AbstractPurpose: An all-oral regimen of etoposide and cyclophosphamide was developed for use in poor-prognosis extensive disease small-cell lung cancer. Limited pharmacokinetic sampling was used to derive a pharmacodynamic model predictive of myelosuppression early in the course of therapy. Patients and methods: Eligible patients were chemotherapy-naive and had extensive disease small-cell lung cancer with either SWOG performance status 2 or serum albumin 3.5g/dl. The first cohort (n=18) received etoposide orally at 50mg daily and cyclophosphamide orally at 50mg daily days 114 every 28days. Due to good hematologic tolerance, the second cohort (n=39) received both agents orally at 50mg twice daily days 114 every 28 days. Plasma etoposide levels were determined in samples drawn at baseline, and at 1h, 2h, and 23.5h (trough) after the first dose. Linear regression analysis was used to determine pharmacokinetic and demographic parameters predictive of myelosuppression. Results: A total of 173 treatment cycles were delivered. Patients on the daily regimen had a 22% response rate (complete and partial), a 22% unconfirmed response rate, and a 5-month median survival, while patients on the twice-daily regimen had a 28% response rate (complete and partial), a 13% unconfirmed response rate, and a 7-month median survival. Granulocytopenia and alopecia were the most common toxicities seen. Significant granulocytopenia could be predicted for the twice-daily regimen according to the formula ln(AGC nadir)=7.801.88(trough), with an increased incidence of granulocytopenia if the etoposide trough value was 1.49g/ml. Conclusion: Oral etoposide and oral cyclophosphamide given days 114 every 28 days is well tolerated and results in an acceptable response rate and median survival in poor-prognosis (poor performance status or low serum albumin) extensive disease small-cell lung cancer. A trough etoposide level obtained within 24h of starting therapy can predict severe granulocytopenia.