The role of dose intensity in determining outcome in intermediate-grade non-Hodgkin's lymphoma

Abstract

To determine whether the dose intensity of chemotherapeutic regimens correlates with the complete remission rate in adult patients with advanced-stage intermediate-grade lymphoma, reports of comparative trials of therapy were reviewed. Reports were identified using MEDLINE, through references from review articles, and through review of selected abstracts. Twenty-two studies including 14 randomized and eight cohort trials were analyzed to assess projected dose intensity. Four other studies were analyzed to assess the role of received dose intensity. Dose intensities were calculated using described methods and correlated with complete remission rates. Individual trials were assessed using "levels of evidence." A metaanalysis of randomized trials and a cross-trial analysis of all comparative trials using a weighted least squares linear regression were performed. Using levels of evidence, support was obtained for the hypothesis that dose intensity correlates with the remission rate from two trials in which dose intensity was "indirectly" tested. As these studies did not "directly" test dose intensity, confounding variables, including those arising from the assumptions made in calculating dose intensity, cannot be excluded. Metaanalysis showed a relative probability of achieving complete remission of 1.34 (95% confidence interval, 1.13 to 1.58) favoring the pooled arm of high dose intensity. Cross-trial analysis showed a relatively weak association between dose intensity and remission rate (r = .49, P = .0001). Two of four reports retrospectively assessing received dose intensity suggested that increased dose intensity is associated with superior remission rates. These analyses suggest that dose intensity may correlate with the remission rate in advanced-stage intermediate-grade lymphoma. However, properly designed trials directly testing dose intensity have not been performed and are needed to confirm this hypothesis.