Visual Abstract
Abstract
Everolimus and peptide receptor radionuclide therapy (PRRT, 177Lu-DOTATATE) are 2 treatments recommended in guidelines for gastroenteropancreatic metastatic neuroendocrine tumors. However, the best treatment sequence remains unknown. Methods: We designed a retrospective multicenter study that included patients from the national prospective database of the Groupe d’Étude des Tumeurs Endocrines who had been treated using everolimus and PRRT between April 2004 and October 2022. The primary aim was to compare the 2 treatments (everolimus and PRRT) in terms of efficacy and safety, and the secondary aim was to evaluate the sequences (PRRT followed by everolimus or everolimus followed by PRRT) based on overall progression-free survival (PFS) (PFS during first treatment + PFS during second treatment) in patients with metastatic neuroendocrine tumors. Results: Both treatments were used for 84 patients. The objective response rate and median PFS were 5 (6.0%) and 16.1 mo (95% CI, 11.5–20.7 mo), respectively, under everolimus and 19 (22.6%) and 24.5 mo (95% CI, 17.7–31.3 mo), respectively, for PRRT. The safety profile was also better for PRRT. Median overall PFS was 43.2 mo (95% CI, 33.7–52.7 mo) for the everolimus–PRRT sequence and 30.6 mo (95% CI, 17.8–43.4 mo) for the PRRT–everolimus sequence (hazard ratio, 0.69; 95% CI, 0.39–1.24; P = 0.22). Conclusion: PRRT was more effective and less toxic than everolimus. Overall PFS was similar between the 2 sequences, suggesting case-by-case discussion if the patient is eligible for both treatments, but PRRT should be used first when an objective response is needed or in frail populations.
Metastatic neuroendocrine tumors (mNETs) of the small intestine and the pancreas represent most mNETs (1), followed by lung NETs. The overall survival (OS) rate at 5 y is higher than or close to 50%. Everolimus (2,3) and peptide receptor radionuclide therapy (PRRT, 177Lu-DOTATATE) (4), as well as somatostatin analogs (SSAs) (5), have been approved for well-differentiated gastroenteropancreatic NETs. For lung mNETs, only everolimus is approved, but SSAs and PRRT (when available) are recommended in guidelines (6).
Most patients with gastroenteropancreatic and lung mNETs, when eligible for PRRT, receive these 3 systemic treatments, but the best treatment sequence remains unknown. The results of randomized clinical trials such as COMPETE (NCT03049189) or COMPOSE (NCT04919226), comparing PRRT with everolimus or chemotherapy, are not yet available. In addition, no randomized clinical trial has assessed the question of treatment sequence; although it was the initial primary endpoint of the SEQTOR study, comparing everolimus and chemotherapy (5-fluorouracil and streptozotocin), its aim changed for a direct comparison. In that study, progression-free survival (PFS) was similar between everolimus and the 5-fluorouracil–streptozotocin combination (7).
Therefore, real-world data with long follow-up reporting information on both treatments, everolimus and PRRT, and on the different sequences are important to improve our knowledge with regard to effectiveness and risks of cumulative toxicities.
MATERIALS AND METHODS
Study Population
For the SeqEveRIV study, we identified patients treated in 1 of the 23 participant centers of the national database of the Groupe d’Étude des Tumeurs Endocrines (GTE; Commission Nationale de l’Informatique et des Libertés number 2219168, January 5, 2021) by everolimus and PRRT between April 2004 and October 2022. Inclusion criteria were patients 18 y or older with advanced or metastatic, unresectable, well-differentiated, and histologically confirmed gastroenteropancreatic or lung NETs and with measurable target lesions according to RECIST 1.1. Patients must have started everolimus and PRRT. Patients who received other systemic antitumoral treatment between everolimus and PRRT were excluded. Patients were not excluded if they were treated using SSAs for functional syndrome during the everolimus–PRRT sequences. Patients with poorly differentiated neuroendocrine carcinomas and mixed tumors and patients treated using a combination of everolimus and PRRT were excluded.
Patients were treated within the Endocan–Reseau National de Prise en Charge des Tumeurs Endocrines (RENATEN) clinical network (expert centers grouped in a single network created under the supervision of the French National Cancer Institute in 2009 to promote care and research in rare cancer). This observational study conformed to the ethical guidelines of the 1975 Declaration of Helsinki and was approved December 20, 2022, by the Medical Ethics Research Committee of the Hospices Civils de Lyon (registered number 22_5068). Written information was given to each patient included in the study. Their consent is not required by French law, but patients are informed about their right to withdraw their data from the cohort.
Data Collection
The following parameters were collected retrospectively at the beginning of the treatment sequence: age, sex, body mass index, date of diagnosis, hereditary syndrome and its type, Eastern Cooperative Oncology Group performance status (0–1 or ≥2), primary tumor location, clinical functioning syndrome, serum or plasma chromogranin A levels in μg/L, World Health Organization classification, Ki-67 index (recording the highest value if multiple specimens were analyzed), number and location of the metastases, uptake at somatostatin receptor imaging according to the Krenning classification or the equivalent for 68Ga-DOTATOC with PET (recording the maximum uptake or SUVmax in the case of heterogeneous uptake), and administration of prior locoregional or systemic treatments.
Treatments
Everolimus was given, according to its label, at 10 mg/d, continuously until diagnosis of progression or discontinuation because of side effects or the patient’s or physicians’ choice. 177Lu-DOTATATE was given, as in the NETTER-1 study (4), at an activity of 7.4 GBq, intravenously, every 8 wk for 4 cycles. The physicians were allowed to administer a reduced dose of 5 mg/d of everolimus or an adjusted activity of PRRT according to the patient’s tolerance. That modification was at the discretion of the physicians, who are all doctors practicing in an Endocan–RENATEN center.
The following parameters of treatments were collected: time under treatment, adjustment of doses or activities, concomitant treatment with SSAs, reason for discontinuation (scheduled, progressive disease, toxicity, death, or patient’s choice), objective response rate of the first treatment (ORR1), ORR2, PFS1, PFS2, and time between the 2 treatments. The locoregional treatment administered between PRRT and everolimus, time to failure of the sequence (TTFS), and OS were also evaluated.
Objectives
The primary aim was to compare everolimus and PRRT in terms of efficacy and safety, not just overall but also at the line level, when given first (ORR1 and PFS1 for everolimus 1 [Eve1] and PRRT1) or second (ORR2 and PFS2 for everolimus 2 [Eve2] and PRRT2). The secondary aim was to evaluate the sequences (PRRT followed by everolimus or everolimus followed by PRRT) based on overall PFS (PFS1 + PFS2) in patients with mNETs, as well as TTFS and OS of both sequences.
ORR and PFS were assessed locally according to RECIST 1.1 by cross imaging (MRI or CT scan) performed every 3–4 mo (8). For each treatment, PFS was calculated from the date of treatment initiation to the date of morphologic progression, the start of the next antitumoral treatment, death, or the last follow-up.
TTFS was calculated from the date of the beginning of the sequence to the date of progression after the second treatment or the start of the next antitumoral treatment, or it was censured at the date of death or the last follow-up. OS was calculated from the date of the beginning of the sequence to the date of death or the last follow-up.
Statistical Analysis
Categoric variables are expressed as percentages and were compared using the χ2 test or the Fisher exact test when appropriate. Continuous variables are presented as median and interquartile range (IQR) and were analyzed by the Mann–Whitney U test. Survival curves for PFS, TTFS, and OS were estimated using the Kaplan–Meier method. Comparisons in univariate analyses were performed using the log-rank test for each variable of interest. For continuous parameters, the threshold was defined as the population median. Multivariate analyses using a Cox proportional hazards regression model were performed to identify factors independently associated with prognosis. All significant factors from the univariate analysis (P < 0.10 with a log-rank test) were included in the multivariate analyses. A P value of more than 0.05 was considered statistically significant. The results from the survival analyses are presented with the effect estimates, hazard ratios, and 95% CI. All statistical analyses were performed using SPSS version 17.0 (IBM).
RESULTS
Patient Characteristics
Among the 23 centers of the national GTE database, 12 centers agreed to participate, in which 124 patients were identified. Among them, we excluded 40 patients: 36 patients because they received systemic antitumoral treatments between everolimus and PRRT, 2 patients because they received a combined treatment of everolimus and PRRT, and 2 patients because the duration of treatment was unknown. Therefore, 84 patients were included in our study who were treated by both everolimus and PRRT, of whom 60 patients started the sequence with everolimus and 24 patients started the sequence with PRRT (Fig. 1).
Flowchart.
The characteristics of the 84 patients were similar between the everolimus and the PRRT groups, because the same patients received both treatments. When we separated them by treatment first given, the characteristics were not significantly different between the 2 groups (Table 1). The median age of patients was 60.4 y (IQR, 54.6–65.5 y), and 52.4% were men. Most patients had a performance status of 0 (59%). The locations of primary tumors were jejunum and ileum (60.7%), pancreas (28.6%), and lung (8.3%). Most patients had carcinoid syndrome (55.4%). According to the World Health Organization classification, 27 patients had a NET grade 1 (33.3%) and 54 patients had a NET grade 2 (66.7%). Most patients had at least 2 metastatic sites, and all had liver metastases. Almost two thirds of patients had a Krenning 4 classification (111In-octreotide) or the equivalent on somatostatin receptor images. Most patients underwent a resection of their primary NET (71.4%). The median of prior systemic treatment was 1 (1,2), and most patients received SSAs (88.1%) before the sequence. SSAs were mostly maintained in the 2 sequences.
Characteristics of Patients at First Treatment
Administration and Safety of Everolimus and PRRT
In total, 24 patients were treated by the PRRT–everolimus sequence. Only 1 patient did not receive all PRRT1 injections because of progression; the others received all PRRT1 injections. The median administration time of PRRT1 was 5.5 mo (IQR, 4.5–6.2 mo). These patients then received Eve2 for a median time of administration of 3.7 mo (IQR, 2.1–7.5 mo). Sixteen patients had an everolimus treatment adjustment (66.7%), and the reasons for discontinuation were toxicity (47.6%) and disease progression (42.9%). In total, 60 patients were treated by the everolimus–PRRT sequence. The median administration time of Eve1 was 9.5 mo (IQR, 2.4–18.7 mo). There was at least 1 dose of everolimus adjustment for 32.2% of patients, and all patients had at least 1 temporary discontinuation of everolimus. The main reasons for everolimus discontinuation were toxicity (58.3%) and disease progression (40.0%). Patients then received PRRT2 for a median time of administration of 5.5 mo (IQR, 4.7–5.9 mo). Seven patients had activity adjustment with PRRT2 (11.9%), and the reasons for discontinuation were toxicity (13.8%) and disease progression (3.4%), which was significantly less than under Eve2 (Table 2).
Progress of Sequences and Results
During everolimus administration, the most common adverse events (AEs) were mucositis (48%), 11% with grade 3–4, and diarrhea (43%), 8% with grade 3–4. The other main AEs were anemia (31%), nausea or vomiting (23%), sepsis (21%), thrombopenia (17%), lymphopenia (16%), noninfectious pneumopathy (14%), and hepatic cytolysis (12%). Patients who received Eve2 seemed to present more diarrhea (54%), stomatitis (46%), nausea or vomiting (33%), and anemia (38%) than those who received Eve1 (27%, 34%, 19%, and 26%), respectively. We also observed 2 noninfectious pneumopathies (8%) and 2 acute renal failures requiring hospitalization (9%) in patients who received Eve2 that were not observed in the patients who received Eve1. The most common AEs due to PRRT injections were lymphopenia (41%), 12% with grade 3–4, and nausea or vomiting (35%). The other AEs were thrombopenia (25%), anemia (21%), chronic renal failure (19%), diarrhea (18%), and hepatic cytolysis (17%). One myelodysplasia occurred in the PRRT2 group (1%). In contrast to everolimus, the numbers of AEs were similar between PRRT1 and PRRT2 (Supplemental Table 2 [supplemental materials are available at http://jnm.snmjournals.org]).
Efficacy of Everolimus and PRRT
Among the 84 patients, ORR was significantly higher under PRRT than under everolimus (19/84, 22.6%, vs. 5/84, 6.0%, P = 0.002). Median PFS was numerically longer (P = 0.072): 24.5 mo (95% CI, 17.7–31.3 mo) under PRRT versus 16.1 mo (95% CI, 11.5–20.7 mo) under everolimus (Figs. 2A and 2B). By line, in the PRRT1 group, ORR1 was observed in 6 patients (6/24, 25.0%) and median PFS1 was 16.4 mo (95% CI, 9.2–23.6 mo). In the PRRT1–Eve2 group under everolimus, none of the patients had ORR2 and median PFS2 was 6.8 mo (95% CI, 3.8–9.8 mo). In the Eve1 group, ORR1 was observed in 5 patients (5/60, 8.5%), which was significantly lower than ORR1 under PRRT1 (P = 0.04). Median PFS1 was 18.2 mo (95% CI, 14.0–22.4 mo). In the Eve1–PRRT2 group, 13 patients (22.4%) had ORR2 under PRRT, which was significantly higher than for the Eve2 of the PRRT1–Eve2 group (P = 0.01), and median PFS2 was 26.4 mo (95% CI, 22.6–30.2 mo; Figs. 2C, 2D, and 3).
Overall PFS for everolimus (A) and PRRT (B), and by line: first PFS1 (C) and second PFS2 (D). Eve = everolimus.
TTFS (A) and OS (B) among both groups. Eve = everolimus.
Median overall PFS was 43.2 mo (95% CI, 33.7–52.7 mo) for the Eve1–PRRT2 sequence and 30.6 mo (95% CI, 17.8–43.4 mo) for the PRRT1–Eve2 sequence (hazard ratio, 0.69; 95% CI, 0.39–1.24; P = 0.22). After univariate analysis, a functioning tumor, a NET grade 1 or typical carcinoid, the absence of bone metastasis, and more than 2 prior systemic treatment lines were significantly associated with longer overall PFS. After the multivariate analysis, a functioning tumor and the number of prior systemic treatment lines (<2) remained significantly associated with longer overall PFS (Supplemental Table 3).
Efficacy of the Sequences
The median time between the 2 treatments was 14.0 mo (IQR, 8.4–21.7 mo) in the PRRT1–Eve2 group and 7.3 mo (IQR, 2.3–21.5 mo) in the Eve1–PRRT2 group; there was no significant difference (P = 0.22). Seven patients (29.1%) in the PRRT1–Eve2 group and 22 patients (36.7%) in the Eve1–PRRT2 group received a locoregional treatment between the 2 treatments (Table 2; Fig. 4; Supplemental Table 4). After a median follow-up of 58.8 mo (IQR, 42.9–78.7 mo) for the Eve1–PRRT2 group and 49.7 mo (IQR, 19.2–64.1 mo) for the PRRT1–Eve2 group, median TTFS was 50.1 mo (95% CI, 40.5–59.7 mo) for the Eve1–PRRT2 group and 33 mo (95% CI, 23.5–42.5 mo) for the PRRT1–Eve2 group (hazard ratio, 0.75; 95 CI%, 0.39–1.30; P = 0.27; Figs. 3A and 4).
Swimmer plot showing details of TTFS with PFS1, time between PFS1 and PFS2, locoregional treatment, and PFS2. Eve = everolimus.
After the treatment sequence, 16 patients (66.7%) in the PRRT1–Eve2 group and 36 patients (60.0%) in the Eve1–PRRT2 group received subsequent systemic treatments. They received these systemic treatments after a median time of 1.5 mo (IQR, 0.4–5.9 mo) and 17.8 mo (IQR, 6.6–29.2 mo) from the end of the sequence, respectively (Supplemental Table 5). Among these 16 patients of the PRRT1–Eve2 group, 10 patients received cytotoxic chemotherapy, 1 patient received SSAs, and 1 patient underwent rechallenge using PRRT. Among the 36 patients of the Eve1–PRRT2 group who received further systemic treatments, it was mainly SSAs (12 patients), cytotoxic chemotherapy (9 patients), rechallenge using PRRT (5 patients), or an antiangiogenic agent (3 patients).
Median OS from the beginning of the sequence was 85.9 mo (95% CI, 52.4–119.4 mo) for the Eve1–PRRT2 sequence and 60.7 mo (95% CI, 43.0–78.4 mo) for the PRRT1–Eve2 sequence (hazard ratio, 0.58; 95% CI, 0.28–1.18; P = 0.13; Fig. 3B).
DISCUSSION
The present study found that PRRT improved ORR and PFS compared with everolimus in patients with mNETs. It did not show a significant difference with regard to overall PFS between the 2 sequences.
This study first showed that when patients were treated using everolimus, ORR was in accordance with those reported in phase III studies: 6% in the present study, 2% in RADIANT-2 and RADIANT-4 (3), and 5% in RADIANT-3 (2). Median PFS seemed longer in the present study (16.1 mo) than in RADIANT-3 and RADIANT-4 (11.0 mo). The RADIANT-3 study was performed only for pancreatic mNETs, whereas in the present study population, less than one third of patients had pancreatic mNETs. However, this was in accordance with the RADIANT-2 study (16.4 mo), which included patients with functioning carcinoid NETs and was thus similar to most of the present population. The patients herein had enough uptake on mNETs assessed by somatostatin receptor imaging to receive PRRT, which is a known favorable prognostic factor of survival and could explain the longer PFS (9). With regard to PRRT, ORR (22.6%) was close to the one reported in the NETTER-1 study (18%). This study included mid-gut mNETs, which represent most primary NETs in the present study (4). Median PFS (24.5 mo) after PRRT seems longer than reported in the OCLURANDOM study (20.7 mo) in pancreatic mNETs (longer than under sunitinib) (10). In a large retrospective cohort of patients with gastroenteropancreatic and lung mNETs, ORR was 39% and median PFS was 29 mo, but patients receiving less than 22.2 GBq of 177Lu-DOTATATE were excluded (11). Lastly, longer PFS could have been found in the present study, but patients who had greater PRRT benefit (long PFS) were not included, because they had not yet been treated by everolimus. Therefore, PRRT seemed more efficient than everolimus, regardless of its position in the sequence. Large prospective comparative studies are ongoing, such as the COMPOSE study (NCT04919226, PRRT vs. capecitabine–temozolomide; folinic acid, fluorouracil, and oxaliplatin; or everolimus) and more specifically the COMPETE study (NCT03049189, PRRT vs. everolimus), although they do not study the sequence.
However, because PRRT has a better outcome than that of everolimus, even if given in a later line, overall PFS and TTFS were not statistically different, whether the sequence started with PRRT or everolimus. Eve1 seemed to allow longer duration of treatment, longer PFS, higher ORR, and probably slightly less toxicity than Eve2, whereas PRRT2 was at least at the same level of efficacy as PRRT1 with similar toxicity. The present study did not identify a statistically better therapeutic sequence. In the case of symptomatic tumors, it seems more appropriate to administer the most efficient treatment to obtain an ORR; PRRT could be preferred as a first-line treatment. Because the factors that significantly affected overall PFS were known prognostic factors, they probably could not contribute to the choice of one sequence over another. The present study reported that overall PFS and TTFS were similar between the sequences; thus, it is important to consider the immediate and long-term tolerance of each treatment when choosing the sequence. As expected, we confirmed that everolimus gives more frequent side effects than PRRT. In addition, there was better quality of life when using PRRT versus the control arm in the NETTER-1 study (12), but there was no gain in quality of life when using everolimus versus placebo (13,14). In frail patients or in patients for whom the quality of life is declining, starting with PRRT may be better tolerated than starting with everolimus. The everolimus treatment should be given before the decline of the quality of life. There is an admitted low risk of myelodysplasia syndrome or acute leukemia related to early PRRT (1 case in the present study) that advocates for the use of PRRT in a later line (15); further studies are warranted to evaluate this long-term risk before recommending large use of early PRRT in all patients.
The present study has several limitations. First, although this is a large cohort of patients treated by everolimus and PRRT, the number of patients remains too small to perform subgroup analyses or propensity analysis; only 12 of the 23 Endocan–RENATEN centers agreed to participate in this study (not enough time to do the work). In addition, this study was not randomized but used the methodology in which each patient was his or her own comparator, so the bias due to different patient characteristics between the 2 treatments was limited. Second, because PRRT was introduced more recently than everolimus, the group of patients treated first by PRRT is relatively small (29%) compared with what can be expected with a study repeated later. This can decrease the power of the statistical analysis. Third, we focused on these 2 approved systemic treatments, given after SSAs, but locoregional treatments, especially for mid-gut mNETs, are also an important tool against mNETs. In the present study, the use of locoregional treatments between the 2 systemic treatments did not affect PFS of each treatment but may have affected TTFS (more patients received locoregional treatment in the everolimus–PRRT group; Fig. 4; Supplemental Table 4). Fourth, it is difficult in a retrospective study to evaluate the selection criteria of initial arm allocation (PRRT1 vs. Eve1) performed by each physician, which could be related to disease factor, treatment availability, safety, and administration profiles of treatments. In addition, we did not collect the imaging performed before the sequence. Therefore, even if everolimus or PRRT is initiated in a case of disease progression, according to France’s national guidelines, we are not able to demonstrate that all patients had a morphologic progression, according to RECIST 1.1, before the sequence. For all of these reasons, results must be interpreted with cautious, because it is not a prospective randomized study.
CONCLUSION
PRRT was more effective and less toxic than everolimus. Overall PFS (PFS1 + PFS2) was the same for both sequences, and PRRT seemed more efficient in a later line. When the patient is eligible for both treatments, the best sequence must be discussed within a dedicated multidisciplinary tumor board case by case, according to the pros and cons of each treatment, but PRRT should be used first when an objective response is needed or in frail populations.
DISCLOSURE
This study did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. However, patients were identified from the French national database of the GTE, which is partially supported by Ipsen Pharma. This study was granted by GTE prize 2022. Karine Bouhier-Leporrier receives financial compensation from Novartis and Servier, Dierickx Lawrence and Thierry Lecomte receive financial compensation from Ipsen and Novartis, Cottereau Anne-Ségolène receives financial compensation from Novartis, and Thomas Walter receives financial compensation from Novartis, Ipsen, Keocyte, and Terumo. No other potential conflict of interest relevant to this article was reported.
KEY POINTS
QUESTION: Is there a better sequence of PRRT and everolimus in the treatment of mNETs?
PERTINENT FINDINGS: In total, 84 patients received everolimus and then PRRT or the reverse sequence, with each patient as his or her own comparator, to evaluate the efficacy and safety of both treatments. ORR, PFS, and safety were greater with PRRT than with everolimus. However, because this was the case in first- or second-line treatment, the efficacies of the 2 sequences were similar.
IMPLICATIONS FOR PATIENT CARE: When a patient is eligible for both treatments, PRRT should be used first in frail populations or when an objective response is needed.
ACKNOWLEDGMENTS
We thank the patients and their association (Association des Patients Porteurs de Tumeurs Endocrines Diverses) for their participation. We also thank the Endocan–RENATEN clinical network constructed and supported by the GTE. We thank Shanez Haouari (Direction de la Recherche en Santé, Hospices Civils de Lyon) for help in manuscript preparation.
Footnotes
Published online Aug. 1, 2024.
- © 2024 by the Society of Nuclear Medicine and Molecular Imaging.
REFERENCES
- Received for publication January 13, 2024.
- Accepted for publication June 5, 2024.
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.