Efficacy and toxicity of reirradiation using intensity-modulated radiotherapy for recurrent or second primary head and neck cancer

Authors


Abstract

BACKGROUND:

Patients with locally recurrent squamous cell cancer of the head and neck (SCCHN) are reported to have a poor prognosis and limited therapeutic options. Optimal management is selectively applied and morbid. Both surgical resection and chemoradiotherapy are reported to result in median survivals of approximately 12 months. Intensity-modulated radiotherapy (IMRT) is a highly conformal approach for delivering RT. This study reported the experience of the Dana-Farber Cancer Institute (DFCI) with IMRT-based chemoradiotherapy with or without surgery for locally recurrent SCCHN.

METHODS:

The current study was a retrospective study of all patients treated at DFCI who were diagnosed with nonmetastatic second primary or recurrent SCCHN and who received reirradiation based on IMRT. The primary endpoint was overall survival (OS), and secondary endpoints were locoregional (LRC) and distant control and acute and chronic toxicity.

RESULTS:

Thirty-five patients were treated from August 2004 until December 2008. Recurrent disease was treated in the oral cavity (4 patients), larynx/hypopharynx (13 patients), oropharynx (7 patients), nasopharynx (2 patients), and neck (9 patients). The median radiation dose was 60 Gray (Gy), and all patients received concurrent chemotherapy. The median follow-up was 2.3 years. The 2-year actuarial OS and LRC rates were 48% and 67%, respectively. Approximately 91% and 46%, respectively, of all patients developed at least 1 acute and late grade 3 toxicity. Four (11%) late deaths occurred in patients with no evidence of disease (2 aspiration events, 1 oropharyngeal hemorrhage, and 1 infectious death).

CONCLUSIONS:

Aggressive chemoradiotherapy with IMRT was found to be feasible and resulted in favorable survival outcomes in comparison with published reports. Acute and late toxicities were substantial. The apparently improved LRC appears to carry a significant risk of developing late complications. Cancer 2010. © 2010 American Cancer Society.

Patients who develop a recurrent or second primary head and neck cancer have a poor prognosis, to a large extent because the initial course of treatment substantially reduces the flexibility and intensity of retreatment. Furthermore, recurrent malignancies often represent tumors that have reduced sensitivity to chemoradiotherapy (CRT). The available palliative treatment options include chemotherapy and surgical debulking. Potentially curative approaches include definitive surgery with or without adjuvant radiotherapy (RT) or CRT, or definitive therapy with radiation-based therapy. Chemotherapy alone is more tolerable than combined modality therapy but offers virtually no chance at long-term tumor control.1 Historically, complete surgical resection when feasible was the only curative option. However, despite aggressive treatment, the vast majority of these individuals die of their disease.2

As nonsurgical primary treatment gains wide acceptance for head and neck cancers, most patients who present with recurrent malignancy have already received a full course of RT. Recurrent disease is often not resectable, and even in resectable cases, some patients decline a surgical approach due to quality-of-life considerations. The radiation tolerance of normal tissues makes reirradiation technically challenging and frequently more toxic than the initial course. The therapeutic ratio in head and neck cancer is narrow, and achieving the proper balance between tumor control and severe toxicity is even more challenging if patients have been previously treated. However, the small body of literature on the efficacy and tolerability of this treatment indicates that, in return for the risks taken, a modest number of long-term survivors will be seen.2

A newer RT technique that facilitates precise dose delivery is intensity-modulated RT (IMRT). This technique allows for dose-escalation while minimizing normal tissue toxicity. Early reports suggest that IMRT can be used in the reirradiation setting.3, 4 We have been treating patients with recurrent head and neck cancer at the Dana-Farber Cancer Institute (DFCI) with IMRT and aggressive concurrent chemotherapy regimens since 2004, often in combination with surgery. The current study presents our initial results.

MATERIALS AND METHODS

Patients

Between August 2004 and December 2008, 35 patients with a history of prior head and neck RT for squamous cell carcinoma were treated with reirradiation using concurrent chemotherapy and IMRT for second primary or recurrent cancer of the head and neck. We restricted this analysis to patients who had nonmetastatic squamous cell carcinoma at both the initial and second cancer presentation, because survival outcomes are generally significantly worse than for other histologies. Data were reviewed under an Institutional Review Board-approved retrospective protocol.

Evaluation

All patients were initially evaluated by a multimodality treatment team, comprised of an otolaryngologist, medical oncologist, and radiation oncologist. A detailed physical examination, including flexible nasopharyngolaryngoscopy, was performed in all patients. All patients underwent neck computed tomography (CT), magnetic resonance imaging, or both. Positron emission tomography (PET) or PET-CT was performed in 26 patients. Histologic confirmation of malignancy was required before initiating treatment.

Treatment

All patients were first considered for primary surgical management, if resectable. The majority of patients were not candidates for resection due to extent of disease, whereas others chose not to undergo surgery. The use of induction chemotherapy was at the discretion of the treating physicians. In general, induction therapy was considered for patients with T4 tumors or N3 lymph node disease. Concurrent CRT was given in all cases, typically with a platinum-based regimen. Patients were strongly urged to undergo prophylactic percutaneous endoscopic gastrostomy (PEG) placement before starting treatment.

Radiotherapy was delivered using IMRT in all cases. One patient also received part of his treatment using 3-dimensional (3D) conformal RT. Patients were immobilized using an S-frame, which secures the shoulders in addition to the head and neck. Patients were first imaged under fluoroscopy to ensure accurate isocenter placement and mask construction and then simulated in the mask using high-resolution (2.5-mm slices) CT. The Eclipse IMRT software (Varian Medical Systems, Palo Alto, Calif) was used for all patients. The clinical target volume (CTV) included areas of macroscopic disease plus microscopic disease margin in all patients. In general, the CTV margin was 1 to 1.5 cm, and the planning target volume (PTV) margin was 0.5 cm. One patient with a history of irradiation for early stage laryngeal cancer received RT to his primary lesion, macroscopic N2 lymph node disease, and bilateral necks. No other patient received elective lymph node irradiation, although 9 patients were treated for neck-only recurrences.

The primary avoidance structure was the spinal cord. The primary goals of inverse planning were to ensure homogenous PTV coverage and limit the additional spinal cord dose to 10 to 12 grays (Gy).

Toxicity

Acute and late toxicities were defined according to the National Cancer Institute Common Terminology Criteria for Adverse Events (version 4.0) and determined through retrospective chart review.5 Acute toxicity was defined as that occurring within 90 days of the completion of treatment. A complication that occurred during treatment (eg, fistula) and that persisted after 90 days was also considered a late toxicity. Local late toxicities in the context of locally recurrent disease were not considered to be related to radiation because the etiology could not be distinguished from tumor progression.

Follow-Up

Patients were typically followed every 4 to 6 weeks in the first year and every 2 to 3 months in the second year by the multimodality treatment team at our institution. Follow-up examinations included flexible nasopharyngolaryngoscopy and/or indirect laryngoscopy. Occasionally, some follow-up examinations were performed locally near the patient's home. Lesions suspicious for disease recurrence were biopsied.

Patterns of Failure

The locations of both first and any failure (locoregional vs distant) were recorded. Locoregional failures were categorized as in-field, marginal, or out-of-field failures using the method of Popovtzer et al.6 In-field failures were defined as >50% of recurrent disease present within the 95% isodose line; marginal failures were defined as <50% of recurrent disease present within the 95% isodose line. The CT scan that demonstrated recurrent disease was not fused to the planning CT, but the designation of in-field failure was so clear that such a process was deemed unnecessary.

Statistical Analysis

Overall survival (OS), progression-free survival (PFS), locoregional failure-free survival (LRFS), and distant metastasis-free survival (DMFS) were calculated according to the method of Kaplan and Meier. PFS was defined as disease progression or death from any cause. Cumulative incidence plots were derived for LRFS and DMFS, censoring patients only for death from any cause. Univariable analyses were performed to determine predictors of OS, LRFS, and DMFS using the log-rank test. The sample size was too small to perform multivariable regression.

RESULTS

Patient Characteristics

The initial diagnosis and treatment regimens for this cohort are listed in Table 1. The median age at first diagnosis was 51 years (range, 46-57 years), and the median age at second treatment was 57 years (range, 51-63 years) (Table 2). Twenty-five patients were male. The median initial radiation dose was 67.5 Gy (interquartile range [IQR], 63-70 Gy). The Eastern Cooperative Oncology Group performance status at retreatment was 0 in 18 patients, 1 in 16 patients, and 2 in 1 patient.

Table 1. Initial Course of Radiotherapy: Patient Characteristics and Treatment Details
Characteristic/DetailNo. (%)
  1. IQR indicates interquartile range; Gy, grays.

Median age at first diagnosis (IQR), y51 (46-57)
Male:female ratio25:10
Diagnosis
 Nasopharynx4 (11%)
 Oropharynx6 (17%)
 Oral cavity9 (26%)
 Larynx/hypopharynx15 (43%)
 Unknown primary1 (3%)
Stage
 I8 (23%)
 II5 (14%)
 III10 (28%)
 IV14 (40%)
 Unknown1 (3%)
Treatment regimen
 Surgery plus radiotherapy9 (25%)
 Surgery plus chemoradiotherapy7 (20%)
 Definitive radiotherapy7 (20%)
 Definitive chemoradiotherapy12 (35%)
Median radiation dose (IQR), Gy67.5 (63-70)
Table 2. Second Course of Radiotherapy: Patient Characteristics and Treatment Details
Characteristic/DetailNo. (%)
  1. IQR indicates interquartile range; ECOG, Eastern Cooperative Oncology Group; PTV, planning target volume; Gy, grays.

Median age at recurrent diagnosis (IQR), y57 (51-63)
ECOG performance status
 018 (51%)
 116 (46%)
 21 (3%)
Median prior disease-free interval (IQR), y2.5 (1.2-9.7)
Diagnosis
 Nasopharynx2 (6%)
 Oropharynx7 (20%)
 Oral cavity4 (11%)
 Larynx/hypopharynx13 (37%)
 Neck failure only9 (26%)
Stage
 I 0
 II6 (17%)
 III7 (20%)
 IV22 (63%)
Treatment regimen
 Surgery plus chemoradiotherapy17 (49%)
 Definitive chemoradiotherapy8 (23%)
 Induction chemotherapy plus chemoradiotherapy10 (28%)
Concurrent chemotherapy
 Platinum alone10 (29%)
 Platinum/taxane10 (29%)
 Platinum/cetuximab6 (17%)
 Platinum/taxane/cetuximab6 (17%)
 Cetuximab alone3 (8%)
Median radiation PTV dose (IQR), Gy60 (60-64)

Disease Presentation at Reirradiation

Table 2 describes the patient characteristics and treatment strategy at reirradiation. Of the 15 patients who initially presented with laryngeal cancer, 12 developed a laryngeal or hypopharyngeal recurrence, 2 developed an oropharyngeal cancer, and 1 failed in the neck. The single patient who presented with an unknown primary tumor failed in the neck. Of the 9 patients who presented with an oral cavity cancer, 5 developed disease recurrence in the neck, 3 failed in the oral cavity, and 1 presented with an oropharyngeal cancer. Six patients initially presented with oropharyngeal carcinoma; 3 patients failed in the oropharynx, 2 developed neck disease, and 1 developed a laryngeal cancer. Of the 4 patients who initially presented with nasopharyngeal carcinoma, 2 developed disease recurrence in the nasopharynx, 1 developed an oral cavity cancer, and 1 developed an oropharyngeal carcinoma.

Radiotherapy

The median interval between the end of the initial course of RT and the start of the second course of RT was 2.5 years (IQR, 1.2-9.7 years). At retreatment, 17 patients (49%) underwent surgery plus adjuvant CRT, 8 patients were treated with definitive CRT, and 10 patients were treated with induction chemotherapy followed by definitive CRT. All patients received concurrent chemotherapy, 25 (71%) of whom were treated with at least 2 drugs. Approximately 92% of patients were treated with at least a concurrent platinum agent, as shown in Table 2.

All patients received IMRT for their reirradiation course. One patient was first treated with 18 Gy using 3D conformal RT, and the remaining 42 Gy by means of IMRT. The median PTV prescription dose was 60 Gy (IQR, 60-64 Gy). All patients except 1 were treated once daily. Twenty-six patients were treated with a fractional dose of 200 centigrays (cGy), 8 patients received a fractional dose of 180 cGy, and 1 patient received a fractional dose of 120 cGy twice daily. The median treatment time was 47 days (IQR, 44-51 days). Only 11 patients had treatment times of >7 weeks, and no patient was treated for longer than 8 weeks.

Disease Recurrence and Survival

The median follow-up period for surviving patients from the time of the initiation of the second course of radiotherapy was 2.3 years (IQR, 1.6-3.6 years). The actuarial 1-year and 2-year OS rates were 59% and 48%, respectively, with a median survival of 1.9 years. The actuarial 1- and 2-year PFS rates were 53% and 45%, respectively, and the median PFS was 1.7 years. Figure 1 displays the Kaplan-Meier OS and PFS curves.

Figure 1.

Kaplan-Meier curves displaying overall survival and progression-free survival are shown.

A total of 14 patients (40%) developed disease recurrence after their second course of RT. Seven (50%) patients first failed locoregionally (defined as above the clavicles), 4 (29%) first recurred distantly, and 3 (21%) progressed simultaneously above and below the clavicles.

Eleven patients eventually developed a locoregional disease recurrence, resulting in 1-year and 2-year actuarial locoregional PFS rates of 67% and 67%, respectively. Of 11 patients, 8 patients failed in field (73%), 2 patients failed out of field (18%), and 1 patient developed both in-field and out-of-field recurrences (9%). In terms of the overall patient population, 23% of patients failed in field, 6% developed an out-of-field disease recurrence, and 3% failed both in and out of field. The 3 out-of-field failures all occurred in unirradiated lymph node basins (Delphian lymph node, parotid lymph node, and supraclavicular fossa). There were no marginal failures reported.

Seven patients eventually developed metastatic disease, resulting in 1-year and 2-year actuarial DMFS rates of 85% and 73%, respectively. Figure 2 shows the cumulative incidence of locoregional and distant recurrences.

Figure 2.

Cumulative incidence plots describing the development of any locoregional failure and distant metastasis are shown. Observations were censored only for death (from any cause).

Acute and Late Toxicity

The acute locoregional toxicities from treatment are listed in Table 3. All but 3 patients, or 91% of the total, developed a grade 3 or higher acute toxicity. There were 5 patients (14%) who developed an acute grade 4 toxicity, including an emergent tracheostomy, and 14 patients (40%) experienced >1 grade 3 or 4 toxicity. No patient died of an acute toxicity.

Table 3. Acute Toxicities During Treatmenta
ToxicityNo. (%)
  • a

    Toxicity was graded according to the National Cancer Institute Common Terminology Criteria for Adverse Events (version 4.0).

Mucositis
 Grade 29 (26%)
 Grade 315 (43%)
 Grade 40
Dermatitis
 Grade 216 (46%)
 Grade 38 (23%)
 Grade 43 (9%)
Respiratory
 Grade 20
 Grade 30
 Grade 41
Esophagitis
 Grade 27 (20%)
 Grade 327 (77%)
 Grade 40
Other
 Grade 20
 Grade 30
 Grade 41 (sepsis)

The late toxicities are listed in Table 4. Four patients never required a PEG, but of the 31 patients who did undergo gastrostomy tube placement, 17 (55%) had it removed at a median time of 3 months (IQR, 1.7-12 months) after treatment completion. Of all patients surviving at least 1 year after treatment, 89% (16 of 18) were free of PEG dependence. Of the 9 patients who survived at least 2 years, 8 (89%) were free of PEG dependence. Sixteen patients (46%) developed a grade 3 or higher late toxicity, and 9 patients (26%) experienced >1 grade 3 or 4 toxicity. All of the grade 4 toxicities were either related to skin or respiratory.

Table 4. Late Toxicity After Treatment (Defined as Occurring >90 Days After the Completion of Radiotherapy)a
ToxicityNo.
  • a

    Toxicity was graded according to the National Cancer Institute Common Terminology Criteria for Adverse Events (version 4.0).

Soft tissue necrosis
 Grade 31 (3%)
 Grade 40
 Grade 51 (3%)
Osteonecrosis
 Grade 31 (3%)
 Grade 41 (3%)
 Grade 50
Fibrosis
 Grade 32 (6%)
 Grade 40
 Grade 50
Trismus
 Grade 34 (11%)
 Grade 40
 Grade 50
Respiratory
 Grade 30
 Grade 43 (9%)
 Grade 52 (6%)
Esophageal
 Grade 317 (49%)
 Grade 40
 Grade 50
Dermatitis
 Grade 31 (3%)
 Grade 43 (9%)
 Grade 50
Other
 Grade 36 (17%)
 Grade 40
 Grade 51 (3%)

Four patients died with late treatment-related events, without evidence of disease. These 4 grade 5 toxicities were 2 aspiration events (at 8 months and 2.6 years, respectively, after treatment), 1 fatal oropharyngeal hemorrhage (10 months after treatment), and 1 persistent infection leading to debilitation and hospice care (6 months after treatment). One of the patients who aspirated had pre-existing dysphagia requiring a gastrostomy tube before retreatment was initiated, and therefore it is difficult to know the extent to which the second treatment contributed to the fatal event. The patient with oropharyngeal hemorrhage was thoroughly evaluated for disease recurrence, but on direct examination, the apparent source of bleeding was soft tissue necrosis.

Predictors of Survival and Toxicity

The following predictors of OS, PFS, locoregional disease recurrence, and grade 5 toxicity were assessed, and none were found to be statistically significant: prior disease-free interval, salvage surgical resection, disease site, treatment time, chemotherapy regimen, and use of induction chemotherapy. There appeared to be a trend toward improved survival with the addition of surgery (median survival, 3.7 years vs 1.9 years; P = .9), but this result may reflect inherently smaller volume and more localized disease. This study was retrospective and did not have adequate power to investigate these prognostic factors.

DISCUSSION

The treatment of patients with locally recurrent head and neck cancer represents a daunting challenge. The intimate anatomic relation between disease and critical structures often makes surgical resection impossible and/or unacceptable to the patient. Reirradiation is technically difficult due to the anatomic relations and normal tissue tolerances. We have been treating patients with recurrent disease with IMRT combined with chemotherapy or with surgery and chemotherapy.To the best of our knowledge, the current study represents 1 of the first publications to date describing outcome and toxicity using this technique. Our survival outcomes compare favorably with previously published results. At a median follow-up of >2 years, this cohort had a 2-year OS rate of 48%, with a median survival of 1.9 years. Our locoregional control rate at 2 years was 68%.

One explanation for this relative success is the routine use of concurrent chemotherapy, which has repeatedly been shown to improve survival in the definitive and adjuvant settings.7-12 In addition, the relatively long disease-free interval before disease recurrence (median, 2.5 years [IQR, 1.2-9.7 years]) suggests that many of these patients may have had second primary tumors rather than recurrent cancers, which portends a more favorable prognosis.13 Furthermore, we provide aggressive nutritional support and supportive care, which helps limit treatment breaks. Our treatment team also has a dedicated social worker, whose interventions may theoretically reduce patient depression and thus improve outcomes.14 Finally, these favorable results may also reflect effective patient selection because we only offer repeat CRT to patients with an excellent performance status who can tolerate treatment with such a high risk of toxicity.

An interesting finding of this analysis is that the pattern of failure after reirradiation is predominantly in-field, suggesting that radioresistant clonogens are likely responsible for the second recurrence as well as the first. Furthermore, the 3 out-of-field failures all occurred in surrounding lymph nodes, which we did not target. It therefore may be warranted to irradiate a high-risk lymph node basin if the additional toxicity would be tolerable.

As expected, these combined-modality regimens did lead to substantial acute and late morbidity, with all patients but 3 developing at least 1 grade 3 acute toxicity, and the vast majority of patients were dependent on PEG by the time of completion of treatment. However, the majority of patients were able to have their PEG removed relatively soon after the completion of RT, and therefore in general the gastrostomy tube was a temporary measure to ensure adequate alimentation during treatment.

The late toxicity and, in particular, the fatal events are concerning and highlight the need for appropriate patient selection and education regarding risk before embarking on such aggressive treatment. We chose a conservative approach in attributing late mortality to retreatment to avoid underestimating the toxicity of reirradiation. However, it is impossible to know whether the 2 aspiration events were truly a function of the second course of treatment or the first, and whether they would have occurred in the absence of any additional treatment. One of the patients was dependent on PEG at the initiation of reirradiation. It is also worth noting that these patients would have almost certainly died significantly sooner without CRT, but it is impossible to know whether the treatment-related side effects outweighed the presumptive increase in survival gained.

Indeed, the burden of both acute and late toxicities of reirradiation must be considered in light of the substantial expected acute and late morbidity and the uniform mortality expected to occur in the setting of disease progression without local treatment. In this study, analyses of predictors of toxicities did not yield further information to refine toxicity risk assessment before embarking on a course of reirradiation, although these analyses were limited by the study's sample size.

A review of the literature revealed a range of results for reirradiation, both with respect to survival and toxicity. In what to our knowledge was 1 of the largest published experiences, the Institut Gustav Roussy reported on 169 patients reirradiated with conventional RT with or without chemotherapy.15 The median survival was only 10 to 11 months, depending on the treatment regimen, and the 2-year DFS rate ranged from 3% to 14%. Late toxicity was substantial, with 41% of patients experiencing grade 2 to 3 cervical fibrosis, 21% of patients developing mucosal necrosis, and 30% of patients developing mild to severe trismus. Although this complication rate appears higher than more modern series, it is worth noting that the RT was planned and delivered in an earlier era, in many cases without CT-based planning.

Salama et al. reviewed the University of Chicago experience with reirradiation comprised of 115 patients from several CRT trials.16 Patients were treated with multiple 2-week cycles of 5 days of CRT, followed by a 9-day break. RT was administered either daily or twice daily, to a mean total dose of 64.8 Gy. The majority of patients were treated with CT-based conformal RT, but no patients received IMRT. The median OS and PFS were 11 months and 7 months, respectively. However, the total 3-year OS rate was 22%, and the 3-year PFS rate was 33%, suggesting there is a cohort of salvageable long-term survivors. Approximately 41% of patients developed a locoregional disease recurrence, which is favorable in comparison with other reports in the literature. Increasing reirradiation dose; surgery before CRT; and the use of cisplatin, paclitaxel, or gemcitabine were found to be significant predictors of improved survival. Toxicity was significant, with 19 patients (17%) dying of treatment-related toxicity. Aspiration rates and severity were not discussed, but 57% of patients had required a gastrostomy at the time of last follow-up. To the best of our knowledge, this study is 1 of the largest series of combined modality therapy, and their results highlight the important components of a successful multimodality therapy (ie, surgical debulking, higher radiation dose, and aggressive chemotherapy), as well as the attendant risks.

Researchers at the University of Michigan have reported their experience using 3D conformal RT and IMRT in an article that to our knowledge is 1 of the largest contour-based RT series published to date.6, 17 Sixty-six patients were retreated to a median dose of 68 Gy, and 71% of patients received concurrent chemotherapy. The actuarial 2-year OS rate was 40%, and 71% of all patients developed a locoregional failure; similar to our study, 96% of these failures occurred within the 95% isodose lines. Mild to moderate late complications were common; 29% of patients experienced at least 1 grade 3 or greater late morbidity. Two patients died of therapy-related complications (1 of cisplatin-induced renal failure, and the other from aspiration). Two patients developed carotid artery blowouts, which were successfully salvaged.

There are only limited data regarding the use of IMRT in the recurrent setting, although physicians from The University of Texas M. D. Anderson Cancer Center have recently described their results with IMRT-based reirradiation for 74 patients.3 Sixty-seven patients were treated with curative intent to a median dose of 63 Gy, and 57 patients were treated for squamous cell carcinoma. Eleven patients received <45 Gy in their first course of RT. In results similar to those presented in the current series, the 2-year OS and locoregional control rates were 61% and 64%, respectively. Fifteen patients, or 20% of the total, developed a “severe late toxicity,” with a 5% risk of treatment-related mortality.

The use of IMRT for reirradiation has been the subject of debate in the literature. Proponents of IMRT point to the ability to optimize the treatment plan and more easily spare critical structures.2 In contrast, some investigators argue that the dose inhomogeneity often noted in inverse planning can lead to inadequate target coverage.18 Nevertheless, because there is gathering evidence that dose escalation is beneficial in these patients, we anticipate that the ability of IMRT to carefully sculpt dose around critical structures and thus increase the total dose will outweigh theoretical concerns of an increased volume of tissue receiving low-dose radiation. Furthermore, the PTV inhomogeneity that is observed more often with IMRT is a parameter that can be addressed and optimized during the planning process. Thus, in our view, the benefits of IMRT and its workable limitations point toward it being indicated in the recurrent setting.

This cohort of patients was treated relatively early in our institutional experience with head and neck IMRT, and our treatment techniques have been refined over time. Treatment is clearly associated with significant acute and late morbidities, but our approach has led to gains in local control and survival in comparison with historic data. Similarly, a review of the literature suggests that the use of more conformal radiation treatment with sensitizing chemotherapy has improved survival outcomes, but life-altering and life-threatening, treatment-related complications are an unfortunate consequence of this paradigm. Although acute and late toxicities are expected with this aggressive treatment, the devastating consequences of local failure mandate rigorous local therapy and often justify its risks in the appropriate patient population.

CONFLICT OF INTEREST DISCLOSURES

The authors made no disclosures.

Ancillary