months and 8.5 months, respectively.

This activity in highly resistant group of patients suggests that targeted therapies respond differently against different mutations, and tumour samples from this study are currently being analyzed to determine mutational status and response to sorafenib.

Improving Response Rates and Determining Clinical Activity in Other Cancers

While targeted therapies may confer good disease control, the induction of complete remissions—the critical pathway to potential cure—has proved an elusive goal. In order to induce more complete and durable responses in advanced malignancy, investigators are combining various targeted therapies. Dr. Jonathan Rosenberg, Assistant Professor of Medicine, University of California, San Francisco, explored such a combination involving sorafenib and everolimus, an mTOR inhibitor, in a phase I study. Twelve patients with metastatic RCC received standard-dose sorafenib and everolimus at either 2.5 mg/day or 5 mg/day. At the lowest everolimus dose there were two relatively durable PRs, one durable PR at the 5-mg dose, and disease stabilized in two others. No dose-limiting toxicity was reported at the 2.5-mg dose level and two grade 3 and 4 toxicities were observed at 5 mg.

Another indication of the feasibility of combination therapy in advanced malignancies was a drug interaction study of sorafenib and rapamycin (also known as sirolimus), another mTOR inhibitor, which indicated that neither agent caused any change in mean concentrations of the other ( Gangadhar et al., abstract 2545).

Dr. Robert Amato, Director, Methodist Hospital Genitourinary Oncology Program, Houston, Texas, reported that patients with metastatic RCC can achieve better clinical responses with doubled doses of sorafenib at 800 mg b.i.d. From a cohort of 21 treated patients, three achieved a CR, four had PRs, and disease stabilized in almost two-thirds of the rest. The majority of adverse events were grade 1 and 2 even at 800 mg b.i.d. without any significant cardiac toxicity observed.

Since patients with metastatic RCC become resistant to targeted agents over time, researchers are investigating whether a combination of these agents can improve the quality and duration of response. The effectiveness of various dose levels of sorafenib and bevacizumab was explored in a phase I/II study led by Dr. Jeffrey Sosman, Professor of Medicine, Vanderbilt University School of Medicine, Nashville, Tennessee. Across all dosing levels, investigators observed PRs in 52% of patients and stable disease in 38%. The investigators determined that the maximum tolerated dose consisted of sorafenib 200 mg q.d. plus bevacizumab 5 mg/kg every two weeks.

In metastatic or advanced gastric and gastroesophageal junction adenocarcinoma, the combination of sorafenib with docetaxel/cisplatin standard chemotherapy demonstrated a response rate among 44 patients of 40.9% and disease stabilized in 31.8% of the group, according to Dr. Weijing Sun, Chemistry Department, Iowa State University, Ames. Median PFS and OS were 5.8 months and 13.6 months, respectively. There were two deaths, one from infection and one from GI hemorrhage, that investigators felt might have been related to the protocol. Still, the triplet clearly had very encouraging activity and the investigators concluded that it merits further exploration in this difficult-to-treat malignancy.

Finally, the role of sorafenib is being explored in advanced soft-tissue sarcoma, including vascular sarcomas and highgrade leiomyosarcoma and liposarcoma, as it has shown some indication of activity in these malignancies.

4584 Efficacy and Safety of Sorafenib in Patients with Advanced Hepatocellular Carcinoma and Vascular Invasion or Extra Hepatic Spread: A Subanalysis from the SHARP Trial M. Sherman, V. Mazzaferro, D. Amadori, J. Seitz, M. Moscovici, M. Shan, A. Nadel, D. Voliotis, J. M. Llovet, J. Bruix, on behalf of the SHARP investigators study group

Background: Liver cancer is the third leading cause of cancer death globally, with most deaths occurring within one year of diagnosis. The multikinase inhibitor sorafenib became the first treatment approved by the FDA and EMEA for the treatment of HCC based on data from the phase III Sorafenib HCC Assessment Randomized Protocol (SHARP) trial, which demonstrated that sorafenib 400mg bid increased overall survival (OS) vs placebo in patients with advanced HCC. The influence of risk factors such as macroscopic vascular invasion (MVI) and/or extrahepatic spread (EHS), on the safety and efficacy of sorafenib has not previously been reported. Here we compare sorafenib versus placebo in patients with and without MVI and/or EHS.

Methods: Patients with advanced, measurable HCC, Eastern Cooperative Oncology Group performance status (ECOG PS) 0-2 and Child-Pugh status A received either sorafenib (400mg bid) or placebo. Endpoints included OS, time to progression (TTP; based on independent tumor assessment), and safety. Presence of MVI was determined through radiological assessment.

Results: In this analysis, 421 patients had MVI and/or EHS (sorafenib=209, placebo=212) and 181 (sorafenib=90, placebo=91) did not. For patients with MVI and/or EHS (sorafenib vs placebo), OS was 8.9 vs 6.7 months (HR: 0.77; 95% CI: 0.60, 0.99) and TTP was 4.1 vs 2.7 months (HR: 0.64; 95% CI: 0.48, 0.84). For patients without MVI and/or EHS, OS was 14.5 vs 10.2 months (HR: 0.52; 95% CI: 0.32, 0.85) and TTP was 9.6 vs 4.3 months (HR: 0.40; 95% CI: 0.23, 0.70). The most common grade 3-4 drug-related, treatment-emergent adverse events (AEs) in the sorafenib arm for patients with vs those without MVI and/or EHS were: diarrhea (5.3 vs 15.7%), hand-foot skin reaction (6.7 vs 10.1%) and fatigue (4.8 vs 1.1%).

Conclusions: Sorafenib prolonged OS and TTP in patients with advanced HCC irrespective of presence or absence of MVI and/or EHS; furthermore, sorafenib was well tolerated and AEs were manageable. These data are consistent with those in the overall SHARP population and confirm that the survival benefit achieved with sorafenib extends to a wide range of patients with advanced HCC, including those with MVI and/or EHS.

Commentary on abstract 4584

Vascular invasion and extrahepatic spread (EHS) are major determinants of outcomes in patients with HCC. Initial results from SHARP included all patients; in the SHARP subanalysis, investigators focused on those patients having either macrovascular invasion (MVI) or EHS at baseline and assessed whether responses to treatment differed from the group overall. The overall response rate in MVI/EHS patients treated with sorafenib was 2.4% vs. 0.9% in placebo patients. Disease stabilized in 65.6% of patients with either MVI or EHS compared with 62.7% of placebo patients, for an overall disease control rate of 41.2% in MVI/EHS patients on sorafenib vs. 27.8% for placebo controls. Median TTP for patients with MVI/EHS was 4.1 months vs. 2.7 months for placebo patients, while median OS for sorafenib MVI/EHS patients was 8.9 months vs. 6.7 months for the placebo arm. There were no notable differences in the distribution or frequency of drug-related adverse events between patients with and without MVI/EHS and the overall SHARP population. This analysis thus confirmed that the efficacy and safety of the multikinase inhibitor in advanced HCC was maintained in patients who had MVI/EHS at baseline.

Questions and Answers with Dr. Morris Sherman, Associate Professor of Medicine, University of Toronto, Ontario

Q: The multikinase inhibitor sorafenib appears to work across all types of patients and in advanced liver disease as well as less advanced disease. Was this surprising?

A: No, it’s not really surprising because the mechanism of action is probably independent of all of these different factors. Sorafenib inhibits all of these various metabolic pathways in the tumour cell and these pathways exist, whether you have more advanced or less advanced disease, whatever your PS is, and so on.

Q: Do patients without MVI/EHS have a survival advantage over those who do?

A: The data suggest but do not prove that patients without MVI/EHS have a better survival advantage than those with MVI/EHS, because the hazard ratio in the non-MVI/EHS group was lower than the hazard ratio in the MVI/EHS-positive group. To confirm this, additional comparisons would be needed.

6527 Economic Evaluation of Sorafenib vs. Best Supportive Care in Hepatocellular CarcinomaN. Muszbek, S. Shah, S. M. Carroll, B. I. Carr, K. Gondek

Background: A randomized phase III trial of sorafenib vs. placebo in hepatocellular carcinoma (SHARP study) demonstrated that sorafenib significantly prolonged overall survival (OS) compared to placebo (median OS = 324 days [95% CI: 286-405] vs. 241 days [95% CI: 206-276]; p<0.001). Sorafenib is the first and only systemic agent to demonstrate survival benefit over the past 30 years. Given the clinical significance of these results, the objective of this study was to demonstrate the cost-effectiveness associated with sorafenib compared to best supportive care (BSC) from the US third party payer perspective.

Methods: A Markov model was developed to evaluate the cost-effectiveness of sorafenib vs. BSC. The model followed survival and time to progression (TTP) in monthly cycles based on the extrapolation of patient level trial data over a patient's lifetime. Health effects were expressed as life years gained (LYG). Resource use included drugs, physician visits, laboratory tests, scans, hospitalizations and treatment of adverse events. Unit costs were gathered from diagnosis related groupings, Medicare charges and the Red Book and were expressed in 2007 US dollars. Costs and effects were evaluated over a lifetime and discounted at 3%. Results were presented as incremental cost/LYG. Deterministic and probabilistic sensitivity analyses were conducted.

Results: LYG was longer for sorafenib compared to BSC (1.59±0.18 vs. 1.06±0.10 LYG/patient for sorafenib and BSC, respectively). The lifetime total costs were $41,782±3,292 for sorafenib and $8,161±1,408 for BSC. The results indicated an incremental cost-effectiveness ratio of $64,301/LYG. The key drivers of the model results were OS, TTP and BSC costs after progression. Sensitivity analyses results showed that the model was robust.

Conclusions: The economic evaluation indicates that sorafenib is cost-effective as compared to BSC, with cost-effectiveness ratios within the established threshold that society is willing to pay (i.e. $50,000- $100,000).

Commentary on abstract 6527

Results from the SHARP trial showed that sorafenib significantly increased OS and TTP in advanced HCC patients, the first systemic therapy ever to provide such advantages. With limited resources available in health-care, it is also important to see if a treatment provides value for money. In this cost-effectiveness analysis, researchers compared sorafenib to best supportive care (BSC) as no other systemic therapy is approved for advanced HCC. The health effects of treatment were expressed as life-years gained (LYG), the total discounted number of years of life with the different treatment choices. Per patient over a lifetime, sorafenib was more effective compared to BSC with a longer LYG/patient at 1.59 vs. 1.06 LYG/patient for BSC. Improvements in effectiveness were associated with a higher total cost, at an incremental costeffectiveness ratio of $63,436/LYG. The key cost drivers were OS, TTP and BSC costs after progression. Researchers concluded that sorafenib is cost-effective vs. BSC, as its cost-effectiveness ratios are within thresholds of $50,000 and $100,000 that society is generally willing to pay.

Questions and answers with Noemi Muszbek, MSc, Senior Research Associate, United BioSource Corporation, London, UK

Q: Could you clarify how sorafenib was more effective than BSC in terms of LYG?

A: The LYGs estimated for both sorafenib and BSC were based on the SHARP study. However, as the SHARP trial was halted early due to the significant OS results observed at the interim analysis, it lasted only 72 weeks. In the cost-effectiveness study, we were interested in the effectiveness and cost consequences of sorafenib over a patient’s lifetime and to include all effects and costs, results had to be extrapolated beyond the trial period using estimated survival functions that best fit the patient-level data from SHARP. Using this extrapolation, mean life-years per patient with BSC was 1.06 years while with sorafenib it was 1.59 years. Thus compared to BSC, sorafenib saved 0.53 life-years per patient.

Q: The higher efficacy of sorafenib vs. BSC was associated with a higher total cost, but was the incremental cost-effectiveness ratio in keeping with accepted thresholds of cost-effectiveness?

A: There are no official thresholds in the US; however, different thresholds have been calculated based on the amount society is willing to pay for a LYG or a quality-adjusted LYG. The most frequently cited figure for threshold is $50,000-$100,000. As the cost of an additional LYG with sorafenib is within these figures, sorafenib can be considered cost-effective.

Q: What are your main conclusions from this analysis?

A: Based on the cost-effectiveness model over a patient’s lifetime, sorafenib is cost-effective in the treatment of patients with unresectable HCC. Thus, as a survival-enhancing agent, sorafenib cannot only help fill an unmet need and extend OS, but within conventional thresholds, is also cost-effective.

4509 Randomized Phase III Trial of Sorafenib vs. Placebo in Asian Patients with Advanced Hepatocellular Carcinoma A. Cheng, Y. Kang, Z. Chen, C. Tsao, S. Qin, J. Kim, K. Burock, J. Zou, D. Voliotis, Z. Z. Guan

Background: In Asian countries, the incidence of HCC is higher than in Western nations, partly attributed to a greater prevalence in Asia of infection with hepatitis B virus - a major HCC risk factor. Sorafenib (Sor), a multikinase inhibitor with anti-angiogenic and proapoptotic activities, significantly increased overall survival (OS) in a randomized, placebo (Pl)-controlled, phase III trial of patients with HCC (SHARP). Here we report efficacy and safety results of a phase III trial of Sor in HCC patients from the Asia-Pacific region.

Methods: Patients with advanced, measurable HCC, ECOG PS 0-2, Child-Pugh status A, no prior systemic therapy for HCC, and life expectancy >12 wks were randomized 2:1 to Sor 400mg bid or Pl. Endpoints were OS, time to progression (TTP), time to symptomatic progression (FSHI8-TSP), disease control rate (DCR; defined as complete or partial response, or stable disease according to RECIST, maintained for >28 days from first demonstration) and safety. No primary endpoint was defined.

Results: 226 patients were randomized; Sor n=150, Pl n=76. Baseline characteristics were similar between the two groups (Sor vs Pl): mean age 51±13 vs 52±12 years, males 85 vs 87%. Efficacy results for time-to-event endpoints are shown in the table. DCR was 35% (95% CI 28-44) and 16% (95% CI 8-26) in the Sor and Pl groups. Most frequently reported grade >3 drug-related AEs were (Sor vs Pl): hand-foot skin reaction (10.1 vs 0.0%), diarrhea (6.0 vs 0.0%), hyperbilirubinemia (3.4 vs 2.7%) and fatigue (3.4 vs 1.3%). Drug-related serious AEs were observed in 13 (9%) and 1 (1%) patients (Sor vs Pl).

Conclusions: Sor significantly prolonged OS and TTP in HCC patients in the Asia-Pacific region compared with Pl, and was well tolerated. These results mirrored those of SHARP, despite Asia- Pacific patients having more advanced disease based on ECOG PS, number of tumor sites and lung metastases. These findings confirm the survival advantage of Sor, and demonstrate its broad applicability in Asian patients with advanced HCC.

Commentary on abstract 4509

The landmark SHARP trial reported at ASCO last year showed that for the first time, treatment of advanced HCC with sorafenib 400 mg b.i.d. prolonged OS by 44% and median TTP by 73% over placebo. On the strength of this study, it became the first approved systemic treatment for advanced HCC. However, investigators cautioned that their results might not apply to liver cancer from other etiologies or to other ethnic groups. Some of these concerns were addressed by the Asia-Pacific (AP) study. With over 70% of AP patients having hepatitis B (HBV) as the predominant etiology, the agent still improved median OS by 47% and median TTP by 74%, virtually identical findings as those from SHARP. Disease control rates were also significantly higher at 35% in the sorafenib arm vs. 16% for placebo patients. Since inclusion and exclusion criteria for both the SHARP and AP studies were identical, findings confirm the safety and efficacy of the agent in patients with HCC across geographic regions around the world.

Questions and Answers with Dr. Ann-Lii Cheng, National Taiwan University Hospital, Taipei

Q: These were young patients with Child-Pugh Class A disease and good PS. Can you explain why the median OS in the placebo arm was only 4.2 months?

A: I think this clearly indicates that doctors in the AP regimen enrol their patients with more advanced stages of liver disease, so it is simply a reflection of practice patterns in this part of the world. Almost half of the patients also had an ECOG PS of 1 and they were quite symptomatic, so it was not a good prognosis population of patients.

Q: When you compare AP patients with those in SHARP, what were the main differences?

A: Over 70% of patients in AP had HBV while just 8% of them had hepatitis C (HCV), whereas in SHARP, 18% had HBV and 28% had HCV. Patients in AP also had more symptomatic disease, more EHS, more BCLC stage C disease, a greater number of tumour sites and more lung metastases. So AP patients were at a more advanced stage of their disease than those in SHARP.

Q: How much pretreatment did the patients have?

A: Exactly 69% of patients had received prior therapy but no prior systemic therapy; any patient who had received prior systemic therapy was excluded from the study. So patients had received some kind of local therapy for their primary disease.

Commentary on abstract 4587

The SHARP study in advanced HCC patients demonstrated that sorafenib was well tolerated with almost no patient experiencing serious side effects. This is particularly important for patients who have a poor PS while receiving treatment, for whom standard chemotherapy can prove very difficult to tolerate. In a subanalysis of responses to sorafenib among SHARP patients with an ECOG PS 1-2 vs. those with a PS of 0, the disease control rate was 46.6% among PS 0 patients and 39.9% in PS 1-2 patients. Median TTP for PS 0 patients was 5.5 months compared with 5.3 months for PS 1-2 patients, while median OS for PS 0 patients was 13.3 months compared with 8.9 months for PS 1-2 patients. Slightly higher rates of serious drug-related adverse events were seen in PS 1-2 patients at 17.5% vs. 10% in PS 0 patients. There were no notable differences in the distribution or frequency of drug-related adverse events between PS 0 and PO 1-2 patients and those that did occur were generally of mild to moderate severity. Results of this subanalysis confirm the utility of this agent in a broad range of patients with HCC.

4587 Efficacy and Safety of Sorafenib in Patients with Advanced Hepatocellular Carcinoma According to ECOG Performance Status: A Subanalysis from the SHARP Trial J. Raoul, A. Santoro, M. Beaugrand, J. A. Marrero, M. Moscovici, M. Shan, A. Nadel, D. Voliotis, J. Bruix, J. M. Llovet

Background: Most patients with HCC present with advanced disease; until recently, there were no proven treatment options in this setting. In late 2007, the FDA and EMEA granted approval of sorafenib, an oral multikinase inhibitor, in the treatment of advanced HCC following demonstration of a significant overall survival (OS) benefit in the phase III SHARP trial. SHARP recruited patients with ECOG PS 0-2, but the influence of PS on safety and efficacy of sorafenib has not previously been reported. Here we report a subanalysis of SHARP, analyzing the effect of ECOG PS on outcomes in this trial.

Methods: Patients with advanced measurable HCC, ECOG PS 0-2 and no cirrhosis or Child-Pugh status A, received either sorafenib (400mg bid) or placebo. Treatment groups were stratified according to region, ECOG PS, and extrahepatic spread or vascular invasion. Endpoints included OS, time to progression (TTP; based on independent tumor assessment), and safety. Patients were grouped according to ECOG PS 0 or 1-2 for this analysis.

Results: A total of 602 patients were eligible, 325 (sorafenib=161, placebo=164) had ECOG PS 0 at baseline and 277 (sorafenib=138, placebo=139) had PS 1-2. Among patients with PS 0, median OS for sorafenib vs placebo was 13.3 vs 8.8 months (HR: 0.68; 95% CI: 0.50, 0.95) and median TTP was 5.5 vs 2.9 months (HR: 0.55; 95% CI: 0.40, 0.77). Among patients with PS 1-2, median OS for sorafenib vs placebo was 8.9 vs 5.6 months (HR: 0.71; 95% CI: 0.52, 0.96) and median TTP was 5.3 vs 2.8 months (HR: 0.61; 95% CI: 0.42, 0.88). There were no notable differences in the adverse event (AE) profile of sorafenib in patients with PS 0 compared with PS 1-2. The incidence rates for grade 3-4 drug-related AEs in the sorafenib arm for patients with PS 0 vs PS 1-2 were: diarrhea (8.8 vs 8.0%), fatigue (2.5 vs 5.1%), hand-foot skin reaction (8.8 vs 6.6%), hypertension (1.9 vs 1.5%) and pain (any event) (3.1 vs 2.9%).

Conclusions: Compared with placebo, sorafenib prolonged OS and TTP in patients with advanced HCC irrespective of ECOG PS 0 or 1-2. Treatment was well tolerated. Overall, these data suggest that sorafenib extends survival in HCC patients with ECOG PS 1-2 in addition to PS 0, and further support sorafenib as the new standard of care in this setting.

Commentary on abstract 5122

This study was undertaken to investigate whether higher sorafenib doses were associated with improved clinical outcomes. Twenty-two patients with clear-cell RCC and progressive metastatic disease received sorafenib 400 mg b.i.d. on days 1 to 28, followed by 600 mg b.i.d. on days 29 to 56 and then 800 mg b.i.d. from day 57 onward. Doses were modified if patients developed grade 3 or 4 toxicity and treatment was discontinued if it proved intolerable or disease progressed. A CR was achieved in three patients (14%), a PR in four patients (19%) and disease stabilized in 13 others (62%). Median PFS was 7.7 months. The majority of adverse events were grade 1 and 2, even at double the standard dose, and no significant cardiac toxicity was observed, even on repeat echocardiography. Interestingly, there was no statistical difference in plasma concentrations between the three doses tested. The fact that the main metabolic pathways remained the same at all three doses was confirmed by the constant relative proportion of the different metabolites at all three doses.

5122 A Phase II Trial of Intra-patient Dose-escalated Sorafenib in Patients with Metastatic Renal Cell Cancer R. J. Amato, J. Jac, P. Harris, M. Dalton, S. Saxena, F. Monzon, J. Zhai, J. Brady, J. P. Willis

Background: In an earlier phase II study ( J Clin Oncol. Vol 25, No. 18S, 2007: Abstract 5026), we were able to show that pts who were able to tolerate doses higher than the FDA-approved dose have significantly improved clinical outcomes, including remission, tumor shrinkage, growth arrest and will have lesser side effects over time. We have expanded the phase II study to determine if an increase in dose will result in higher AUC and Cmax level. Response rate, progression free survival (PFS) and overall survival (OS) was assessed.

Methods: Pts had a confirmed component of clear cell RCC, progressive measurable metastatic disease, adequate organ/marrow function, and no more than 1 prior therapy. The initial dose of sorafenib was 400 mg PO, BID, daily days 1 to 28. Dose escalation was defined as; on day 29 through day 56, an increase to 600 mg BID and on day 57 throughout as tolerated 800 mg BID daily. Dose modifications were performed for toxicity per the National Cancer Institute Common Toxicity Criteria Version 3.0. Pts were evaluated every 2 cycles (8 weeks) using RECIST. PFS and OS were determined from entry into the study. Blood and urine samples are obtained on days 28, 56, and 84.

Results: Twenty-two pts were enrolled to date, 19 are evaluable for a response and toxicity, and 3 for toxicity only. One pt had a complete response (CR), 5 pts partial response (PR) and 7 pts stable disease (SD) for 3 months or more. One pt had progressive disease and it is still too early to assess 5 pts. Most common treatment related adverse events; hand/foot syndrome, skin rash, diarrhea, alopecia, fatigue and hypertension. Noted laboratory abnormalities; hypophosphatemia.

Conclusions: 14/14 pts were escalated to 1,600 mg per day. Dose escalated sorafenib has promising anti-tumor activity in pts with MRCC as demonstrated by a 32% CR/PR rate and is further suggested by a prolonged PFS >3 months for an additional 50% of pts. Intra-patient dose escalation data in association with anti-tumor activity and toxicity in correlation with pharmacokinetic studies will be presented, in addition to an update of the ASCO 2007 data.

5045 Safety and Efficacy of Sorafenib in Elderly Patients >65 Years: A Subset Analysis from the Advanced Renal Cell Carcinoma Sorafenib (ARCCS) Expanded Access Program in North America R. M. Bukowski, W. M. Stadler, R. A. Figlin, J. J. Knox, N. Gabrail, D. F. McDermott, L. Cupit, W. H. Miller, J. D. Hainsworth, C. W. Ryan, Advanced Renal Cell Carcinoma Sorafenib (ARCCS) Expanded Access Program in North America

Background: Sorafenib, an orally active multikinase inhibitor, has been approved by the US Food and Drug Administration for the treatment of advanced RCC. Prior to the drug's approval for RCC, the North American ARCCS expanded access program was initiated to make sorafenib available to patients (pts) who were not eligible for or did not have access to sorafenib clinical trials in the United States and Canada. Because sorafenib is likely to be used in elderly pts with advanced RCC, we evaluated its safety and efficacy profile in a subset of pts <u>></u>65 years (yrs) who enrolled in the ARCCS program.

Methods: Eligible pts were <u>></u>15 yrs of age with advanced RCC, Eastern Cooperative Oncology Group performance status of 0-2 (with waivers granted for pts with status 3-4), and adequate prior therapy of brain metastases. Major exclusion criteria included treatment with other investigational drugs in the 4 weeks prior to enrollment; life expectancy <2 months; active coronary artery disease, ischemia or hypertension; and renal impairment requiring dialysis. Pts received oral sorafenib 400 mg bid as a single agent until disease progression or intolerance.

Results: Of 2,504 pts assessed for safety, 1,135 (45%) pts were <u>></u>65 yrs. Grade 2 and Grade <u>></u>3 adverse events (AEs) observed in <u>></u>5% of pts <u>></u>65 yrs and <65 yrs are listed in the table. Grade <u>></u>3 cardiacrelated AEs were reported in 8% of pts <u>></u>65 yrs and in 7% of pts <65 yrs. Of the 762 pts >65 yrs of age who were evaluable for response, partial response was reported in 3% and stable disease in 81%.

Conclusions: In the ARCCS program, sorafenib 400 mg bid was associated with similar rates of clinically manageable side effects in elderly pts with advanced RCC as compared to younger pts. Response rates to sorafenib in elderly pts were comparable to previously reported results for total ARCCS pts and phase III data
bid in advanced RCC.

<img1977|center>

Commentary on abstract 5045

The ARCCS study was an expanded-access, open-label treatment protocol for use of sorafenib in patients with advanced RCC. Enrolment criteria were less strict than usual for clinical trials and so patient responses were more realistic in terms of how patients in general oncology practice would respond to the same treatment. In a subset analysis, disease control rates were 84% in those 65 years of age and older and 83% for those under 65 years of age (CR+PR+SD). Median PFS at 38.1 weeks was similar for patients over the age of 65 and those 65 years of age and younger, at a median PFS of 34.9 weeks. OS rates were also similar for those over the age of 65, at a median of 46.6 weeks, vs. a median of 50.3 weeks for patients 65 years of age and younger. Rates of grade 3 and higher adverse events were also low and comparable between the two age cohorts, the most common being hand-foot skin reactions and rash. Importantly, there were no differences in cardiac-related events in the two age groups. These data suggest that response to the agent and safety are similar regardless of age.

Questions and answers with Dr. Jennifer Knox, Assistant Professor of Medicine, University of Toronto, Ontario

Q: Why did investigators feel it was necessary to exclude patients with active coronary artery disease, cardiac ischemia or uncontrolled hypertension from this study?

A: If patients are unstable at the time treatment is initiated, there is the potential for exacerbating hypertension or theoretical problems with cardiac function. Anybody who is unstable should not be started on these drugs, that is just good clinical practice. But patients in this study would have had comorbidities, so if they were stable, we could treat them.

Q: Why are the toxicity data from ARCCS important?

A: The toxicity assessment in this study was very thorough and the number of patients involved in ARCCS was huge, including a high proportion of elderly patients (n=1135). This is the reason we keep bringing the study back to talk about it. The toxicity assessment is quite reassuring, and is telling us that the drug is consistently safe in a more generalized oncology practice and indeed safe as well in older patients.

Q: Why is this an important study?

A: The expanded access program has produced data that is very consistent with data from TARGET [Treatment Approaches in Renal Cancer Global Evaluation Trial], where patients were much more tightly selected, so I think most people see this data as being quite reassuring, as sorafenib does what it says it does and is quite safe.

5109 Phase I Study of Sorafenib and RAD001 for Metastatic Clear Cell Renal Cell Carcinoma J. E. Rosenberg, V. K. Weinberg, C. Claros, C. Ryan, A. M. Lin, L. Fong, D. Brocks, E. J. Small

Background: Both RAD001 and sorafenib have activity against advanced clear cell renal cell carcinoma (ccRCC). Simultaneous angiogenesis and mTOR inhibition may lead to improved outcomes. A phase I trial combining sorafenib and RAD001 was undertaken.

Methods: Cohorts of 3 or 6 patients with ccRCC and good organ function were treated with a 7 day runin period of sorafenib 400 mg PO BID continuously followed by RAD001 (dose level I: 2.5 mg; dose level II: 5 mg) PO QD and sorafenib 400 mg PO BID continuously. Pharmacokinetic sampling of sorafenib was obtained on day -1, and of both RAD001 and sorafenib at day 15 of combination therapy. Doselimiting toxicity (DLT) was defined as occurring within the first 28 days of therapy.

Results: Eleven patients (8 male, 3 female; 3 good and 7 intermediate MSKCC risk) with a median age of 64 (range 51-75) were enrolled. One patient was noncompliant with therapy and was not evaluable for response or dose limiting toxicity (DLT) evaluation. None of the 6 patients in dose level 1 experienced a DLT. Two of 4 evaluable patients treated at dose level II experienced protocol-defined DLT's (grade 4 uric acid, and grade 3 lipase with grade 2 pancreatitis). Other grade 3 or greater toxicities over all cycles related to therapy included lymphopenia, syncope, and hypophosphatemia for dose level 1, and hypophosphatemia, diarrhea, edema, hypoalbuminemia, hyponatremia, dental infection, and photosensitivity for dose level 2. Grade 2 rash was observed in 6 patients, and grade 2 hypertension was observed in 4 patients. Independently-reviewed best objective responses in 10 evaluable patients include 3 confirmed partial responses, 2 stable disease (6+ and 9 months), and 5 progressive disease. Steady state dosing of RAD001 demonstrated dose-proportional increases in AUC between dose levels, and no evidence of a pharmacokinetic interaction between RAD001 and sorafenib.

Conclusions: Combination therapy of sorafenib 400 mg BID is safe and feasible in combination with RAD001 2.5 mg daily. Additional patients are being accrued to dose level II to further explore toxicity. No clinically relevant pharmacokinetic interaction was observed. Activity of the combination was observed, and deserves further evaluation in a phase II study.

Commentary on abstract 5109

Clear-cell RCC is the most common type of renal cell cancer and has become the target for a number of novel agents in the metastatic setting. As the effects of a multikinase inhibitor such as sorafenib and the mTOR inhibitor everolimus may be synergistic rather than additive, such combinations are being explored. In this study, investigators determined the maximum tolerated dose and dose-limiting toxicities of sorafenib plus everolimus in 12 patients with metastatic RCC. Three dose levels of everolimus were planned: 2.5 mg/day, 5 mg/day and 10 mg/day, while sorafenib was given as 400 mg b.i.d. in all treatment arms. There were two relatively durable PRs at dose level 1, one durable PR at dose level 2 and disease stabilized in two others. Results for dose level 3 were not reported. There were no dose-limiting toxicities at dose level 1, but at dose level 2 there was one grade 4 hyperuricemia and one grade 3 lipase. Asymptomatic hypophosphatemia was also frequent. Patients are continuing to be accrued to the 5-mg arm to further explore toxicity at this level.

Questions and Answers with Dr. Jonathan Rosenberg, Assistant Professor of Medicine, University of California, San Francisco

Q: Were you concerned about drug interactions with this particular combination?

A: Not really, because while everolimus is a major substrate of the CYP3A4 pathway, sorafenib is not metabolized by the same pathway so clinically significant drug interactions would be less likely. And so far, we haven’t seen any drug interactions.

Q: Were you surprised by the responses you did see?

A: It’s very premature to talk about response rates in a phase I study. However, we did see responses and probably more than we expected with either sorafenib or with [everolimus], so it is certainly provocative, to say the least. Whether or not it means there is true synergy between the two drugs we can’t tell from this data yet.

Q: What are you planning to do next?

A: We are expanding the 5-mg cohort and we’re hoping to see whether that is the tolerable dose of [everolimus] or not. If it is, there are two natural places to take it: one is in front-line combination and the other is second-line therapy.

Commentary on abstract 5011

Patients with metastatic RCC develop uniform resistance to targeted agents. Combination targeted strategies are being pursued in an effort to induce more complete and durable remissions. The combination depends in part on the agent’s adverse-event profile, as those with overlapping toxicities may prove too toxic to be used together. A combination that appears to have significant activity in the setting of metastatic RCC is sorafenib and bevacizumab. In this phase I/II study, various dose levels and dose schedules of both agents were tested. The maximum-tolerated dose appeared to be a combination of sorafenib 200 mg q.d. plus bevacizumab 5 mg/kg every two weeks. On this dose, patients were able to receive multiple cycles of therapy with no dose reductions. Across all dosing levels, investigators saw PRs in 25 out of 48 patients (52%), while disease stabilized in 18 others (38%). Tumours shrank in almost all patients (92%) and 11 patients remained on treatment between 13 and 32 months after study entry. Median PFS was 14 months. This combination appears to have good clinical activity in patients with metastatic RCC. Bevacizumab use increased known toxicities from sorafenib but only four out of 48 patients discontinued treatment.

5011 Updated Results of Phase I Trial of Sorafenib and Bevacizumab in Patients with Metastatic Renal Cell Cancer J. A. Sosman, K. T. Flaherty, M. B. Atkins, D. F. McDermott, M. L. Rothenberg, W. L. Vermeulen, K. Harlacker, A. Hsu, J. J. Wright, I. Puzanov

Background: We conducted a Phase I trial of S and B to study inhibition of the vascular endothelial growth factor (VEGF) pathway at both the ligand (B) and receptor (S), and platelet derived growth factor receptor (S) in mRCC.

Methods: Pts with measurable mRCC, adequate organ function, and PS 0-1 were eligible. Cohorts of at least 6 pts were enrolled to define the MTD and DLT of B given IV q 2wks and S po QD on 28-day cycles. Response and toxicity were assessed after 2 cycles. Initial doses were S at 200mg BID and B at 5 mg/kg with modification based on observed toxicities. High dose pyridoxine was administered in some cohorts attempting to ameliorate hand-foot syndrome (HFS) symptoms.

Results: 47/48 pts completed their 1st response evaluation. Median age was 61 yrs (35-83 ); M/F: 40/8; PS: 0/1= 32/16; 41 clear cell, 7 non-clear cell; 41 had prior nephrectomy; 6 prior IFN and 9 prior IL-2. The table summarizes dose levels explored and toxicities. Toxicities included PUD, rash, weight loss, proteinuria, hypertension, and HFS with no evidence of microangiopathy. Three pts had late vascular events after >12 mos of treatment (TIA, AMI). Pyridoxine did not ameliorate HFS. Only 4/48 pts stopped therapy due to toxicities. MTD was S 200 mg PO QD and B 5 mg/kg IV Q 2 wks. 46/48 pts were evaluable for response. Twenty-one of 46 (46%) had PR by RECIST, including 2 pts with sarcomatoid RCC. Additionally, 23 pts had SD and 1 PD at 1st eval. with 1 dying from disease progression prior to 1st eval. Median time to progression (TTP) was 11.2 months with 10 pts (21%) progressionfree at 18 mos.

Conclusions: While toxicities led to a MTD with lower doses of each agent, B + S had impressive antitumor activity based on both OR and TTP. B appears to significantly increase S-related toxicity, especially HFS. We have opened a Phase II trial with correlative and PK studies for S and B to investigate the mechanism of enhanced S-related toxicity. The combination has been incorporated into E2804 (BeST) a multi-arm Phase
n regimens. The study was supported by U01 CA099177 and CTEP-NCI.

<img1978|center>

Commentary on abstract 2545

Both sorafenib, a multikinase inhibitor, and rapamycin (also known as sirolimus), an mTOR inhibitor, have activity in similar tumour types and may therefore be synergistic. Investigators thus assessed a drug-drug interaction study in patients with solid tumours and progressive disease for which no standard therapy exists. Twenty patients took part in the study. Arm A received sorafenib 400 mg b.i.d. days 1 to 28 and rapamycin 3 mg/day was added on days 15 to 28 in the first cycle. Arm B received rapamycin 3 mg/day from days 1 to 14 and then the combination of both drugs from day 15 onward. At a median of two 28- day cycles, stable disease occurred in 45% of patients. Grade 3 and 4 toxicity did occur but was relatively limited, most commonly diarrhea at 20% and rash at 15%. With no significant drug interaction between rapamycin and sorafenib, and no unexpected toxicity, investigators suggested that this combination could be safely administered and warrants further study.

2545 A Drug Interaction Study of Sorafenib (S) and Rapamycin (R) in Patients with Advanced Malignancies T. Gangadhar, E. E. Cohen, L. Janisch, L. K. House, S. D. Undevia, G. F. Fleming, M. L. Maitland, J. Ramirez, M. J. Ratain

Background: R inhibits the mammalian target of rapamycin (mTOR) and is active in preclinical tumor models. S inhibits multiple kinases. S and mTOR inhibitors have activity in similar tumor types and may be synergistic.

Methods - Design: A drug-drug interaction study of oral S and R in patients with PS 0-2 and histologically confirmed solid tumors with progressive disease after standard therapy or for which no standard therapy exists. Patients were assigned to one of two treatment arms in sequence. Arm A: S 400 mg bid starting day (D) 1, then S + R 3 mg qd starting D15. Arm B: R 3 mg qd starting D1, then R + S 400 mg bid starting D15.

Primary endpoint: determine whether a significant pharmacokinetic interaction exists between R and S. Samples for pharmacokinetic analysis were collected on D1, 14, 15 and 29 (A) or D1, 8, 15, 22 (B). Paired t tests were used to compare mean drug concentrations.

Results: 20 patients were enrolled from 4/07- 10/07. Median age 54 (range 36-74); 55% men; PS 0: 60%, 1: 40%. 58 4-week cycles were administered (median 2; range 1-8). Grade 3 toxicity (% patients): rash 15%, hypophosphatemia 10%, diarrhea 20%, hypertension 5%, bleeding 5%, vomiting 5%. 2 patients died: 1 from infection, the other by arrhythmia. Pharmacokinetics: 6 patients per arm were evaluable. Initiation of R caused no significant change in the mean concentration of S at similar timepoints on D14 vs. D15: (7664, 5704 ng/mL; p=0.20) or D15 vs. D29: (5704, 5350 ng/mL; p=0.68). Initiation of S caused no change in the mean concentration of R on D8 vs. D15 (14.6, 16.4 ng/mL; p=0.43) or D15 vs. D22: (16.4, 16.1 ng/mL; p=0.91). The mean concentration R 20 minutes after drug was lower on D22 vs. D15: (9.1, 15.1 ng/mL; p=0.01) but there was no difference at 5 hours. Best response was stable disease in 45% of patients enrolled.

Conclusions: S lowered the concentration of R at 20 minutes, which may merit further study. No other interactions between R and S were observed. Combination therapy was well tolerated and may be active in advanced cancer.

Commentary on abstract 10532

Soft-tissue sarcomas are a heterogeneous group of diseases and there are few treatment options when patients reach advanced stages. Investigators studied certain sarcomas by targeting the VEGF pathway. These include vascular sarcomas and high-grade leiomyosarcoma and liposarcoma, all of which were metastatic, unresectable or recurrent in this study. Of the 38 patients who received standard-dose sorafenib, 20 had a leiomyosarcoma, 10 had a liposarcoma and eight had vascular sarcomas. The median duration of treatment was three cycles, during which 24% of patients required dose modification during the first cycle. There were no CRs and no PRs but disease stabilized in nine patients with leiomyosarcoma, two patients with liposarcoma and six with vascular sarcoma. The median OS was 17 months, while the median PFS was three months. Seventy-six per cent of the cohort discontinued treatment due to clinical progression, while 5% discontinued because of adverse events. However, two patients remained on treatment and one patient had stable disease for two years with radiographic changes showing necrosis of the tumour much like that seen in GIST with imatinib. Investigators suggested that further study of sorafenib, possibly in combination with chemotherapy, is warranted in leiomyoscarcoma and vascular sarcomas.

10532 Phase II Intergroup Study of Sorafenib (S) in Advanced Soft Tissue Sarcomas (STS): SWOG 0505 C. W. Ryan, M. von Mehren, C. J. Rankin, J. R. Goldblum, G. D. Demetri, V. H. Bramwell, E. C. Borden

Background: S is an oral multikinase inhibitor of VEGFR, PDGFR, Raf, Flt-3 and c-KIT. STS of vascular derivation (VS: angiosarcoma, hemangiopericytoma) express VEGFR and VEGF mRNA. Prognosis for high grade liposarcomas (LPS) and leiomyosarcomas (LMS) has been correlated with levels of serum VEGF. We therefore studied S in patients (pts) with advanced VS, LPS and LMS.

Methods: Pts with advanced VS, LPS and LMS with 0- 1 prior chemotherapies for advanced disease were treated with S 400 mg BID. Solitary fibrous tumors were included as VS because of morphologic similarity to hemangiopericytoma. Pts were required to have adequate blood parameters, PS of 0-1, and no uncontrolled hypertension or history of thromboembolic disease. Prohibited medications included therapeutic anticoagulation, combination anti-retroviral therapy, and prior VEGFR or MAPK inhibitors. 15 pts were accrued to each of three strata: VS, LPS, LMS. If >1 RECIST responses were seen in a stratum, an additional 10 pts were to be added to that stratum.

Results: 51 pts were accrued between 3/2006 and 2/2007. 12 pts were ineligible due to: abnormal lab data (6), >1 prior chemotherapy (2), and incorrect diagnosis on pathology review (4). One pt did not receive any therapy. 38 pts are evaluable for toxicity and 37 for response. Most common AEs were hand-foot, fatigue and diarrhea. Grade 3-4 AEs included amylase/lipase (7), hand foot/rash (6), liver function tests (4), diarrhea (4), electrolytes (4), fatigue (2), hypertension (2), and nausea, vomiting, constipation, abdominal pain, pancreatitis, hemoglobin, platelets (1 each). There were no RECIST responses, but unconfirmed responses were observed. Of note, 7/9 patients in the VS cohort had SD or better and 2 remained on treatment >1 year. With 13 patients deceased, overall median PFS and OS were 3 and 13 months, respectively.

Conclusions: S monotherapy did not result in REC
LPS or LMS. Favorable PFS was observed in pts with advanced VS and further study in these tumors is warranted.

<img1979|center>

Commentary on abstract 4535

Adenocarcinoma of the stomach and gastroesophageal junction cancers are very virulent cancers that are increasing at a rate of about 4% a year in the Western world. Docetaxel (D) plus cisplatin (C) are commonly used treatments that are sometimes used with 5-FU, which increases the toxicity of the regimen. With D/C, approximately 25% of patients respond and median OS is about nine months. Since outcomes between D/C and D/C/5-FU are similar, investigators replaced 5-FU with sorafenib to assess efficacy and tolerability of the novel triplet. Out of 44 patients, the response rate was 40.9%, while disease stabilized in 31.8%. Medium PFS was 5.8 months, while median OS was 13.6 months. Hematologic toxicities from the D/C component of the regimen included grade 3 and 4 leukopenia and neutropenia, and there were two grade 5 toxicities—one from infection and another from GI hemorrhage—that investigators felt might be related to the study protocol. Predictable grade 3 or higher toxicities associated with sorafenib included hand-foot syndrome and fatigue. Investigators considered responses to the novel triplet very encouraging and suggest further study of the combination in a phase III trial.

4535 A Phase II Study: Combination of Sorafenib with Docetaxel and Cisplatin in the Treatment of Metastatic or Advanced Unresectable Gastric and Gastroesophageal Junction (GEJ) Adenocarcinoma (ECOG 5203) W. Sun, M. E. Powell, P. O’Dwyer, R. H. Ansari, A. B. Benson

Background: Adenocarcinoma of the stomach and gastroesophageal junction (GEJ) is a highly virulent disease and remains a leading cause of death worldwide. Alterations of the ras/raf pathway may contribute to the pathogenesis of the disease. Sorafenib is a potent inhibitor of raf tyrosine kinase and of several receptor tyrosine kinases (RTKs) that are involved in tumor progression (e.g. VEGFR-2, VEGFR-3, PDGFR-ß). The combination of docetaxel and cisplatin is commonly used in the treatment of gastric cancer. This study tested the efficacy and tolerability of combining sorafenib with docetaxel and cisplatin in patients with metastatic or advanced unresectable gastric/GEJ adenocarcinoma.

Methods: Sorafenib 400 mg bid orally, docetaxel 75 mg/m2 IV on day 1, cisplatin 75 mg/m2 IV on day 1, administered on 21 day cycles.

Results: Forty-four patients (35 pts [79.5%] with metastatic and 9 pts [20.5%] with locally advanced unresectable disease) were included with median age of 60 (28-86) and M/F of 37/7. The median number of treatment cycles was 4.5 (1-20+). The median PFS was 5.8 months (90% CI [5.4 - 7.2]), and the median OS 14.9 months (90% CI [8.6 - 15.2]). Seventeen pts have shown an objective response (RR 38.6%, 90% CI [26.3%, 52.2%]), including one complete response (CR 2.3%). The most common grade 3 or higher toxicity was neutropenia (26 pts [59%], 90% CI [45.6%, 71.6%]). Two patients died on study with an event attributed to treatment as the 'probable' cause (one died from infection but with only mild [grade 1] neutropenia, the other died of GI hemorrhage). Other common grade 3 or higher toxicities were fatigue, anorexia, hand-foot reaction, nausea, diarrhea and dehydration.

Conclusions: The results of the combination of sorafenib with docetaxel and cisplatin in metastatic/advanced gastric/GEJ adenocarcinoma are very encouraging. Further phase III study with the combination should be considered.

10502 Activity of Sorafenib in Patients with Imatinib- and Sunitinib-resistant (RES) Gastrointestinal Stromal Tumours (GIST): A Phase II Trial of the University of Chicago Phase II Consortium L. Wiebe, K. E. Kasza, R. G. Maki, D. R. D'Adamo, W. A. Chow, J. L. Wade, E. Agamah, W. M. Stadler, E. E. Vokes, H. L. Kindler

Background: Therapeutic options are limited in GIST pts who have progressed on IM and SU. Sorafenib inhibits KIT, VEGFR and PDGFR-ß kinases. In vitro, SOR inhibits KIT kinase activity and cell growth of multiple IM-RES Ba/F3 KIT mutants (Guo, CCR 2007). In a randomized phase II discontinuation trial of SOR, 2 pts with IM-RES GIST developed partial responses (PR) lasting 11 months (mo).

Methods: We are conducting a multicenter phase II trial of SOR in PS 0-2 pts with unresectable, c-kitexpressing GIST that progressed on IM per RECIST criteria. In 2/07, after FDA approval of SU for IM-RES GIST, the study was amended to require progression after both IM and SU.

Primary endpoint: response. Simon minimax 2-stage design requires 1 response in 18 pts for 2nd stage, and 4 responses in 32 IM/SU RES pts. Pts receive SOR 400 mg orally twice daily. CT scans: Q2 28- day cycles.

Results: 26 pts (6 IM-RES, 20 IM/SU-RES) enrolled 1/06-12/07 at 5 sites. Pt characteristics: male 54%; median age 57 (range 42-85); PS 0/1/2: 38%/54%/8%. Cycles: 119 (median 4, range 1- 14). 25 pts are evaluable for toxicity (1 too early) and 24 for response (2 too early) in this ongoing trial. 3 pts (13%) (1 IM-RES, 2 IM/SU-RES) have PR (1 unconfirmed); 14 pts (58%) (3 IM-RES, 11 IM/SU-RES) have stable disease (SD). Disease control rate (PR + SD): 71%. Median progressionfree survival: 5.3 mo (95% CI: 3.2, ). Median survival 13.0 mo (95% CI: 5.1, ), estimated 1-year survival 62%. Grade 3 toxicities (% pts): hand-foot syndrome 28%, hypertension 24%, rash 20%, diarrhea 12%, GI bleed 8%, fatigue 8%, perforation 4%, anemia 4%. Grade 4 toxicity: thrombosis 4%.

Conclusions: These preliminary data suggest that SOR is active and well-tolerated in pts with IM-and SU-resistant GIST.

Commentary on abstract 10502

Response to first-line imatinib (IM) in advanced GIST is 80% with a median PFS of 19 months but response to second-line sunitinib (SU) at 7% is considerably lower and median PFS at 6.2 months considerably shorter than to first-line treatment. After progression, therapeutic options are limited. This trial involved 29 patients enrolled to date divided into two cohorts: six resistant to IM alone and 23 resistant to both agents. After a median of four cycles with sorafenib, four out of 29 patients (14%) had a PR and disease stabilized in 62% of the rest, for a disease control rate of 76%. Many patients with stable disease also had tumour shrinkage, although they did not reach RECIST criteria for PR. Median PFS and OS in the IM-resistant group were 3.4 months and 13.6 months, respectively. For the double-resistance group, they were 5.7 months and 8.5 months, respectively. Treatment with sorafenib was reasonably well tolerated, but dose reductions were required in 59% of patients. As different tyrosine kinase inhibitors have activity against different mutations, an important molecular analysis of tumour samples to determine the relationship between mutational status and response to sorafenib is ongoing.

Questions and Answers with Dr. Lauren Wiebe, University of Chicago, Illinois

Q: Were there differences in the PS in IM-resistant vs. double-resistance patients?

A: Almost twice as many IM-resistant patients at 67% were ECOG PS 0 vs. only 35% of the doubleresistance group, and more patients were ECOG 1 and 2 in the double-resistance group at 56% for PS 1 patients and 9% PS 2. There were no PS 2 patients in the IM-resistant arm, so the rest of that arm were PS 1.

Q: Your eligibility criteria consisted of KIT-expressing GIST with documented disease progression on or after imatinib. Was that to imply sunitinib resistance or resistance to any tyrosine kinase inhibitor before they went on to sorafenib?

A: No, the entry criteria were meant to imply there was progression on imatinib and on sunitinib.

Q: Regarding grade 3 and 4 toxicities, did you see any major toxicities in this cohort of patients?

A: We saw only one grade 4 toxicity and that was thrombosis in 3% of patients. Otherwise, about onequarter of patients developed grade 3 hypertension, one-quarter developed hand-foot syndrome, 17% had rash, and 10% had diarrhea, all grade 3.