Background
The capecitabine/temozolomide (CAPTEM) regimen has shown significant activity in neuroendocrine tumors (NETs), particularly in pancreatic NETs, for which objective radiographic response rates (ORRs) have ranged from roughly 30% to 70%.1–5 A recent randomized phase II study sponsored by the ECOG compared CAPTEM with temozolomide monotherapy in 144 patients with advanced, progressive, low- and intermediate-grade pancreatic NETs.2 The study showed a clinically and statistically significant improvement in progression-free survival (PFS) with the combination regimen, with a median PFS of 22.7 versus 14.4 months (hazard ratio [HR], 0.58; P=.023). Moreover, the trial also showed a significant improvement in overall survival (OS) with CAPTEM (not reached vs 38.0 months; HR, 0.41; P=.012).
The CAPTEM regimen has also showed efficacy in other types of NETs, albeit with lower response rates. Retrospective studies of CAPTEM in lung NETs have shown response rates of 18% to 30% in lung NETs and anecdotal responses in thymic NETs.6–9 Several studies have enrolled heterogeneous populations of patients, generally showing the highest ORRs in pancreatic NETs, an intermediate response in lung/thymic NETs, and the lowest response in gastrointestinal NETs.
Although moderately powered studies are sufficient to determine critical outcomes such as ORR and PFS, larger databases are key to answering other important questions. These include the frequency of rare toxicities, risk factors for toxicities, and efficacy across all subtypes of NETs. A real-world clinical database with long-term follow-up is also important to address other questions, such as duration of treatment, dose adjustments, and other patterns of practice.
With respect to the CAPTEM regimen, central questions are as follows: is the regimen active in nonpancreatic NETs? Does temozolomide (an alkylating agent) cause myelodysplastic syndromes (MDS), and is this risk associated with long-term use? Is the risk of Pneumocystis pneumonia (PCP) and other opportunistic infections similar to that seen in the brain tumor population?10 What are the rates of grade 4 cytopenias? Are there clinical risk factors for severe myelotoxicity? Can patients be safely treated for >1 year (as was the strategy in the randomized ECOG study)? What is the treatment-associated mortality rate? To address these and other clinical questions, we analyzed a database of 462 patients with advanced gastroenteropancreatic and lung NETs treated with CAPTEM at Moffitt Cancer Center.
Methods
We reviewed the records of patients with advanced neuroendocrine neoplasms seen at Moffitt Cancer Center between January 2008 and June 2019 who received treatment with CAPTEM. Patients who initiated treatment at outside institutions were included if they were prescribed treatment at appropriate doses and if complete records were available. The local Institutional Review Board approved the study protocol, and a waiver of consent was granted due to the retrospective nature of the study.
Demographic and pathologic data included age, primary disease site, Ki-67 proliferation index, histologic grade and differentiation, prior oncologic therapy, date of treatment initiation, and date of last follow-up and death, if applicable. We collected data on outcomes (ORR, PFS, OS, and disease control rate), duration of treatment, starting doses, dose reductions and interruptions, toxicities associated with treatment, and reasons for discontinuation.
PFS was defined as the time from initiation of treatment to either clinical or radiographic progression (whichever was shortest) or death of any cause. Assessments of radiographic response and progression were based on chart review rather than formal RECIST analysis. Partial response (PR) was defined as significant tumor regression on radiology reports and noted by the treating physician. Stable disease was defined as overall no significant change on radiology reports or minor disease improvement. Increase in tumor burden (size and/or number of tumors) and significant clinical progression as assessed by the treating physician defined progression of disease. OS was measured from the date of treatment initiation until death of any cause or last known follow-up date. Data were analyzed using SPSS Statistics, version 25 (IBM Corp). Survival curves were estimated using the Kaplan-Meier method, and categorical variables were analyzed using logistic regression or categorical response models. A P value set at 0.05 was used for Pearson correlations and chi-square analyses.
Results
Patient Characteristics
Supplemental eTable 1 presents patient demographics (available with this article at JNCCN.org), and Table 1 presents tumor characteristics. A total of 462 patients met eligibility criteria for analysis, including 252 men (55%) and 210 women (45%) with a median age of 59 years (range, 16–88 years) at the time of treatment initiation. Ten percent of patients began their treatment at an outside institution. Most patients had a primary pancreatic tumor (n=330; 71.4%). A total of 365 patients (79%) had well-differentiated tumors, 39 (8.4%) had poorly differentiated neuroendocrine carcinomas (NECs), and 58 (12.6%) had undefined differentiation status. A plurality of patients (n=189; 41%) had grade 2 histology. A total of 132 patients (28.6%) were treatment-naïve, and 187 (40.5%) had received 1 prior line of therapy. Tables 2 and 3 outline additional patient treatments before and after CAPTEM.
Tumor Characteristics
Pre-CAPTEM Treatment Summary
Post-CAPTEM Treatment Summary
Regimen
The target dose of capecitabine was 750 mg/m2 twice daily on days 1 through 14, and the target dose of temozolomide was 200 mg/m2 at bedtime on days 10 through 14 every 28 days, with ondansetron prophylactically administered 30 to 60 minutes before temozolomide. However, starting doses were frequently rounded down or reduced mildly at baseline. Consequently, the actual average starting dose of capecitabine was 675 mg/m2 twice daily, and the average starting dose of temozolomide was 180 mg/m2. The median duration on treatment was 8 months (range, 0–136 months). Median time to maximal response was 6 months. The median treatment-free interval for patients who discontinued treatment for reasons other than toxicity or disease progression was 14 months. A dose reduction was required for 113 patients (25%), and 72 (16%) discontinued treatment because of toxicity of any grade. A total of 193 patients (42%) discontinued treatment because of progressive disease, and 129 (28%) completed their course of treatment before progression because they achieved maximal response or reaching an arbitrary time point such as 1 year. The remaining patients discontinued because of other complications (nononcologic health issues, n=10; 2%) or insurance coverage issues or by personal request (n=16; 3%). A total of 42 patients (9%) remained on active treatment at the time of data cutoff.
Efficacy
Radiographic responses were assessed by review of patient progress notes and radiology reports, including tumor measurements provided. As the best response, 8 patients (2%) had a complete response, 204 (44%) had a PR, 161 (35%) had stable disease, and 76 (16%) had progressive disease. In 13 patients, response was not assessable. Median PFS was 18 months (95% CI, 14.0–21.9 months) and median OS was 51 months (95% CI, 42.8–59.2 months) (eFigure 1A, B).
When comparing patients with pancreatic versus nonpancreatic primary tumors, we found that median PFS was 23 versus 10 months (P<.0001) and median OS was 62 versus 28 months (P<.0001) (eFigure 2A, B). Patients with a pancreatic primary tumor had an ORR of 51.5%, whereas those with a nonpancreatic primary tumor had an ORR of 31.8% (P<.0001). Of note, all complete responses were in patients with pancreatic primary tumors. Patients with well-differentiated tumors had significantly better outcomes than those with poorly differentiated carcinomas, with a median PFS of 24 versus 7 months (P<.0001) and a median OS of 62 versus 14 months (P<.0001) (Figure 1A, B). Patients with grades 1, 2, and 3 tumors had a median PFS of 30, 23, and 10 months, respectively (P=.081) and median OS of 81, 57, and 23 months, respectively (P<.0001) (Figure 2A, B). Among patients with grade 3 tumors (n=77; 38 NETs and 39 NECs), those with well-differentiated grade 3 NETs had significantly better outcomes than those with poorly differentiated NECs, with median PFS of 28 versus 7 months (P=.005) and median OS of 36 versus 14 months (P<.001); however, no difference was observed in ORR (42% and 33%, respectively; P=.388). Patients with pancreatic primary or well-differentiated tumors were more likely to have PR as the best response (P<.001 and P=.001, respectively). Table 4 outlines ORR, PFS, and OS by patient subgroup.
(A) Progression-free and (B) overall survival, stratified by tumor differentiation.
Citation: Journal of the National Comprehensive Cancer Network 20, 1; 10.6004/jnccn.2021.7017
(A) Progression-free and (B) overall survival, stratified by tumor differentiation.
Citation: Journal of the National Comprehensive Cancer Network 20, 1; 10.6004/jnccn.2021.7017
(A) Progression-free and (B) overall survival, stratified by tumor differentiation.
Citation: Journal of the National Comprehensive Cancer Network 20, 1; 10.6004/jnccn.2021.7017
(A) Progression-free and (B) overall survival, stratified by tumor grade.
Citation: Journal of the National Comprehensive Cancer Network 20, 1; 10.6004/jnccn.2021.7017
(A) Progression-free and (B) overall survival, stratified by tumor grade.
Citation: Journal of the National Comprehensive Cancer Network 20, 1; 10.6004/jnccn.2021.7017
(A) Progression-free and (B) overall survival, stratified by tumor grade.
Citation: Journal of the National Comprehensive Cancer Network 20, 1; 10.6004/jnccn.2021.7017
Response Data per Primary Site of Disease
Treatment-Related Adverse Effects per CTCAE Version 5.0
Adverse Effects
Treatment-related adverse effects were graded per CTCAE version 5.0 and are outlined in Table 5. Seventy percent (n=322) of patients experienced at least grade 1 treatment-related toxicity. Particular attention was paid to grade 4 hematologic events, opportunistic infections, and the development of hematologic malignancies or disorders. The incidence of grade 4 thrombocytopenia was 7% (10% in women and 5% in men; P=.02), and 4 cases were complicated by bleeding (0.8%). The incidence of grade 4 neutropenia was 3% (5% in women and 1% in men; P=.004), and the incidence of grade 4 lymphopenia was only 2%. There were no differences in prescribed doses between men and women. There was no difference between male and female patients with regard to prior exposure to chemotherapy or duration on CAPTEM (P=.850 and P=.536, respectively).
Only 1 case (0.2%) of suspected PCP was observed in a patient receiving corticosteroids. There were 5 cases of herpes zoster virus and no other opportunistic infections. Three patients developed myelodysplastic disease, all of whom had also received peptide receptor radiotherapy (PRRT) with 177Lu-DOTATATE (P<.0001). Two of the patients had received PRRT approximately 5 years earlier and developed MDS during or shortly after completing CAPTEM. The third patient received PRRT roughly 20 months after completing CAPTEM and developed MDS 3 months after PRRT ended. There was no statistically significant increase in the risk of developing grade 3 or 4 hematologic toxicity among patients who received prior PRRT (P=.480). There were no acute treatment-related deaths, although 1 patient died 2 months after a thrombocytopenic bleed.
Of the 113 patients who required dose reductions, 15 (13%) underwent dose reduction of temozolomide alone, and 1 (0.8%) was unable to tolerate dose reduction and continued on capecitabine monotherapy. Thirty-eight patients (34%) had a dose reduction of capecitabine alone, 7 (6%) of whom continued on temozolomide monotherapy. A total of 60 patients (53%) required dose reduction of both drugs, 4 (4%) of whom could not tolerate dose reductions and permanently discontinued therapy.
Discussion
Our analysis of a large database of patients with NETs treated with CAPTEM yielded many new insights. One key finding is that the regimen is safe and that treatment-related deaths are exceedingly rare: only 1 patient (0.2%) death was likely attributable to treatment. Another finding is that the opportunistic infection risk is negligible and that serious opportunistic infections such as PCP are absent among otherwise immunocompetent patients. This stands in contrast to the brain tumor literature, in which prophylactic PCP therapy is recommended, likely because many patients are receiving chronic immunosuppressive corticosteroids.10 It is important to note that previous data on opportunistic infections in patients with neuroendocrine neoplasms involved patients who were receiving temozolomide at a higher dose intensity (every other week), which may contribute to the higher risk observed.11
Alkylating cytotoxic drugs, as a class, are associated with development of MDS.12 However, with our study’s median follow-up of 2.4 years, we identified no cases of treatment-associated MDS or acute leukemia, apart from 3 patients who had also received PRRT with 177Lu-DOTATATE, a known risk factor for myelodysplasia. Consequently, it appears that temozolomide in itself does not present a significant risk for subsequent development of MDS. It is unclear, however, whether sequential treatment with CAPTEM and 177Lu-DOTATATE increases the risk of MDS significantly above the risk of 177Lu-DOTATATE alone. Of note, in a recent study of 38 patients, researchers evaluated long-term outcomes of patients treated with a CAPTEM-PRRT regimen in which no long-term cases of MDS or leukemia were reported.13
The rates of grade 4 acute myelotoxicities were similar to those reported in the randomized ECOG study.2 However, the size of our database allowed us to show significant sex disparities in grade 4 thrombocytopenia and neutropenia, with women at substantially higher risk. We conclude that mild sex-based dosing adjustments should be considered, particularly among older female patients, in whom risks of grade 4 thrombocytopenia or neutropenia can be quite significant.
Response rates in our cohort of patients support those previously reported in the literature, with well-differentiated tumors and patients with pancreatic primary tumors exhibiting significantly higher PFS, OS, and ORR.14–16 We also note that our average starting doses were 10% lower than the target doses of 750 mg/m2 twice daily of capecitabine and 200 mg/m2 of temozolomide and that our prior analyses of outcomes using these doses showed highly favorable results in the pancreatic and lung NET population.
Limitations of the study include lack of Ki-67 percentage in approximately one-third of patients, primarily due to the fact that many were treated during a time when mitotic activity was the primary determinant of grade. Although the average follow-up duration was relatively long (median, 2.4 years), cases of treatment-related MDS occurring many years after onset of therapy may have been missed.
Conclusions
The CAPTEM regimen is exceptionally safe, with a treatment-associated mortality rate of 0.2%. It is particularly active in well-differentiated pancreatic NETs but appears to have significant activity in other NET subtypes. Severe myelotoxicity is rare, but risks of grade 4 thrombocytopenia and neutropenia are significantly increased in women compared with men. Sex-based dosing should be considered. The risk of MDS or acute leukemia appears to be negligible in patients who are unexposed to other known risks, such as PRRT, although cases occurring many years after onset of therapy may have been missed. The risk of a serious opportunistic infection likewise appears to be negligible in otherwise immunocompetent patients.
References
- 1.↑
Cives M, Ghayouri M, Morse B, et al. Analysis of potential response predictors to capecitabine/temozolomide in metastatic pancreatic neuroendocrine tumors. Endocr Relat Cancer 2016;23:759–767.
- 2.↑
Kunz PL, Catalano PJ, Nimeiri H, et al. 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) [abstract]. J Clin Oncol 2018;36(Suppl):Abstract 4004.
- 3.↑
Strosberg JR, Fine RL, Choi J, et al. First-line chemotherapy with capecitabine and temozolomide in patients with metastatic pancreatic endocrine carcinomas. Cancer 2011;117:268–275.
- 4.↑
Chatzellis E, Angelousi A, Daskalakis K, et al. Activity and safety of standard and prolonged capecitabine/temozolomide administration in patients with advanced neuroendocrine neoplasms. Neuroendocrinology 2019;109:333–345.
- 5.↑
Fine RL, Gulati AP, Krantz BA, et al. Capecitabine and temozolomide (CAPTEM) for metastatic, well-differentiated neuroendocrine cancers: the Pancreas Center at Columbia University experience. Cancer Chemother Pharmacol 2013;71:663–670.
- 6.↑
Al-Toubah T, Morse B, Strosberg J. Capecitabine and temozolomide in advanced lung neuroendocrine neoplasms. Oncologist 2020;25:e48–52.
- 7.↑
Owen DH, Alexander AJ, Konda B, et al. Combination therapy with capecitabine and temozolomide in patients with low and high grade neuroendocrine tumors, with an exploratory analysis of O6-methylguanine DNA methyltransferase as a biomarker for response. Oncotarget 2017;8:104046–104056.
- 8.↑
Papaxoinis G, Kordatou Z, McCallum L, et al. Capecitabine and temozolomide in patients with advanced pulmonary carcinoid tumours. Neuroendocrinology 2020;110:413–421.
- 9.↑
Saranga-Perry V, Morse B, Centeno B, et al. Treatment of metastatic neuroendocrine tumors of the thymus with capecitabine and temozolomide: a case series. Neuroendocrinology 2013;97:318–321.
- 10.↑
De Vos FY, Gijtenbeek JM, Bleeker-Rovers CP, et al. Pneumocystis jirovecii pneumonia prophylaxis during temozolomide treatment for high-grade gliomas. Crit Rev Oncol Hematol 2013;85:373–382.
- 11.↑
Schwarzberg AB, Stover EH, Sengupta T, et al. Selective lymphopenia and opportunistic infections in neuroendocrine tumor patients receiving temozolomide. Cancer Invest 2007;25:249–255.
- 12.↑
Bhatia S. Therapy-related myelodysplasia and acute myeloid leukemia. Semin Oncol 2013;40:666–675.
- 13.↑
Parghane RV, Ostwal V, Ramaswamy A, et al. Long-term outcome of “sandwich” chemo-PRRT: a novel treatment strategy for metastatic neuroendocrine tumors with both FDG- and SSTR-avid aggressive disease. Eur J Nucl Med Mol Imaging 2021;48:913–923.
- 14.↑
Chan D, Bergsland EK, Chan JA, et al. Temozolomide in grade III neuroendocrine neoplasms (G3 NENs): a multicenter retrospective review [abstract]. J Clin Oncol 2019;37(Suppl):Abstract 321.
- 15.↑
Spada F, Maisonneuve P, Fumagalli C, et al. Temozolomide alone or in combination with capecitabine in patients with advanced neuroendocrine neoplasms: an Italian multicenter real-world analysis. Endocrine 2021;72:268–278.
- 16.↑
Thomas K, Voros BA, Meadows-Taylor M, et al. Outcomes of capecitabine and temozolomide (CAPTEM) in advanced neuroendocrine neoplasms (NENs). Cancers (Basel) 2020;12:206.
