A phase I study of gefitinib, capecitabine, and celecoxib in patients with advanced solid tumors

Abstract

This phase I study was designed to determine the maximum tolerated dose (MTD) and toxicity profile of the combination of gefitinib, capecitabine, and celecoxib in patients with advanced solid tumors. Patients were treated with escalating doses of gefitinib once daily, capecitabine twice daily (14 of 28 days), and celecoxib twice daily. Plasma samples for biomarkers were obtained at baseline and weekly for the first 2 cycles. Pharmacokinetic variables were correlated with toxicity and presence of biological effect. Tumor biopsies from 5 patients were analyzed for changes in tumor metabolic activity by nuclear magnetic resonance spectroscopy. [18F]fluorodeoxyglucose positron emission tomography was done as a correlate in 6 patients at the MTD. Thirty-nine patients received 168 cycles of therapy. The dose-limiting toxicities observed included nausea, dehydration and nausea, diarrhea, and stomatitis. The MTD was 250 mg/d gefitinib (days 1-14) and 2,000 mg/m2/d capecitabine divided twice daily (days 8-21) every 28 days. Celecoxib was eliminated due to concerns of increased risk for cardiovascular toxicity, although no patients in this study had cardiac events. One patient with cholangiocarcinoma had a confirmed partial response. Fourteen of 39 (36%) patients maintained prolonged stable disease for a median of 4 months (range, 3-24 months). [18F]fluorodeoxyglucose positron emission tomography scan and metabolomic analyses revealed differences in metabolic response to gefitinib versus capecitabine. The combination of gefitinib and capecitabine is well tolerated and appears to have activity against certain advanced solid tumors, providing a rationale for further evaluation in advanced solid malignancies.

Figure 1

A, plasma pharmacokinetic interactions between capecitabine and gefitinib. Top, no significant effect of capecitabine dose on gefitinib plasma trough levels when gefitinib is administered at 250 mg/d; bottom, dose-proportional increase of the plasma trough levels of gefitinib given at 250 and 500 mg/d when administered in combination with the recommended dose of capecitabine at 2,000 mg/m²/d. B, plasma gefitinib trough levels in the absence (run-in day 1) and presence [cycle 1, day 1 (C1D1), day 8 (C1D8), and day 15 (C1D15)] of capecitabine therapy (1,000 mg/m²). Gefitinib dose was 250 mg/d and plasma trough levels on cycle 1 days 8 and 15 are significantly elevated when compared with the run-in period and cycle 1 day 1 levels (P < 0.05).

Figure 2

A, PCA scores on metabolomic data sets from 5 patients before (black triangles; baseline) and 14 d after starting treatment (gray circles; cycle 1, day 14). Each point was derived from a metabolic data set of 37 endogenous metabolites. B, time-sensitive changes in metabolites from treated patients with breast cancer. Six metabolites were distinguished by the partial least-squares discriminant analysis to differ among patients 1 to 3 (prolonged stable disease) and patient 5 (progressive disease immediately after cycle 2). All values are represented as normalized metabolite concentrations and expressed as micromole per gram biopsy. Patient 4 (melanoma) was excluded from the partial least-squares discriminant analysis due to the entirely different metabolic profile observed. C, metabolic data on ¹⁸FDG uptake from PET scans (patients 1 and 3–6) in conjunction with intratumoral concentrations of glucose and lactate (patients 1–5) from ¹H NMR metabolomics studies. For patients 1 to 5, the SUV was reported for left axillary nodes to compare with NMR data on biopsies from same anatomic regions. For patient 6, the mean SUV on left lower neck (cycle 1), left axillary lymph nodes (cycle 2), and mediastinal lesions (cycle 4) are reported.