Has dose dense chemotherapy met an ICONic end?

Ovarian cancer remains the leading cause of death among gynecologic malignancies. Despite significant research, carboplatin and paclitaxel have remained the primary frontline treatment for several decades [1,2]. Multiple investigators have explored alternative dosing strategies in efforts to improve survival outcomes or patient quality of life (QOL) and toxicity.

Dose-dense chemotherapy was initially lauded as a superior alternative to every three-week dosing, following the Japanese Gynecologic Oncology Group (JGOG) 3016 study demonstrating improvement in progression free survival with weekly paclitaxel (80/mg/m²/wk) [3]. Subsequent studies, including Multicenter Italian Trials in Ovarian cancer (MITO7) and Gynecologic Oncology Group (GOG) 262, failed to demonstrate similar survival benefits [4,5]. Notably, each trial refuting the results of JCOG3016 had unique dosing regimens, resulting in ongoing debate regarding optimal regimen and schedule. MITO7 tested weekly administration of carboplatin and paclitaxel (60mg/m²), while GOG 262 allowed for addition of bevacizumab to the JCOG3016 regimen and bevacizumab maintenance.

The International Collaborative Ovarian Neoplasm (ICON) 8 study compared standard three-week carboplatin (AUC 5/6) and paclitaxel (175mg/m²) to the JGOG 3016 regimen of weekly dose-dense paclitaxel (80mg/m²) or a modified MITO7 regimen using weekly paclitaxel (80mg/m²) and weekly carboplatin (AUC 2) [6]. Neither weekly regimen improved survival compared to the three-week combination.

In this Lancet Oncology, Blagden and colleagues report on QOL outcomes in the ICON8 trial (Blagden, et al). Patients completed validated surveys prior to each cycle, every 6 weeks until 9 months, 3-monthly to 2 years, then 6-monthly to 5 years from randomization. Co-primary endpoints were the cross-sectional global QOL score at 9 months and the mean longitudinal global QOL score from randomization to 9 months. A change in global health of 5 or more points was considered clinically significant.

The cross-sectional snapshot of global QOL score between study arms was not significantly different. Compared with the three-weekly regimen, weekly paclitaxel had a global QOL score difference of 2.3 (p=0.094), while the weekly carboplatin/paclitaxel combination had a score difference of −0.8 (p=0.61). The longitudinal analysis revealed that both weekly paclitaxel arms had marginally lower global QOL scores, enough to reach statistical significance, but short of the clinical significance cut off (group 2 vs 1: diff= −1.8, p=0.043; group 3 vs 1, diff= −2.9, p=0.002).

This study has several strengths, including large size and comparatively complete data, with greater QOL survey completion compared to JGOG 3016 and MITO7. As mentioned by the authors, the definition of minimal clinically significant difference is critical in QOL analyses, and the 5-point cut off which has traditionally been used in cross-sectional analyses may be overly conservative for a longitudinal analysis.

Importantly, lack of measurable QOL difference between ICON8 trial arms is concordant with prior QOL assessments comparing similar regimens. In JGOG 3016, QOL assessments were completed at baseline, after cycles 3 and 6, and at 12 months, though only 30% of patients completed the final evaluation [8]. Overall, QOL did not differ significantly between weekly and three-weekly treatment groups (p=0.46). Similarly, after adjusting for clinical factors and bevacizumab in GOG 262, those receiving dose dense paclitaxel reported lower scores after completion of 6 cycles of chemotherapy (p=0.02), but the maximum difference of 2.7 points was not clinically significant [5].

In contrast, weekly assessments of QOL scores within MITO7 identified a difference between regimens over the treatment duration [4]. This study demonstrated the QOL scores in the three-weekly regimen worsened with every chemotherapy cycle. In the weekly schedule, while an initial decrease in QOL was observed at week 1, scores subsequently remained stable, thus, the treatment-by-time interaction resulted in favor of the weekly chemotherapy (p< 0.0001).

Taken together, the results from ICON8 support the lack of clinically significant differences in QOL between weekly and three-weekly administration of carboplatin/paclitaxel in ovarian malignancy. Without QOL improvement, given multiple studies demonstrating equivalent survival, this study may be the final nail in the coffin for dose-dense chemotherapy in ovarian cancer. However, outlier results should not be ignored. Survival differences observed in JGOG3016 may indicate population-level pharmacogenetic differences in ovarian cancer that merit further investigation. In terms of QOL, only MITO7 demonstrated differences between dosing schedules, potentially due to more frequent weekly assessment and more equivalent cumulative paclitaxel doses. Perhaps certain frail patient populations may benefit from weekly administration of the MITO7 regimen with fewer fluctuations in QOL and a more shallow QOL nadir. As we strive to understand the nuanced impact on QOL of these dosing regimens, the current weight of evidence with the addition of ICON8 supports three-weekly administration as the standard of care.