Cancer Chemother Pharmacol (1989)23:252-254 Cisplatin with high-dose infusions of hydroxyurea to inhibit DNA repair ancer hemotherapyand harmacology © Springer-Verlag 1989 A phase II study in non-small-cell lung cancer Brian M. J. Cantwell 1, Daniel Veale 2, Christine Rivett 3, Sarah Ghani 1, and Adrian L. Harris ! University Department of Clinical Oncology, Regional Radiotherapy Centre, Newcastle General Hospital, Newcastle upon Tyne NE4 6BE, England 2Department of Respiratory Medicine, Freeman Hospital, Newcastle upon Tyne NE77 DN, England 3Pharmaceutical Department, Newcastle General Hospital, Newcastle upon Tyne NE4 6BE, England Summary. A total of 45 patients with locally advanced and/or metastatic non-small-cell lung cancer (NSCLC) were treated in a phase II trial with high-dose i.v. infu- sions of 24 g hydroxyurea over 24 h, with 50 mg/m 2 i.v. cisplatin 8 h after the start of hydroxyurea infusion. Hy- droxyurea, a cell-cycle-specific inhibitor of ribonucleotide reductase, inhibits DNA repair by depleting nucleotide pools. We gave hydroxyurea to achieve steady-state levels of > 1 mM and to potentiate therapy by inhibiting repair of DNA damage produced by cisplatin. Among 21 pa- tients with squamous cell lung cancer, there were 1 com- plete response (CR), 2 partial responses (PR) and 3 minor responses (MR). Of 13 patients with adenocarcinoma of the lung, 2 had MRs; of I 1 patients with large-cell ana- plastic lung cancer, none responded. The dominant toxici- ty was nausea and vomiting, which was manageable and mainly related to cisplatin. The response rate in squamous cell lung cancer was similar to responses obtained with cis- platin alone. The relative ineffectiveness of high-dose 24-h infusions of hydroxyurea in inhibiting repair of DNA damage produced by cisplatin may be due to the low growth fraction of human NSCLC. The high-dose hy- droxyurea approach may be more applicable in tumours with a high growth fraction. Introduction Hydroxyurea is an S-phase cell-cycle-specific agent that inhibits DNA synthesis by inhibiting ribonucleotide reduc- tase and depletes cells of nucleotide triphosphates [19]. At higher doses, it inhibits DNA repair [7, 12, 17], probably by preventing the filling of gaps in DNA with nucleotides. Cisplatin has moderate single-agent anti-tumour activity against non-small-cell lung cancer (NSCLC), [1] and in hu- man cell lines it induces DNA inter-strand cross-links reaching a maximal level 6-12 h after exposure to the drug [5]. Hydroxyurea rapidly affects the nucleotide pools within 1 h after its addition to cultured cells [6]. We there- fore designed a clinical study combining cisplatin with high-dose intermittent infusions of hydroxyurea to prevent repair of DNA damage induced by the former. Since hy- droxyurea is a cell-cycle-specific agent, it would therefore be most effective when given such as to ensure continuous exposure of tumour cells to the drug [10]. We [15] and others [2] have shown that high-dose i.v. infusions of hy- Offprint requests to: Adrian L. Harris droxyurea, achieving levels that inhibit repair in vitro, are possible with mild toxicity in humans. Patients and methods A total of 45 patients with histologically diagnosed, locally advanced and/or metastatic NSCLC were treated with hy- droxyurea and cisplatin. All patients had assessable dis- ease, and their performance status was graded by WHO criteria [18]. Patient characteristics are given in Table 1, and Table 2 shows details of the courses of treatment. Hy- droxyurea powder was aseptically prepared for i.v. use by the Pharmacy, Newcastle General Hospital, and was given at a dose of 24 g dissolved in 3 1 4% glucose and 0.18% so- dium chloride by i.v. infusion over a 24-h period. In addi- tion, 8 h after the start of the hydroxyurea infusion 50 mg/m 2 cisplatin in 100 ml 0.9% sodium chloride solu- tion was infused i.v. over 15 min. Since steady-state levels of hydroxyurea are achieved within 8 h of a constant i.v. infusion [15], cisplatin was given at thistime and hydroxy- urea was continued over the time course predicted for sub- sequent repair of DNA adducts, i.e. 16 h. Courses were re- peated every 3 weeks, to a maximum of six courses in any one patient. Before each treatment course, haemoglobin, WBC count, serum urea, creatinine, electrolyte and liver func- tion tests were carried out. Response was assessed clini- cally and by serial chest X-rays (CXR), and some patients also underwent computerized axial tomography. Routine anti-emetics were given with each chemotherapy course, usually i.v. metoclopramide at an initial dose of 2 mg/kg followed by 100 mg by 8-h infusion. Alternatively, patients were given 4mg qd oral dexamethasone for 1 day and 30 mg qd oral domperidone for 2 days or longer. Results One patient had a complete response (CR, complete disap- pearance of all pre-treatment evidence of tumour), and two others had >_50% decreases in tumour volume desig- nated as partial responses (PRs). These three objective re- sponses occurred in patients with squamous cell lung can- cer. In addition, three patients with squamous cell lung cancer had < 50% regression in tumour volume designated as minor responses (MRs). Thus, the percentage of re- sponse rate for conventionally defined objective re- sponses, i.e. CRs + PRs, was 14.2% (95% confidence lim- its, 3.0%-36.3%). If CRs, PRs and MRs are combined, the