Abstract
A rechallenge is common after the initial efficacy of alkylating-based chemotherapy (ALK) in pancreatic neuroendocrine tumors (PanNET). High MGMT expression seems associated with a lower response to ALK. We aimed to evaluate the efficacy and toxicity of ALK rechallenge in PanNET, and to assess the evolution of MGMT expression under ALK. All consecutive patients with advanced PanNETs who received initial ALK (achieving tumor control) followed by a pause of > 3 months, then an ALK rechallenge (ALK2) upon progression were retrospectively studied (cohort A). The primary endpoint was progression-free survival under ALK2 (PFS2). The MGMT expression was retrospectively assessed by immunohistochemistry (H-score) in consecutive PanNET surgically resected following ALK (cohort B). We found that Cohort A included 62 patients (median Ki67 8%), for whom ALK1 followed by a pause achieved an objective response rate of 55% and a PFS1 of 23.7 months (95% IC, 19.8–27.6). ALK2 achieved no objective response and stability in 62% of patients. The median PFS2 was 9.2 months (IC 95% 7.1–11.3). At multivariable analysis, a hormonal syndrome (P = 0.032) and a pause longer than 12 months (P = 0.041) were associated with a longer PFS2. In cohort B (17 patients), the median MGMT H-score increased from 45 (IQR 18–105) before ALK to 100 (IQR 56–180) after ALK (P = 0.003). We conclude that after the initial efficacy of ALK treatment, a pause followed by ALK rechallenge might be appropriate to prolong tumor control, improve quality of life and limit long-term adverse events. Increased MGMT expression under ALK might explain the low efficacy of ALK rechallenge.
Introduction
Pancreatic neuroendocrine tumors (PanNET) are a heterogeneous group of rare neoplasms. However, in the past few decades, their incidence has increased for all grades and stages (Dasari et al. 2017). Approximately 20% of PanNET are functioning, defined by hormone hypersecretion of mainly insulin, glucagon, gastrin or vasoactive intestinal peptide. More than half of PanNET are diagnosed with synchronous metastases. Nevertheless, due to their relatively slow growth, patient survival is generally long even at the metastatic stage with a 5-year overall survival rate of approximately 50% (Ter-Minassian et al. 2013, Nuñez-Valdovinos et al. 2018).
The therapeutic strategy of advanced PanNET is guided by the main prognostic factors, which include the histopathological grade and Ki67 index, location of the metastases, volume and growth rate, the existence of a functioning syndrome and performance status (Pavel et al. 2016, de Mestier et al. 2020a). Systemic chemotherapy is recommended in patients with PanNET and progressive and/or symptomatic metastases, liver invasion > 50% and/or Ki67 > 10%, especially if the main therapeutic objective is reducing tumor bulk (Pavel et al. 2016, de Mestier et al. 2020a). Combinations of alkylating agents (dacarbazine (DTIC) or temozolomide (TEM)) and fluoropyrimidines (5-fluorouracil (5FU) or capecitabine (CAP)) have been found to be effective in this setting, with an objective response rate of between 40 and 50% and median progression-free survival (PFS) between 12 and 18 months (Lu et al. 2018, de Mestier et al. 2019).
Extended survival of patients with metastatic PanNET includes numerous successive treatments with potential cumulative long-term toxicity. TEM and DTIC can cause hematotoxicity, and prolonged treatment may favor long-term myelosuppression (Armstrong et al. 2009). Because this potential toxicity must be taken into account in clinical practice, a therapeutic pause is usually observed after obtaining tumor control with alkylating-based chemotherapy. Alkylating-based chemotherapy (ALK) is then generally rechallenged when progression occurs because it was initially effective to achieve PanNET control again. While this strategy has been validated in malignant gliomas (Wick et al. 2009, Perry et al. 2010, Kuczynski et al. 2013), it has never been evaluated in advanced PanNET.
Cytotoxicity induced by alkylating agents is mainly due to methyl adducts on O6-guanines, which are repaired by the O-6-methylguanine-methyltransferase (MGMT) enzyme (Koumarianou et al. 2015, de Mestier et al. 2020b). In PanNET, high expression of MGMT seems to be associated with a lower response and shorter PFS under ALK (Walter et al. 2015, Cros et al. 2016, Campana et al. 2018, de Mestier et al. 2020b). However, the impact of ALK on MGMT expression in PanNET remains unknown.
The main objective of this study was to assess the efficacy and tolerance of the rechallenge of ALK (5FU-DTIC or TEM-CAP) in patients with advanced PanNET and to explore potential factors associated with its efficacy. The secondary objective was to explore the impact of ALK on MGMT expression in PanNET.
Patients and methods
Patients
The records of all consecutive patients with advanced unresectable PanNET treated with an ALK (TEM or DTIC) between 2006 and 2019 in three French expert centers (Robert Debré University Hospital, Reims; Edouard Herriot University Hospital, Lyon and Beaujon University Hospital, Clichy) were retrospectively reviewed (cohort A). The inclusion criteria required histological proof of locally advanced unresectable well-differentiated PanNET G1, G2 or G3 or with metastatic extension at the time of initial treatment with ALK (ALK1). ALK1 included initially effective (at least stabilization) TEM ± CAP or DTIC ± LV5FU2 followed by a therapeutic pause (for whatever reason) of at least 3 months with no cytotoxic treatment. At disease progression, this pause was followed by rechallenge with TEM ± CAP or DTIC ± LV5FU2 (ALK2). Exclusion criteria were poorly differentiated neuroendocrine carcinomas, a therapeutic pause of less than 3 months or cytotoxic treatment received during the pause (excepted for primary or metastatic surgery, radiofrequency ablation or somatostatin analogs).
In a separate cohort (cohort B), tumor expression of MGMT before and after treatment with ALK (surgical specimen or biopsy) was determined in a single-center retrospective surgical cohort (Beaujon hospital). We included all consecutive patients with advanced PanNET who underwent surgical resection between 2010 and 2019 (primary PanNET and/or metastases), after neoadjuvant ALK treatment consisting in TEM +/− CAP, with a pathological sample taken before ALK treatment. In this cohort B, the objective of ALK treatment was tumor control or downstaging as part of a curative-intent surgical resection. No cytotoxic treatment, other than ALK, could be received between the dates of the two pathological samples.
Treatments
DTIC was administered intravenously at a dose of 400 mg/m2, alone or combined with leucovorin (200 mg/m2 on days 1–2), bolus 5FU (400 mg/m2 on days 1–2) and continuous 5FU (1200 mg/m2 on days 1–2), every 21 days. TEM was administrated orally at the dose of 200 mg/m2 on days 10–14 (150 mg/m2 for cycle 1), alone or combined with CAP (750 mg/m2 twice a day on days 1–14) every 28 days. All patients were prescribed ondansetron on the days chemotherapy was administered to prevent nausea/vomiting (de Mestier et al. 2019).
The indications for ALK1 and ALK2 were decided during NET tumor board meetings. A standard clinical and biological evaluation was performed before each treatment cycle. All patients underwent high-quality contrast-enhanced CT scan within 6 weeks before ALK1 initiation and then every 3 months during ALK1, pause and ALK2, or earlier in case of symptoms suggesting progression. Imaging data were systematically reviewed by experienced radiologists during multidisciplinary NET tumor board meetings.
Data collection
Clinical, therapeutic, radiological and pathological data prior to ALK1 and ALK2, as well as follow-up data were collected. The histopathological grade was assessed centrally according to the 2019 WHO classification (G1 with a Ki-67 index < 3%, G2 with a Ki-67 index of 3–20%, or G3 with a Ki-67 index > 20%) (WHO Classification of Tumours 2019). The ALK1 and ALK2 doses received were calculated as the average dose rate of the optimal recommended dose. Radiological response and progression according to RECIST 1.1 were reviewed for all patients in NET multidisciplinary meetings. The severity of adverse events was graded according to CTCAE v4. Anonymous data collection was performed after receiving the patient agreement, according to the Declaration of Helsinki and following Institutional Review Board approval (CEERB Paris Nord, IRB no. 00006477-15-073).
Immunochemistry
In surgical or biopsy specimens from patients included in cohort B, nuclear expression of the MGMT protein was assessed by immunohistochemistry (clone MT3.1, 1/25, Thermo Scientific®) using the H-score, which combines the intensity of expression (0: absent, 1: low, 2: medium, 3: high) and the rate of positive cells (0–100%), and ranges from 0 to 300 (Cros et al. 2016).
Statistical analyses
Quantitative variables were expressed as medians (25–75 interquartile range (IQR)) and compared using the Wilcoxon test. Qualitative variables were expressed as frequencies (percentages) and compared using the Chi-square test. Progression-free survival from ALK1 (PFS1) was defined as the time between the first ALK1 administration and the date of progression (and thus included the pause). PFS2 was defined as the time between the first ALK2 administration and the date of documented progression or death whichever occurred first. Patients alive and with no disease progression at the final follow-up were censored. The cut-off date was December 31, 2019. Survival rates were estimated using the Kaplan–Meier method. Factors associated with a risk of tumor progression after ALK rechallenge (shorter PFS2) were analyzed using Cox proportional hazard models. All clinically relevant baseline variables were entered into univariate analysis, and all non-collinear variables with a P-value < 0.20 in univariate analysis were entered into multivariate analysis. Factors associated with disease stabilization after rechallenge were studied using a logistic regression model with the same methodology. Tolerance (Common Terminology Criteria for Adverse Events, version 4) to initial treatment and at reintroduction was compared by the Mc Nemar test.
In cohort B, we compared MGMT expression (H-scores) of tumor samples taken before and after ALK treatment by a Wilcoxon rank test for paired samples.
All analyses were two-sided. P-values < 0.05 were considered to be significant. Analyses were performed with SPSS® (version 20; IBM™) and Prism® (version 6; Graphpad™) softwares.
Results
Patients (cohort A)
Sixty-two patients with advanced PanNET were included (Fig. 1). The sex ratio was balanced, 16% of patients were presented with a hormonal syndrome including five gastrinomas, four glucagonomas and one insulinoma (Table 1). All PanNETs were well-differentiated and were mostly G2 (71%) or G3 (11%), with a median Ki67 of 8%. Eighteen patients (29%) had extra-hepatic metastases. Thirty-five patients had received antitumor treatment before ALK1, mostly chemotherapy or somatostatin analogs, and 44% of patients were treatment-naive. Surgery (primary PanNET and/or metastases) had been previously performed in 27 patients. When ALK1 was initiated, most patients (62%) had a performance status of 0.
Flow chart of the study – cohort A and B.
Citation: Endocrine-Related Cancer 28, 7; 10.1530/ERC-21-0034
Flow chart of the study – cohort A and B.
Citation: Endocrine-Related Cancer 28, 7; 10.1530/ERC-21-0034
Flow chart of the study – cohort A and B.
Citation: Endocrine-Related Cancer 28, 7; 10.1530/ERC-21-0034
Baseline characteristics of the 62 patients with PanNET who underwent alklyating agent rechallenge.
| Parameter | Value | |
|---|---|---|
| Female gender, n (%) | 28 (45) | |
| Performance status, n (%) | PS 0 | 37 (62) |
| PS 1–2 | 23 (38) | |
| Age (year), median (IQR) | 62 (52–68) | |
| Ki67, median (IQR)* | 8% (5–15) | |
| Grade, n (%) | G1 | 5 (8) |
| G2 | 44 (71) | |
| G3 | 7 (11) | |
| Not available | 6 (10) | |
| Metastatic sites, n (%) | Bilobar hepatic metastasis | 50 (81) |
| Extra-hepatic metastasis | 18 (29) | |
| Prior treatments, n (%) | Primary tumor surgery | 26 (42) |
| Metastasis surgery or ablation | 14 (23) | |
| Chemotherapy** | 22 (36) | |
| Somatostatin analogs | 18 (29) | |
| Targeted therapy | 8 (13) | |
| Transarterial liver embolization | 6 (10) | |
| None (excluding surgery) | 27 (44) | |
*Seven missing data; **previous chemotherapy was streptozotocin-based (n = 9), oxaliplatin-based (n = 5), etoposide-platinum (n = 5) or unknown (n = 3).
ALK1 included TEM or DTIC in 39 (63%) and 23 (37%) patients, respectively, with a median number of eight cycles (Fig. 2). There were no differences between patients receiving DTIC or TEM as ALK1, except that the former were more frequently pretreated (73.9% vs 46.2%, P = 0.03) and had less frequently a performance status of 0 (39.1% vs 71.8%, P = 0.016). Among the patients included in this study, ALK1 resulted in an objective response (complete or partial) rate of 55% (74.4% with TEM and 21.7% with DTIC, P < 0.001). The pause consisted of a planned treatment holiday in all patients and lasted a median of 14.7 months, during which seven patients underwent a tumor resection/destruction procedure, and 16 patients received somatostatin analogs. The median total follow-up was 44.6 months (IQR 29.8–66.5) from the beginning of ALK1. Median PFS1, including the therapeutic pause, was 23.7 months (95% IC, 19.8–27.6) (Fig. 3).
Initial alkylating-based chemotherapy (ALK1), pause and alkylating rechallenge (ALK2) in 62 patients with advanced PanNET.
Citation: Endocrine-Related Cancer 28, 7; 10.1530/ERC-21-0034
Initial alkylating-based chemotherapy (ALK1), pause and alkylating rechallenge (ALK2) in 62 patients with advanced PanNET.
Citation: Endocrine-Related Cancer 28, 7; 10.1530/ERC-21-0034
Initial alkylating-based chemotherapy (ALK1), pause and alkylating rechallenge (ALK2) in 62 patients with advanced PanNET.
Citation: Endocrine-Related Cancer 28, 7; 10.1530/ERC-21-0034
Progression-free survival after first(ALK1) and second (ALK2) alkylating based-treatment
Citation: Endocrine-Related Cancer 28, 7; 10.1530/ERC-21-0034
Progression-free survival after first(ALK1) and second (ALK2) alkylating based-treatment
Citation: Endocrine-Related Cancer 28, 7; 10.1530/ERC-21-0034
Progression-free survival after first(ALK1) and second (ALK2) alkylating based-treatment
Citation: Endocrine-Related Cancer 28, 7; 10.1530/ERC-21-0034
Rechallenge: treatment received and tolerance
ALK2 included TEM or DTIC in 40 (64%) and 22 (36%) patients, respectively, with a median of four cycles (IQR 3–8) (Fig. 2). ALK2 introduction was motivated by RECIST-defined tumor progression in all patients. When ALK2 was initiated, 46% of patients had a performance status of 0 and 27% had extra-hepatic metastases. Toxicity was similar with both TEM or DTIC whether it was ALK1 or ALK2. The most frequent adverse events observed with ALK1 and ALK2 were nausea (48 and 44%, respectively), asthenia (52 and 40%, respectively) and thrombocytopenia (both 20%) (Table 2). Nevertheless, grade 3–4 adverse effects were only observed in seven (12%) patients. There was no significant difference in the individual tolerance to ALK2 and ALK1.
Adverse events (CTCAE 4) during initial (ALK1) and second (ALK2) alkylating chemotherapy for advanced PanNET.
| Grade 1–2 ALK1 | Grade 3–4 ALK1 |
Grade 1–2 ALK2 | Grade 3–4 ALK2 | P-value (ALK2 vs ALK1) |
|
|---|---|---|---|---|---|
| Nausea | 27 (45%) | 2 (3%) | 25 (42%) | 1 (2%) | 0.267 |
| Diarrhea | 6 (10%) | 3 (5%) | 9 (15%) | 0 | 1 |
| Neutropenia | 9 (15%) | 1 (2%) | 3 (5%) | 2 (3%) | 0.386 |
| Thrombopenia | 9 (15%) | 3 (5%) | 10 (17%) | 2 (3%) | 1 |
| Anemia | 11 (18%) | 0 | 6 (10%) | 0 | 0.444 |
| Mucositis | 8 (13%) | 0 | 3 (5%) | 0 | 1 |
| Asthenia | 31 (52%) | 0 | 23 (38%) | 1 (2%) | 0.286 |
| Palmar-plantar erythro-dysesthesia | 5 (8%) | 0 | 3 (5%) | 1 (2%) | 1 |
Best morphological response under ALK2
Among the 60 patients evaluated for best morphological response, ALK2 achieved tumor stabilization in 37 patients (62%) but with no objective response (Fig. 2). On univariate analysis, the probability of disease stabilization under ALK2 treatment was significantly higher in case of a longer pause between ALK1 and ALK2 (OR 1.05; IC 95% (1–1.01); P = 0.038) (Table 3). The presence of a hormonal syndrome and a performance status of 0 were also associated with a higher probability of disease stabilization under ALK2, but this was not statistically significant. Disease stabilization under ALK2 was not influenced by the type of ALK received either as ALK1 or ALK2. On multivariate analysis, a longer pause between ALK1 and ALK2 was the only variable independently associated with a significantly higher probability of disease stabilization under ALK2 (for each added month of the pause, OR 1.06; 95% CI (1–1.12; P = 0.034).
Univariate and multivariate analysis of factors associated with disease stabilization after alkylating rechallenge in patients with PanNET.
n |
Univariate | Multivariate | |||||
|---|---|---|---|---|---|---|---|
| OR | IC 95% | P-value | OR | IC 95% | P-value | ||
| Hormonal Syndrome | 60 | 7.07 | 0.83–60.1 | 0.073 | 7.47 | 0.81–69.2 | 0.077 |
| ALK1 duration (each additional month) | 60 | 0.92 | 0.81–1.03 | 0.146 | 0.92 | 0.79–1.07 | 0.293 |
| Pause duration (each additional month) | 60 | 1.05 | 1.00–1.01 | 0.038 | 1.06 | 1.00–1.12 | 0.034 |
| Age at ALK2 initiation (each additional year) | 60 | 0.97 | 0.92–1.02 | 0.206 | 0.98 | 0.93–1.04 | 0.554 |
| Performance status 1–2 at ALK2 initiation | 59 | 0.36 | 0.12–1.08 | 0.067 | 0.47 | 0.13–1.69 | 0.250 |
| Extra-hepatic metastases at ALK2 initiation | 60 | 0.48 | 0.16–1.44 | 0.192 | 0.93 | 0.23–3.67 | 0.912 |
| Ki-67 index (each additional 1%) | 55 | 0.98 | 0.93–1.03 | 0.449 | – | – | – |
| DTIC1 vs TEM1 | 60 | 0.84 | 0.29–2.47 | 0.755 | – | – | – |
| DTIC2 vs TEM2 | 60 | 1.39 | 0.46–4.22 | 0.560 | – | – | – |
Bold indicates statistical significance, P < 0.05.
Factors associated with prolonged PFS2
The median follow-up was 17.2 months (IQR 7.3–31.2) from the beginning of ALK2. Median PFS2 was 9.2 months (IC 95% 7.1–11.3) (Fig. 3). On univariate analysis, factors significantly associated with a prolonged PFS2 were the presence of a hormonal syndrome (HR 0.40; 95% CI (0.18–0.88); P = 0.023) and a pause longer than 12 months (HR 0.57; 95% CI (0.33–1.00); P = 0.049), while extra-hepatic metastases were associated with an increased risk of progression (HR 1.89; 95% CI (1.05–1.43); P = 0.035) (Table 4). The duration of ALK1 showed no influence on PFS2, whether it was explored according to the median or quartiles (data not shown). PFS2 was not influenced by tumor response achieved with ALK1, duration of ALK1, nor by the type of ALK (TEM or DTIC) received as ALK1 or ALK2.
Univariate and multivariate Cox proportional hazard analyses of factors associated with prolonged progression-free survival after alklyating agent rechallenge.
n |
Univariate | Multivariate | |||||
|---|---|---|---|---|---|---|---|
| HR | 95% CI | P-value | HR | 95% CI | P-value | ||
| Hormonal syndrome | 62 | 0.40 | 0.18–0.88 | 0.023 | 0.4 | 0.17–0.93 | 0.032 |
| Pause duration > 12 months | 62 | 0.57 | 0.33–1.00 | 0.049 | 0.53 | 0.28–0.98 | 0.041 |
| Age at ALK2 initiation (each additional year) | 62 | 1.02 | 0.99–1.04 | 0.210 | 1.02 | 1.00–1.05 | 0.100 |
| Performance status 1–2 at ALK2 initiation | 61 | 1.45 | 0.83–2.54 | 0.188 | 1.07 | 0.59–1.95 | 0.818 |
| Extra-hepatic metastases at ALK2 initiation | 62 | 1.89 | 1.05–3.43 | 0.035 | 1.66 | 0.88–3.12 | 0.117 |
| Ki-67 index (each additional 1%) | 55 | 0.99 | 0.96–1.02 | 0.403 | – | – | – |
| Objective response obtained with ALK1 | 62 | 1.15 | 0.66–2.01 | 0.613 | – | – | – |
| Bilobar hepatic metastasis | 62 | 0.83 | 0.41–1.67 | 0.600 | – | – | – |
| Prior treatment by chemotherapy | 62 | 1.29 | 0.73–2.28 | 0.375 | – | – | – |
| Prior treatment by somatostatin analogs | 62 | 0.85 | 0.46–1.55 | 0.589 | – | – | – |
| Tumor grade | |||||||
| Grade 1 | 56 | 1 | – | – | – | – | – |
| Grade 2 | 56 | 1.65 | 0.51–5.39 | 0.404 | – | – | – |
| Grade 3 | 56 | 1.43 | 0.38–5.46 | 0.600 | – | – | – |
| DTIC1 vs TEM1 | 62 | 1.16 | 0.66–2.04 | 0.607 | – | – | – |
| DTIC2 vs TEM2 | 62 | 0.99 | 0.56–1.75 | 0.958 | – | – | – |
Bold indicates statistical significance, P < 0.05.
On multivariate analysis, the presence of a hormonal syndrome (HR 0.40; 95% CI (0.17–0.93); P = 0.032) and a pause of more than 12 months (HR 0.53; 95% CI (0.28–0.98); P = 0.041) were independently associated with a significantly decreased risk of progression.
Impact of ALK on PanNET expression of MGMT
Increasing evidence has reported that TEM is more efficient in PanNET with low MGMT expression. Hence, we then sought to explore whether MGMT expression may secondarily increase under the effect of ALK, to potentially explain, at least partly, why ALK rechallenge is less effective. Only three patients included in the above-described cohort were operated on during the therapeutic pause with an available pathological specimen. In those patients, the MGMT expression h-score increased between the pre-ALK1 and pre-ALK2 specimens (30 to 90, 100 to 180 and 0 to 200, respectively).
Therefore, the influence of ALK on MGMT expression was evaluated in a distinct series of 18 patients (cohort B) operated on for PanNET after ALK treatment with TEM +/− CAP (Fig. 1). The median H-score of MGMT significantly increased from 45 (IQR 18–105) on pre-ALK specimens to 100 (IQR 56–180) on post-ALK specimens (P = 0.003) (Fig. 4). Conversely, no significant variation in median Ki67 was observed (5.0%, IQR (3–10) to 5.8%, IQR (2–11), P = 0.188). Patients with increased MGMT expression were characterized by a shorter delay between ALK exposure and surgery than patients with stable MGMT expression over time (13.9 months vs 3.6 months, respectively, P = 0.038).
Comparison of MGMT expression before and after TEM +/- CAP in 18 consecutive patients with advanced PanNET
Citation: Endocrine-Related Cancer 28, 7; 10.1530/ERC-21-0034
Comparison of MGMT expression before and after TEM +/- CAP in 18 consecutive patients with advanced PanNET
Citation: Endocrine-Related Cancer 28, 7; 10.1530/ERC-21-0034
Comparison of MGMT expression before and after TEM +/- CAP in 18 consecutive patients with advanced PanNET
Citation: Endocrine-Related Cancer 28, 7; 10.1530/ERC-21-0034
Discussion
This study shows that rechallenge with ALK is feasible in patients with advanced PanNETs, which concerned 37.8% of all patients treated by ALK in the three centers in the study period. Although it was less effective than the initial treatment with ALK and did not result in an objective response, this strategy resulted in a stabilization rate of 62% and a PFS of 9.2 months, with increased efficacy in case of a pause > 12 months and acceptable tolerance. The global strategy including ALK-based chemotherapy, a treatment holiday and rechallenge of ALK may delay the time until another treatment is introduced, which is relevant when the goal of the management strategy is to obtain the longest tumor control possible. Still, these results must be interpreted with caution as we only included patients who achieved tumor control with initial ALK.
To the best of our knowledge, this is the first study to evaluate the efficacy of rechallenge of ALK in a large cohort of advanced PanNET. Recently, Wu et al. 2020 reported results in 42 patients with PanNET who achieved an 'exceptional response' to ALK, defined as tumor control for at least 9 months after discontinuation of the agent. Eleven of these patients underwent ALK rechallenge at progression, with prolonged tumor control (and another treatment pause) in two of them. These results, along with ours, show that (1) some patients with PanNET can achieve sustained responses to ALK even after discontinuing therapy and (2) a treatment pause could be beneficial for some patients and could probably improve their quality of life and reduce the risk of the effects of cumulative toxicity such as myelosuppression. ALK2 might stand as an alternative to switching to another treatment. One way to confirm the relevance of the ALK rechallenge would be to compare the global tumor control duration between (1) ALK rechallenge and then new treatment or (2) a new treatment, after progression during a pause following ALK1 treatment. Again, the median PFS1 was 23.6 months in our study, which is longer than that in the literature (from 13 to 16 months (Cives et al. 2016, Campana et al. 2018, de Mestier et al. 2020b)) because only patients with prolonged tumor control following ALK – who are the only appropriate candidates for a rechallenge – were included.
The choice of the alkylating agent (DTIC or TEM) could have depended on the center. Nevertheless, neither PFS2 (our main outcome) nor disease stabilization under ALK2 was influenced by the type of ALK received either as ALK1 or ALK2. The optimal duration of ALK in patients has not been determined (Chatzellis et al. 2019). In clinical trials testing chemotherapy agents, chemotherapy is usually administered until progression or toxicity (Mitry et al. 2014, Kunz et al. 2018) except that in the REMINET study which recently explored the interest of maintenance treatment with lanreotide after a short course of chemotherapy (Lepage et al. 2020). Moreover, the toxicity of prolonged ALK and the potential risk of secondary malignancies are important factors, as seen in TEM-treated glioblastoma patients (Noronha et al. 2006). It is very important to preserve the bone marrow of patients with PanNET for future treatments due to prolonged survival and many treatment options, such as PRRT. In a series of 60 patients treated with TEM-CAP, Lamarca et al. (Lamarca et al. 2020) reported that achieving an objective response significantly prolonged PFS (multivariate HR 0.2; 95% CI (0.1–0.6); P = 0.001). However, maintenance of TEM-CAP beyond six cycles did not prolong PFS (HR 1.12; 95% CI (0.40–3.11); P = 0.826). In addition, most partial responses were achieved during the first 6 months of treatment (median time to maximal response, 3.7 months), further suggesting that six cycles of maintenance therapy may not significantly improve the antitumor efficacy of these agents. Additional studies are needed to define the optimal initial duration and rechallenge strategy of ALK and to validate predictive markers of efficacy for a maintenance strategy or a pause-rechallenge strategy.
A pause longer than 12 months was associated with prolonged tumor control at rechallenge. This might be the reflection of the more indolent natural course of the tumor. However, tumor features reflecting biological aggressiveness, such as Ki67 and grade, did not influence PFS2 or stabilization under ALK2, suggesting that other factors may be implicated. The efficacy of ALK1 was not predictive of tumor stabilization after rechallenge or prolonged PFS2. Among possible explanations, an increase in MGMT expression during ALK could be a mechanism of acquired resistance. MGMT repairs O6-methylguanine DNA lesions induced by alkylating agents; its low expression is associated with the efficacy of alkylating agents in PanNET (de Mestier et al. 2020b). In cohort B, MGMT expression increased after ALK administration in PanNET. This could be due to ALK-induced clonal selection, mutagenicity and/or changes in tumor methylation profile, as reported in glioblastomas (Zehir et al. 2017, Feldheim et al. 2019). In cohort B, the delay between ALK exposure and the surgical sample was shorter for PanNET with increased MGMT expression over time, compared to those with stable MGMT expression over time. This appears to support the results in cohort A, in which a longer therapeutic pause was associated with a more effective alkylating rechallenge. The other predictive factor associated with longer PFS2 was the presence of a functioning syndrome. Although there is no clear explanation for the association of hormonal syndrome with prolonged tumor control of ALK2, hypotheses include specific biological and methylation profiles (How-Kit et al. 2015). Still, this result should be interpreted with caution due to the low number of patients with functioning PanNET and the possibility of unconsidered cofounding factors.
A few studies have reported on the effect of TEM over MGMT regulation in glioblastoma, for which TEM is a validated therapeutic option. In glioblastoma as in PanNET, the methylation of MGMT gene promoter has been associated with increased sensitivity to TEM, as for decreased MGMT protein expression. A decrease in the rate of MGMT gene methylation has been reported in recurrent glioblastoma, in comparison with primary tumors initially treated with TEM suggesting that the development of reduced methylation in the MGMT promoter is one of the mechanisms for acquiring therapeutic resistance after TEM treatment in glioblastoma (Park et al. 2012, Storey et al. 2019). However, mechanisms of acquired TEM resistance in tumor recurrences remain unclear since multiple levels of epigenetic regulation of the MGMT gene may exist, including promoter methylation as well as chromatin remodeling (Kitange et al. 2012).
Conversely, to our knowledge, only one preliminary study has reported that exposure to TEM induced increased MGMT expression and further resistance to TEM in initially sensitive PanNET cell lines (Blazevic et al. 2018). In PanNET, the most important mechanism regulating MGMT expression might not be amplification but methylation of the MGMT promoter (de Mestier et al. 2020b). This process is dynamic and might explain a quick increase in MGMT expression following TEM exposure and hypothetically, a decrease with increasing time elapsed from TEM exposure. The accessibility of chromatin, depending on histone modifications, might also be implicated in the capacity to acquire resistance to TEM.
The main limitation of this study is its retrospective design. Our study carries inherent selection biases and could be at best hypothesis-generating. In particular, the duration of ALK before a therapeutic pause and the choice of alkylating agents DTIC or TEM were not standardized. We included patients with various tumor grades and previous treatments, and we believe that our real-life cohort of consecutive patients is comprehensive and reflects the practices of expert centers. Although our total follow-up was nearly 4 years, longer follow-up is needed to evaluate the potential long-term toxicity (especially myelosuppression) of this strategy.
After initial ALK treatment (achieving at least tumor stabilization), a pause followed by ALK rechallenge was well-tolerated and might prolong tumor control or delay the time until another treatment is required. Should it be confirmed prospectively, this strategy may improve quality of life and limit long-term adverse events. Increased MGMT expression under ALK is a hypothesis that might explain the low efficacy of ALK rechallenge.
Declaration of interest
Ophélie De Rycke: none; Thomas Walter: Novartis, IPSEN, Keocyt, Roche, Abbvie, SIRTEX; Marine Perrier: none; Olivia Hentic: AAA, Keocyt, Ipsen, Novartis, Pfizer; Catherine Lombard-Bohas: AAA, Ipsen, Novartis, Pfizer; Romain Coriat: Ipsen, Novartis, AAA, Pfizer; Speaker: AAA, Ipsen, Novartis; Guillaume Cadiot: AAA, Keocyt, Ipsen, Novartis, Pfizer; Anne Couvelard: none ; Philippe Ruszniewski: Ipsen, Novartis, AAA, ITM, Keocyt; Jérôme Cros: Ipsen; Louis de Mestier: Ipsen, Keocyt, Novartis, Pfizer.
Funding
Dr Louis de Mestier received financial support from ITMO Cancer AVIESAN (Alliance Nationale pour les Sciences de la Vie et de la Santé/National Alliance for Life Sciences & Health) within the framework of the Cancer Plan.
References
Armstrong TS, Cao Y, Scheurer ME, Vera-Bolaños E, Manning R, Okcu MF, Bondy M, Zhou R & Gilbert MR 2009 Risk analysis of severe myelotoxicity with temozolomide: the effects of clinical and genetic factors. Neuro-Oncology 11 825–832. (https://doi.org/10.1215/15228517-2008-120)
Blazevic A, Dogan-Oruc F, Dedeci M, van Koetsveld PM, Feelders RA, De Herder WW & Hofland LJ 2018 The effect of temozolomide on pancreatic neuroendocrine tumors in vitro and role of MGMT and MMR system in temozolomide resistance. Neuroendocrinology 106 42.
Campana D, Walter T, Pusceddu S, Gelsomino F, Graillot E, Prinzi N, Spallanzani A, Fiorentino M, Barritault M, Dall’Olio F, et al.2018 Correlation between MGMT promoter methylation and response to temozolomide-based therapy in neuroendocrine neoplasms: an observational retrospective multicenter study. Endocrine 60 490–498. (https://doi.org/10.1007/s12020-017-1474-3)
Chatzellis E, Angelousi A, Daskalakis K, Tsoli M, Alexandraki KI, Wachuła E, Meirovitz A, Maimon O, Grozinsky-Glasberg S, Gross D, et al.2019 Activity and safety of standard and prolonged capecitabine/temozolomide administration in patients with advanced neuroendocrine neoplasms. Neuroendocrinology 109 333–345. (https://doi.org/10.1159/000500135)
Cives M, Ghayouri M, Morse B, Brelsford M, Black M, Rizzo A, Meeker A & Strosberg J 2016 Analysis of potential response predictors to capecitabine/temozolomide in metastatic pancreatic neuroendocrine tumors. Endocrine-Related Cancer 23 759–767. (https://doi.org/10.1530/ERC-16-0147)
Cros J, Hentic O, Rebours V, Zappa M, Gille N, Theou-Anton N, Vernerey D, Maire F, Lévy P, Bedossa P, et al.2016 MGMT expression predicts response to temozolomide in pancreatic neuroendocrine tumors. Endocrine-Related Cancer 23 625–633. (https://doi.org/10.1530/ERC-16-0117)
Dasari A, Shen C, Halperin D, Zhao B, Zhou S, Xu Y, Shih T & Yao JC 2017 Trends in the incidence, prevalence, and survival outcomes in patients with neuroendocrine tumors in the United States. Journal of the American Medical Association Oncology 3 1335–1342. (https://doi.org/10.1001/jamaoncol.2017.0589)
de Mestier L, Walter T, Brixi H, Evrard C, Legoux JL, de Boissieu P, Hentic O, Cros J, Hammel P, Tougeron D, et al.2019 Comparison of temozolomide-capecitabine to 5-fluorouracile-dacarbazine in 247 patients with advanced digestive neuroendocrine tumors using propensity score analyses. Neuroendocrinology 108 343–353. (https://doi.org/10.1159/000498887)
de Mestier L, Lepage C, Baudin E, Coriat R, Courbon F, Couvelard A, Do Cao C, Frampas E, Gaujoux S, Gincul R, et al.2020a Digestive neuroendocrine neoplasms (NEN): french intergroup clinical practice guidelines for diagnosis, treatment and follow-up (SNFGE, GTE, RENATEN, TENPATH, FFCD, GERCOR, UNICANCER, SFCD, SFED, SFRO, SFR). Digestive and Liver Disease 52 473–492. (https://doi.org/10.1016/j.dld.2020.02.011)
de Mestier L, Couvelard A, Blazevic A, Hentic O, de Herder WW, Rebours V, Paradis V, Ruszniewski P, Hofland LJ & Cros J 2020b Critical appraisal of MGMT in digestive NET treated with alkylating agents. Endocrine-Related Cancer 27 R391–R405. (https://doi.org/10.1530/ERC-20-0227)
Feldheim J, Kessler AF, Monoranu CM, Ernestus RI, Löhr M & Hagemann C 2019 Changes of O6-methylguanine DNA methyltransferase (MGMT) promoter methylation in glioblastoma relapse—A meta-analysis type literature review. Cancers 11 1837. (https://doi.org/10.3390/cancers11121837)
How-Kit A, Dejeux E, Dousset B, Renault V, Baudry M, Terris B, et al. 2015 DNA methylation profiles distinguish different subtypes of gastroenteropancreatic neuroendocrine tumors. Epigenomics 7 1245–1258.
Kitange GJ, Mladek AC, Carlson BL, Schroeder MA, Pokorny JL, Cen L, Decker PA, Wu W, Lomberk GA, Gupta SK, et al.2012 Inhibition of histone deacetylation potentiates the evolution of acquired temozolomide resistance linked to MGMT upregulation in glioblastoma xenografts. Clinical Cancer Research 18 4070–4079. (https://doi.org/10.1158/1078-0432.CCR-12-0560)
Koumarianou A, Kaltsas G, Kulke MH, Oberg K, Strosberg JR, Spada F, Galdy S, Barberis M, Fumagalli C, Berruti A, et al.2015 Temozolomide in advanced neuroendocrine neoplasms: pharmacological and clinical aspects. Neuroendocrinology 101 274–288. (https://doi.org/10.1159/000430816)
Kuczynski EA, Sargent DJ, Grothey A & Kerbel RS 2013 Drug rechallenge and treatment beyond progression--implications for drug resistance. Nature Reviews: Clinical Oncology 10 571–587. (https://doi.org/10.1038/nrclinonc.2013.158)
Kunz PL, Catalano PJ, Nimeiri H, Fisher GA, Longacre TA, Suarez CJ, Yao JC, Kulke MH, Hendifar AE, Shanks JC, et al.2018 A randomized study of temozolomide or temozolomide and capecitabine in patients with advanced pancreatic neuroendocrine tumors: a trial of the ECOG-ACRIN Cancer Research Group (E2211). Journal of Clinical Oncology 36 4004–4004. (https://doi.org/10.1200/JCO.2018.36.15_suppl.4004)
Lamarca A, Barriuso J, McNamara MG, Hubner RA, Manoharan P, Mansoor W & Valle JW 2020 Temozolomide-capecitabine chemotherapy for neuroendocrine neoplasms: the dilemma of treatment duration. Neuroendocrinology 110 155–157. (https://doi.org/10.1159/000503392)
Lepage C, Phelip JM, Lièvre A, Le Malicot K, Tougeron D, Dahan L, Toumpanakis C, Di Fiore F, Bohas CL, Borbath I, et al.2020 Lanreotide as maintenance therapy after first-line treatment in patients with non-resectable duodeno-pancreatic neuroendocrine tumours (NETs): an international double-blind, placebo-controlled randomized phase II trial (1136P). Annals of Oncology 31 S774. (https://doi.org/10.1016/j.annonc.2020.08.1376)
Lu Y, Zhao Z, Wang J, Lv W, Lu L, Fu W & Li W 2018 Safety and efficacy of combining capecitabine and temozolomide (CAPTEM) to treat advanced neuroendocrine neoplasms: a meta-analysis. Medicine 97 e12784. (https://doi.org/10.1097/MD.0000000000012784)
Mitry E, Walter T, Baudin E, Kurtz JE, Ruszniewski P, Dominguez-Tinajero S, Bengrine-Lefevre L, Cadiot G, Dromain C, Farace F, et al.2014 Bevacizumab plus capecitabine in patients with progressive advanced well-differentiated neuroendocrine tumors of the gastro-intestinal (GI-NETs) tract (BETTER trial)--a phase II non-randomised trial. European Journal of Cancer 50 3107–3115. (https://doi.org/10.1016/j.ejca.2014.10.001)
Noronha V, Berliner N, Ballen KK, Lacy J, Kracher J, Baehring J & Henson JW 2006 Treatment-related myelodysplasia/AML in a patient with a history of breast cancer and an oligodendroglioma treated with temozolomide: case study and review of the literature. Neuro-Oncology 8 280–283. (https://doi.org/10.1215/15228517-2006-003)
Nuñez‐Valdovinos B, Carmona‐Bayonas A, Jimenez‐Fonseca P, Capdevila J, Castaño‐Pascual Á, Benavent M, Pi Barrio JJ, Teule A, Alonso V, Custodio A, et al.2018 Neuroendocrine tumor heterogeneity adds uncertainty to the World Health Organization 2010 classification: real‐world data from the Spanish tumor registry (R‐GETNE). Oncologist 23 422–432. (https://doi.org/10.1634/theoncologist.2017-0364)
Park CK, Kim JE, Kim JY, Song SW, Kim JW, Choi SH, Kim TM, Lee SH, Kim IH & Park SH 2012 The changes in MGMT promoter methylation status in initial and recurrent glioblastomas. Translational Oncology 5 393–397. (https://doi.org/10.1593/tlo.12253)
Pavel M, O’Toole D, Costa F, Capdevila J, Gross D, Kianmanesh R, Krenning E, Knigge U, Salazar R, Pape UF, et al.2016 Enets consensus guidelines update for the management of distant metastatic disease of intestinal, pancreatic, bronchial neuroendocrine neoplasms (NEN) and NEN of unknown primary site. Neuroendocrinology 103 172–185. (https://doi.org/10.1159/000443167)
Perry JR, Bélanger K, Mason WP, Fulton D, Kavan P, Easaw J, Shields C, Kirby S, Macdonald DR, Eisenstat DD, et al.2010 Phase II trial of continuous dose-intense temozolomide in recurrent malignant glioma: RESCUE study. Journal of Clinical Oncology 28 2051–2057. (https://doi.org/10.1200/JCO.2009.26.5520)
Storey K, Leder K, Hawkins-Daarud A, Swanson K, Ahmed AU, Rockne RC & Foo J 2019 Glioblastoma recurrence and the role of O6-methylguanine-DNA methyltransferase promoter methylation. JCO Clinical Cancer Informatics 3 1–12. (https://doi.org/10.1200/CCI.18.00062)
Ter-Minassian M, Chan JA, Hooshmand SM, Brais LK, Daskalova A, Heafield R, Buchanan L, Qian ZR, Fuchs CS, Lin X, et al.2013 Clinical presentation, recurrence, and survival in patients with neuroendocrine tumors: results from a prospective institutional database. Endocrine-Related Cancer 20 187–196. (https://doi.org/10.1530/ERC-12-0340)
Walter T, van Brakel B, Vercherat C, Hervieu V, Forestier J, Chayvialle JA, Molin Y, Lombard-Bohas C, Joly MO & Scoazec JY 2015 O6-Methylguanine-DNA methyltransferase status in neuroendocrine tumours: prognostic relevance and association with response to alkylating agents. British Journal of Cancer 112 523–531. (https://doi.org/10.1038/bjc.2014.660)
WHO Classification of Tumours 2019 Digestive System Tumours. Lyon: International Arctic Research Center.
Wick A, Pascher C, Wick W, Jauch T, Weller M, Bogdahn U & Hau P 2009 Rechallenge with temozolomide in patients with recurrent gliomas. Journal of Neurology 256 734–741. (https://doi.org/10.1007/s00415-009-5006-9)
Wu YL, Raj N & Reidy-Lagunes D 2020 Exceptional responses after cessation of therapy With alkylating agents for pancreatic neuroendocrine tumors. Pancreas 49 e14–e16. (https://doi.org/10.1097/MPA.0000000000001451)
Zehir A, Benayed R, Shah RH, Syed A, Middha S, Kim HR, Srinivasan P, Gao J, Chakravarty D, Devlin SM, et al.2017 Mutational landscape of metastatic cancer revealed from prospective clinical sequencing of 10,000 patients. Nature Medicine 23 703–713. (https://doi.org/10.1038/nm.4333)
