Radiation Dose to Otologic Structures During Head and Neck Cancer Radiation Therapy†
Presented at the Meeting of the Middle Section of the American Laryngological, Rhinological and Otological Society, Inc., Minneapolis, Minnesota, January 25, 1998.
Background: Otologic structures are often contained within head and neck cancer radiation treatment ports. The dosimetry to otologic structures has not been routinely analyzed and radiation treatment planning does not currently attempt to specifically avoid the inner ear structures when dosimetry is calculated. Recent studies demonstrate that up to 30% of patients experience sensorineural hearing loss on multimodality therapy with cisplatin and radiation.
Methods: In the current case series, radiation dosimetry to otologic structures was calculated from computed tomogram treatment plans on patients. Fifteen nasopharyngeal, oral cavity, oropharyngeal, and hypopharyngeal cancer patients were analyzed.
Results: Between 8% and 102% of the total dose is delivered to the petrous bone/cochlea, with 4 of 15 patients getting more than 50% of the dose to at least one cochlea. The mastoid air cells received between 3% and 75% of the total dose, with higher doses being delivered to patients with bulky high neck metastases or nasopharyngeal tumors. The eustachian tubes received between 2% and 102% of the total dose, with 10 of 15 patients receiving more than 50% of the dose to this anatomic site.
Conclusion: We conclude that the cochlea and eustachian tubes receive significant radiation during treatment, particularly in nasopharyngeal cancer patients. Careful design of radiation treatment ports may allow for the reduction of radiation to hearing structures.
In the current treatment of head and neck squamous cell carcinoma, strategies for organ preservation using chemotherapy and radiation therapy have been subject to intense investigation. These treatments usually include radiation in combination with a chemotherapeutic agent, most often cisplatin. These agents in combination have been shown to achieve similar cure rates to standard surgical extirpation of the larynx, when salvage surgery is performed early for treatment failures. 1 Additionally, there is recent evidence that the combination of radiation and cisplatin provides survival advantage over radiation alone in nasopharyngeal carcinoma. 2 Both radiation and cisplatin alone have been shown to be ototoxic, but less is known about their combined ototoxicity. However, a recent report indicates that one third of patients undergoing combined therapy for nasopharyngeal cancer experience hearing loss as a consequence of their treatment. In this longitudinal study, 132 patients with nasopharyngeal cancer underwent serial audiologic testing and it was ascertained that either exposure to radiation therapy after cisplatin or radiation therapy alone resulted in significant sensorineural hearing loss. 3 When similar treatments are given in combination for brainstem gliomas, as many as half of the patients experience ototoxicity. 4 Other ear-related toxicities that are associated with radiation to the nasopharynx and parotid include chronic radiation otomastoiditis, osteoradionecrosis of the temporal bone, and eustachian tube dysfunction. 5,6 These toxicities also may contribute to hearing loss and have been considered unintended consequences of treatment.
In standard radiation therapy for head and neck cancer, certain strategies have been developed to minimize the radiation exposure to important central nervous system structures including the spinal cord, brain stem, optic chiasm, and pituitary gland. 7 Additionally, in the radiation treatment for paranasal sinus carcinoma, shielding is adopted to minimize radiation exposure to the eye, a sensory end-organ. 8 Current radiation treatment strategies have not been adopted to similarly protect the cochlea.
In the present study, an examination of radiation doses to specific otologic structures including the mastoid, cochlea, and eustachian tube was attempted. These calculations were based on the computer-generated treatment plans for 15 patients on an organ preservation phase I intramural trial of the National Institute on Deafness and Other Communication Disorders and the National Cancer Institute.
MATERIALS AND METHODS
Fifteen consecutive patients with stage III and stage IV head and neck squamous cell carcinoma in an internal review board–approved organ preservation/unresectable patient pilot study using chemotherapy and radiation therapy (NCI 95-CO192) were included in the current analysis. All patients received full-course, standard radiation therapy (180 cGy/fraction per day) as a portion of their treatment. Three patients received intraoral boosts with radioactive implants, but brachytherapy dosimetry was performed on a different treatment planning system at a second institution and these fields were not included in the dosimetry calculation in this study. Tumor sites included the oropharynx, nasopharynx, oral cavity, hypopharynx, and larynx. Computer tomogram–generated treatment plans were made for all patients.
The Treatment Planning System (TPS) used in these studies was developed at the National Cancer Institute. It is based on the Macintosh Desktop computer and is the subject of a continuing development program. The basic algorithm for dose calculation from external beam irradiation was described originally by van de Geijn and Fraass 9 and subsequently extended by van de Geijn. 10 Recent versions of the TPS read directly both computed tomography (CT) images and magnetic resonance (MR) images, and allow the contouring (outlining) of body organs within these images. The CT scans of our 15 patients were done at a variety of slice spacings: two at 3-mm spacing, three at 5-mm spacing, and 11 at 10-mm spacing. The organs studied were the cochlea, the eustachian tubes, and the mastoid air cells. At a CT spacing of 10 mm, these organs are mostly contained in one slice and in our small sample did not appear in more than two slices, but at spacings of 3 and 5 mm, the mastoid air cells can appear in as many as nine or seven slices respectively. Thus it is clear that the identification of these otologic organs depends on the spacing of the CT slices. For each patient, the otologic organs were outlined in each CT slice in which they were visible and dose calculations were performed for each field/block configuration used. In these calculations, the density was assumed to be unity throughout the head and neck. The estimated mean dose for each organ (left and right separately) on each slice was read from the TPS using the “point dose” feature, i.e., on each slice on which that organ appeared, the cursor was used to read the dose at various positions across the organ. From these measurements a mean was estimated; in addition, the area of the organ on that slice was estimated. From these measurements an average dose over the whole organ (left and right separately) was calculated. Average doses for each field/block configuration were added to determine the total dose for each organ. Doses from the posterior cervical chain electron fields were not included. Using this system, the cochlea and petrous portion of the temporal bone, the pneumatized mastoid cavity, and the region of the bony and soft tissue areas of the eustachian tubes were identified on CT treatment planning scans. The scans were performed for radiation planning purposes and did not contain complete visualization of all structures in every case. For that reason, several structures were not visualized on the scans and were excluded from study. In the present analysis, 22 cochleae, 14 eustachian tubes, and 16 mastoid air cells were fully visualized and evaluable. As noted below, future studies should include a higher quality temporal bone CT (perhaps standard 1-mm spacing) to provide better visualization of the otologic organs. However, our treatment planning CT system is gauged to provide accurate treatment plans based on standard radiation fields, and algorithms for obtaining typical high-resolution CT scans of the temporal bone are not currently a part of this system.
Patients enrolled on this study were simultaneously given paclitaxel during their radiation treatment. Audiometry was obtained only in the patients with nasopharyngeal cancer.
All patients underwent full-dose radiation therapy to the tumor and lymph node fields. Because the radiation fields differed from case to case, they were separated into treatment groups based on the site of the lesion. These groups included laryngeal/hypopharyngeal sites, oropharyngeal sites, oral cavity sites, and nasopharyngeal sites of disease. All patients had stage III or stage IV disease, so very limited fields for treatment (e.g., small fields for T1 laryngeal lesions) were not utilized. No specific attempts were made to exclude any otologic structures from the treatment fields.
There were four patients with hypopharyngeal and laryngeal tumor sites, and radiation dose to their otologic structures is shown in Table I. All four patients received 7000 cGy or more to their tumors. All patients in this group had minimal mastoid irradiation on average (less than 2000 cGy). However, two patients with T3 pyriform sinus lesions (patients 2 and 3) received 4000 cGy or more to their eustachian tubes on the side ipsilateral to the tumor. At 1 year, neither patient had significant eustachian tube dysfunction requiring polyethylene (PE) tube placement or other intervention. None of the four patients in this group received more than 4000 cGy to the area of the cochlea, a dose that has been considered as a minimal toxic dose to other structures that contain neural tissue. 11 However, one patient with bulky neck disease (patient 4) did receive 3500 cGy to the cochlea ipsilateral to the neck disease.
Table Table 1.. Dose to Temporal Bone Structures in Patients With Hypopharyngeal or Laryngeal Tumors.
Three patients underwent treatment for oropharyngeal disease (Table II). The mastoid structures of these patients received no more than 3300 cGy of radiation. The patient with a 3300-cGy dose to the mastoid had high, bulky neck disease (N3) ipsilateral to the radiated mastoid (patient 6). Two patients with significant lateral oropharyngeal extension of their primary disease received more than 4000 cGy to the eustachian tube area, on the side ipsilateral to the tumor (patients 6 and 7). The patient with high N3 neck disease extending around the sternocleidomastoid muscle at the skull base (patient 6) received 3900 cGy of radiation to the ipsilateral cochlea. Interestingly, one patient who presented with a T4 tonsillar cancer (patient 5) experienced minimal cochlear radiation on the ipsilateral side (650 cGy) compared with the patients who had more extensive neck disease.
Table Table 2.. Oropharyngeal Cancer Patients and Radiation Doses.
Five patients were treated and their plans evaluated for primary tumors of the oral cavity (Table III). All but one of these patients all had stage IV disease. Two patients had tumors involving the mobile tongue, while the other patients had primary disease in the anterior floor of mouth. Three of the patients (patients 8, 9, and 12) received 5400 cGy or less of external beam radiation to their primary sites. These patients received interstitial implant brachytherapy after external beam radiation therapy. The other two patients received 7020 and 7200 cGy to their tumor sites and necks. All patients receiving radiation to their anterior floor of mouths also received more than 4000 cGy to their eustachian tubes (patients 10, 11, and 12). This represents at least 55% of the total dose of radiation. The mastoid areas received 1100 cGy or less in all cases, representing 20% or less of the total dose to the tumor. Only one patient received more than 3000 cGy to the cochlea. This patient had a T4 tumor of the floor of mouth (patient 11). One patient died at 8 months after treatment, but the other patients (4/5) are currently alive more than 6 months after treatment. None of these patients experienced eustachian tube problems after radiation treatment. These data are presented in Table III.
Table Table 3.. Radiation Dose to Structures in Patients With Oral Cavity Tumors.
The final group of patients consists of three patients who were irradiated for tumors involving the nasopharynx. Two of the three patients had nasopharyngeal primary tumors and the other patient had a soft palate tumor with posterior nasopharyngeal wall involvement. These three patients received 7200 cGy to their primary sites and in excess of 6200 cGy to their cochleas on at least one side. One patient (patient 14) actually received a calculated dose of radiation to the petrous apex/cochlea in excess of that calculated to the primary site. All patients in this group essentially received full-dose radiation to their eustachian tubes, as they were all involved with tumor. The patients received less radiation to their mastoid cavities than to the aforementioned sites. Two patients received in excess of 4000 cGy to the mastoid on at least one side (patients 14 and 15). Two of the three patients (patients 14 and 15) experienced middle ear fluid collections that subsided without PE tube placement. Pretreatment audiometry was performed in these patients. Audiometric monitoring at 6 months after treatment revealed no changes in sensorineural hearing levels. The other 12 patients did not undergo pretreatment audiologic testing. Those patients alive at 6 months after treatment did not complain of hearing loss, exception for patient 6, who had a PE tube placed for an effusion 6 months after the completion of treatment. No other patients in the study experienced symptoms that would prompt auditory evaluations for the first 6 months after the completion of treatment.
To our knowledge, this is the first study that has attempted to quantify the radiation to otologic structures based on CT-guided treatment plans. Recently, it has been shown that patients undergoing radiation for head and neck tumors experience sensorineural hearing loss after treatment. 3,12,13 This has also been the subject of a recent review. 14 Additionally, the combination of cisplatin with radiation, especially for nasopharyngeal cancer, will likely become the treatment standard in the future. Both of these agents are individually ototoxic, and it is not known whether their combination will produce additive or synergistic ototoxicity beyond what might be expected from either agent alone.
We have found that otologic structures are most at risk for receiving high-radiation doses in patients with cancer arising in or involving the nasopharynx. In one case the patient actually received a greater calculated dose to the cochlea than to the primary tumor site. For cancers of this site, the eustachian tubes always receive essentially the full tumor dose (Table IV), and this is unavoidable. In tumors at other head and neck sites, the only patients that received a significant dose of radiation to the cochlea were patients who had high bulky jugulodigastric adenopathy. Interestingly, several patients had more than 4000 cGy to the area of the eustachian tubes, even if their tumors involved anatomic sites relatively distant from the cochlea. One patient with a pyriform sinus primary received full-dose radiation therapy to the eustachian tube on the involved side. No patients received a massive dosage of radiation to the mastoid air cells and no patients have yet to present with any symptoms of radiation mastoiditis or osteoradionecrosis of the temporal bone. This series of patients, however, did not include patients with parotid or temporal bone malignancy, two anatomic sites that would likely receive a greater dose of radiation to the external ear canal and mastoid. Also, none of the patients in the series had temporomandibular joint involvement with tumor, a condition in which a higher dose of radiation therapy might be delivered to the external auditory canal and mastoid.
Table Table 4.. Nasopharyngeal Radiation Sites.
One difficulty that we encountered in this study is that the treatment planning CT images included only one to three sections through the petrous apex including the cochlea. Some of the CT images were not evaluable because the quality of the image, the angle at which the image was taken, or the lack of clear imaging through the petrous apex precluded further analysis. In the future, a standard temporal bone CT, involving thin sections through the cochlea, could be performed at the time of treatment planning. The CT treatment plans are designed to give good images through the tumor and the immediately surrounding important anatomic structures (e.g., spinal cord) for the design of radiation delivery. To get pinpoint accuracy on other structures (e.g., the cochlea vs. the labyrinth vs. the ossicles), a higher quality treatment planning CT is necessary. This would involve greater time and cooperation from the patient, and perhaps increased overall cost. However, this additional information may be very useful if a significant number of patients are long-term survivors and experience hearing loss as a result of their treatment.
Currently, many combined modality plans under investigation involve cisplatin or paclitaxel individually or in combination with radiation delivery. At present it is not known whether there will be increased short-term or long-term hearing morbidity in patients undergoing treatment with radiation in combination with these compounds. Platinum compounds are both ototoxic and neurotoxic. 15–17 Paclitaxel has the ability to cause neuropathy as well. 18 Cisplatin in combination with simultaneous radiation has qualities that may be radiosensitizing. 19 Thus it may be important to assess hearing before and after treatment in patients undergoing these regimens. In addition, many of the current treatment plans for patients in these national protocols do involve treatment planning CTs and therefore could be evaluated to ascertain the dose to the otologic structures. If there is an increase in ototoxicity as a result of these treatments (which correlates with dose to those structures), and it becomes a significant morbidity for long-term survivors, it may be important to closely reevaluate the port design for the delivery of radiation. It may be possible in some cases to redesign the ports to specifically exclude the cochlea from the field, if there is no therapeutic gain to including it. The current technology of CT guided treatment planning already allows this. 9,10 For example, with CT treatment plans, both the direction of the beam and the use of lead blocking could be employed to more accurately configure a field to exclude the petrous apex. Clearly, further study of the radiation dose to the otologic structures and its relation to hearing loss is necessary before the value of any potential field changes can be assessed.
Otologic structures were at risk for radiation exposure when the nasopharynx was the primary tumor site in our clinical trial. Patients in this study with large adenopathy in zone 2 and upper zone 5 of the neck may experience up to 50% of the total radiation dose to their cochleas. Even patients with primary tumor sites anatomically far removed from the nasopharynx and petrous apex (hypopharynx) may receive high doses of radiation in selected cases (patient 2). With the increasing survival of patients with these diseases due to more intense combined modality treatment (chemotherapy and chemoradiation), and the increasing availability of sophisticated three-dimensional treatment planning systems, we believe further study is warranted of the radiation dose to the otologic organs and any correlation of this dose with hearing loss.