Capecitabine Plus Temozolomide vs Temozolomide Alone in Advanced Pancreatic Neuroendocrine Tumors

Get Permission

In the updated final analysis of the phase II ECOG-ACRIN E2211 trial, patients with advanced pancreatic neuroendocrine tumors experienced a significant improvement in progression-free survival with capecitabine plus temozolomide over temozolomide alone.1 Although the 5-month difference in overall survival was not statistically significant, it was “clinically meaningful,” according to Pamela L. Kunz, MD, Associate Professor of Medicine at Yale School of Medicine and Director of the Center for Gastrointestinal Cancers at Smilow Cancer Hospital and Yale Cancer Center.

“MGMT deficiency is predictive of response to temozolomide in glioblastoma but is a matter of debate for pancreatic neuroendocrine tumors.”
— Pamela L. Kunz, MD

Tweet this quote

“The combination of capecitabine and temozolomide should be included as a standard treatment option for patients with advanced pancreatic neuroendocrine tumors and is a reasonable comparator arm in future randomized studies,” Dr. Kunz said during her presentation of the data at the 2022 ASCO Annual Meeting.1

The study also explored outcomes according to MGMT (O[6]-methylguanine-DNA methyltransferase) status. There is a biologic rationale for MGMT-deficient tumors to respond better to temozolomide, as they do in glioblastoma multiforme, she said.

About ECOG-ACRIN E2211

As Dr. Kunz noted, strong treatment response is uncommon in patients with advanced pancreatic neuroendocrine tumors. Retrospective studies and small, prospective studies have reported high response rates with capecitabine plus temozolomide, along with relatively long progression-free survival.

The randomized E2211 trial was conducted to establish a role for this combination. The trial evaluated capecitabine/temozolomide (capecitabine at 750 mg/m2 twice daily on days 1–14, temozolomide at 200 mg/m2/d on days 10–14) vs temozolomide (200 mg/m2/d on days 1–5) in 144 patients with metastatic or unresectable low- or intermediate-grade pancreatic neuroendocrine tumors. Eligible patients had to have experienced disease progression within the preceding 12 months and to have had no prior treatment with temozolomide, dacarbazine, capecitabine, or fluorouracil.

Baseline characteristics were balanced, with a few notable exceptions: more Black patients in the temozolomide arm, longer time from diagnosis in the combination arm, and more grade 2 tumors in the temozolomide arm (62% vs 49%).

The primary endpoint was progression-free survival. Investigators also evaluated MGMT by immunohistochemistry (IHC) and by promoter methylation to look for associations with outcomes.

Why Look at MGMT?

E2211 isthe largest prospective study to examine MGMT expression as related to the treatment of pancreatic neuroendocrine tumors. Of the 144 patients, 97 underwent MGMT testing. MGMT deficiency was indicated by low IHC or positive promoter methylation.

As Dr. Kunz explained, pancreatic neuroendocrine tumors are known to respond to temozolomide alone; it is believed that capecitabine may be synergistic with temozolomide, perhaps by downregulating MGMT. Temozolomide induces DNA methylation of the O(6) position of guanine, leading to DNA damage and cell death, which is usually repaired by MGMT. In glioblastoma multiforme, MGMT is silenced by promoter methylation, rendering cells more sensitive to alkylating agents. In neuroendocrine tumors, mechanisms other than methylation may be involved, since promoter methylation is less common, yet MGMT is still lost, according to Dr. Kunz.

“MGMT deficiency is predictive of response to temozolomide in glioblastoma but is a matter of debate for pancreatic neuroendocrine tumors,” she said. “This study was not designed to test MGMT as a predictive biomarker, as both arms contain temozolomide. However, we can examine the association of MGMT overall with response and survival by arm.”

Combination Improves Outcomes

This study met its primary endpoint, with capecitabine/temozolomide improving median progression-free survival by more than 8 months, from 14.4 months with temozolomide alone to 22.7 months (hazard ratio [HR] = 0.58; P = .022). In the updated and final analysis, a trend toward overall survival was observed, with a median of 58.7 months with the combination vs 53.8 months with temozolomide alone (HR = 0.82; P = .42). “While this difference did not reach statistical significance, this is a 5-month clinically meaningful difference between the two arms,” she maintained.

Objective response rates were high in both arms: 40% with the combination and 34% with temozolomide alone. The median duration of response was 16.6 months and 12.6 months, respectively, and disease control rates were 84% and 74%.

MGMT Deficiency Associated With Response

Although the findings on MGMT deficiency and outcomes are only exploratory and hypothesis-generating, the investigators reported MGMT deficiency was associated with greater odds of response as follows:

For MGMT deficiency defined by low IHC, the odds ratio was 6.38; 52% of patients with low H-scores (ie, 1–2) responded, vs 15% with high H-scores (P = .0004).

For MGMT defined by positive promoter methylation, the odds ratio was 9.79; responses were observed in 85% of these patients vs 38% lacking promoter methylation (P = .04).

“We also observed trends in MGMT deficiency associated with progression-free survival and overall survival,” she added. Combination treatment led to a median progression-free survival of 27.3 months for MGMT-deficient patients and 16.6 months for patients not deemed MGMT-deficient (HR = 0.57). In MGMT-deficient patients treated with the combination, median overall survival was not reached but was 48.6 months with the single agent (HR = 0.55).


  • The phase II ECOG-ACRIN E2211 trial evaluated capecitabine plus temozolomide vs temozolomide alone in patients with advanced pancreatic neuroendocrine tumors.
  • The combination significantly improved progression-free survival by about 8 months; a clinically meaningful but not statistically significant improvement was seen in overall survival.
  • MGMT deficiency was associated with response, although the study was not designed to test MGMT as a predictive biomarker. More data are needed to support the assessment of MGMT for therapy selection.

“Combined with a strong biologic rationale, these results are provocative and suggest that MGMT may, in fact, be predictive in pancreatic neuroendocrine tumors, as it is in glioblastoma. However, confirmatory testing is needed with studies that have non–temozolomide-containing arms,” she said.

Capecitabine/temozolomide was associated with higher rates of grade 3 or 4 toxicities (44% vs 22%; P = .005). 

DISCLOSURE: Dr. Kunz has reported relationships with Guardant Health, Acrotech Biopharma, Advanced Accelerator Applications/Novartis, Amgen, Crinetics Pharmaceuticals, Genentech/Roche, HUTCHMED, Ipsen, Lexicon, Natera, RayzeBio, and Sun Pharma.


1. Kunz PL, Graham N, Catalano PJ, et al: A randomized study of temozolomide or temozolomide and capecitabine in patients with advanced pancreatic neuroendocrine tumors: Final analysis of efficacy and evaluation with MGMT (ECOG-ACRIN E2211). 2022 ASCO Annual Meeting. Abstract 4004. Presented June 5, 2022.

Related Articles

Expert Point of View: Marianne E. Pavel, MD

Marianne E. Pavel, MD

Marianne E. Pavel, MD

Based on the final results of ECOG-ACRIN E2211, invited discussant Marianne E. Pavel, MD, of Friedrich-Alexander University Erlangen-Nürnberg, Erlangen, Germany, concluded: “Capecitabine/temozolomide is a preferred regimen…, but temozolomide can still be an option in...