1

Case reports

A 54-year-old Caucasian male initially presented in January of 2007 with a T1N2bM0 squamous cell carcinoma (SCC) of the left tongue base (not P16 tested). His medical history was significant for moderately controlled type 1 diabetes mellitus (DM). He denied a history of alcohol or tobacco use. A preference for CCRT was advocated by the head and neck oncological tumor board. He elected not to pursue conventional treatment in favor of complementary and alternative medicine remedies. He was lost to follow up until May of 2008 when he re-presented with disease progression. On clinical and radiological staging he was now T4aN2cM0 (P16 positive) with cancer invading into the deep muscles of the tongue. The bilateral cervical bulky adenopathy extended to levels IV and V in the supraclavicular fossae. His upper airway caliber was significantly diminished.

CCRT was again recommended as the preferred treatment which he elected. Prior to treatment he received a percutaneous gastrostomy (PEG) and a tracheotomy. The chemotherapy regimen consisted of two cycles of cisplatin and 5-fluorouracil. The RT plan was designed to expose all gross disease to 70.2 Gy and the radiographically negative (but clinically-at-risk) necks with 50.4 Gy. This necessitated inclusion of the clavicular heads and superior sternum in the treatment field. The radiation dose was delivered in 39 fractions. The high dose administered to the neck was considered necessary to achieve sterilization of all gross malignancy. Therefore (as is often the case) within the tumor volume there was a greater acceptance of ‘hot spots’ than ‘cold spots’. This fact together with the need to protect vital structures led to some dose inhomogeneity ( Fig. 1 ). He tolerated treatment well which was completed in November, 2008. Subsequent clinical and radiographic examination including positron emission tomography (PET) at 12 weeks demonstrated a complete response.

Coronal (A) and sagittal (B) views showing the gross tumor volume (shaded red) with the 70 Gy prescription isodose line (purple), as well as the 75 (blue line) and 79 Gy (green line) accepted dose inhomogeneity.

Nearly eight months following treatment the patient was decannulated. However, the tracheostoma was not healing as expected and within weeks granulation tissue was seen emanating from the stoma. This regressed into a persistent and painful tracheo-cutaneous fistula which emitted a fetid odor. Biopsies of this region were twice obtained during this process and revealed inflammatory tissue with no evidence of persistent neoplastic disease. Exposed bone and erosion of the anterior and lateral tracheal walls with purulent drainage were subsequently noted, prompting imaging studies.

A CT of the neck and chest revealed lytic lesions involving both clavicular heads while PET scan revealed significant hypermetabolism in the medial heads of the clavicles and sternum (manubrium) extending into adjacent soft tissues. With biopsies consistently negative for carcinoma this clinical picture was consistent with ORN and osteomyelitis. The patient had no known underlying condition that would have made him more susceptible to radiation toxicity.

Therefore, with the patient sedated, a more extensive debridement of the stoma and periostomal tissues was performed with conservative bony debridement. All specimens were negative for malignancy. The procedure was followed by 6 weeks of intravenous antibiotics and hyperbaric oxygen treatment (HBOT) consisting of 30 dives. This temporarily arrested the disease process. However, several weeks following treatment, malodorous discharge from the tracheostoma as well as sloughing of necrotic tissue resumed. Therefore, an aggressive debridement was performed with bilateral partial claviculectomies, partial sternectomy, tracheal debridement with reconstruction of the soft tissue defect using a pectoralis myocutaneous flap (tissue was again negative for malignancy). A permanent tracheostomy was constructed ( Fig. 2 ) for reasons related to airway safety. He responded well to treatment with excellent tissue healing and elimination of the ORN ( Fig. 3 ). He is over 3 years post treatment for ORN and 50 months post-CCRT and is disease free though he has persistent and diffuse CCRT-induced neck soft tissue induration. Speech is fully intact but he has frank aspiration with thin liquids and continues to require a PEG.

(A) The surgical field is shown after much of the devitalized bone and soft tissue has been removed. The inferior skin flap has been retracted for better exposure. The arrowheads indicate the limits of the superior flap skin resected. The endotracheal tube can be seen passing through the tracheal lumen (below the middle arrowhead). Much of the anterior and lateral walls of the trachea in the vicinity of the prior tracheotomy site were not viable and have been removed. The arrows point to the clavicular stumps. The edges were subsequently smoothed out with a bur prior to reconstruction. (B) A pedicled pectoralis myocutaneous flap was used to reconstruct the defect. A permanent tracheostoma was constructed.

The reconstructed surgical site over 3 years following the procedure described in Fig. 2 .

2

Comment

The average time to diagnosis of ORN following radiation therapy is generally 1–2 years, however, individuals can range from 4 to 228 months . The most common initial presentation of ORN is foul odor, exposed bone, pain, and a discharging fistula or sinus tract, all of which can also be seen in patients with cancer. The patient described here presented with all four of these symptoms. The diagnosis of ORN, the symptomatology of which often resembles malignancy, is made with a combination of clinical presentation, radiographic evidence of bony necrosis, and the absence of malignancy after tissue biopsies. PET or bone scanning, generally cannot be relied upon for differentiating ORN from malignancy. Hao reported in his series that 21% of ORN patients had recurrent cancer noted after wide debridement of devitalized tissue.

Though ORN of the sternoclavicular joint is a known complication of radiotherapy for breast or pulmonary malignancy , it is rare in HNC patients. For a variety of reasons ORN is much more common in the mandible with an incidence as high as 18.4% . There are at least three case reports of ORN of the clavicles after external beam radiation for head and neck cancer . The primary lesion in these cases was either the supraglottis or hypopharynx. The patient in this case report had oropharyngeal cancer which is significant given the rise of documented HPV related cancer in this site. HPV related oropharyngeal SCC is pathologically distinct and more responsive to treatment than non-HPV HNC raising the possibility of ‘treatment deintensification’ which would lower the risk of ORN.

With recent developments in radiation therapy including intensity modulated radiation therapy (IMRT) the incidence of ORN has dropped and in one study was reportedly zero (median follow-up: 34 months) . However, most would agree that even with advances in technique, ORN continues to be a potential complication of treatment. Given that the complications of using sensitizing chemotherapeutic regimens together with RT appear to increase the incidence of late complications such as dysphagia and pneumonia , it is reasonable to question whether CCRT also increases the risk ORN relative to RT either alone or following surgery. In the oral cavity Stenson reports an 18.4% incidence of mandibular ORN in patients treated with CCRT (median follow-up 3.25 years). There are no dedicated studies specifically comparing CCRT vs. RT with respect to ORN in HNC sites.

Cooper et al. and Bernier et al. reported on landmark studies that showed the superiority of CCRT over RT alone for disease control in patients having previously been treated with surgery who were at high risk for persistent or recurrent cancer . Of these two studies only Cooper et al. showed a potentially elevated risk of late skeletal complications secondary to CCRT relative to RT. However, given the length of follow up no definitive conclusions can be drawn from this. Nguyen et al. compared patients with locally advanced HNC receiving CCRT vs. surgery followed by RT and found no difference in the incidence of ORN . However, the cohort size was small, the study design retrospective, and the follow-up insufficient. Schratter-Sehn also reported no difference but with a mean follow-up of 26 months and using earlier treatment protocols . Focused studies investigating the incidence of ORN as a function of CCRT versus RT alone (or following surgery) would need to follow post RTOG treatment protocols and have sufficient follow-up.

Treatment of patients with ORN of the sternoclavicular joints is approached using a similar algorithm applied to patients with ORN of other sites in the head and neck. Initial treatment includes a local debridement of necrotic tissue and appropriate use of antibiotic therapy. If the disease is persistent, then local debridement can be accompanied by HBOT and, if osteomyelitis is present, intravenous antibiotics for up to 6 weeks. If these measures fail then a comprehensive resection of the diseased tissue is necessary in the operating room often with revascularization using a locoregional or microvascular free tissue transfer.

In conclusion it is vital for all practitioners who care for the HNC patient to be aware of the potential presentation of ORN in the sterno-clavicular complex particularly given the similarity to stomal recurrence—a difficult-to-treat and often lethal disease. All non-radiation oncologists who care for the head and neck cancer patient must understand and be fully aware of the radiation fields used to treat each patient. The strategy for preventing this outcome in at-risk patients focuses on the use of IMRT, good tracheotomy hygiene, tight control of medical comorbidities, and possibly, treatment deintensification for HPV related oropharyngeal disease. However, if ORN does develop, it is treatable using the appropriate treatment paradigm. Further studies are necessary to determine if CCRT increases the incidence of ORN in HNC patients as a late outcome relative to RT alone.