Clinical course of a patient with high-grade mucoepidermoid carcinoma of the parotid gland. A, Fluorine-18 fluorodeoxyglucose positron emission tomographic scan demonstrating mediastinal and abdominal lymph node metastases at 3 months after completion of adjuvant concurrent chemoradiation. B, Pathologic features of this patient showing frequent mitoses, areas of anaplasia, neural invasion, and a markedly decreased intracystic component, characteristic of high-grade mucoepidermoid carcinoma (hematoxylin-eosin, original magnification × 100).
Survival outcomes of all patients. A, Progression-free survival (n = 24). B, Locoregional recurrence-free survival (n = 24). C, Overall survival (n = 24). The plus signs on the curves indicate censored data.
Survival outcomes among patients treated with chemoradiation and those treated with radiation alone. A, Progression-free survival by treatment group (P = .14). B, Locoregional recurrence-free survival by treatment group (P = .06). C, Overall survival by treatment group (P = .05). The plus signs on the curves indicate censored data.
Tanvetyanon T, Qin D, Padhya T, McCaffrey J, Zhu W, Boulware D, DeConti R, Trotti A. Outcomes of Postoperative Concurrent Chemoradiotherapy for Locally Advanced Major Salivary Gland Carcinoma. Arch Otolaryngol Head Neck Surg. 2009;135(7):687-692. doi:10.1001/archoto.2009.70
To investigate the potential value of postoperative concurrent chemoradiation among patients with high-risk salivary gland carcinomas.
Case control study based on retrospective medical record review.
A tertiary care comprehensive cancer center.
A total of 24 patients, 12 with major salivary gland carcinoma who were treated with postoperative concurrent chemoradiotherapy from 1998 to 2007 (chemoradiation group), and a control group of 12 patients treated with postoperative radiation alone.
Main Outcome Measures
Overall survival, progression-free survival, toxic effects.
All but 1 patient had stage III or IV disease; close or positive surgical margins were identified in 20 patients (83%). The median radiation dose was 63 Gy. In the chemoradiation group, platinum-based regimens were used in all. Treatment was well tolerated, but toxic effects, predominantly hematologic, were increased in the chemoradiation group. To date, 8 patients have died; the median overall survival was 53 months. The overall survival in the chemoradiation group was significantly better than in the radiation-alone group: 3-year survival rates were 83% and 44%, respectively (P = .05).
Locally advanced or high-grade salivary gland carcinomas follow an aggressive clinical course. Based on our limited experience, postoperative chemoradiation with a platinum-based regimen seems to be effective in selected patients and warrants further investigation.
Patients with locally advanced or high-grade carcinomas of the major salivary glands have a poor prognosis. Although most salivary gland carcinomas, which comprise less than 5% of all head and neck cancers, have a favorable prognosis, there is a sharp increase in the risk of recurrence and metastasis for patients with high T stage and lymph node involvement.1 In addition, a high tumor grade is also associated with a poor outcome. In high-grade mucoepidermoid carcinoma, long-term remission is observed in only about 30% of patients regardless of stage.2 Other adverse prognostic features include positive surgical margins, soft tissue, or perineural invasion.3- 5 For these patients, complete surgery followed by postoperative radiation is considered the standard of care. Based on the Surveillance Epidemiology End Results database,6 the 5-year survival rates of patients with locally advanced salivary gland carcinoma range from 50% to 70% for stage III disease and 30% to 40% for stage IV disease. Clearly, more effective treatment strategy is needed. Unfortunately, little improvement has been made in the treatment of salivary gland carcinomas.
Significant progress has, however, been made in the treatment of carcinomas originating from other sites in the head and neck. Several large randomized trials have demonstrated the benefit of postoperative concurrent chemoradiation over postoperative radiation alone.7,8 Postoperative concurrent chemotherapy, particularly with cisplatin, delays time to progression and improves survival among patients with high-risk features such as positive margins, multiple lymph node involvement, or extracapsular extension.
Nevertheless, the feasibility of postoperative concurrent chemoradiation has never been evaluated in salivary gland carcinomas, to our knowledge, and no such clinical trial has been conducted. Concurrent chemoradiation is associated with greater toxic effects than radiation alone.7,8 Excess toxic effects can result in an interruption of radiation, which in turn may impair overall treatment outcomes. Furthermore, certain salivary gland carcinomas are not sensitive to chemotherapy.9 It is therefore unknown if chemotherapy will benefit these patients.
Over the years, selected patients with locally advanced or high-grade salivary gland carcinomas have been treated with adjuvant concurrent chemoradiation at our institution (H. Lee Moffitt Cancer Center, Tampa, Florida). Chemoradiation is offered based on the treating physician's discretion and undertaken based on individual patients' consent. Herein, we summarize our experiences. To understand the outcomes associated with standard therapy, we also identified a group of matched case controls, who were treated with adjuvant radiation alone during this same period.
After approval by our institutional review board, the H. Lee Moffitt Cancer Center tumor registry was searched for patients with a diagnosis of major salivary gland carcinoma, who had undergone surgery and had been treated with postoperative concurrent chemoradiation during the period of March 1998 to March 2007 (chemoradiation group). Those with gross residual disease after surgery or those with recurrent disease at the time of radiation were excluded. To identify a matched control group of patients who had been treated with adjuvant radiation alone (radiation-alone group), a computerized search of the same database was performed for cases with matched histologic characteristics and stage.
Electronic medical records of eligible patients were systematically reviewed for patient demographic, histologic, and treatment characteristics and tumor stage. For patients whose tumor slides were available, pathology specimens were rereviewed for high-risk pathologic features by 1 of us (D.Q.), a head and neck pathologist, who was unaware of previous pathologic interpretation. Staging was based on the staging manual issued by American Joint Committee on Cancer.10 Nodal staging was based on pathologic findings, except among patients who had not undergone a neck dissection. Tumor grading was based on the 1996 criteria by the Armed Forces Institute of Pathology recommendations.11 Deaths were verified by the Social Security death index.12 Grading of toxic effects was assessed based on Common Toxicity Criteria, version 2.0.
Overall survival was calculated from the date of surgery to the last follow-up date or death. Recurrence-free survival was calculated from the date of surgery to the time of locoregional recurrence, distant metastasis, or death. Locoregional recurrence-free survival was calculated from the date of surgery to the time of locoregional recurrence or death. Time to progression and time to locoregional progression were calculated from the date of surgery to the date of progression or locoregional progression. The Kaplan-Meier method was used to calculate survival estimates and times to progression. Difference in the survival across groups was tested using the log-rank test. All P values were for 2-tailed tests. The statistical analyses were performed on SAS statistical software (version 9.1; SAS Institute Inc, Cary, North Carolina).
A total of 24 patients were reviewed: 12 in the chemoradiation group and 12 in the radiation-alone group (Table 1). All, except 1 patient, had stage III or IV disease. The most common presenting symptom was a mass in the parotid or submandibular gland. Patients with a T4a tumor typically presented with symptoms related to facial nerve dysfunction. One patient was found to have carotid involvement (T4b). Poor prognostic pathologic features were prevalent in both groups. Overall, 20 of 24 patients (83%) had close or positive surgical margins.
Comparing the chemoradiation group with the radiation-alone group, patient characteristics were balanced, although there was a slightly higher proportion of patients with N2 disease or positive surgical margins in the chemoradiation group. Patients in the radiation-alone group were slightly younger and there was a higher proportion with a close surgical margin than those in the chemoradiation group.
Treatment characteristics were also comparable between the 2 groups. The most common surgical procedure was total parotidectomy (Table 2). Neck treatment consisting of either nodal dissection or nodal irradiation occurred in all except 3 patients.
Intensity-modulating radiation technique (IMRT) or tomotherapy was more frequently used in the chemoradiation group, although this constituted a small number of patients. In the chemoradiation group, chemotherapy regimens were platinum-based in all, given along with standard antiemetics. Chemotherapy was started on the first day of radiation. Single-agent cisplatin was given in 8 patients at a dosage of 100 mg/m2 every 3 weeks in 6 patients at 25 mg/m2 weekly in 2 patients. Carboplatin, administered weekly at area under the curve = 2, was given in 3 patients. Finally, in 1 patient, cisplatin was given at a dosage of 25 mg/m2 weekly along with fluorouracil, 750 mg/m2, in a continuous infusion for 96 hours every 4 weeks.
Concurrent chemoradiation was well tolerated. There was no interruption of radiation treatment owing to toxic effects observed. In both groups, the most adverse effects were grade I or II radiation dermatitis and mucositis. Grade 3 or higher toxic effects, mostly hematologic, were more frequent in the chemoradiation group than in the radiation-alone group (Table 3). In the concurrent chemoradiation group, grade 3 or higher toxic effects were present in 8 patients (67%), but in the radiation-alone group, they were observed in 2 patients (17%). The most common toxic effects were hematologic toxic effects, followed by mucositis and dermatitis. Long-term adverse events, primarily dry mouth, were mild in both groups. Among those with facial nerve involvement, symptoms related to facial dysfunction included ectropion and epiphora. One patient developed a nonhealing ulcer, resulting in an exposed graft at 6 months after the completion of chemoradiation. This patient subsequently underwent surgical reconstruction with good cosmetic result.
To date, progressive disease has occurred in 10 patients. Overall, all types of recurrent disease were observed among 5 patients in each group. Locoregional recurrence occurred in 5 patients (3 patients in the chemoradiation group and 2 patients in the radiation-alone group), and metastatic disease occurred in 7 patients (4 in the chemoradiation group and 3 in the radiation-alone group). One patient developed both local recurrence and distant metastasis 3 months after completion of chemoradiation (Figure 1). Overall in this series, the observed sites of distant metastasis included lung, liver, brain, bone, mediastinal, and para-aortic lymph nodes.
We performed analyses on time to progression and time to locoregional progression, which did not consider death as an event of interest. The median time to progression and time to locoregional progression seemed better in the chemoradiation group than in the radiation-alone group, although this was not statistically significant (P = .24 and .76, respectively). The median time to progression and time to locoregional progression were both not reached in the chemoradiation group, compared with both 40.3 months in the radiation-alone group.
The median progression-free survival of all 24 patients was 40.3 months (Figure 2), corresponding to 1- and 3-year recurrence-free survival rates of 77% and 52%, respectively. The median locoregional recurrence-free survival was also 40.3 months, corresponding to the 1- and 3-year locoregional recurrence-free survival rates of 86% and 55%, respectively. The median overall survival has not been reached, corresponding to the 1- and 3-year overall survival rates of 86% and 69%, respectively.
When survival was analyzed based on treatment, there was a separation of the survival curves in favor of the chemoradiation group (Figure 3). First, the median progression-free survival was 53.2 months in the chemoradiation group compared with 40.3 months in the radiation-alone group. The estimated 1- and 3-year progression-free survival rates were 83% and 54% in the chemoradiation group but were 69% and 55% in the radiation-alone group, respectively. This difference, however, was not statistically significant (P = .14 by log-rank test).
Second, the median locoregional progression-free survival was 53.2 months in the chemoradiation group compared with 23.3 months in the radiation-alone group. The estimated 1- and 3-year locoregional progression-free survival rates were 92% and 61% in the chemoradiation group but were 80% and 44% in the radiation-alone group, respectively. This difference just approached a level of statistical significance (P = .06 by log rank test).
Finally, to date, death has occurred in 8 of 24 patients: 3 in the chemoradiation group and 5 in the radiation-alone group. Median overall survival was significantly better in the chemoradiation group than in the radiation-alone group: survival not reached vs reached in 23.3 months, respectively (P = .05 by log-rank test). The 1- and 3-year survival rates were 92% and 83% in the chemoradiation group compared with 80% and 44% in the radiation-alone group, respectively. Owing to an increased number of deaths in the radiation-alone group, the median follow-up period was longer in the chemoradiation group than in the radiation-alone group: 31.6 months (range, 20.0-78.5 months) vs 14.9 months (range, 2.8-53.5 months), respectively.
First, we performed an exploratory subgroup analysis of survival among patients with mucoepidermoid carcinoma. There were 11 patients: 5 in the chemoradiation group and 6 in the radiation-alone group. No difference in median progression-free survival was observed: 15.6 months in the chemoradiation group vs not reached in the radiation-alone group (P = .63). Similarly, no difference in median overall survival was observed: both not reached (P = .84).
Second, we performed another subgroup analysis of survival among patients with salivary duct carcinoma or adenocarcinoma. There were 7 patients: 4 in the chemoradiation group and 3 in the radiation-alone group. Although there was no notable difference in the survival between groups, a favorable trend toward survival was observed in the chemoradiation group. The median progression-free survival was 41.8 months in the chemoradiation group, compared with 28.7 months in the radiation-alone group (P = .32). Median overall survival was 53.2 months in the chemoradiation group compared with 36.0 months in the radiation-alone group (P = .16).
We describe our experiences with postoperative chemoradiation or radiation alone for the treatment of patients with high-risk salivary gland carcinoma, using a matched case-control approach. Adjuvant concurrent chemotherapy with a platinum-based regimen was well tolerated, and there was no observed interruption in the delivery of planned radiation therapy, although toxic effects were increased. The survival outcomes of patients treated with chemoradiation compared favorably with those of control patients who were treated with radiation alone. About half of patients in the chemoradiation group attained a durable progression-free survival at 3 years, whereas in the radiation-alone group, recurrence and death continued to be observed.
This report confirms the aggressive clinical course of locally advanced or high-grade salivary gland carcinomas. Over 60% of patients with high-risk major salivary gland carcinomas ultimately developed recurrence or metastatic disease. Recently, Carrillo et al13 proposed a model to predict the probability of recurrence. In that model, T stage, the presence of facial palsy, tumor grade, surgical margin status, and age are the most important determinants of survival. For patients with multiple adverse features (including positive surgical margins, high T stage, facial palsy, age >55 years, and high tumor grade), the estimated 5-year recurrence-free survival is less than 10%.
A somewhat better survival than expected was observed among patients treated with adjuvant chemoradiation. However, this observation should be interpreted with caution owing to the small number of patients with follow-up periods of limited duration. This report cannot delineate the difference or lack of difference in the 2 treatment modalities. In addition, long-term follow-up is crucial because salivary gland carcinomas may present with late recurrence. Chemotherapy may delay, but not prevent, the unfavorable outcomes. Other limitations of our report include a documentation bias, which is inherent to a retrospective study. This is especially problematic when calculating time to progression because patients with recurrent disease might not always return for follow-up at our institution. In addition, treating clinicians often assign patients with more advanced disease to receive adjuvant chemoradiation, although we attempted to find perfectly matched controls. One notable imbalance between the 2 groups was the use of IMRT, which was more frequent in the chemoradiation group. Although to date, IMRT has never been shown to improve survival in head and neck cancers compared with conventional radiation technique, it does allow radiation to be delivered at a high conformity, thereby limiting toxic effects to adjacent organs and theoretically reducing the risk of local recurrence.14
Contrary to the belief that salivary gland carcinomas are resistant to chemotherapy, several reports have confirmed the sensitivity of these tumors, with the exception of adenoid cystic carcinoma, to various chemotherapeutic agents. Cisplatin, carboplatin, paclitaxel, and vinorelbine produce tumor responses in 20% to 30% of patients with metastatic high-grade mucoepidermoid, salivary duct, or adenocarcinomas.15- 17 Interestingly, our exploratory subgroup analysis suggested a particularly favorable outcome with adjuvant chemoradiation for patients with salivary duct carcinoma or adenocarcinoma. Salivary duct carcinomas are among the most aggressive tumors of the salivary gland and are automatically classified as high-grade tumors by the World Health Organization.18 Preliminary data have suggested an efficacy of trastuzumab, a HER-2/neu receptor inhibitor, in combination with chemotherapy for salivary duct carcinoma.19 Further studies are still needed to better understand the effect of biological therapy, particularly the inhibition of epidermal growth factor receptors on salivary gland carcinomas. At our institution, postoperative chemoradiation is offered to selected, fit patients with high-risk pathologic features after a detailed discussion of risks and benefits of treatment.
In summary, our experience suggests that combining radiation with chemotherapy seems to be promising and may improve long-term survival among patients with locally advanced salivary gland carcinomas. This treatment approach warrants further investigation in a prospective clinical trial. Further directions should include investigation of biomarkers specific to salivary gland cancer as well as integration of targeted agents into combined modality management.
Correspondence: Tawee Tanvetyanon, MD, 12902 Magnolia Dr, Tampa, FL 33612 (email@example.com).
Submitted for Publication: July 17, 2008; final revision received October 20, 2008; accepted November 22, 2008.
Author Contributions: Dr Tanvetyanon had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Tanvetyanon, DeConti, and Trotti. Acquisition of data: Tanvetyanon, Qin, and DeConti. Analysis and interpretation of data: Tanvetyanon, Qin, Padhya, McCaffrey, Zhu, Boulware, DeConti, and Trotti. Drafting of the manuscript: Tanvetyanon, Boulware, and Trotti. Critical revision of the manuscript for important intellectual content: Tanvetyanon, Qin, Padhya, McCaffrey, Zhu, and Trotti. Statistical analysis: Tanvetyanon, Zhu, and Boulware. Administrative, technical, and material support: Tanvetyanon, Qin, Padhya, McCaffrey, DeConti, and Trotti. Study supervision: Tanvetyanon, McCaffrey, DeConti, and Trotti.
Financial Disclosure: None reported.