Bilateral osteonecrosis of the external auditory canal caused by bone-modifying agents for cancer bone metastases

Abstract

A 51-year-old woman with a 54-month history of treatment with bone-modifying agents for thyroid cancer bone metastases, namely zoledronic acid for three months and denosumab for 51 months, presented with jaw pain, trismus, and ear discharge. Physical examination, bacterial culture, and radiological findings led to a diagnosis of osteonecrosis of the external auditory canal. Since the discontinuation of denosumab did not improve the diagnosed condition, she underwent surgical reconstruction. Bone-modifying agents such as zoledronic acid and denosumab, which are commonly used to prevent skeletal-related events in patients with bone metastases from various cancers, have been used for several years, especially for patients with slow-growing cancers. Osteonecrosis of the external auditory canal is an infrequent but serious adverse event caused by prolonged use of bone-modifying agents. Therefore, clinicians should be fully aware of its potential risks in the long-term management of bone metastasis.

Introduction

Patients with cancer bone metastases (BM) are at risk of skeletal-related events (SREs) such as bone pain, pathologic fractures, spinal cord compression, and hypercalcemia [1]. As SREs negatively affect the activities of daily living, quality of life, and survival of patients with cancer [2–7], multidisciplinary measures to avoid SREs are a crucial part of care for patients with various cancer types [1–8]. Denosumab and some bisphosphonates, including zoledronic acid (ZA), are categorized as bone-modifying agents (BMA) and are used to prevent SREs by inhibiting overactivated bone resorption. The clinical benefit of BMA for many solid cancers has been demonstrated [1–9]. Osteonecrosis of the external auditory canal (ONEAC) is a potential adverse event associated with BMA and other bisphosphonates [10–23]. However, as ONEAC is a rare condition compared to the other side effects of BMA, such as medication-related osteonecrosis of the jaw (MRONJ) and atypical fractures [24, 25], it has only recently gained attention. Here, we report a case of bilateral ONEAC in a patient who received prolonged BMA treatment for unresectable BM from thyroid cancer.

Case report

A 51-year-old woman with BM from follicular thyroid cancer undergoing denosumab therapy presented with increasing pain in her left jaw and trismus. She was diagnosed with follicular thyroid cancer at the age of 20 years and underwent thyroidectomy. Subsequently, she underwent multiple resections of metastases to her lung, brain, and bones, as her tumor was resistant to radioactive iodine therapy. The last surgery, which involved curettage and cementing of the bone metastasis in the ilium, was performed when she was 40 years of age. Five years later, the tumor regrew with apparent osteolytic changes. Therefore, monthly ZA administration was initiated. With no improvement in her localized pain after three doses, we decided on external beam irradiation and replaced ZA with denosumab, which successfully resolved both the pain and tumor growth. After 51 months of monthly denosumab treatment, the patient experienced discomfort in the left jaw. Considering this symptom and her history of long-term BMA use, we initially suspected MRONJ; however, no oral findings were identified to confirm the MRONJ diagnosis. Simultaneously, the patient noticed consistent discharge from the left ear and consulted the otorhinolaryngology, head, and neck surgery department. Otoscopic examination revealed ulceration in the anterior wall of the left external auditory canal (EAC) that exposed the underlying bony surface (Fig. 1a). Fluorodeoxyglucose-positron emission tomography (FDG-PET) revealed slight-to-moderate uptake adjacent to the EAC, suggesting inflammation (Fig. 1b). A computed tomography (CT) scan at that time showed an intact temporal bone. She was initially diagnosed with otitis externa and underwent oral antibiotic and topical treatments. After 3 months, an edematous and a hemorrhagic mass had grown in the left EAC, and an ulcer was observed in the right EAC. Histopathological examination of the biopsy specimen revealed inflammatory granulation tissue without evidence of thyroid cancer metastasis. No bacterial infection was detected. Based on the above findings, she was diagnosed with bilateral ONEAC. Therefore, denosumab was discontinued because of its possible contribution to the condition.

Fig. 1.

Otoscopy of the left ear at disease onset (a) shows ulceration of the external auditory canal wall (white arrowheads), and fluorodeoxyglucose-positron emission tomography scan (b) shows slight but apparent uptake around the bilateral ear canals (black arrowheads), suggesting local inflammation. TM, tympanic membrane

Eleven months after the initial presentation of her symptoms, a hypertrophic granulation mass in her left EAC continued to grow until it almost completely obstructed the canal (Fig. 2a). Bone exposure in the right EAC did not improve (Fig. 2b). A CT scan indicated osseous destruction within the bilateral temporal bones (Fig. 3a, b).

Fig. 2.

Otoscopy performed 11 months after disease onset shows an edematous granulation tissue mass extending toward the left external auditory canal (EAC) (a, white arrowheads) and ulceration and bone exposure in the right EAC (b, the exposed bone is marked with black arrowheads)

Fig. 3.

Computed tomography images of the temporal bone 11 months after the onset show soft tissue growth within the right external auditory canal (a, arrowhead) and destructive changes in the bilateral temporal bones (a, b)

The patient’s condition did not improve even after denosumab discontinuation and repetitive minor surgical debridement of the granulation mass at an outpatient clinic of the otorhinolaryngology, head, and neck surgery department (Fig. 4a, b). The patient consequently developed impaired hearing due to canal obstruction. Finally, after a 48-month conservative management period, a major surgical debridement and canaloplasty of the bilateral EAC were performed. Histopathological examination of the curetted necrotic bone revealed sequestrum with loss of nuclei, accompanied by granulation tissue. Thirty-six months postoperatively, no ONEAC recurrence was observed, and her hearing recovered completely. In the last 7 years, she has not received denosumab and has been undergoing only thyroid-stimulating hormone suppression treatment. The iliac metastasis has remained asymptomatic, with no progression. Brain metastasis was newly diagnosed during the follow-up period but was successfully treated with radiosurgery.

Fig. 4.

Otoscopy of the left (a) and right (b) external auditory canals 18 months after disease onset shows no improvement after conservative treatment, which included minor debridement and occasional oral antibiotics. The granulation mass is marked with white arrowheads (a) and bone exposure is marked with black arrowheads (b)

Discussion

Modern cancer treatment strategies have markedly improved the survival of patients with cancer, including those with BM [6]. The importance of SRE prevention throughout the extended lives of these patients has been widely acknowledged. BM management may become imperative over time, including the use of BMA, which are potent systemic medications commonly used for patients with BM, in combination with other anticancer therapeutics. In differentiated thyroid cancer (DTC), the incidence of BM is reportedly 2–15% [2, 7, 26]. The 5-year and 10-year overall survival of patients with DTC after initial BM diagnosis is reported as 61–64% and 27–45%, respectively [26, 27], whereas 37–78% of patients with DTC diagnosed with BM experience SREs [7, 26, 28, 29]. In the management of unresectable BM caused by DTC, a systemic approach is essential to prevent SREs [29, 30], thereby necessitating prolonged treatment with BMA, which may increase the risk of adverse events.

ONEAC associated with ZA and pamidronate use for multiple myeloma was first reported in 2006 [10]. Although rare, other antiresorptive drugs, including BMA and anti-osteoporosis medications, have been reported to cause ONEAC. Through a search of the literature, we identified an additional 24 cases (Table 1) [10–22]. Among these 25 ONEAC cases, 19 used antiresorptive agents for osteoporosis treatment, and six neoplastic cases used antiresorptive agents as BMA for SRE prevention. Neoplastic cases have been reported in patients with myeloma and breast cancer, despite the limited number of ONEAC cases. Although the risk of ONEAC with denosumab use is assumed to be similar to that of ONEAC with bisphosphonate use, 76% of the 25 cases were associated with bisphosphonate. In our patient, denosumab seemed more likely than ZA to have caused ONEAC because she had been on denosumab for much longer (51 months) than she was on ZA (3 months). To our knowledge, this is the seventh reported case of ONEAC associated with denosumab treatment. Furthermore, given the limited number of ONEAC cases caused by BMA in the context of cancer treatment, our case is the second-reported case of ONEAC occurring in a patient with cancer treated with denosumab. The incidence of ONEAC associated with BMA should be more widely recognized by physicians in various oncology fields, as well as by head and neck surgeons and otorhinolaryngologists.

Table 1.

Summary of reports of osteonecrosis of the external ear canal that was associated with bone-modifying or other antiresorptive agents [10–22]

	Sex	Age	Indication	Antiresorptive medication	Duration	Clinical manifestation	Treatment	Clinical course	Report (Year)
1	M	64	Multiple myeloma	BP	9 years	Asymptomatic	Conservative	Improved	Polizzoto et al. (2006)
2	F	51	Breast cancer	BP	5 years	Asymptomatic	Surgery	Healed	Froelich et al. (2011)
3	F	70	Breast cancer	BP	11 years	Asymptomatic	Surgery	Healed
4	F	64	Multiple myeloma	BP	6 years	Delayed wound healing after ear surgery	Surgery	Healed after revision surgery
5	F	74	Osteoporosis	BP	Not described	Painful infection	Conservative	Unchanged	Kharazmi et al. (2013)
6	F	67	Osteoporosis	BP	2 years	Changes in the auditory canal	Conservative	Unchanged
7	F	88	Osteoporosis	BP	several years	Chronic otitis	Not described	Not described
8	F	79	Osteoporosis	BP	10 years	Pain and otorrhea	Conservative	Partly improved	Salzman et al. (2013)
9	M	64	Multiple myeloma	BP	6 years	Ear infection	Conservative	Unchanged	Wickham et al. (2013)
10	F	72	Osteoporosis	BP	4 years	Ear infection	Surgery	Not described
11	F	76	Osteoporosis	BP	14 years	Ear pain	Surgery	Partly improved	Thorsteinsson et al. (2015)
12	5 F 1 M		Osteoporosis	BP	20 months – 15 years	Otitis externa in 2 patients, wax impaction in 4 patients	Conservative	Healed	McCadden et al. (2018)
13							Conservative	Improved
14							Conservative	Healed
15							Conservative	Healed
16							Conservative	Healed
17							Surgery	Healed
18	F	81	Osteoporosis	Den	1.5 years	Otorrhea, hearing loss	Surgery	Healed	Takeda et al. (2021)
19	F	79	Osteoporosis	BP → Den	> 10 years → 4 months	Hearing loss, otalgia	Conservative	Healed	True et al. (2021)
20	F	68	Breast cancer	Den	3 years	Otorrhea	Surgery	Healed	Perina et al. (2021)
21	F	79	Osteoporosis	Den	2 years	Asymptomatic	Conservative	Unchanged	Kumar et al. (2023)
22	F	64	Osteoporosis	BP → Den	5 years → 5 years	Ear Pain	Conservative	Unchanged
23	F	84	Osteoporosis	BP → Den	Unknown → 2 years	Ear pruritis, otorrhea	Surgery	Healed	Van Dolah et al. (2023)
24	M	69	Osteoporosis	BP	4 years	Ear fullness, dizziness, pain	Conservative	Unchanged
25	F	65	Osteoporosis	BP	Not described	Ear blockage, itchiness	Conservative	Healed	Mayer et al. (2024)
26	F	51	Thyroid cancer	BP → Den	3 months → 4 years	Jaw pain, trismus	Surgery	Healed	Present study, (2024)

F female, M male, BP bisphosphonates, Den denosumab

EAC cholesteatoma is another ear disorder, with a reported possible association with bisphosphonate use [23, 31, 32]. Two Japanese case series introduced 16 such cholesteatoma cases, all of which were associated with bisphosphonate use [31, 32]. A Danish national register-based cohort study on bisphosphonate-related EAC cholesteatoma reported an incidence of 0.026%, i.e., 34 cases among 131,794 patients with osteoporosis undergoing bisphosphonate treatment [23]. The study aimed to identify latent bisphosphonate-induced ONEAC cases that had been misdiagnosed as cholesteatomas. ONEAC and EAC cholesteatoma share similar clinical features, such as bone surface exposure and bone destruction [23, 32]. Clinical diagnoses of these two diseases can easily be confusing, and we speculate that the real-world occurrence of ONEAC in association with antiresorptive drugs is likely higher than what previous case reports have suggested.

While ONEAC commonly presents initially as otorrhea or chronic local inflammation of the EAC, several cases have an asymptomatic onset that is detected incidentally [10, 11, 20]. Medication-related ONEAC has not yet been precisely defined. ONEAC is diagnosed during its clinical course, often beginning with the diagnostic criteria of MRONJ. The criteria include 1) current or previous treatment with antiresorptive therapy alone or in combination with immune modulators or antiangiogenic medications, 2) exposed bone that has persisted for more than 8 weeks, and 3) no history of radiation therapy to or metastasis of the lesion [10–23, 33, 34]. In our case, the patient’s initial symptoms were pain in the jaw and trismus, which were later accompanied by ear symptoms. This pattern led to our initial suspicion of MRONJ, as it is the most common concern with long-term BMA use. Subsequently, we diagnosed our patient with ONEAC because the findings from the otoscopic examination and imaging studies and the clinical course of the ear condition were consistent with the diagnostic criteria for MRONJ.

Differential diagnoses are cholesteatoma, malignant otitis externa, osteoradionecrosis, and primary malignant or metastatic tumor of the EAC [11, 13–15]. The patient had no history of local irradiation to the temporal bone and had undergone radioactive iodine therapy. Radioactive iodine is radiopharmaceutical, and it utilizes the radioactivity of iodine-131 (I-131). After I-131 is absorbed selectively by thyroid cells or thyroid cancer cells, it emits therapeutic doses of radiation. Its selective absorption is a significant advantage of this therapy. In contrast, radium-223 is another radiopharmaceutical drug used to treat painful BM from prostate cancer, and is broadly absorbed by bone tissue during active turnover. While radium-223 is recognized as a risk factor for MRONJ, I-131 is not [34]. Therefore, the history of radioactive iodine therapy in this patient is not considered as related to the onset of ONEAC. The histological findings from the biopsy, which showed only granulation tissue formation, and negative bacterial cultures ruled out the possibility of cholesteatoma, malignant otitis and metastatic tumor. In the final histopathological analysis of the surgical specimen, sequestrum and granulation tissue were observed, aligning with the pathological features of ONEAC [35].

Imaging studies, such as CT, are useful for evaluating bony destruction. However, CT changes are not typically observed until ONEAC macroscopically progresses to involve the temporal bone, with other significant findings remaining absent. Direct examination using otoscopy offers greater advantages over imaging studies during ONEAC diagnosis. In our case, slight-to-moderate FDG uptake around the bilateral EACs was detected on FDG-PET, preceding the CT-detected destructive changes in the bony wall of the EACs. The use of FDG-PET for detecting and predicting MRONJ at an early stage has been documented and has already been integrated into clinical practice [36]. Since FDG-PET is widely employed to assess metastatic disease progression in various malignancies associated with BM, FDG uptake in the temporal bone may facilitate the early detection of ONEAC, even in asymptomatic patients. Further data on FDG-PET findings in medication-related ONEAC cases are needed for diagnostic accuracy.

ONEAC treatment, which can be either conservative or surgical, including effective sequestration of the necrotic bone and reconstruction, is similar to that of MRONJ [10–22, 34]. Of the 25 previously reported ONEAC cases (Table 1), 9 cases required surgical treatment, while 15 cases received only conservative treatment. The treatment course for the remaining case was not described. Of 24 patients, 58% did not respond to conservative management alone. In treating stage 2 or 3 MRONJ, characterized by exposed and necrotic bone with infection or inflammation, surgical intervention is recommended nowadays, which replaces conservative therapy. In the literature, high success rates for surgical management have been reported across all stages [34]. Complete removal of the necrotic lesion is essential for resolving MRONJ, as conservative management alone is insufficient to achieve adequate sequestration of the necrotic bone. ONEAC severity in our patient was comparable to stage 2 or 3 of the MRONJ classification criteria. Accordingly, we hypothesize that the persistent remnants of necrotic bone or sequestrum associated with chronic inflammation hindered improvement. Complete removal of necrotic bone tissue seemed critical for the healing process, considering the extent to which ONEAC had progressed. Detection of ONEAC at an early stage is also important to reduce the treatment burden on patients.

The cause of ONEAC is unclear. Nevertheless, the similarity between the environments in the oral cavity and ear canal with regard to high bacterial loads and frequent minor mechanical stresses [13] suggests that MRONJ and ONEAC have similar underlying mechanisms [33, 34]. Accumulated doses of bisphosphonates or BMA are speculated to increase the risk of ONEAC development, although little evidence exists to support this idea [23]. All patients with neoplastic disease who reportedly developed BMA-associated ONEAC were treated with BMA for more than 3 years (Table 1). Long-term BMA use is a risk factor for MRONJ in patients with cancer [25, 33, 34]. BM is managed using BMA to avoid or minimize SREs until the cancer progresses lethally. BMA use can be extended to several years, especially in some cancers with favorable survival prognoses, such as breast, prostate, and thyroid cancers, and can cause adverse events. BMA-associated ONEAC, despite its rarity, is one such adverse effect.

In addition to detecting adverse events at an early stage, measures to lessen the risk by de-escalating BMA use, for instance, are currently under discussion. The recent report from a randomized trial for breast and castration-resistant prostate cancers revealed that SREs prevention by 12-weekly BMA treatments showed non-inferiority to 4-weekly treatments, consistent with previous reports [9, 25, 37]. SREs-free survival rates (SSR) at 1-year in these groups were reportedly 72.4% (12-weekly) and 79.1% (4-weekly), which were not statistically different [37]. Incidence of MRONJ was not shown to be decreased by the longer treatment interval in this study. Furthermore, when focusing on breast cancer patients, 4-weekly BMA achieved greater SSR than 12-weekly administration [37]. Whether BMA de-escalation results in both satisfactory SREs prevention and a lower incidence of adverse events may depend on the cancer type. Optimal dose intensity should be proposed according to the risk–benefit balance for each patient, predicted efficacy of the systemic cancer treatment and estimated risk of SREs in metastatic bone lesions. Although BMA use is conceived as beneficial with the aim of preventing SREs in patients with BM, clinicians should always be aware of the risks associated with prolonged BMA use, as BMA-related adverse events can severely deteriorate a patient’s quality of life.

Author contributions

All authors contributed to the diagnosis and treatment of the patient. All authors contributed to drafting the article and approved the final version of the manuscript.

Funding

No funding was received for conducting this study.

Data availability

The data regarding this study are not openly available due to reasons of sensitivity and are available from the corresponding author upon reasonable request.

Declarations

Conflict of interest

The authors have no relevant financial or non-financial interests to disclose.

Ethical approval and consent to participate

This study was approved by the ethics committee of Keio University School of Medicine. The patient provided signed informed consent for an Institutional Review Board-approved protocol for research use of the medical records of pathological specimens obtained as part of routine clinical care and publication.

Consent for publication

Written informed consent to publish the clinical data was obtained from the patient.