A B S T R A C T PurposeSince our individual patient data (IPD) meta-analysis (MA) of supportive care and chemotherapy for non-small-cell lung cancer (NSCLC), published in 1995, many trials have been completed. An updated, IPD MA has been carried out to assess newer regimens and determine conclusively the effect of chemotherapy. MethodsSystematic searches for randomized controlled trials (RCTs) were undertaken, followed by central collection, checking, and reanalysis of updated IPD. Results from RCTs were combined to calculate individual and pooled hazard ratios (HRs). ResultsData were obtained from 2,714 patients from 16 RCTs. There were 1,293 deaths among 1,399 patients assigned supportive care and chemotherapy and 1,240 among 1,315 assigned supportive care alone. Results showed a significant benefit of chemotherapy (HR, 0.77; 95% CI, 0.71 to 0.83; P Յ .0001), equivalent to a relative increase in survival of 23% or an absolute improvement in survival of 9% at 12 months, increasing survival from 20% to 29%. There was no clear evidence that this effect was influenced by the drugs used (P ϭ .63) or whether they were used as single agents or in combination (P ϭ .40). Despite changes in patient demographics, the effect of chemotherapy in recent trials did not differ from those included previously (P ϭ .77). There was no clear evidence of a difference or trend in the relative effect of chemotherapy across patient subgroups. ConclusionThis MA of chemotherapy in the supportive care setting demonstrates conclusively that chemotherapy improves overall survival in all patients with advanced NSCLC. Therefore, all patients who are fit enough and wish to receive chemotherapy should do so.

This study did not show a wide difference in activity and toxicity between CAV-E and CAV-T. The number of patients who entered the second randomization was too small to reach the second study endpoint.

This regimen appears to be active and to have a favourable toxicity profile.

A simple instrument for self-assessment of quality of life (QL) in patients with cancer was elaborated using a linear analogue scale (LAS). The instrument was based on five questions, exploring different functional areas; the same questions were also addressed in a parallel format, where problems were seen from an opposite point of view (positive/negative). The LAS was given to 222 patients, for a total of 372 tests collected. Internal consistency was satisfactory (Cronbach's alpha = 0.75); QL score was significantly correlated to parameters of disease. Concordance between scales, as judged by comparison of parallel formats, was statistically significant but poor. A questionnaire was then elaborated with similar items, based on a categorical scale. A direct comparison between LAS and our questionnaire was made on a group of 41 patients. Internal consistency was poor for the LAS (alpha = 0.58) and good for the questionnaire (alpha = 0.93); Spearman's rank correlation coefficients were disappointing for the LAS and good for the questionnaire; the questionnaire was judged reliable in 82.9% of cases, the LAS in 29.3% only; the questionnaire score, and not the LAS score, was significantly correlated with PS and disease status. In conclusion, many patients appeared unable to correctly interpret the visual-analogue scale; the categorical scale was more immediate and correctly understood by the large majority of patients; the correlation between score and important parameters of QL was maintained, and internal consistency was excellent, indicating a satisfactory reliability of this instrument.

Fifteen of 146 (10%) adult patients with non-Hodgkin's lymphoma showed clinical and pathologic evidence of involvement of the central nervous system (CNS): in 6 patients, the CNS lymphoma was present at the onset of disease, in 3 of them it was the only sign detected. In the remaining 9 cases, CNS involvement appeared during the course of systemic disease. In all cases symptoms related to infiltration of the CNS were associated with advanced disease (stage IV); bone marrow or bone involvement was found in 9 patients (60%). The histologic subtypes were mostly of high-grade malignancy according to the Kiel classification: immunoblastic (3), centroblastic (3), Burkitt type (2), lymphoblastic (1), LP immunocytoma in polymorphic variant (3), unclassifiable (3). The prominent signs and symptoms of CNS lymphoma are listed: the cranial nerve palsies are the most common finding. The principal means of detecting CNS involvement are discussed: cerebrospinal fluid cytology, brain scan and CAT scan were the most useful diagnostic procedures. The reported data allow identification of patients at high risk of CNS lymphoma: this includes histologies of high-grade malignancy, advanced stage of disease, and bone marrow or bone infiltration. Therefore, either intensive systemic chemotherapy or CNS prophylaxis are recommended for patients with high risk of CNS disease.