Thirty Years Later: We've Only Just Begun
Thirty Years Later: We've Only Just Begun
Over the past 30 years, thoracic oncologists have sought therapies for use in the adjuvant setting that might mimic the survival advantage offered by adjuvant chemotherapy in other solid tumors such as breast and colon cancers.[1,2] For decades, the mainstay of treatment for early-stage lung cancer has been resection. With increasing stage, though, escalating percentages of patients relapse with distant metastases. In the 1980s and 1990s, multiple regimens were tested in clinical trials as adjuvant therapy for lung cancer. None of these regimens in isolation provided convincing evidence that adjuvant chemotherapy would be beneficial in non-small-cell lung cancer (NSCLC).
In a 1995 meta-analysis, the British NSCLC Collaborative Group combined the results of many early trials, finding that cisplatin-based chemotherapy appeared to offer a 5% survival advantage at 5 years; however, the confidence intervals around that figure were not narrow enough for the result to be statistically significant (P = .08). This sparked interest in the next generation of clinical trials assessing the use of platinum-based therapy in the adjuvant setting in NSCLC. Looking back, we will never know if the P value sparked the current trials or if the time was finally right in the thoracic oncology community for serious trials on adjuvant chemotherapy.
IALT, BLT, and ALPI
In 2003, the International Adjuvant Lung Cancer Trial (IALT) Collaborative Group published the largest trial designed to address the question of the efficacy of cisplatinum-based chemotherapy in NSCLC. This study found a 4.5% improvement in overall 5-year survival, which was statistically significant and consistent with the magnitude of response seen in the meta-analysis. These results clearly were viewed with great enthusiasm from the thoracic oncology community.
The results of two other trials, the Big Lung Trial (BLT) and the Adjuvant Lung Project Italy (ALPI), were published shortly thereafter.[5,6] Several points must be taken into consideration when comparing these seemingly similar large simple randomized trials. ALPI was stopped after 5 years of accrual with only 1,088 evaluable patients, accounting for only 84% of the recruitment goal. The BLT only enrolled 381 patients (76% of their goal) and clearly stated in the methods section that the power to detect a 5% difference in 5-year survival based on this sample size was only 20%.
Other significant differences that appear in these studies are the treatment choices utilized (Table 1). First, ALPI included twice as many patients who received adjuvant radiation therapy (43% in both the experimental and control arms). The effects of radiation therapy on survival in adjuvant lung cancer have yet to be fully determined. Previous studies have shown detrimental effects using older radiation delivery techniques. Use of modern radiation equipment may not confer the detrimental effects previously seen, but modern studies such as the trial published in 2000 by the Eastern Cooperative Oncology Group (ECOG), which assessed the use of sequential adjuvant chemotherapy and radiation therapy as compared to radiation therapy alone, did not show a survival benefit. The patients in these studies who received radiation therapy may not be comparable to those in the studies who received either chemotherapy alone or best supportive care. The BLT did not have as many patients who received radiation therapy (14% in both arms), but they prescribed a lower targeted cisplatin cumulative dose (150-240 mg/m2 vs 300-400 mg/m2 in IALT) with a lower compliance.
Stating that either BLT or ALPI are negative studies when compared to IALT is an oversimplification of the data. Studies that have been reported as positive or negative may be wrong in their conclusions since a study may not have been appropriately designed to detect the difference that was proposed. The benefit of using point estimates and confidence intervals is that readers are able to determine the weight of evidence a study provides either in support of the experimental arm or in support of the standard. Taking a dichotomizing approach and deeming a study positive or negative misleads readers to assume that the study design was appropriate and the study result was definitive.
More Recent Trials