Radiotherapy for advanced adenoid cystic carcinoma: neutrons, photons or mixed beam?

Abstract

Purpose: To compare retrospectively radiotherapy with neutrons, photons, and a photon/neutron mixed beam in patients with advanced adenoid cystic carcinoma of the head and neck. Local control, survival, distant failure, and complications were analyzed.

Materials and methods: Between 1983 and 1995, 75 patients with inoperable, recurrent, or incompletely resected adenoid cystic carcinoma of the head and neck received radiotherapy that consisted of either fast 14.1 MV DT neutrons (median dose 16 neutron Gy), linac-based photon irradiation (median dose 64 photon Gy), or both (median dose 8 neutron Gy and 32 photon Gy). Follow-up ranged from 1 to 160 months (median 51 months), and the surviving patients had a minimum follow-up of 3 years at the time of analysis.

Results: The actuarial 5-year local control was 75% for neutrons, and 32% for both mixed beam and photons (P=0.015, log-rank). This advantage for neutrons in local control was not transferred to significant differences in survival (P>0.1). The survival is dictated by the tumor diseases due to distant metastases occurring in 29 (39%) of the 75 patients. Positive lymph nodes were the only significant factor (P=0.001) associated with the development of distant metastases although negative lymph nodes did not predict absence of distant metastases, but predicted a delay of occurrence. In multivariate analysis postoperative radiotherapy (P=0.003) and small tumor size (P=0.01) were associated with high local control, while primary therapy (P=0.006) and negative lymph nodes (P=0.01) were associated with longer survival. While acute toxicity was similar in all three radiotherapy groups, severe late grade 3 and 4 toxicity tended to be more prevalent (P>0.1) with neutrons (19%) than with mixed beam (10%) and photons (4%).

Conclusion: Fast neutron radiotherapy provides higher local control rates than a mixed beam and photons in advanced, recurrent or not completely resected adenoid cystic carcinoma of the major and minor salivary glands. Neutron radiotherapy can be recommended in patients with bad prognosis with gross residual disease (R2), with unresectable tumors, or inoperable tumors. The type of radiation does not impact survival, which is dominated by the high number of distant metastases.

Introduction

Adenoid cystic carcinomas of salivary glands in the head and neck are relatively rare tumors. With an incidence of approximately 1 in 1 000 000, adenoid cystic carcinomas comprise 5–10% of all salivary gland tumors, which account for 2–4% of all head and neck malignancies [4], [23], [24], [28]. Nevertheless, adenoid cystic carcinomas are a therapeutic challenge because of their slow but aggressive local nature, persistent recurrence patterns, and frequent systemic spread after long-term survival [6], [21], [24], [25], [30]. Surgery is the mainstay of treatment for adenoid cystic carcinomas; however, Guillamondegui et al. [10] described a local failure rate of 50% for adenoid cystic carcinoma in the parotid with surgery alone. On the basis of this early experience, it was shown that postoperative radiotherapy had a favorable impact on local control rates [7], [19], [23] and survival rates [20], [26] in advanced stages. The results of conventional radiotherapy with photons for advanced salivary gland tumors have been reported to be suboptimal in the setting of inoperable, not completely resected and recurrent tumors [9], [17]. Therefore the question arose whether other forms of radiotherapy, in particular neutron irradiation, could improve the results for these tumors.

Compared to X-rays, neutrons are characterized by a reduced oxygen enhancement factor (OER), by less or no repair of sublethal or potentially lethal cell damage; and by less variation of sensitivity through the cell cycle [11]. Neutrons were first used in cancer therapy in the 1930’s at the Lawrence Berkeley Laboratory, California, without a clear rationale for hopeless cases. However, clinical trials to date have shown that neutrons do not offer an advantage over X-rays across the spectrum of tumor types. Nevertheless, evidence that neutrons may give better results for certain tumors, has altered the rationale for their use. The earlier rationale was the reduced OER, whereas the revised rationale is based on their higher relative biological effectiveness (RBE) for slowly cycling tumors [11]. In this context, Batterman measured the RBE of fast neutrons relative to megavolt photons for metastases in the lung [1]. With growth delay as the endpoint, neutrons had the highest RBE for adenoid cystic carcinomas from salivary gland primaries. The RBE for fractionated radiotherapy was 8.0, compared with 3.0–3.5 for most normal tissues, with a therapeutic gain factor of 2.0–2.5.

In accordance with these findings, the efficacy of neutron radiotherapy for patients with advanced salivary gland tumors has been demonstrated clinically by several groups [3], [16], [18]. Pooled data of European studies [15] in the treatment of advanced salivary gland tumors resulted in local control rates of 26% for photons compared with 67% for neutrons. In these studies 15–60% of the patients had adenoid cystic carcinomas. The results of these nonrandomized clinical studies were in agreement with the only prospective randomized study [9], [17], comparing neutron vs. photon irradiation of unresectable salivary gland tumors of all histologies in 32 patients. This randomized study revealed higher local control rates in the neutron arm than in the photon arm (56 vs. 17%, P=0.009). In contrast, long-term survival was not improved in the neutron arm, while incidence of severe morbidity was greater in the neutron arm.

Despite a variety of favorable reports using neutron radiotherapy for salivary gland tumors including adenoid cystic carcinoma [5], [8], there is little information on direct comparison among neutrons, photons, or a mixed-beam schedule. The aim of the present study was therefore to assess the efficacy of fast neutron radiotherapy alone vs. mixed-beam photon/neutron irradiation and photon irradiation alone in patients with locally advanced adenoid cystic carcinoma of the head and neck. In particular, we hypothesized that a mixed-beam schedule would result in a reasonable compromise combining the neutrons’ high local control rates and the photons’ less severe late morbidity. Our analysis, based on the relative uniform treatment modalities of a single institution, was prepared with emphasis on survival, local control, distant failure, and radiation-related complications.