Key Points
Question For patients with advanced neuroendocrine carcinoma of the digestive system, which of the 2 community standard regimens is more effective: etoposide plus cisplatin (EP) or irinotecan plus cisplatin (IP)?
Findings In this randomized clinical trial of 170 patients who were chemotherapy naive and had recurrent or unresectable neuroendocrine carcinoma of the digestive system, median overall survival was 12.5 months in the EP arm and 10.9 months in the IP arm.
Meaning Both EP and IP therapy remain standard first-line chemotherapy options.
Abstract
Importance Etoposide plus cisplatin (EP) and irinotecan plus cisplatin (IP) are commonly used as community standard regimens for advanced neuroendocrine carcinoma (NEC).
Objective To identify whether EP or IP is a more effective regimen in terms of overall survival (OS) in patients with advanced NEC of the digestive system.
Design, Setting, and Participants This open-label phase 3 randomized clinical trial enrolled chemotherapy-naive patients aged 20 to 75 years who had recurrent or unresectable NEC (according to the 2010 World Health Organization classification system) arising from the gastrointestinal tract, hepatobiliary system, or pancreas. Participants were enrolled across 50 institutions in Japan between August 8, 2014, and March 6, 2020.
Interventions In the EP arm, etoposide (100 mg/m2/d on days 1, 2, and 3) and cisplatin (80 mg/m2/d on day 1) were administered every 3 weeks. In the IP arm, irinotecan (60 mg/m2/d on days 1, 8, and 15) and cisplatin (60 mg/m2/d on day 1) were administered every 4 weeks.
Main Outcomes and Measures The primary end point was OS. In total, data from 170 patients were analyzed to detect a hazard ratio (HR) of 0.67 (median OS of 8 and 12 months in inferior and superior arms, respectively) with a 2-sided α of 10% and power of 80%. The pathologic findings were centrally reviewed following treatment initiation.
Results Among the 170 patients included (median [range] age, 64 [29-75] years; 117 [68.8%] male), median OS was 12.5 months in the EP arm and 10.9 months in the IP arm (HR, 1.04; 90% CI, 0.79-1.37; P = .80). The median progression-free survival was 5.6 (95% CI, 4.1-6.9) months in the EP arm and 5.1 (95% CI, 3.3-5.7) months in the IP arm (HR, 1.06; 95% CI, 0.78-1.45). A subgroup analysis of OS demonstrated that EP produced more favorable OS in patients with poorly differentiated NEC of pancreatic origin (HR, 4.10; 95% CI, 1.26-13.31). The common grade 3 and 4 adverse events in the EP vs IP arms were neutropenia (75 of 82 [91.5%] patients vs 44 of 82 [53.7%] patients), leukocytopenia (50 of 82 [61.0%] patients vs 25 of 82 [30.5%] patients), and febrile neutropenia (FN) (22 of 82 [26.8%] patients vs 10 of 82 [12.2%] patients). While incidence of FN was initially high in the EP arm, primary prophylactic use of granulocyte colony-stimulating factor effectively reduced the incidence of FN.
Conclusions and Relevance Results of this randomized clinical trial demonstrate that both EP and IP remain the standard first-line chemotherapy options. Although AEs were generally manageable, grade 3 and 4 AEs were more common in the EP arm.
Evidence for treatment strategies of advanced NEC is lacking owing to the rarity of the disease. Recently, the results of a randomized phase 2 trial for NEC were reported16; however, enrollment was terminated early in 66 patients owing to the premature analysis, and no definitive conclusion was obtained. The present study represents the final results of a well-designed and, to our knowledge, the first phase 3 trial with high-quality pathological diagnoses and detailed information provided by a CPR.
The primary analysis of OS revealed no statistically significant difference between arms. Therefore, it is reasonable to continue using both regimens as the standard treatment for NEC. We were interested in the effect of pathological findings and primary organs on treatment efficacy. Although the results showed no statistically significant differences in any subgroups, the overall impression had a greater interaction by primary organ than by pathological findings in terms of OS and PFS (Figure 3 and eFigure 5 in Supplement 1). Because platinum regimens have been reported to be less effective in grade 3 NETs,5,8 and the WHO 2017 classification system9 and 2019 classification system10 formally declared grade 3 NETs as a separate entity from NECs,9,10 subgroup analyses of PDNECs only are of relevant clinical interest. Subgroup analysis of OS in PDNECs only indicated weak tendency of preferred results in the EP arm over the IP arm. A subgroup analysis of only PDNECs in each primary tumor indicated that the OS of the EP arm was better than that of the IP arm in the subset of pancreatic PDNEC. Although post hoc subgroup analyses should be interpreted with caution, these results have important implications. The CPR indicated that pancreatic NECs are characterized by a higher proportion of small cell carcinoma and a lower proportion of tumors with non-NEC components (eTable 2 in Supplement 1). Although these characteristics may have influenced the unique results in this subgroup, they cannot fully explain this phenomenon.
In terms of AEs, grade 3 and 4 AEs such as myelosuppression and FN were more common (>2 times) in the EP arm. Although these AEs were generally manageable, the high incidence of FN remains a point of caution. We took note of the high incidence of FN during this trial in the monitoring report; therefore, we revised the protocol to recommend primary prophylaxis with G-CSF32,33 after October 31, 2017. In the exploratory analysis of this study, prophylaxis with G-CSF seemed to effectively decrease the incidence of FN. However, considering higher planned/actual DI and lower relative DI in the EP arm, and higher rate of treatment termination owing to AE or patient’s refusal (third dose reduction owing to neutropenia), lower initial dose of EP therapy may be appropriate.
In addition to the outcome of the clinical trial, this study is very valuable in its provision of information on high-quality pathological diagnoses made by consistent CPR panel members throughout the entire period. Notably, the pathological diagnosis was not consistent with that of the participating institutions in nearly 10% of cases, indicating the difficulty of an accurate pathological diagnosis of NEC. Additionally, detailed information about the pathological findings was obtained for the cross-organ manner, which itself is an important asset in the field of NEC. In the forest plots of OS and PFS, the difference between EP and IP among the CPR subgroups was smaller than that of the primary organs (Figure 3 and eFigure 5 in Supplement 1).
This study has some limitations. Although the primary analyses revealed no statistically significant difference in OS between the 2 arms, the study was not designed to assess the equivalence of the 2 regimens. Therefore, the 2 regimens should not be recognized as equivalent, but rather as they do not differ beyond a certain level. In addition, in patients diagnosed via results of biopsy specimens, accurate information of the entire tumor could not be obtained with respect to proliferative activity or the proportion of non-NEC components. In this study, patients diagnosed by biopsy harboring any proportions of non-NEC components, as long as a harboring NEC component was allowed to be enrolled, means that theoretically a certain percentage of patients with mixed adenoneuroendocrine carcinomas were enrolled. These limitations were unavoidable considering the diagnostic procedure, and this is also a universal issue in daily practice. Additionally, because further effects of immune checkpoint inhibitors on platinum regimens has been proven in small cell lung carcinoma,34,35 it is the next important clinical question in NECs.
Results of this phase 3 randomized clinical trial demonstrate that both EP and IP remain the standard first-line chemotherapy regimens for advanced digestive NECs. Post hoc subgroup analyses pointed to the superiority of EP in pancreatic PDNECs, and EP was safely given along with the use of primary prophylactic G-CSF.
Source: JAMA
ARYAN'S BLOG 