• clinical response
  

• progression free survival
  
• 1 year survival
  
• safety and tolerability of regimen
  

Lung cancer is the leading cause of cancer death in men and women in the USA, and account for 28% of all cancer deaths in 2002 (1). In stage IV and some patients with stage III disease chemotherapy is widely used with the primary aim of prolonging survival and /or palliating symptoms. A meta analysis of clinical studies that randomized patients between the best supportive care and chemotherapy concluded that Cisplatin based chemotherapy resulted in a potential gain in the median survival and an increase of the 1-year survival rate compared to the supportive care alone (2). Combination chemotherapy regimens (Platinum/Taxanes, Platinum/Gemcitabine, Platinum/Vinca alkaloids) have demonstrated a response rate of 30-50% with median survival of approximately 10 months, one year survival rates around 40%, and 2 year survival of only 10 % in patients with advanced Non Small Cell Lung Cancer (NSCLC) (3,4). Therefore it is important to continue to search for new agents/combinations that can be used to improve the overall prognosis of these patients.

Rationale for using alternating cycles of chemotherapies:

Emergence of progressive or recurrent disease is a consequence of selection and overgrowth of pre-existent, drug resistant cells during treatment. Based on the mechanism of somatic cell mutation, Goldie and Coldman (5,6) proposed a model for the emergence of drug resistant cells in tumors. Their model proposed that alternating non-cross-resistant combinations would prevent the overgrowth of resistant cancer cells and improve the probability of tumor control or cure. Their recommendation assumed that the two non-cross-resistant regimens were of equal or similar efficacy and that the drugs contained in the two combinations could not be administered together in one single regimen.

There are few instances in Clinical Oncology where two regimens of similar efficacy exist for the same tumor type. Since apparently equivalent chemotherapy regimens exist for the treatment of NSCLC, the Goldie and Coldman hypothesis could also be tested in this tumor type. Historically, SWOG conducted a phase III trial randomizing patients between FOMi (5-Fluorouracil/Vincristine/MitomycinC), CAP (Cyclophosphamide/Doxorubicin/Cisplatin) and a third arm FOMi/CAP, which alternated the two combinations. Response rates were 26%, 17% and 22%, respectively, and were not statistically significantly different (P=0 .247). Survival was reported as 20, 24 and 23 weeks, respectively. Statistically survival of FOMi/CAP treated patients was superior to FOMi treated (P=0.024) but not CAP treated (P=0.23) patients (7). Since then, several studies have attempted to improve upon the response rate and survival in advanced NSCLC by using regimens with alternating cycles of non-cross-resistant therapy (8,9,10,11). These regimens yielded overall response rate of 21-49%, median survival of 20-48 wks with acceptable toxicity.

Though several studies have been done to evaluate clinical response in lung cancer using alternating chemotherapies, none of these studies have evaluated clinical response with alternating Paclitaxel and Gemcitabine based treatment which are two of the most effective chemotherapy regimens in use currently. In studies of Platinum combinations that included one of these agents, results were better than those achieved with Platinum alone or combined with earlier generation agents such as Vindesine or Etoposide (12). Current evidence suggests that combinations of Paclitaxel or Gemcitabine with Cisplatin or Carboplatin are among the most active palliative treatments for advanced NSCLC. With combination of Cisplatin and Gemcitabine a response rate in phase II trials ranged from 30-54%, with median survival times of approximately 8-15 months and I year survival rates from 34-61% (13,14). Myelosuppression was the major toxicity, while Cisplatin associated nonhematologic gastrointestinal, renal and neurologic effects were also problematic. Attempts to reduce toxicity associated with the Cisplatin/ Gemcitabine regimen have included replacing Cisplatin with Carboplatin, a platinum analog possessing similar efficacy and hematologic effects, but lacking the non hematologic toxicity commonly experienced with Cisplatin therapy (15,16). Furthermore Carboplatin can be administered with no need for prehydration to avoid renal toxicity.

In several phase II studies of the Carboplatin/Gemcitabine regimen in patients with stages IIIB and IV NSCLC, objective response rate ranged from 25%-59%, with median survival of 10-16 months (17,18). In a phase II trial, Carrato et al (19) compared 21 day and 28-day schedules of the Gemcitabine/Carboplatin regimen in 75 advanced NSCLC patients. Patients received Gemcitabine 1000mg/m2 on days 1,8 and 15 with Carboplatin AUC 5 on day 1 every 28 days, or a modified regimen omitting the day 15 Gemcitabine dose and repeating cycles every 21 days. Incidence of grades 3 and 4 neutropenia were 52% and 38%, respectively with the 28 day regimen, and decreased to 39% and 23% with the modified regimen. Significant reductions in rates of grades 3 and 4 thrombocytopenia were noted from 45% and 61%, respectively, in the 28 day cohort to 24% (P=0.04) and 17% (P=0.002) respectively, in the 21 day group. Response rate (46% and 37%, respectively) and median survival time (38 wks for both schedules) were maintained in both groups of patients. Furthermore, higher dose intensities of both Gemcitabine and Carboplatin were achieved every 3 wk vs every 4 wk schedules (Gemcitabine 1,133 v 1,002 mg/wk: Carboplatin 162 v 124 mg/wk) (20). Results from current phase III trials (21,22) are also consistent in showing good activity and tolerability of the Carboplatin/Gemcitabine combination in advanced NSCLC, and establish this regimen as an appealing approach for treating this disease. On the basis of above discussion we have chosen Carboplatin/Gemcitabine as one of the doublets we will use in our study. The other doublet we are using in our study is Carboplatin/Paclitaxel. Several studies have shown response rate of approximately 30% with median survival 8-11 months (23,24,25). Toxicity was mild. Grades 3,4 neutropenia, thrombocytopenia and anemia were seen in 15%, 2% and 5% respectively (25). Dose finding studies (26,27) of Paclitaxel and Carboplatin in advanced NSCLC have shown that doses of Paclitaxel higher than 175 mg/m2 prolongs the median time to progression but causes more neurotoxicity and leucopenia. Also, better response rate, the longer overall and better one-year survival seen with the higher dose of Paclitaxel were not statistically significant.

Rationale for using Gemcitabine and Paclitaxel (alternatively in the same regimen) for NSCLC is provided by their antitumor activity, different mechanism of cytotoxicity and different toxicity profiles. In an in-vitro study (28) Jensen PB et al has shown collateral sensitivity between Paclitaxel and Gemcitabine in seven resistant small cell lung cancer cell lines.

The importance of evaluating the impact of any therapy on the quality of life (QOL) of patients with malignancy has been demonstrated in recent literature (29). Patients with advanced NSCLC suffer from a variety of treatment related challenges and do not have an expectation of cure. Hence, it is particularly important to compare treatment alternatives not only on survival endpoints but also on QOL. Therefore, we propose to incorporate QOL evaluations in the proposed clinical trial. QOL instrument we are using is the FACT-L (version 4). The FACT-L is a validated 36-item Likert instrument that combines both relative frequency of symptomatic/quality of life problems with the patients perceived relative importance of each issue (30).

Can we use Taxotere as second line agent if patient had received Paclitaxel as first line agent?

In comparison with best supportive care, Docetaxel (Taxotere) has shown a benefit in overall survival for the second line treatment of NSCLC (31,32). Patients with prior exposure to paclitaxel were excluded from TAX 317 trial (29). But in TAX 320 trial (30) many patients had received Paclitaxel prior to enrollment (31% of patients in Docetaxel 100 mg arm, 42% of patients in the Docetaxel 75 mg arm and 41% of patients in the control arm. Authors were able to evaluate retrospectively, what impact, if any, prior Paclitaxel exposure may have had on response and survival with Docetaxel. Partial response rates were equivalent in the cohort of 91 patients who had received prior Paclitaxel (10.5%) and the group of 157 patients who had not received prior Paclitaxel (8.5%). In a similar analysis of survival data, prior Paclitaxel therapy had no bearing on the survival advantage seen with Docetaxel.