41.1 Principles of Radiation Toxicity

Adjuvant radiotherapy is a mainstay in the management of oral cavity cancers as its use decreases the risk of cancer recurrence. ^{1 , 2} However, these benefits come at the cost of various side effects. Similar to surgical procedures, radiotherapy is a locally directed therapy and most (though not all) related sequelae are anatomically directed and dependent on the volumes a radiation oncologist decides to treat. Therefore, deciding what to treat and what not to treat is a crucial part of radiotherapy design. Advances in technology such as the development of intensity-modulated radiotherapy (IMRT), which is standard of care for most head and neck cancers, ³ allows for tighter treatment margins resulting in better sparing of normal tissues. ⁴ Despite this, toxicity still remains significant and impacts the quality of life. ⁵ Key research goals include anatomically guiding which normal tissues can be safely spared and identifying what are the meaningful radiotherapy dose volume constraints to the normal tissues.

Side effects of radiotherapy can be viewed as a continuum of acute (during radiotherapy and several weeks after recovery), intermediate (within several months post radiotherapy), and late term (1 year or later) effects in their manifestation. Acute-term effects start at a delayed rate (about 2 weeks into the course). In the oral cavity, these may include progressively worsening oral mucositis or ulceration, thickened saliva, radiation dermatitis, which may result in desquamation at its peak, dysgeusia or ageusia (dysfunction in taste), lymphedema, cosmetic alteration in skin and tissue integrity, and overgrowth of opportunistic pathogens, such as candida, or soft-tissue infections. ⁶ The underlying pathophysiology of these side effects are thought to reflect direct target cell injury ⁷ and loss of function followed by the expected inflammatory response for wound healing.

Late manifesting complications in the neck include skin and muscle and fibrovascular tissue fibrosis, fistula formation, chronic orofacial/neck pain, and decreased range of mobility in the neck. Endocrinopathies in the thyroid and/or pituitary can also occur. In the oral cavity and oropharynx, complications include xerostomia, osteoradionecrosis (ORN), scarring in the tongue/soft tissues worsening speech quality and swallow ability potentially resulting in risk of aspiration or feeding tube dependence, trismus which can make speech and chewing difficult, and risk of dental complications including dental caries. ^{8 , 9} The pathophysiology of late manifesting complications is less clear, but believed to reflect a progressive local fibroproliferative process and possibly the effectiveness of stem cells from the bone marrow or local microenvironment. Management is largely symptom directed at this time and requires significant hands-on counseling by the radiation oncologist in conjunction with a multidisciplinary team including surgeons, medical oncologists, nurses, speech language pathologists (SLPs), and physical therapists. This chapter will focus on identifying and effectively managing many of these radiation side effects in order to optimize efficacy of cancer-related treatments while minimizing their impact on a patient’s overall quality of life. The future directions of reducing and managing radiation toxicities will also be addressed.

41.2 Diagnosis and Evaluation

The first step in assessing radiation-associated side effects is proper evaluation of the patient, beginning with a history and physical examination. Evaluation of the oral cavity can reveal dry mucous membranes, loss of enamel, or tooth decay suggesting xerostomia. Ulceration of the mucosa represents mucositis and if necrotic bone is exposed it could be indicative of ORN. Thrush can be diagnosed clinically as a white covering on the tongue. Externally, lymphedema can be characterized by painless swelling of soft-tissue structures, while radiation fibrosis presents with reduced tissue elasticity and flexibility and reduced range of motion, usually of the neck.

Visualization of structures further along the aeordigestive tract, including those of the larynx and pharynx, can be done using a rhinolaryngoscope allowing for assessment of anatomic integrity. Complementary to an anatomic assessment is a functional assessment often done by an SLP through a fiberoptic endoscopic evaluation of swallowing (FEES). FEES allows the clinician to evaluate an individual’s management of secretions, spontaneous swallow, and physiologic swallow through presentation of food and liquid while the rhinolaryngoscope is still in place. ¹⁰ Physiologic evaluation of swallow function is also done with a modified barium swallow (MBS). During a MBS, the patient is given liquid with radio-opaque barium, which is swallowed while taking fluoroscopic images allowing for visualization of the liquid as it moves through the upper digestive tract. A particular strength of the MBS is the ability to identify penetration and potential aspiration of the swallow bolus. A penetration aspiration score can then be assigned to denote its severity. Emerging strategies to quantify swallow function include the use of high-resolution manometry and impedance that has been traditionally limited to the esophagus. Recent investigations have demonstrated that the pressure and flow data that is also measured in the pharyngeal portion of a manometry study is not only reproducible, but can yield swallow function variables that robustly describe swallow dysfunction and the risk of aspiration. ^{11 – 14}

Evaluation of the severity of toxicities can be done by either the clinician or patient. Clinician-driven evaluations are usually through qualitative and semi-quantitative grading systems such as the common terminology criteria for adverse events (CTCAE), which typically grades side effects on a scale of 1 to 5 using a combination of history, physical examination, and laboratory data. The second approach is the use of patient-reported outcomes (PROs) typically through the completion of paper-based validated instruments, which offer the benefit of understanding symptoms from a patient’s perspective. ¹⁵ It is important to recognize that PROs are not necessarily synonymous with quality of life assessments which further reflect additional dimensions such as a patient’s expectations and potentially their coping skills. For example, dysphagia is often assessed with the M. D. Anderson Dysphagia Inventory (MDADI) which was developed to assess patient perspective views of their swallowing ability and its effects on quality of life. ¹⁶ In contrast, the Sydney Swallow Questionnaire (SSQ) was developed to measure patient-reported oral-pharyngeal swallow function. ¹⁷ Coadministration of these two instruments has demonstrated that while correlated, they also provide complementary information with the SSQ scores separating patients by swallow function and the MDADI providing insight into the impact the dysphagia had on their quality of life. ^{18 , 19}

41.3 Mucositis

Mucositis describes an inflammatory state that reflects a complex interplay of repeated mucosal injury, subsequent inflammatory response, and the effects of a wound healing response that can be incomplete after a given radiation treatment. As such, progressive clinically evident mucosal injury can be observed as fractionated radiotherapy progresses. As with head and neck carcinomas ^{20 , 21} the proliferative kinetics of the normal mucosa can also increase with increasing radiation doses leading also to accelerated wound healing (▶ Fig. 40.1). ²²

Studies have demonstrated the importance of the dose delivered during each fraction of radiation (and likely to how large of a volume of mucosa) and the complex interplay with net healing occurring during weekends when treatment is not administered. If insufficient healing of mucosal injury during treatment should occur, persistent mucositis leading to chronic mucosal ulceration has been described as seen in randomized studies evaluating the benefit of continuous daily irradiation including weekend days. ^{23 , 24} Similarly, prolonged interruptions in the delivery of radiotherapy can impact tumor control rates due to accelerated tumor repopulation in response to the mucosal injury. For daily fractionation (i.e., conventional fractionation) the peak of the mucosal injury typically occurs during the latter half of the course of the radiotherapy. For more accelerated radiotherapy schedules involving multiple fractions per day, depending on the size of the radiotherapy fraction, the peak of the injury can occur after completion of the course of the radiotherapy. These considerations may be relevant in the scheduling of care for patients receiving radiotherapy and offer insights into the management of acute mucositis and reducing the risk of late consequential mucosal ulceration.

In the upper aerodigestive tract, mucositis typically is associated with functional effects on swallow, speech, and taste. The severity of mucositis is often qualitatively scored using the CTCAE, graded from 1 to 5 representing mild, moderate, severe, life threatening, and death due to toxicity and is based on pain, dietary alterations, extent/confluence, and depth of the mucositis.

The risk of developing oral mucositis increases as the radiation dose to the oral cavity increases. Narayan et al ²⁵ found that oral mucositis could be limited to mild (grade 1 or less) and short term (one week or less) when point doses to the oral cavity were kept less than 32 Gy. Rosenthal et al ²⁶ similarly found maximum point dose of 33.5 Gy or higher to the anterior mandible was associated with the development of grade = 1 oral mucositis. As dose to the oral cavity rises, so does the risk of higher-grade mucositis. Narayan et al found more moderate (grade 2) and longer lasting (3 weeks or longer) mucositis was associated with oral cavity point doses of 39.1 Gy or higher, and Bhide et al ²⁷ reported mean dose of = 51 Gy to the oral cavity was correlated with the development of grade 3 oral cavity mucositis. In addition to the impact of radiotherapy dose, the volume of mucosa irradiated is also an important factor in the development of mucositis. Higher ratios of involved to normal mucosa, indicating a large primary tumor, places an individual at higher risk for severe (grade 3) mucositis. ²⁸

Though multimodality therapy with concurrent chemotherapy (CRT) improves cancer survival outcomes, it also significantly increases toxicity. Individually, chemotherapy and RT are known independent risk factors for the development of mucositis. However, when combined they act synergistically to increase the risk of severe mucositis in a multiplicative rather than additive fashion. ²⁹ Rates of mucositis were significantly higher in the combination CRT of GORTEC 94-01 (71 vs 39%) as were rates of severe (grade = 3) mucositis in the combination CRT arm of the intergroup study (43 vs. 32%). ^{30 , 31} There was optimism that replacing traditional chemotherapeutics with targeted agents (e.g., cetuximab) might result in equally efficacious oncologic outcomes while reducing toxicity. ³² However, RTOG 1016, a trial of over 800 patients randomized to cetuximab versus cisplatin with concurrent accelerated fractionated radiotherapy, has demonstrated that oncologic outcomes for human papillomavirus-associated oropharyngeal carcinomas are inferior when patients are treated with concurrent cetuximab. ³³ It is also noteworthy that an incompletely accrued phase II study randomizing patients with head and neck squamous cell carcinoma including oral cavity carcinomas to cisplatin versus cetuximab with concurrent radiotherapy raised concerns that patients receiving cetuximab demonstrated a higher rate of acute toxicities necessitating enteral support, raising caution regarding its use as a potentially less toxic radiosensitizer. ³⁴

The first step in managing mucositis is minimizing the impact of any mucositis that would be expected based on the radiotherapy dosimetry. In the oral cavity, radiotherapy target volumes that approximate the lips present unique challenges to reduce the impact of mucositis. Where significant mucositis may be expected, the use of dental rolls or spacers placed at the time of radiotherapy simulation and immobilization allows for the displacement of the lips, offering an opportunity to reduce the severity of mucositis on the lips (▶ Fig. 41.1). Similarly, the introduction of a mouth piece with a spacer displacing the palate from the tongue can be beneficial to reduce the risk of mucositis on the palate. This is a particularly beneficial strategy for oral cavity/tongue malignancies where the tongue may be more immobilized due to the effects of surgery limiting its movement and ensuring its position is reproducible with each radiotherapy fraction. However, in patients with a mobile tongue and where it is part of the target volume, judgment is recommended as to whether the benefits with the introduction of a mouth piece displacing it from the palate may be offset by the risk of day-to-day variability in the tongue position due to the presence of the mouth piece.

During radiotherapy, ongoing close surveillance of the oral cavity mucosa to identify areas of mucositis is beneficial for several reasons. It not only ensures additional quality assurance of the delivery of the radiotherapy plan, but also allows for several interventions that can reduce its impact on oral cavity function. Mucosal injury arising from bite-line trauma offers opportunities for dental solutions to minimize the damage. Often, it is the secondary mucosal edema that may be exacerbated by the prior oncologic surgery that contributes to the risk of developing dental trauma and requires ongoing evaluation. In addition, dental amalgam can create scatter beta-radiation contributing to premature mucositis beyond what is expected based on the planned dosimetry. In the oral cavity, the buccal and oral tongue mucosa are particularly sensitive to this given the proximity to the teeth and their movement in the oral swallow phase preparing the food bolus. Identifying these areas offers opportunities to either replace the amalgam or to introduce a spacer with either a dental roll or wax, provided this does not alter the reproducibility of the target volumes.

Lastly, ongoing surveillance of areas of mucosal ulceration is important to determine if sufficient normal tissue repair response is occurring based on the severity of the acute mucositis. This is important as it has been demonstrated that severe acute mucosal ulceration in the setting of insufficient normal mucosal repair during the radiotherapy can lead to an increased risk of consequential late mucosal side effects ^{24 , 35} due to injury to the mucosal stem cell population. ³⁶ Reducing the severity of the acute injury has been proposed as a strategy to reduce this risk. ³⁷ Minimizing trauma, replanning of radiotherapy dosimetry, and judicious use of treatment interruptions are potential strategies to consider.

Maintaining oral hygiene is imperative to reduce microbial overgrowth that can lead to an aggravation of mucositis. ³⁸ Salt and baking soda rinses can be used several times a day and if microbial overgrowth is suspected, appropriate antimicrobial agents (Nystatin for yeast) should be prescribed. Pain control is essential for patients to maintain nutrition, engage the muscles to reduce the risk of late swallowing complications, ^{39 , 40} and avoid treatment breaks. Topical anesthetics, such as lidocaine, can be used alone or in combination with diphenhydramine solution and simethicone-based antacid products for initial pain management. High-dose gabapentin (900 mg TID) can be slowly titrated up prior to beginning treatment and maintained for the duration of radiotherapy. Its role in the management of mucositis pain remains an active area of investigation and one with significant promise as it has been associated with decreased narcotic use and treatment breaks. ⁴¹ As treatment progresses, additional pharmacologic interventions can be used as pain dictates typically with the introduction of short-acting opioids (i.e., oxycodone, hydrocodone) titrated to maintain pain control. Long-acting narcotics can be used as well, but with extreme caution if short-acting narcotics are still to be titrated up given the risk of respiratory depression and sedation these agents can cause. Several other drugs, including amifostine ^{42 , 43} and granulocyte macrophage stimulating factor ^{44 , 45} have demonstrated mixed results.

41.4 Xerostomia

Xerostomia is the subjective feeling of dry mouth as well as an objective decline in salivary flow, which when caused by radiotherapy is due to damage to the major and minor salivary glands. Beyond a perceived feeling of discomfort, xerostomia can also result in pain, dental caries, changes in voice quality, and swallowing dysfunction.

A mean dose of 10 to 15 Gy to the parotid typically results in minimal impairment of gland function. However, as mean RT dose increases to 20 to 40 Gy, gland function reduces and at doses of > 40 Gy severe (> 75%) reduction in gland function is observed. ^{46 , 47} IMRT is now frequently used in order to spare the contralateral parotid with the hopes of avoiding long-term xerostomia. Quantitative Analyses of Normal Tissue Effects in the Clinic (QUANTEC) guidelines recommend sparing at least one parotid gland to a mean dose of < 20 Gy or both parotids to a mean dose of < 25 Gy in order to avoid long-term severe (< 25% of baseline) salivary dysfunction ⁴⁸ and a study by Portaluri et al observed that a mean dose of < 30 Gy to the contralateral parotid was associated with only mild subjective xerostomia. ⁴⁹ These investigations typically summarized the parotid dosimetry as the average dose and assumes that the spatial function of the parotid gland is uniform. However, evidence in both animal models and more recently in humans supports the conclusion that parotid gland function is not uniform and offers unique opportunities to develop new strategies to further reduce this complication. Rat xerostomia models have demonstrated that irradiation of only the cranial half of the parotid glands results more permanent xerostomia compared to the caudal half where recovery was observed (▶ Fig. 41.2). ⁵⁰ Further preclinical studies have identified stem cells which are speculated to be responsible for the improvement in the xerostomia. ⁵¹ Investigations by our group using a growing clinical big data inventory of National Cancer Institute (NCI) CTCAE grading of xerostomia further supports not only improvement in the irradiated head and neck cancer patient with time, but has also demonstrated the importance of parotid sub-volumes that are more at risk of injury and are associated with the severity of xerostomia at 3 to 6 months after completion of radiotherapy (▶ Fig. 41.3). ^{52 , 53}

Applying a data science approach to understand which factors are associated and predictive of the risk of severe xerostomia (CTCAE grade = 2) at different time points (i.e., 3-6 months vs. 18 months or greater) following radiotherapy has further demonstrated that the risk of severe xerostomia is also dependent on the dosimetry to sub-volumes in the parotid glands. The importance of the dosimetry to these parotid sub-volumes offers the potential to develop future radiotherapy strategies to further reduce this risk. While it is clear that the parotid dosimetry to these sub-volumes (especially the cranial portions) dictates a great deal of the risk of severe xerostomia, patient factors along with the dosimetry to the oral cavity minor salivary glands ^{54 , 55} and submandibular glands ⁵⁶ appear to be secondarily modifying this risk. For example, the application of radiomic analysis of parotid and submandibular glands based on the baseline computed tomography (CT) images used for radiotherapy planning allows one to quantify the “health state” of the glands. In doing so, our group has demonstrated that these salivary gland radiomic features further add to the accuracy of xerostomia prediction models suggesting that they may reflect baseline salivary gland function or regions at greater risk for RT-injury. ⁵³

Symptomatic xerostomia is often managed with salivary substitutes, which can provide temporary hydration and lubrication of the oral cavity and be used several times per day. Lozenges with acidic, bitter, or sweet flavors can be used as gustatory salivary stimulants of saliva production. Pharmacologic salivary stimulants, such as pilocarpine or cevimeline, can also be used in symptom management. Pilocarpine has been evaluated in randomized studies and when compared to placebo in patients treated with historic non-parotid sparing radiotherapy techniques, 5 mg of pilocarpine three times per day improved symptomatic oral dryness (44 vs. 25%) with similar results seen with 10 mg three times per day. ⁵⁷ Side effects seen with pilocarpine are typically anticholinergic (sweating, flushing, nausea, headache, diarrhea) and were moderate enough to cause 5 and 20% of patients on the 5 mg and 10 mg TID arms respectively to stop treatment.

Acupuncture-like transcutaneous nerve stimulation in the treatment of xerostomia was recently compared to pilocarpine treatment in a randomized fashion. ⁵⁸ Both demonstrated subjective improvements in xerostomia-related quality of life, though no assessment could be made in terms of superiority. Acupuncture has also been studied in the prevention of radiation-induced xerostomia. When administered three times per week concurrently with radiotherapy, acupuncture improved both subjective xerostomia as well as objective salivary flow rates. ⁵⁹ Finally, the use of submandibular gland transfer to the submental space has been explored in nasopharyngeal carcinoma with success; however, the dosimetric benefit for oral cavity cancer is not as clear given the physical limitations to which the submandibular gland can be transferred relative to the radiotherapy doses that are typically prescribed for oral cavity malignancies. ⁶⁰

41.5 Osteoradionecrosis

ORN ⁶¹ is a serious complication of radiotherapy resulting in nonhealing and possibly devitalized bone. Affected tissues can become hypovascular, hypocellular, and hypoxic reducing the chances of future healing. ⁶² Rates of ORN have declined from 11.8% before 1968 to 5.4% from 1968 to 1992 to less than 3% today with improvements in radiation dose conformality as well as dental care. ^{63 – 65}

Radiation dose to the mandible is a strong risk factor for ORN and ideally the maximum dose is kept below 66 to 70 Gy (RTOG 0225, RTOG 1016) as doses greater than 70 Gy increase the risk of developing ORN. ⁶⁶ However, the volume of mandible receiving slightly lower doses of radiation (40-60 Gy) is also strongly correlated with risk of developing ORN. In a matched cohort of patients, each percentage increase in the volume of the mandible receiving 50 Gy resulted in a 9% increased risk of ORN (odds ratio: 1.09, p = 0.02). ⁶⁷ Underlying dental health is another probable risk factor, with individuals requiring dental extractions prior to radiotherapy and those already edentulous prior to treatment are at higher risk of developing ORN. ⁶⁷ The importance of dental health is highlighted by the higher than expected rates of ORN in clinical trials such as RTOG 0022 ⁶⁸ where there was not rigorous use of prophylactic dental care. This is in contrast to published series where stringent dental evaluation was required, including extractions of decaying or nonrestorable teeth prior to therapy and daily topical fluoride use where no cases of ORN were observed. ⁶⁴ Other noted risk factors include a history of smoking and alcohol use, which is often associated with poor baseline dental health, as well as larger primary tumors, likely a surrogate for larger volumes of the mandible receiving radiation ⁶⁷ and the addition of a brachytherapy boost. ⁶⁶ Whether it is necessary to remove healthy teeth that could require future extractions prior to radiotherapy is not clear at this time.

Management of ORN is often based on its severity. Conservative management includes local irrigation (with saline, bicarbonate, or chlorhexidine), avoidance of irritants such as smoking and alcohol, and continued maintenance of oral hygiene. ⁶⁹ Pharmacologic interventions include the antitumor necrosis factor agent pentoxifylline (400 mg BID) and the antioxidant tocopherol (1,000 IU daily) also known as vitamin E. ⁷⁰ The PENTACLO regimen, consisting of 800 mg pentoxifylline, 1,000 IU vitamin E, and 1,600 mg clodronate 5 days per week alternating with 20 mg prednisone and 1,000 mg ciprofloxacin 2 days per week has proven safe and effective in the treatment of ORN. ⁷¹ The use of hyperbaric oxygen in patients with ORN is controversial. The only randomized placebo controlled trial conducted on the subject did not show any benefit to its use, though the study of often criticized for its design and definition of ORN. Surgical management is the most invasive intervention, but has benefited from technological advancements. Myocutaneous flaps and the use of microvascular free bone replaces dead bone with a vascularized bone-containing flap, allowing for reconstruction of the mandible and transposition of nonirradiated soft tissue with intact blood supply to help the healing process. ⁶⁹ A general rule of thumb is the more extensive the ORN, the less likely conservative measures will be effective. For example, Notani et al reported that 70% of patients with ORN confined to the alveolar and/or mandible above the level of the alveolar canal healed with conservative therapy alone. However, only 10% of the patients with ORN that extended to the mandible below the level of the alveolar canal, ORN with a skin fistula, or pathologic fracture healed with conservative therapy, but 40% of those were salvaged by surgery. ⁷²

Poor dental health after radiotherapy requiring tooth extraction or major work is difficult situation in follow-up. Hyperbaric oxygen has been suggested as a prophylactic measure to decrease the risk of ORN in these situations. Marx et al performed a randomized trial comparing the use of prophylactic hyperbaric oxygen to penicillin and found the use of hyperbaric oxygen decreased the risk of ORN (5 vs. 30%). ⁷³ However, the generalizability of this study has been questioned given the very high rate of ORN in the penicillin arm. A pooled analysis of studies since 1986 by Wahl demonstrated a rate of ORN of only 3.5% without hyperbaric oxygen ⁶⁵ and therefore the utility of prophylactic hyperbaric oxygen is not clear. Whether there are specific doses to the mandible, above which would preclude tooth extraction due to untenably high ORN risk, is not well established either. Several authors posit that doses of > 50 to 60 Gy to the area undergoing extraction puts an individual at a > 10% risk for development of ORN; however, in one of the largest experiences of 528 in-field tooth extractions, only 4 developed ORN. ^{63 , 74 , 75}