Irinotecan-based chemotherapy in extrapulmonary neuroendocrine carcinomas: survival and safety data from a multicentric Italian experience

Neuroendocrine carcinomas (NECs) are a rare subgroup of neuroendocrine neoplasms that occasionally originate from gastro-entero-pancreatic (GEP) tract. Evidence of the effectiveness of chemotherapy is scarce. Platinum plus Etoposide regimens are currently the standard treatment in first-line, while little data are available on second-line treatments. The aim of this study is to evaluate the efficacy and safety of irinotecan (IRI)-based chemotherapy in a series of extrapulmonary NECs. Patients with NEC diagnosis treated at University Hospitals of Modena, Florence, Pisa, and European Institute of Oncology of Milan with an IRI-based regimen (FOLFIRI or XELIRI) after progression to a first-line platinum-based therapy were enrolled. Objective responses were assessed according to RECIST criteria. Progression-free survival (PFS) and overall survival (OS) were calculated. Thirty-four patients, 16 males, and 18 females, median age of 59 years (range 32–77), with metastatic NEC were included. Twenty-seven patients had Ki-67 ≥ 55% and four patients Ki-67 of <55% (for three patients data were not available). The median number of treatment cycles of the IRI-based regimen was 7.5 (range 1–16). Six partial responses (17.6%) and 9 stable diseases (26.5%) were observed, with a disease control rate of 44.1%. Median PFS and OS were 4.4 and 5.9 months, respectively. Neutropenia, anemia, and nausea were the only G3–G4 toxicities reported. Despite the relatively small sample size, IRI-based therapy demonstrated to be a valid option for patients with pretreated extrapulmonary NEC.


Introduction
Neuroendocrine carcinomas (NECs) are a rare, aggressive variant of neuroendocrine neoplasms (NENs), with poorly differentiated histology and high proliferative index (Ki-67 > 20%). Roughly 35-55% of NEC originate from lungs but occasionally they can also arise from the gastro-entero-pancreatic (GEP) tract [1,2], from bladder, and, in a further minimal percentage, from other districts like head and neck, breast, and prostate. GEP-NECs represent about 10% of all GEP-NENs and show a dismal prognosis with a median survival of~5 months without treatment [3,4].
Due to their low incidence, therapies for NECs have not robust evidence beyond: the available data derive from old, retrospective analyses with small samples, different schedules, and not designed to answer this specific unmet medical need. Therefore, platinum plus etoposide is the most common chemotherapeutic regimen used in clinical practice as first-line treatment resulting in a gain in median overall survival (OS) of 15 months [5]. Unfortunately, in the second-line setting, there is no consensus regarding further treatment. Hentic et al. from the French group has proposed irinotecan-based chemotherapy (IRI-CT) as a valid second-line strategy [6] as compared to other agents tested like amrubicin [7], temozolomide [8,9], S-1 [7], FOLFOX [10], and taxanes [7].
However, the exiguous sample size of the study prevents us from drawing definitive conclusions about the efficacy of this regimen.
With this study, we retrospectively evaluate the efficacy and safety of IRI-CT in a cohort of patients with metastatic extrapulmonary (EP) NEC progressing after first-line platinum-based treatment.

Study population
This is a retrospective multicentric population-based study. We collected data from four oncological centers across Italy (Department of Oncology Thirty-four consecutive patients with a diagnosis of EP-NEC according to 2010 WHO Classification as reported by expert pathologists in this field in each center and treated with an IRI-CT after platinum plus etoposide chemotherapy failure were included. The study was approved by the local Ethic Committee (n°473/2020). All patients provided written informed consent.

Selection criteria and treatment regimens
Patients with a diagnosis of EP-NEC who received at least one cycle of IRI-CT were considered eligible for the study.
All patients were treated with one of the following regimens: -FOLFIRI: Irinotecan 180 mg/mq over 60 min on day 1, followed by calcium levofolinate 100 mg/mq over 120 min, FU 400-mg/mq bolus, and FU 600-mg/mq continuous infusion over 22 h on days 1 and 2 every 14 days.
-Modified FOLFIRI plus or minus bevacizumab: Irinotecan 180-mg/mq IV over 60 min, calcium levofolinate 200-mg/mq IV over 2 h, followed by FU 400-mg/m2 bolus on day 1 and FU 2400-mg/mq continuous infusion over 46 h every 14 days. Bevacizumab was administered at the dose of 5 mg/Kg over 30 min before modified FOLFIRI.
All patients received intravenous premedication with dexamethasone, chlorphenamine, and ondansetron. Administration of granulocyte clone-stimulating factors was considered only as secondary prophylaxis. Treatment was continued until disease progression, patient refusal, or occurrence of unacceptable toxicity.

Efficacy and safety assessment
Although a follow-up program was not standardized due to the retrospective design, response rate (RR) was assessed by computer tomography (CT) scan of chest, abdomen, pelvis, and brain approximately every 3 months according to RECIST criteria. Progression-free survival (PFS) was calculated from the date of the first IRI-CT administration to progression or death for any cause, whichever occurred first. OS was defined as the time from starting IRI-based chemotherapy until death. Disease control rate (DCR) was defined as the percentage of patients who have achieved complete response (CR), partial response (PR), and stable disease (SD) achieved to IRI-CT. Adverse events were evaluated using common terminology criteria for adverse events version 5.0 [11].

Statistical analysis
Quantitative data were expressed as medians (range). Survival rates were calculated according to the Kaplan-Meier method. All statistical analyses were performed using SPSS statistical software.
Other patients' characteristics are resumed in Table 1.

Efficacy outcomes
The median OS from starting IRI-based chemotherapy until death was 14 months (range 5-46 months) as reported in  Table 2.
Overall, six patients achieved a PR (17.6%) and nine patients showed SD (26.5%) as the best response to treatment, with a DCR of 44.1%. Four of six patients with PR to irinotecan had a high proliferative index (Ki-67 ≥ 55%), although there was no statistically significant difference in terms of RR according to Ki-67 cutoff (OR = 0.69; 95% CI: 0.084-5.642).
Patients presenting with liver involvement showed worse survival outcomes: PFS to IRI-CT was 3.5 months compared to 5.2 months (p = 0.290, 95% CI: 0.27-1.48) observed in the patients without liver metastases; also OS was modestly different (5.7 vs. 7.9 months; p = 0.293, 95% CI: 0.27-1.49). However, the data did not reach statistical significance, probably due to an exiguous sample of patients without liver involvement. In our analysis, we evaluated the association of selected inflammatory biomarkers with risk of death and progression: while lymphocyte-platelet ratio (PLR) and neutrophil-lymphocyte ratio (NLR) seems not to impact OS and PFS, low lymphocyte-monocyte ratio (LMR) demonstrate to be associated with a significantly worse PFS (p = 0.041) (Fig. 3), with no effects on OS.
Furthermore, primary tumor location seems to partially impact survival: colorectal NECs tend to show worse OS to IRI-CT than NECs arising from other districts (3.4 vs. 7.3 months; p = 0.102).
It should be noted that, at the time of the drafting of this paper, two patients were still receiving irinotecan-based treatment.

Discussion
Patients with advanced EP-NECs are usually treated with platinum-based chemotherapy regimens in first-line setting. However, many options have been evaluated in second-line setting, but none with robust literature evidence, with data mainly coming from several small retrospective studies in which amrubicin [7], temozolomide [8,9], S-1 [7], oxaliplatin [10], irinotecan [6], EpiCO (epirubicin, cyclophosphamide, and vincristine) [12], and taxanes [7] have been tested.   CR complete response, PR partial response, SD stable disease, PD progressive disease, ORR objective response rate, DCR disease control rate, PFS1 progression-free survival from start of first-line treatment Hentic et al. described the efficacy of irinotecan in this setting of patients retrospectively analyzing a population of 19 patients affected by PD-NEC and observing DCR of 62% with 31% PR, 31% of SD, PFS and OS of 4 and 18 months, respectively. However, in that context, only five patients presented a Ki-67 ≥ 55%, while the majority of the population (14 patients, 74%) had a Ki-67 < 55% [6]. A recent analysis demonstrated Ki-67 = 55% as the best cutoff value concerning RR and survival. In addiction, data from the NORDIC NEC trial showed that patients with Ki-67 < 55% had a lower RR (15% versus 42%, P < 0.001), but paradoxically better survival than those with Ki-67 ≥ 55% [13]. Based on this evidence, no consistent data concerning the impact of IRI-CT on EP-NECs with higher Ki-67 are available in the second-line setting.
To the best of our knowledge, this is the largest study aiming at evaluating the efficacy and feasibility of IRI-CT in a cohort of patients with pretreated NEC. In the last 12 years, in the four cancer centers that participated in this retrospective analysis, 34 patients affected by EP-NECs received IRI-CT after progression to first-line platinum plus etoposide chemotherapy. Twenty-seven of them had a Ki-67 ≥ 55% while four had Ki-67 < 55% (for three patients data was not available). Our population showed a DCR of 44.1% with 17.6% PR and 26.5% SD with median PFS and OS of 4.4 and 5.9 months, respectively. According to the data of this retrospective study, survival outcomes appear significantly worse than literature evidence [6]. This difference should be probably attributed to the different populations of patients enrolled, with a higher proliferative index and probably a more aggressive disease. Nevertheless, our results show the antitumor activity of irinotecan also in patients with higher Ki-67, but without an important gain in OS, underlining the worse prognosis of this subgroup of patients.
Besides the Ki-67 value, also liver involvement and colorectal primary location tend to negatively impact survival: data showed a moderately difference in OS between patients with and without liver metastases and worse prognosis in patients with NEC arising from the colorectal district, as already demonstrated by Sorbye et al. [13]. However, data did not reach statistical significance, probably due to an exiguous sample of patients analyzed.
This is the first study in which the prognostic significance of serum inflammatory biomarkers is evaluated in patients with NECs. While PLR and NLR did not impact OS and PFS, in our casuistry, lower LMR seems to be associated with poorer PFS, although it did not correlate with poorer OS. This data must be deeply explored in a larger and dedicated trial in order to assess possible major impact on outcome and clinical medical history.
As far as safety is concerned, adverse events were predominantly graded 1-2, and no treatment-related deaths were observed. As described by Sugiyama et al. [14], our data support the good tolerance of IRI-CT with only minor treatment delay. However, the retrospective design of this study could justify the relatively small number of drug-induced a. Kaplan-Meier curves of PFS. Curves include patients with PD and CR/PR/SD to second-line therapy as the best response, respectively. An interesting point is that, despite a cohort of patients affected by such an aggressive disease, 66% of patients received further systemic treatments, thus demonstrating the feasibility of a third-line treatment.

PFS
We have also to outline that the limitation of our study may be related to the retrospective nature of the analysis with relatively small sample size and without an independent pathological review.
However, we presume that the reported data may have value due to the rarity of the disease prompting larger studies on irinotecan as second-line treatment for PD-NECs, particularly in those patients with high Ki-67, which still have an unacceptable prognosis.
Currently, there are several ongoing prospective trials aiming at evaluating the most efficacious chemotherapy schedule for second-line PD-NEC: SENECA trial (Clin-icalTrials.gov Identifier: NCT03387592) in which FOLFIRI is compared to a combination of capecitabine and temozolomide; BEVANEC trial (ClinicalTrials.gov Identifier: NCT02820857) testing the efficacy of bevacizumab in combination with FOLFIRI vs. FOLFIRI alone; NET-02 trial (ClinicalTrials.gov Identifier: NCT03837977) in which the standard second-line for SCLC (docetaxel) is compared to liposomal irinotecan (nal-IRI) and 5-fluorouracil/folinic acid; the EVINEC trial (ClinicalTrials.gov Identifier: NCT02113800) designed to test everolimus in NEC and NET G3 after the failure of first-line platinum-based chemotherapy; cabozantinib in combination with nivolumab/ ipilimumab (ClinicalTrials.gov Identifier: NCT04079712) and avelumab (AVENEC ClinicalTrials.gov Identifier: NCT03147404) have been recently evaluated. Final data of the above studies should clarify the best clinical approach to such aggressive carcinomas.

Conclusion
In conclusion, IRI-CT seems to be a feasible and welltolerated therapeutic option in second-line treatment of EP-NECs, but these findings warrant further confirmation in larger prospective studies.

Data availability
The data sets used and/or analyzed during the current study are available from the corresponding author on reasonable request.