Quality of life (QoL) as a prognostic factor for survival in previously treated advanced gastric or gastroesophageal junction (GEJ) cancer: Analysis of pooled data from two phase 3 studies (REGARD and RAINBOW)

Konference: 2015 40th Congress ESMO a 18th ECCO - účast ČR

Kategorie: Gastrointestinální nádory

Téma: Postery

Číslo abstraktu: P305/2343

Autoři: M.D. Ian Chau; M.D. Charles S. Fuchs, MPH; Kei Muro; MUDr. Jiří Tomášek, Ph.D.; Eric Van Cutsem; Jae-Yong Cho, M.D., Ph.D.; Prof. M.D. Sang Cheul Oh; Prof. Dr. Howard Safran; M.D. György Bodoky, Ph.D.; Yasuhiro Shimada; Dr. Filip Dumitru; MD Rodolfo Passalacqua; Dr. Atsushi Ohtsu; Yanzhi Hsu, Ph.D.; Astra Liepa; Kumari Chandrawansa; MD Michael Emig; Prof. M.D. David R. Ferry, Ph.D.; Prof. MUDr. Hansjochen Wilke; MD Salah-Eddin Al-Batran

Background: QoL has previously been identified as prognostic for overall survival (OS) for advanced gastroesophageal cancer in the 1st-line setting. Here we explore the prognostic role of QoL in the 2nd-line setting. Patients (pts) were enrolled in two phase 3, randomized, double-blind studies of ramucirumab (RAM) in advanced gastric/GEJ cancers following progressive disease (PD) on 1st-line platinum- and/or fluoropyrimidine-containing therapy: REGARD (RAM+best supportive care [BSC] vs placebo [PL]+BSC) and RAINBOW (RAM+paclitaxel [PTX] vs PL+PTX).

Material and Methods: Pts in both trials completed the EORTC QLQ-C30 (v3) at baseline. For this analysis, all 15 QoL scales were evaluated, along with 12 baseline clinical characteristics and lab parameters that had been previously identified as independent prognostic factors for OS in this pooled dataset. QoL scales were scored 0–100 and dichotomized by median values (or 0 vs >0 if median=0). To identify QoL prognostic factors, univariate Cox models were first used to select covariates with p≤0.05. For these QoL factors, a multivariate Cox model was used for stepwise selection with entry and exit p=0.01; previously selected 12 factors had forced inclusion in the model. Once QoL factors were selected, the model was refitted to include only selected factors, to avoid exclusion of pts with missing data in unselected variables who were included in the selection pool. Also, a model with entry and exit p=0.05 was explored. All models were stratified by treatment and geographic region.

Results: Of 1020 pts, 927 (91%) were included in the final Cox regression, after excluding pts with missing covariate values. In univariate analyses, 13 QoL factors had p≤0.05. In the multivariate analysis using entry and exit p=0.01, only one QoL factor (ie. appetite loss) was selected, along with all the previously identified factors.

Poor Prognostic Factors for OS Hazard Ratio (99%CI)
Peritoneal metastasis 1.44 (1.17, 1.77)
Time to PD on prior therapy <6 months 1.31 (1.06, 1.61)
ECOG performance status ≥1 1.28 (1.03, 1.60)
Tumor differentiation (poor/unknown) 1.32 (1.07, 1.64)
Primary tumor present 1.31 (1.05, 1.62)
Alkaline phosphatase (high) 1.28 (1.02, 1.59)
Sodium (low) 1.97 (1.47, 2.63)
Lactate dehydrogenase (high) 1.30 (1.04, 1.63)
Aspartate aminotransferase (high) 1.32 (1.02, 1.70)
Albumin (low) 1.30 (1.04, 1.61)
Lymphocytes (low) 1.35 (1.08, 1.69)
Neutrophils (high) 1.48 (1.13, 1.95)
Appetite loss (high) 1.50 (1.20, 1.86)

 

The multivariate analysis using entry and exit p=0.05 also selected nausea/vomiting, with scores >0 as a poor prognostic factor.

Conclusions: In addition to clinical characteristics and lab parameters, baseline pt-reported appetite loss was independently prognostic for OS for previously treated advanced gastric/GEJ cancer. This symptom had more prognostic value than pt-reported functioning domains including global health status.

Conflict of interest: Ownership: Hsu: Lilly (ownership/stock) Liepa: Lilly (ownership/stock) Chandrawansa: Lilly (ownership/stock) Emig: Lilly (ownership/stock) Ferry: Lilly (ownership/stock). Advisory Board: Chau: Sanofi Oncology, Lilly, Bristol Meyers Squibb, Merck Serono, Gilead Science, Bayer Fuchs: Eli Lilly, Genentech, Sanofi, Pfizer, Bristol Myers Squibb, Merck, Amegn, Entrinsic Health, Bayer, Medimmune, Celgene, Gilead, Macrogenics, Acceleron Tomase: Roche, Amgen, Lilly Ferry: Sanofi, Roche, Lilly Wilke: Lilly Al-Batran: Lilly. Corporate-sponsored Research: Chau: Novartis, Roche, Merck-Serono (grant to institution) Muro: Lilly, BMS, Merck Tomasek: Amgen, Novartis, Lilly Van Cutsem: Lilly Shimada: Lilly Japan Ferry: Roche, Sanofi, Astra-Zeneca.

Keywords:
QoL
Gastric cancer
prognostics
 

Datum přednesení příspěvku: 28. 9. 2015