This is an early access version, the complete PDF, HTML, and XML versions will be available soon.
Molecular Subtype-Associated Response to Cyclophosphamide–Epirubicin–Cisplatin Regimen in Recurrent or Metastatic Adenoid Cystic Carcinoma: A Retrospective Single-Center Study
Department of Oncology, Shanghai East Hospital, School of Medicine, Tongji University, Shanghai 200123, China
*
Authors to whom correspondence should be addressed.
†
These authors contributed equally to this work.
Cancers 2026, 18(11), 1847; https://doi.org/10.3390/cancers18111847
Submission received: 11 May 2026
/
Revised: 29 May 2026
/
Accepted: 2 June 2026
/
Published: 4 June 2026
(This article belongs to the Special Issue Head and Neck Cancer Therapies: Current and Innovative Options)
Simple Summary
Adenoid cystic carcinoma is a rare cancer of the salivary glands. Once it spreads, no drug has been approved for it, and current options help only modestly. Drugs that target tumor blood vessels work for a while but stop working over time, and there is no clear standard for what to give next. We looked back at 31 such patients treated at our hospital with a three-drug chemotherapy combination. Prior blood-vessel-targeting therapy did not reduce how well chemotherapy worked. Tumor shrinkage was seen only in one of the two molecular subtypes—none in the other. In 5 patients, a gene alteration was linked to shorter survival. These numbers are small. Larger prospective studies are needed before any of this should change how patients are treated.
Abstract
Background/Objectives: Recurrent or metastatic adenoid cystic carcinoma (R/M ACC) has no standard systemic therapy. VEGFR-targeting tyrosine kinase inhibitors (TKIs) are commonly used first-line, though no international standard exists; cisplatin-based chemotherapy is an alternative. We retrospectively reviewed the cyclophosphamide–epirubicin–cisplatin (CEP) regimen to determine whether prior TKI exposure compromises subsequent chemotherapy efficacy. Methods: We studied 31 patients given CEP for progressive R/M ACC (2018–2023). Tumor response was assessed by RECIST 1.1. Molecular subtype was determined by c-MYC/p63 immunohistochemistry (ACC-I, c-MYC-positive/p63-negative; ACC-II, p63-positive/c-MYC-low or negative). Multivariable models used Firth’s penalized likelihood Cox regression. Next-generation sequencing (NGS) was available in 21 of 31 patients. Results: Thirty-one patients were enrolled (median age 48; 17 [54.8%] with prior TKI). At median follow-up of 22.6 months, the objective response rate (ORR) was 19.4%, disease control rate 71.0%, median progression-free survival (PFS) 5.3 months, and median overall survival (OS) 10.3 months. Prior TKI did not lower efficacy: ORR 17.6% vs. 21.4%, PFS hazard ratio 0.76 (p = 0.519). All six partial responses occurred in ACC-I tumors (35.3% vs. 0% in ACC-II, p = 0.021). In the NGS subset (5 PIK3CA-mutant), PIK3CA mutation (OS HR 6.19, p = 0.024) and bone metastasis (OS HR 5.84, p = 0.027) remained associated with shorter OS after adjustment. No treatment-related deaths occurred. Conclusions: CEP is active in R/M ACC, and prior TKI exposure did not appear to reduce efficacy. Higher response rates in ACC-I tumors and the apparent PIK3CA-related survival deficit are exploratory observations that need prospective testing before they can guide treatment.
