Edited by: Anatoly Dritschilo, Georgetown University, USA
Reviewed by: Stephan Bodis, Kantonsspital Aarau, Switzerland; Wenyin Shi, Thomas Jefferson University, USA
Specialty section: This article was submitted to Radiation Oncology, a section of the journal Frontiers in Oncology
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
The late effects of RT are not well reported in patients with oral tongue cancer (OTC). This study reports the incidence of late effects and factors associated with the development of late effects in OTC patients.
Patients with OTC treated in our institution from 2003 to 2013 were evaluated. The association between RT doses, including mandible maximum and minimum doses and total 3D maximum dose, and late toxicity, defined as development of osteoradionecrosis (ORN), percutaneous endoscopic gastrostomy (PEG) tube dependence for >6 months after treatment, and narcotic dependency >6 months posttreatment were assessed using both univariate and multivariable (MV) analysis.
Seventy-six patients with OTC (45% males and 55% females) were treated with definitive surgical resection followed by adjuvant RT. The median follow-up was 4.3 years. Combined late toxicities were reported in 38% of patients. Thirty-four percent of the patients had narcotic dependency and, 3.9% of the patients had ORN of the mandible. Thirteen percent of patients developed PEG tube dependency that was significantly associated with a higher 3D maximum radiation dose on univariate analysis (
Patients with OTC treated with adjuvant RT are at significant risk for development of late toxicities. Increasing maximum dose is associated with long-term PEG tube dependence, and care should be taken to reduce the “hot spot” within radiation treatment plans as much as possible.
Cancers of the oral tongue represent the most common primary site of oral cavity cancer (OCC), the majority of which are squamous cell carcinomas. In 2014, approximately 28,030 patients were diagnosed with OCC, of which 13,590 had oral tongue cancer (OTC) (
Advances in reconstructive surgery have led to better functional outcomes following primary surgical resection (
Given such high rates of grade 3 or higher late toxicities in postoperative head and neck cancer patients, the goal of this study was to retrospectively analyze clinical and treatment-related variables that may contribute to the development of late side effects within OTC patients. Specific late toxicity endpoints in this study include osteoradionecrosis (ORN) of the mandible, long-term percutaneous endoscopic gastrostomy (PEG) tube dependence, and long-term narcotic dependency.
After Institutional Review Board (IRB) approval at the Winship Cancer Institute of Emory University, a retrospective chart review of OTC patients treated between 2003 and 2013 was performed (IRB code number: 00016211). Inclusion criteria for this study included a confirmed diagnosis of squamous cell carcinoma of the oral tongue and surgical resection followed by RT delivered at Winship Cancer Institute. Patients treated by surgical resection alone, patients receiving adjuvant treatment outside of Emory, and patients with any distant metastases at the time of diagnosis were excluded from this study. All patients had routine pretreatment evaluations consisting of a complete history, physical examination, blood tests, computed tomography (CT), or positron emission tomography (PET) of the head and neck region and chest X-ray or chest CT as clinically appropriate.
All data were collected using electronic medical records and departmental radiation oncology charts. The clinical parameters of interest included tumor characteristics, stage, age, radiation techniques and dose, treatment dates, and late toxicities following radiation. In this study, data for late toxicity including ORN, PEG tube placement, and narcotic dependency rates were retrieved from clinical notes at each follow-up visit. All radiological examination reports were reviewed for general pathological changes and for changes specific to ORN lesions. PEG tube dependency was reported as continuation of tube feedings 6 months after completion of treatment and narcotic dependency as continued use after 6 months of completion of treatment.
Computed tomography simulation was completed following the surgery using a GE Lightspeed Large Bore CT Scanner (GE Healthcare, Waukesha, WI, USA). A thermoplastic head and shoulder mask (WFR/Aquaplast Corporation, Wyckoff, NJ, USA) was used for each patient during simulation and treatment for immobilization. The treatment planning CT and intravenous contrast was matched with PET (if obtained) in the treatment position to better define target volumes. The RTOG guidelines were used to contour organs at risk (OARs) (
Statistical analysis was performed using SAS 9.4 statistical software (SAS Institute Inc., Cary, NC, USA). Categorical variables were compared across toxicity endpoints using chi-squared tests or Fisher’s exact tests, where appropriate, while continuous variables were compared across endpoints using ANOVA. For the multivariable (MV) logistic regression models, we modeled the probability of an event (long-term narcotic dependence, PEG tube dependence, or any adverse event) as a function of 3D maximum dose (%) and the three most significant predictors in each univariate comparison. For ORN, we modeled the probability of ORN as a function of mandible minimum dose (%) and mandible maximum dose (%). The number of variables in each MV model was restricted due to the low number of toxicity events.
Seventy six patients (45% males and 55% females) met the inclusion criteria. The median follow-up time was 4.3 years. Patient characteristics are demonstrated in Table
Characteristics | No. of patients (%) |
---|---|
Age (y) | |
Median | 56 |
Range | 15–89 |
Race | |
White | 59 (78) |
Other | 17 (22) |
Gender | |
Female | 42 (55) |
Male | 34 (45) |
Histology | |
SCC | 76 (100) |
T stage | |
T1 | 35 (46) |
T2 | 21 (28) |
T3 | 8 (10) |
T4 | 12 (16) |
N stage | |
N0 | 42 (55) |
N1 | 19 (25) |
N2 | 15 (20) |
Bone invasion | 3 (4) |
PNI | 42 (55) |
ALI | 23 (30) |
Table
No. of patients (%) | |
---|---|
Radiology | |
CT | 57 (75) |
PET/CT | 19 (25) |
Chemotherapy | |
None | 51 (67) |
Cisplatin | 18 (24) |
Carboplatin and paclitaxel | 6 (8) |
Cetuximab | 1 (1) |
Concurrent chemoradiotherapy | 25 (33) |
Surgical procedure | |
Primary alone | 7 (9) |
Primary and ND | 69 (91) |
Radiation therapy | |
Prescribed dose | |
Median | 63 Gy |
Range | (59.4–70.29 Gy) |
Fraction size | |
Median | 2 Gy |
Range | (1.8–2.2 Gy) |
3D dose max (%) | |
Mean | 109 |
Median | 109 |
Range | (104–129) |
Mandible min. dose | |
Mean | 3.5 Gy |
Median | 2 Gy |
Range | (0–13 Gy) |
Mandible max. dose | |
Mean | 63 Gy |
Median | 66 Gy |
Range | (10–85 Gy) |
Mandible median dose | |
Mean | 47 Gy |
Median | 50 Gy |
Range | (8–76 Gy) |
Late toxicity was reported in 38% of patients. Three patients developed ORN alone or with other late toxicities, 10 patients had PEG tube dependency alone or with other late toxicities, and 26 patients had narcotic dependency alone or with other late toxicities (Table
No. of patients with late toxicity | % of patients with late toxicity | |
---|---|---|
ORN | 3 | 3.9 |
PEG dependence at 6 mo | 10 | 13 |
Narcotic dependency | 26 | 34 |
Total patients with late toxicity | 29 | 38 |
Individual late toxicities—ORN, PEG tube dependency, and narcotic dependency—were then analyzed separately (Table
Parameters | Yes | No | |
---|---|---|---|
3D dose max (%) | 0.002 | ||
Mean | 112.5 | 108.8 | |
Median | 110.7 | 108.7 | |
Number of patients | 10 | 66 | |
3D dose max (%) | 0.53 | ||
Mean | 108.9 | 109.5 | |
Median | 108.3 | 109.4 | |
Number of patients | 26 | 50 | |
3D dose max (%) | 0.12 | ||
Mean | 106.0 | 109.4 | |
Median | 106.9 | 109.2 | |
Number of patients | 3 | 73 |
Odds ratio (95% CI) | ||
---|---|---|
3D dose max | 1.27 (1.00–1.61) | 0.05 |
T stage (reference = T4) | ||
T1 | 0.46 (0.03–6.29) | 0.08 |
T2 | 1.12 (0.12–9.97) | |
T3 | 0.62 (0.03–13.45) | |
Margins (reference ≤5 mm) | 0.26 | |
Positive | 1.21 (0.09–16.84) | |
Negative | 0.26 (0.05–1.43) | |
Radiology (reference = CT) | ||
PET/CT | 1.88 (0.32–10.96) | 0.48 |
Mandible min. dose (%) | 0.87 (0.57–1.32) | 0.51 |
Mandible max. dose (%) | 0.97 (0.93–1.01) | 0.18 |
3D dose max (%) | 0.97 (0.83–1.12) | 0.65 |
PNI (reference = negative) | 0.97 (0.83–1.12) | 0.21 |
ALI (reference = negative) | 0.61 (0.18–2.11) | 0.44 |
Treatment group (reference = post-CRT) | 0.66 | |
Surgery | 1.46 (0.33–6.55) | |
Postoperative RT | 0.83 (0.21–3.20) |
In the present study, we demonstrate that 38% of patients with OTC developed late toxicities (inclusive of long-term PEG dependence, ORN, and long-term narcotic dependence) after surgical resection and adjuvant radiation treatment. When analyzing all late toxicities together, there were no specific variables associated with the development of late toxicities. However, when analyzing individual late toxicities, the maximum radiation dose, or “hot spot,” was significantly higher in patients with long-term PEG tube dependency on both univariate and multivariate analysis.
Mitigation of toxicities and mechanisms to improve quality-of-life are areas of active research for patients with OTC. While the addition of chemotherapy to postoperative radiation improved 5-year overall survival rate from 40 to 53% in EORTC trial, it also resulted in increased toxicity (
Designing radiation plans that limit dose heterogeneity and maximum dose is a critical part of the radiation treatment planning process. RTOG 0920, an active phase III trial of surgery and postoperative radiation ± cetuximab mandates a maximum dose (or hot spot) outside of the PTV to be limited to ≤110% of the prescribed dose. Keeping the maximum dose to ≤110% is ideal, as hot spots above 110% put normal tissue at risk for significant side effects of radiation. In order to minimize the maximum dose, or hot spot, dosimetrists create optimization planning target volumes (OPTI-PTV) and create avoidance structures for OARs. However, for some difficult cases where the initial optimization meets all given parameters but has a hot spot greater than 110%, the dosimetrist will convert an isodose line of less than 110% (between 106 and 108%) to a control point structure and reoptimize the plan starting from the last iteration and give a constraint of equivalent dose of the structure with a high priority. In the majority of cases this will minimize the hot spot below 110%. The dosimetrist can create another control point structure and repeat the same process or they can manipulate the normal tissue optimization settings and reoptimize the plan. That can be done alone or combined with the second control point structure. In most cases, these steps will drop the hot spot to less than 110%.
Prophylactic gastrostomy tube placement has been routinely offered to patients with locally advanced head and neck cancer prior to initiating definitive chemoradiation therapy at our institution. During treatment, patients are closely assessed by speech pathologists to determine swallowing function. After completion of therapy, PEG tube removal is recommended if patients are able to tolerate all nutrition orally for at least 2 weeks and adequately maintain weight. In the past, we have reported that 25% of elderly patients with oropharyngeal squamous cell carcinoma develop PEG tube dependency (
Pain following head and neck cancer therapies often leads to significant psychological and physical suffering resulting in long-term narcotic dependency. Radiation-induced mucositis typically improves 4–6 weeks after completion of treatment (
In our study, 3.9% of patients developed ORN, within previous reports in the literature. The incidence of ORN was estimated to be 4.74–37.5% historically; however, less than 5% incidence has been reported in recent studies, likely attributed to the improvement of dental care and radiation techniques (
There are several limitations of this study. First, the retrospective nature of this study hinders the ability to collect data on other relevant late toxicity endpoints for this patient population, including fibrosis and more detailed data regarding swallowing function and xerostomia. Additionally, other risk factors for developing late toxicities, including tobacco use, were not consistently reported and therefore were not included in this analysis. Second, the cohort included in this study included oral tongue patients only; therefore, the number of patients was small with a median follow-up time of 25.8 months. This small sample size is likely underpowered to detect potential significant differences. Furthermore, improvements in treatment planning techniques during the time period of this study could have also accounted for variation in treatment-related toxicity. There is certainly a learning curve with IMRT treatment planning; however, in order to generate a large cohort of oral tongue patients, the study time period is somewhat extensive and thus variability in treatment planning over time remains an issue. Larger prospective studies are required to overcome such limitations in order to provide improved insights into the development of late toxicities.
In conclusion, this study demonstrates that the risk of long-term narcotic dependence, ORN, or PEG tube dependence is 38% after completion of multimodality treatment for OTC. Furthermore, the development of long-term PEG tube dependency in these patients was significantly associated with higher RT maximum dose. Further evaluation of late toxicities in the randomized setting is needed in order to better understand patient-specific and treatment-related risk factors that may increase the likelihood of late side effects after treatment. Care should be taken to minimize the “hot spot” in radiation planning for patients with OTC treated postoperatively in order to reduce the risk of long-term PEG tube dependency.
MS, RC, JS, OK, NS, CS, DS, JW, ME-D, MP, JB, and KH all contributed to study concept, design, and/or acquisition of data. MS and RC completed the data collection. JW contributed to the data analysis. MS and KH were responsible for drafting the manuscript. All authors contributed to revising and giving final approval to the manuscript. All authors agree to be accountable for all aspects of the work including its accuracy and integrity.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.