Non small cell lung cancer patients with ECOG PS2: unsolved questions and lessons from clinical trials

Abstract

In the last two decades the results of medical treatment of advanced non-small cell lung cancer (NSCLC) have constantly improved even if they are still far from being considered satisfactory. Today systemic cisplatin-based chemotherapy (CT) is able to increase survival and improve cancer-related symptoms in patients with advanced ‘wet’ stage III and metastatic stage IV NSCLC, but it not clear if the benefits of CT also apply to patients with poor performance status (PS) [1, 2]. PS is the most powerful independent prognostic factor in advanced NSCLC since it is a reliable measure of functional independence, ability to perform daily activities and work, and a strong predictor of survival and adverse events as well [3]. The vast majority of prospective phase III trials had been conducted stratifying patients according to stage (III versus IV) and PS (PS 0–1 versus 2). In a retrospective analysis by the Veteran Administration Lung Group including more than 5000 patients, PS recorded according to the Karnofsky Index was the strongest prognostic factor followed by extent of disease and weigh loss [3]. The same conclusion has been reached by more recent large trials using new regimens. The survival analysis of 1960 patients treated with cisplatin-based CT in five ECOG trials showed a median survival of 3.3 months for PS2 patients as compared to 9.4 months and 6.4 months recorded for PS0 and PS1 patients, respectively [4]. A similar analysis of 2531 patients included in fourteen SWOG trials also showed that a poor PS was the strongest independent predictor of shorter median survival (PS 0–1: 6.4 months; PS >_2. 3.3 months) and lower 1-year survival rate (20% versus 9%; P _2 was 3.8 months. Data from 1052 patients treated with cisplatin-based CT in seven EORTC trials showed that, after disease extent, PS was the most important determinant of outcome [6]. In all studies a better outcome was also associated to less disease extent, presence of less than 1 metastatic lesion, absence of liver, skin or bone metastases, basal haemoglobin levels >11 gr%, normal LDH and alkaline phosphatase levels, absence of symptoms due to metastatic disease, and a preserved appetite. Recently an analysis of 1436 patients treated with third-generation doublet regimens (ECOG E5592 and E1594 trials) allowed the identification of six independent poor prognostic factors including PS2, presence of skin or liver metastases, loss of appetite, presence of less than four metastatic sites, and no prior surgical treatment. A normogram using these six pretreatment prognostic factors was elaborated and validated to predict 1-and 2-year survival in CT-naive patients. This prognostic model could be potentially useful in clinical decision making and research planning [7]. Despite the negative influence of poor PS on survival parameters, the relationship between PS score and control of cancer-related symptoms deserves some comments. A significant proportion of patients with PS2 treated with CT may experience an improvement in tumor-related symptoms even if a better PS is associated with better control. A study on nearly 300 patients showed that 48% of PS2 patients and even 30% PS3 ones may experience an improvement of cancerrelated symptoms. However a better PS was a strong prognostic factor for higher symptomatic response [8]. Only one study specifically analysed quality of life (QoL) in the different PS sub-groups [9]. Patients with PS2 reported the worst scores at baseline assessment. The drop-out rate in PS2 patients was greater than in the other PS levels (PS2 35% versus PS1 23% versus PS0 18%). Nevertheless, PS2 pts had significant benefit from CT and, with the greater potential for palliation determined by worse baseline condition, showed an improvement in QoL even higher than that of patients with PS 0–1. Overall, median survival of PS2 patients is substantially shorter than that of PS 0–1 patients, being usually <_5 months with a 1-year survival rate lower than 20% independently of the type of regimen employed [2]. This poor outcome had–and probably still has–a negative influence on the management of *Correspondence to: Dr Vittorio Gebbia, Via Alessandro Paternostro n. 48, 90133 Palermo, Italy. Tel. +39-0916806906; Fax: +39-0916806906; E-mail: vittorio.gebbia@tin.it Annals of Oncology16 (Supplement 4): iv123–iv131, 2005 doi:10.1093/annonc/mdi921