O-187 ZD1839 (iressa) in advanced bronchioloalveolar carcinoma (BAC): A preliminary report of SWOG SO126

S56

Oral Sessions/Practice Guidelines

computer-assisted approach is being compared with RECIST in assessing rate of measurable disease and rate of response of BAC to ZD1839 in the SWOG phase II trial.

O-1 87

ZD1839 (Iressa) in Advanced Bronchioloalveolar Carcinoma (BAC): A Preliminary Report of SWOG SO126

Howard L. West’, Wilbur A. Franklins, Paul Gumerlock3, Ralph B. Vance4, Derick H.M. Lau3, Kari Chansky’, John Crowleys, Jason McCoys, David Ft. Gandara3. f Swedish Cancer InstitufePuget Sound Oncology Consortium, Seattle, USA; ’ University of Colorado Health Sciences Center, Denver, USA; 3 UC Davis Cancer Center, Sacramento, USA; 4 Univ. of Mississippi School of Medicine, Div. of Oncology: Jackson, USA; 5 Southwest Oncology Group, Seatt/e, USA BAC is a non-small cell lung cancer (NSCLC) subtype with distinctive clinical, histologic, and radiographic characteristics. No optimal therapy has yet been defined for recurrent or advanced multifocal disease, and few prospective studies have been performed. The response rate of BAC to conventional chemotherapy is generally considered to be low compared to other histologic subtypes, with radiographic interpretation limited by the diffuse and poorly circumscribed nature of typical lesions. In the largest therapeutic trial in BAC performed to date (S9714), SWOG observed a 14% response rate, median survival of 12 months, a one year survival of 50%, and a three year survival of 10% among 58 patients treated with a 96 hr infusion of paclitaxel. These 59714 clinical data serve as a historical control for the current study. Preclinical studies demonstrate expression of the epidermal growth factor receptor (EGFR) in most NSCLC tumors and antitumor activity of EGFR-targeted agents. Based on preclinical data of EGFR expression in BAC from the SWOG Lung Pathobiology group (Franklin, ASCO 2003, submitted) and anecdotal reports of complete response of BAC to the EGFR tyrosine kinase inhibitor ZD1839 (lressaTM), SWOG initiated a phase II trial to test this agent among untreated and previously treated patients with stage IIIB (with pleural effusion) or stage IV BAC. At present, 129 eligible patients (101 untreated, 29 previously treated) have been registered. The median age is 68.0 (range 34.1-88.5); 50% are male; 86% have a performance status (PS) of 011 and 14% have a PS of 2. Patients received ZD1839 500 mg PO CID until progression of disease or prohibitive toxicity. Both complete and partial responses by RECIST criteria have been seen. Toxicity has primarily consisted of the acneiform rash and diarrhea previously described, although grade 5 toxicity has also been observed. In addition to response evaluation by RECIST criteria, the present study is also assessing response using an investigational computerassisted image analysis (CAIA) system (Lau, Proc AACR#1231, 1999) to determine whether a CAIA- or RECIST-based system better predict survival. Response, survival, and toxicity data will be presented. (CA32102, CA38926)

I

0 188

Encouraging Activity and Durable Responses Demonstrated by the Epithelial Growth Factor Receptor-Tyrosine Kinase Inhibitor, Erlotinib (Tarceva TM, OSl774), in Patients with Advanced Bronchioloalveolar (BAC) Cell Carcinoma

Jvoti D. PateI’, Vincent A. Miller’, Mark G. Kris’ , Neelam T. Shah’, Barbara Pizza’, Leslie Tyson’, Maureen Zakowski’, Natalie Memoli’, Robert Heelan’, David H. Johnson”. ’ Memorial Sloan-Kettering Cancer Center, New York, USA; ’ Vanderbilt Ingram Cancer Cenfer, Nashville, USA

Background: Bronchioloalveolar cell carcinoma or its variants are an increasingly common subtype of non-small cell lung cancer (NSCLC). When compared to other NSCLC, BAC is characterized by less frequent responses to cytoxic therapy, a predilection for affecting women, and a weaker association with smoking. Erlotinib has shown promising results in the treatment of NSCLC, and anecdotally, some of the most dramatic responses have occurred in individuals with BAC. We therefore conducted a Phase II trial of erlotinib in BAC to more precisely define activity in this patient population. Methods: Patients with clinical presentations or pathologic findings suggestive of BAC are first screened for pathologic review. Patients with pathology assessed to be pure BAC (PBAC), BAC with focal invasion (BACFI), or adenocarcinoma with BAC features (A-BAC) [Ebright, Ann Thor Surg, 2002. 74(5); 1640-61 are then screened for trial participation. Results: Between 6/02 and 2/03, 86 patients underwent pathologic review. of these, 56 were felt to have BAC or a variant. 44 patients have been treated to date. 42 patients have completed 4 weeks of therapy and are therefore assessable for response. Patient characteristics: Male-1W Female-28: KPS: 100-1, 90-l 1, 80-25, 70-5; Prior chemotherapy regimenso-32/i-10; Smoking history: never-II/former or current-31. of 42 patients treated for at least one month, 9 patients have achieved a partial response (8 confirmed, 1 too early), major objective response rate 21% (95% Cl 1 l-37). No responding patient has since progressed. Currently, the longest duration of response is 7.2+ months (range 1.5+ to 7.2~). of the 9 responding patients; Male-P/Female-7; KPS: 100-1, 90-1,

80-7; Prior chemotherapy regimens: O-7/1-2; Smoking history: never-5/farmer or current-4. Conclusions: Erlotinib is an active agent in BAC, and can induce durable responses in some patients, To further understand erlotinib’s impressive activity in subsets of patients, we are prospectively constructing a tissue microarray to discern differences in the EGFR and other signaling pathways between responding and non-responding patients. Supported, in part, by Genentech, Inc.

WEDNESDAY, 13 AUGUST 2003

Practice 0 189 EII

Guidelines

Practmoner Feedback on Lung Cancer Practice :. Guidelines Developed Through Cancer Care Ontario’s Program in Evidence-based Care

William K. Evans’, David Cameron*, Jean Mackay”, Nancy Laetschs, Melissa Brouwersa. ’ Cancer Care Ontario, Toronto, Toronto; 2 McMaster University Hamilton, Canada Cancer Care Ontario, an agency of government in Ontario, Canada supports a Program in Evidence-based Care (PEBC), which develops and disseminates evidence-based recommendations or summaries on specific topics chosen by members of 14 Disease Site Groups (DSG). These multidisciplinary DSGs are supported by research coordinators, who conduct literature searches, collate data from the literature and present it to the content experts of the DSG according to a standardized template. The Lung DSG consists of medical oncologists, radiation oncologists, thoracic surgeons, a medical sociologist, methodologists and two consumer representatives. It has published 10 practice guidelines (PG) and has 4 draft PGs in development, which have been sent for practitioner feedback across Ontario as part of the PG development cycle. The PG is accompanied by a survey questionnaire that asks questions on the acceptability and usefulness of the PG. The original feedback survey included 8 items; the current instrument has 21 questions. Since 1995, 1198 questionnaires regarding 14 lung cancer PGs have been sent to medical oncologists or hematologists (52.3% of total questionnaires), radiation oncologists (16.3%), surgeons (15.9%) respirologists (13.7%) and other specialists (1.8%). Almost 60% of surveys went to community-based practitioners and 83.1% of those surveyed were not members of any of the provincial DSGs. 62.8% of surveys were returned and of those, 80% of respondents indicated the PG was relevant to their practice. Respondents indicated that they agreed or strongly agreed with the need for the PG topic in 86.9% of cases, with the acceptability of the evidence summary (91.9%) and with the guideline recommendation as stated (86.1%). Overall, 77.1% indicated that they were likely to use the guideline in their practice. There were no significant differences between the acceptability of the PGs or the likelihood of using the PGs in practice between the community-based practitioners and those working in specialized cancer treatment facilities. The rates of acceptability of the PGs and the likelihood of using the PGs in practice between the community-based practitioners and those working in specialized cancer treatment facilities were similar. Lung cancer related PGs developed through the Lung DSG have had a high rate of acceptability by practitioners. Whether the stated high rate of likelihood of use has translated into practice will need to be studied to determine whether actual practice is consistent with the evidence-based recommendations.

I

0 190

Early Results of a Model for Prospective Quality Improvement in Thoracic Surgical Oncology

Joe B. Putnam Jr.‘, Arlene M. Correa*, David C. Rices, W. Roy Smythes, Stephen G. Swisher*, Ara A. Vaporciyan’, Garrett L. Walshs, Jack A. Roth’. ’ University of Texas M. D. Anderson Cancer Center, Houston, USA; 2 Univ Texas M. D. Anderson Cancer Centeer, Houston, USA

Background: Continuous quality improvement in surgery must focus on processes of care, and individual patient characteristics. We created and used a prospective quality improvement (Ql) program to identify and monitor outcomes of care for patients undergoing resection for thoracic malignancies, and evaluated these outcomes for trends, prior to establishing programs to attenuate specific postoperative events. Methods: A prospective QI program was created using defined database elements selected by our faculty, prospectively defined, collected concurrently with patient care, and reviewed monthly for individual and population characteristics. Modular data forms were completed at the time of patient care in the preoperative clinic, the operating room, at the time of discharge, in the followup clinic, and after completion of pathology through an InterneVlntranet-based