Reviewers: John Han-Chih Chang, MD
Source: Journal of Clinical Oncology 1999; Volume 17 (Number 1): pages 4 ? 11
Non-small-cell lung cancer (NSCLCa) is usually best treated with surgery as the mainstay of curative therapy, reserving chemotherapy (CHTX) and radiation therapy (RT) as adjuncts. However, NSCLCa has the terrible reputation of being unresectable in a highly significant amount of cases at the time of diagnosis. Many trials have attempted to answer the question as how best to apply CHTX and/or RT in these patients. The Cancer and Leukemia Group B (CALGB) reported on a randomized trial of definitive radiation therapy to the thorax with randomization to either CHTX prior to RT (neoadjuvant) versus no prior therapy at all. The neoadjuvant CHTX clearly demonstrated a survival advantage, albeit modest (13% versus 6% survival at 7 years, respectively). The effectiveness of combined modality (CHTX and RT) therapy was again confirmed in trials performed by the Radiation Therapy Oncology Group (RTOG) and Eastern Cooperative Oncology Group (ECOG).
Despite the gains made by combined modality, we were still far away from winning this battle of unresectable NSCLCa. In a few prior studies, it had been demonstrated that there were perhaps benefits in utilizing concurrent carboplatin with RT. The goal was that carboplatin would act as a synergistic agent with RT in improving tumor cell kill. Thus, the CALGB along with ECOG began a phase III prospective randomized trial on utilizing the neoadjuvant CHTX approach that had been effective in the prior CALGB trial and randomizing patients to the addition or omission of concurrent carboplatin weekly during the RT to the thorax.
Two hundred and eighty-three stage IIIA or IIIB (unresectable) patients were registered onto this trial. Full eligibility criteria were explained on page 5. Induction CHTX consisted of vinblastine 5 mg/m2 as a bolus injection on days 1, 8, 15, 22 and 29 along with cisplatin 100 mg/m2 on days 1 and 29. RT was initiated on day 50 (or 2 ? 3 weeks after induction CHTX). RT fields initially encompassed the tumor with a 2-cm margin along with the ipsilateral hilum and entire mediastinum to 40 Gy followed by a boost of 20 Gy to the tumor itself with a 2-cm margin. During the radiation, for those randomized to receive carboplatin concurrently, it was given at a dose of 100 mg/m2/week on the first day of each RT week. Carboplatin was mixed into solution and intravenously infused over a short time period.
The goal of accrual was 265 patients to detect a difference of 15% in overall 3-year survival with 80% confidence. Survival was determined from the date of randomization into the trial to date of death or last follow-up.
Patients were stratified and randomized, such that 137 patients were placed on the combined modality (CMT) arm and 146 were on the combined modality + weekly concurrent carboplatin (CMT+C) arm. Table 1 displays the patient characteristics, which appear to be equal for each treatment arm on listed factors of age, gender, race, performance status, disease assessment and stage. Thirty-three patients were found to be ineligible for various reasons ? these were also balanced between the two arms. Median follow-up in the surviving patients was 41 months.
After completion of the neoadjuvant CHTX (both arms being equal treatment), the clinical complete response (cCR) rate was only 2% with a clinical partial response (cPR) rate of 32%. They also defined or created the entity called regression as decrease in the size of the lesion agreed upon by two independent investigators in disease that was assessable but not measurable. Overall response rate to neoadjuvant CHTX was 37% if you add the cCR + cPR and the regression rate (4%). Reassessment of response after RT alone demonstrated increase of cCR rate to 10% and cPR to 48% with a 5% regression rate. In the RT with added concurrent carboplatin group, cCR rate was increased to 18% and cPR rate was 40% with a regression rate of 5%. Thus, it seems that there was no change in overall response with the addition of carboplatin. The cCR rates are modestly better with concurrent carboplatin, but this did not reach statistical significance (p = 0.101). The Kaplan-Meier curves are nearly identical for the 2 treatment arms in overall survival and failure free survival. The median survival was 13 months (similar to the prior CALGB trial) in both arms. The 3-year survival rate was 19%, but dropped at 4-years to 10% for CMT versus 13% for CMT+C. Median failure free survival was 8 months for both. Analysis of subgroups on the basis of stratification factors demonstrated that stage at diagnosis had a bearing on overall treatment outcome (median survival of 10.5 months in stage IIIB patients versus 16.5 months in the stage IIIA patients.)
Severe toxicities (grade 3 and 4) were seen more often in the CMT+C group in pancytopenia (lowered blood counts). Two deaths from pulmonary complications of treatment were seen in both arms.
The attempt of this phase III trial was to improve upon our dismal track record with unresectable stage III NSCLCa. We have had randomized data to support combined modality with CHTX and RT is better than RT alone. How do we improve upon that? Concurrent carboplatin did not seem to provide that answer. The wave of current chemotherapy methodology lies in the taxane drugs. Perhaps utilizing concurrent Taxol could improve our results. Perhaps the newer modalities or techniques of drug delivery (liposomal preparation or continuous infusion) could enhance tumor cell kill, while decreasing normal tissue toxicity ? improving the therapeutic ratio. Radiation techniques and technology can perhaps be improved to decrease local regional relapse, which ultimately can lead to distant disease, if not already present. There are many avenues to pursue; hopefully we will discover the path to improvement.