Prognostic Factors and Prognostic Index for Chemonaïve and Gemcitabine-Refractory Patients with Advanced Pancreatic Cancer

Raphaël Maréchal, Anne Demols, Fabien Gay, Viviane de Maertelaere, Marianna Arvanitaki, Alain Hendlisz, Jean‐Luc Van Laethem

Oncology · 2007 · 58 citations · 32 references

Concepts

Abstract

BACKGROUND: Gemcitabine monotherapy is the cornerstone of the treatment of patients suffering from advanced pancreatic cancer (PC). For a few years, new chemotherapeutic agents and combinations have been under validation. The use of such treatment makes it necessary to determine factors that could predict survival time. PATIENTS AND METHODS: To identify factors that predict survival time in chemonaïve patients with advanced PC and after gemcitabine failure, a retrospective analysis was performed on patients with advanced PC coming from phase II and III studies and treated with gemcitabine-based first-line chemotherapy. RESULTS: Ninety-nine patients (median age 66 years, range 27-87) suffering from pathologically proven unresectable or metastatic adenocarcinoma of the pancreas were reviewed. Median overall survival time for the whole population was 251 days and progression-free survival in first- and second-line treatment was 108 and 67 days, respectively. The Cox regression analysis identified aspartate transaminase >53 IU/l, weight loss > or =10% and Karnofsky performance status <90 as significant independent negative prognostic factors in first-line and CA 19-9 >400 IU/ml and albumin < or =3.5 mg/dl in second-line chemotherapy. A prognostic index was calculated from the regression coefficients for each independent prognostic factor and used to classify the patients in 3 different groups with good, intermediate and poor prognosis. The prognosis index in chemonaïve and gemcitabine-refractory patients was (Karnofsky performance status x 0.52) + (weight loss x 1.10) + (aspartate transaminase x 0.82) and (albumin x 1.40) + (CA 19-9 x 0.74), respectively. CONCLUSIONS: Predictive factors could be identified in first- and second-line treatments, although they require prospective validation before they could be used in the design and analysis of future clinical trials.

References

32