Definitive External Beam Radiotherapy

Two major series examine the use of EBRT alone or in conjunction with planned neck dissection. Erkal et al.²⁹ analyzed 107 patients with SCC of the soft palate treated at the University of Florida between 1965 and 1996. Ninety-nine patients received external beam irradiation alone, while 8 patients (7%) received external beam RT followed by an interstitial brachytherapy boost. Patients received a median dose of 62.4 Gy via once-daily fractionation or 76.8 Gy via hyperfractionated techniques. The 5-year local control rates for T1 through T4 tumors were 86%, 91%, 67% and 36%, respectively. Ultimate 5-year local-regional control (LRC) after surgical salvage for Stages I through IV was 90%, 92%, 84%, and 60%, respectively. On multivariate analysis, overall treatment time significantly affected local control (P = .0002), regional control (P = .01) and overall survival (P = .04). Overall, 3% of patients treated with RT and 11% (3/27) patients treated with RT followed by neck dissection sustained severe late complications. Weber et al.³⁰ reviewed the experience at the University of Texas M.D. Anderson Cancer Center. Treatment to the primary site consisted of RT for 150 patients, surgery alone for 28 patients, and combined therapy for 10 patients. Local control rates for T1 through T4 lesions were 91%, 77%, 77%, and 35%, respectively. Local control was obtained in 88% of patients with N0 necks and in 77% of patients with nodal involvement. In addition, these investigators found that patients with tumor extension to the tongue base, midline tumors, or tumors that extended across the palatine arch had inferior survival compared with those who did not exhibit those features (P < .05).³⁰ Similar results have been reported in other single-institution retrospective series.³¹

Combined External Beam Radiation Therapy and Brachytherapy Implant

The use of a brachytherapy implant for small soft palate lesions has been reported extensively by French investigators. Esche et al.³² describe 43 patients with carcinoma of the soft palate and uvula who were treated by interstitial implant and EBRT. Patients received 5000 cGy external irradiation to the oropharynx and neck, followed by 2000 to 3500 cGy delivered by a low-dose rate (40 to 100 cGy/hour) interstitial iridium-192 (192Ir) implant. This therapy yielded a local control of 92%, with no local failures in 34 T1 primary tumors. One serious complication was seen. Overall actuarial survival was 60% at 3 years and 37% at 5 years, but cause-specific survivals were 81% and 64%, respectively. The leading cause of death was other aerodigestive cancers, with an actuarial rate of occurrence of 10% per year after treatment of a soft palate cancer.

Mazeron et al.³³ reported on a subset of patients with early stage tumors who received EBRT to the primary tumor and neck nodes to a dose of 4500 cGy, followed by 3000 cGy delivered by a 192Ir implant to the primary tumor. Local control was 85% for soft palate tumors. Regional control was 97% for patients with N0 disease and 88% for patients with N1 through N3 disease. Soft tissue ulceration occurred in 17 patients, all of whom healed spontaneously.

Pernot et al.³⁴ reported on 277 patients with velotonsillar cancer (oropharyngeal cancer excluding base of tongue and valleculae) who were treated by brachytherapy either alone (14 patients) or combined with external beam irradiation (263 patients) using an after-loading 192Ir technique. Thirty-five percent of the patients had soft palate lesions. Of the patients treated for early lesions in the soft palate, the local control rates for T1 to T2N0 and T1 to T2N1 to N3 were 90% and 86%, respectively. No local recurrence was detected after 3 years.

External Beam Radiation Therapy versus External Beam Radiation Therapy Plus Brachytherapy

The need for an implant was analyzed retrospectively by Mazeron et al.³⁵ who reported T1 and T2 carcinomas of the soft palate and uvula treated definitively by EBRT alone (16 patients), 192Ir implant alone (14 patients), or a combination of the two methods (29 patients). Two techniques of implantation were used: the guide-gutter technique (33 patients) and the plastic-tube technique (10 patients). Local failure was 25% after EBRT alone, 0% after 192Ir implant alone, and 18% after combined therapy. No local failures were seen with the plastic tube technique compared with 15% for guide gutters. Only two nodal failures were observed (3%). Crude 5-year disease-free survival (DFS) was 33%. Severe complications were limited to one incident of osteonecrosis, one soft-tissue necrosis, and one case of partial palatal incompetence. Xerostomia was reduced when implantation was used for part or all of the treatment. This is not conclusive proof that an implant is required, and these results may be difficult to duplicate without the requisite expertise in brachytherapy. Clearly, patient selection factors are important in these results.

Fractionated High Dose Rate or Pulsed Dose Rate Brachytherapy

In an effort to minimize occupational radiation exposure and decrease the patient isolation period while minimizing normal tissue complications, Levendag et al. reported an initial experience combining EBRT with high dose rate (HDR—100 cGy/minute) or pulsed dose rate (PDR) brachytherapy in 38 patients—19 soft palate tumors (14 T1 to 2) and 19 tonsillar cancers.³⁶ Twice-a-day fractions of 3.0 to 5.4 Gy HDR to a dose of 15 to 27 Gy or 4 × 2 Gy to 8 × 1 Gy per day of PDR to a dose of 20 to 28 Gy was delivered within 1 to 2 weeks after completion of EBRT (46 to 50 Gy, median cumulative dose of 66 Gy). At a mean follow-up of 2.6 years, 87% of soft palate tumors were locally controlled. The incidence of grade 3 mucositis or ulceration was no different between patients treated with PDR versus HDR and similar to those reported after low dose rate brachytherapy.

Clearly, RT is very effective for patients with early disease but is suboptimal for patients with advanced disease. The need for a brachytherapy implant has not been proven. However, there are anecdotal comments related to improved salivary function in those patients who received an implant, due to the decreased external beam dose to the major salivary glands. Patient selection is important, and a radiation oncologist with experience in palatal implants must perform the therapy.

Radiation Therapy

The use of RT as the definitive treatment for tumors of the tonsillar fossa is appropriate for T1, T2, and T3 (exophytic) tumors. For infiltrative or endophytic T3 or T4 lesions, either an organ-preserving approach involving concurrent chemotherapy or, in selected instances, surgery combined with postoperative RT (or concurrent chemoradiation) should be used. The target volume of the ipsilateral treatment should include the primary lesion with at least 2-cm margins, the ipsilateral jugular vein, and lateral retropharyngeal lymph nodes. If disease extends to the base of tongue, then EBRT alone is not as effective as EBRT plus implant to the tongue.³⁷ In this case, full-dose EBRT is delivered, followed by an interstitial 192Ir implant as a boost to the tongue portion of the target.

For early-stage, ipsilateral lesions with N0 or N1 nodal involvement, the primary site and ipsilateral neck can be treated while sparing the contralateral side, thus, optimizing the patient’s oncologic treatment and treatment-related morbidity. In the modern era, this is achieved with either three-dimensional conformal radiation therapy (3DCRT) or an intensity-modulated radiation therapy (IMRT) plan (see Radiation Therapy Techniques). In general, most T1 to T2 lesions in patients with an N0 or N1 neck can be treated with ipsilateral fields. Lesions that cross the midline, involve the tongue base, or involve N2 or more advanced neck disease should be treated to both the ipsilateral and contralateral neck. Treatment planning must ensure protection of critical normal structures including the spinal cord and larynx. A computed tomographic (CT) scan is performed for treatment planning.

Surgery Alone

The use of surgery as the sole treatment for early tonsillar disease is not reported frequently. However, excellent local control rates ranging from 80% to 90% have been reported.^34,38,39 When there is extension to the lateral pharyngeal wall or base of tongue, local recurrence approaches 33% and 47%, respectively.^40–42 The degree to which local control can be obtained depends on the extension of disease outside the tonsillar fossa.

External Beam Radiation Therapy Alone

No definitive randomized studies comparing surgery versus EBRT have been reported. However, no obvious differences between these modalities in LRC or absolute survival can be determined based on the reported literature.^38,43–54 EBRT alone has become the treatment of choice for most early stage lesions. If local control is maintained in patients by EBRT alone, the greatest risk to these patients is the development of future aerodigestive malignancies.⁵⁵ In general, the results using conventionally fractionated EBRT alone (1.8 to 2.0 Gy/dose during 6.5 to 7 weeks to doses of 65 to 70 Gy) are excellent for early stage tumors.^{39,47–49,56}

Mendenhall updated the University of Florida experience using definitive radiation to the primary site for treatment of 503 tonsillar cancers during a 38-year period, with a minimum 2-year follow-up.⁵⁷ Patients were treated with continuous course radiation (n = 177, median total dose 66 Gy) or using hyperfractionation (n = 326, median total dose 76.8 Gy). Eighteen patients received induction chemotherapy, 39 received concomitant chemotherapy, 198 underwent planned neck dissection, and 106 underwent an interstitial brachytherapy boost. Five-year rates of local control were as follows: T1 (n = 75) 88%; T2 (n = 191) 84%; T3 (n = 151) 78%; and T4 (n = 86) 61%. Of the 95 local recurrences, 26 patients (27%) were successfully salvaged. The ultimate local control rates at 5 years were T1 94%, T2 91%, T3 82%, and T4 68%. Five-year cause-specific survivals by 1998 AJCC stage (5th ed.) were as follows: I 100%, II 86%, III 84%, IVA 73%, and IVB 46%. Absolute overall survival (OS) at 5 years was as follows: I 54%, II 61%, III 62%, IVA 57%, and IVB 33%. Only three patients required hospitalization during or shortly after RT; however, 9% (46/503) developed severe late complications including osteoradionecrosis requiring mandibulectomy (n = 15), dysphagia requiring gastrostomy (n = 18), bone exposure necessitating debridement and hyperbaric treatment (n = 10), orocutaneous fistula (n = 2), fatal radiation-induced sarcoma (n = 1) and fatal aspiration pneumonia (n = l).

Remmler et al.⁴⁰ reported the results in 160 patients in whom EBRT was the sole therapy for the majority of patients. Primary tumor control rates were 100% for T1 lesions, 89% for T2, 68% for T3, and 24% for T4. In addition, control of cervical metastases by RT was excellent (95%). When a planned neck dissection was performed 5 weeks after RT, the control of cancer in the neck was 100%. The incidence of distant metastases was 10% and was not affected by the selection of therapy. The 2- and 5-year determinate survival figures for 112 patients treated with RT alone were 67% and 48%, respectively, whereas 31 patients treated with RT followed by neck dissection achieved a 5-year survival rate of 48% (Table 28-4).

Lower doses or split-course radiation results in poorer LRC.⁵⁸ Using RT alone, Bataini et al.⁴⁷ reported that in tumors arising from the glossopalatine sulcus, characterized by involvement of the tongue, inferior local control is achieved compared with similar therapy for tumors arising from other sites within the tonsillar region.

Ipsilateral External Beam Radiation Therapy

For patients with T1 and T2 lesions, treatment of the ipsilateral neck alone usually is possible without having to irradiate the contralateral neck, which minimizes irradiation to the contralateral salivary gland and reduces the incidence of xerostomia. Eisbruch has demonstrated that patients treated unilaterally report less xerostomia and better quality of life compared with those treated with intensity-modulated RT delivery of bilateral radiation with contralateral parotid sparing.⁵⁹ However, careful patient selection is required to minimize the risk for contralateral neck failure. Two recent major series document excellent outcomes using ipsilateral neck EBRT.

O’Sullivan et al. reported the results of 228 tonsillar carcinomas treated with ipsilateral RT during a 20-year period at Princess Margaret Hospital.⁶⁰ Eligible patients typically had T1 or T2 tumors (191 T1-2, 30 T3, 7 T4) with N0 (133 N0, 35 N1, 27 N2 to 3) disease. During this period, only 16 patients were treated surgically. Radiation was typically delivered with wedged-pair cobalt beams and ipsilateral low anterior neck field delivering 50 Gy in 4 weeks (90% isodose line) to the primary volume. At a median follow-up of 5.7 years, the 3-year local control rate was 77%, regional control was 80%, and cause-specific survival was 76%. Contralateral neck failure occurred in 3% (8/228). All patients with T1 lesions or N0 neck status had 100% contralateral neck control. Patients with a 10% or greater risk for contralateral neck failure included those with T3 lesions, lesions involving the medial one third of the soft hemipalate, tumors invading the middle third of the ipsilateral base of tongue, and patients with N1 disease.

Jackson et al. reported an 18-year experience of 178 patients receiving ipsilateral treatment using limited fields for tonsillar cancers.⁶¹ Patients presented primarily with T1 to T2 (117/178 [66%]) and N0 (101/178 [57%]) disease, but 29% (52/178) had T3 tumors and 30% had N1 disease, with 63% presenting with stage III to IV disease. Sixty Gy per 25 fractions (50 to 66 Gy) was delivered to gross tumor volume with a 1-cm margin and first echelon lymph nodes using two or three wedged fields. The length of follow-up was not stated. The rates of LRC and contralateral neck recurrence by stage were as follows after ipsilateral RT: I (n = 23), 91%, 0%; II (n = 43), 74%, 2%; III (n = 82), 51%, 4%; IV (n = 30), 53%, 0%. Patients with N0 (n = 101) or N1 (n = 54) disease had contralateral failure rates of 0% and 4%, respectively. None of the 23 patients with N2 to N3 disease had contralateral failure; however, the determinate risk for contralateral failure may have been obscured by the 52% incidence of ipsilateral failure. The authors were unable to relate the risk for contralateral neck failure according to the degree of tumor extension along the glossopalatine fold due to the retrospective nature of the study. The overall rate of local control was 75% (T1-2, 84% and T3-4, 58%) and OS was 56%.

External Beam Radiation Therapy and Brachytherapy

The goal of combining brachytherapy with EBRT is to improve local control while preserving salivary function and lessening muscular fibrosis.

Pernot et al.³⁴ reported on 277 patients, 101 of whom had advanced disease (T3). The 5-year local control, disease-specific survival, and OS rates by T stage (T1, T2, T3) were as follows: local control, 89%, 86%, and 69%, respectively; disease-specific survival, 78%, 62%, and 46%, respectively; and OS, 62%, 53%, and 43%, respectively. No local recurrence was detected after 3 years. In a later update of 361 cases, the 5-year outcomes of patients with tonsil, posterior pillar, or soft palate cancers (group A) were compared with those involving the anterior pillar and glossopharyngeal sulcus (group B).⁴⁶ Local control was better in group A patients compared with those in group B as follows: T1 94% versus 75%; T2 93% versus 67%; and T3 71% versus 51%, respectively. Disease-specific survival and OS was also better in group A as follows: T1: 88%, 65% versus 55%, 48%; T2 78%, 63% versus 43%, 38%; and T3 53%, 49% versus 27%, 27%, respectively. Multivariate analysis revealed that a treatment interval of less than 20 days between EBRT and brachytherapy and an overall treatment time of less than 55 days yielded better outcomes. Complications in this patient population appeared to be related to a total dose greater than 80 Gy, dose rate greater than 70 cGy per hour, treated surface area greater than 12 cm², treatment volume greater than 30 cc, and absence of leaded protection.⁶²

Levendag et al. reported on 248 patients with Stage T1-T3, N0-N+ SCC of the tonsillar fossa and/or soft palate treated at Erasmus Medical Center between 1986 and 2001.⁶³ One hundred four patients were treated with external RT (median dose 46 Gy) followed by a brachytherapy implant, primarily using high-dose-rate (15 to 30 Gy at two daily fractions of 3 to 4 Gy to 15 to 30 Gy, n = 31) or pulsed-dose-rate (eight daily fractions of 1 to 2 Gy to 18 to 30 Gy, n = 62) treatment schedules. Patients who were node-positive underwent unilateral or bilateral neck dissection at the time of brachytherapy implant. Eighty-six patients who were felt to be unsuitable for brachytherapy at presentation (e.g., parapharyngeal extension) underwent surgery of the primary tumor and neck dissection, followed by postoperative RT to 50 to 70 Gy, depending on pathologic risk factors. Local control rates at 5 years for the definitive RT and surgical cohorts were 88% and 88%, respectively. Regional control, DFS, and OS rates were 93% versus 85%, 57% versus 52%, and 67% versus 57%, respectively. Significant late side effects in the brachytherapy group versus the surgery group included and trismus (1% versus 21%, P = .005) ulceration (39% versus 7%, P = .001). Ulceration healed spontaneously in 88% after a mean period of 6 months.

External Beam Radiation Therapy versus External Beam Radiation Therapy Plus Brachytherapy

Mazeron et al.⁶⁴ reported on 165 T1 to T2 SCCs of the faucial arch. Because of institutional policy changes, these authors were able to compare patients who received EBRT alone with those with EBRT and implantation. Those who received an implant were first treated by EBRT to the tumor site and neck areas (4500 cGy in 25 fractions during 5 weeks) and then received a 3000-cGy low-dose-rate 192Ir implant. For patients with clinically positive nodes, either additional 2500- to 3000-cGy electron beam irradiation to the nodes or neck dissection was added. Both local control (77% versus 94% at 5 years; P < .01) and DFS (56% versus 71%; P = .03) were improved for the implant group. No randomized study has shown whether this combined approach is superior to EBRT alone. Nonetheless, even advocates of the use of EBRT alone agree that the addition of an implant can improve local control when disease extends into the tongue.³⁷

High Dose Rate Brachytherapy

High dose rate brachytherapy as boost treatment with external beam radiation has been reported in two small series to offer excellent local control for early stage tonsil, base of tongue, and soft palate tumors with local control of 83% to 87% but lower with T3 to T4 tumors, 42% to 47%.^36,65 Fractionated high dose rate brachytherapy using b.i.d. treatments of 1.2 to 3.0 Gy fractions was combined with EBRT to total cumulative doses of 66 to 72 Gy. Successful surgical salvage was 50% to 60% among those who failed locally with no obvious increase in surgical complications such as fistula or flap necrosis. Rates of serious complications (primarily soft tissue ulcer and mandibular osteoradionecrosis) were reported in 10% to 16% of patients and were similar to those reported by low dose rate brachytherapy reports.

Radiation Therapy

Primary RT is the preferred definitive treatment for most tumors of the base of tongue (see Radiation Techniques). It is the sole primary therapy for most T1, T2, and many T3 (exophytic) tumors. For infiltrative or endophytic T3 or T4 lesions, an organ-preservation approach using RT and chemotherapy should be used. The target volume should include the base of tongue and margin that will incorporate a portion of the oral tongue, vallecula, pharyngeal walls, and suprahyoid epiglottis, as well as the superior portion of the preepiglottic space. Owing to the high probability of lymph node metastasis, these portals also will include the bilateral regional and retropharyngeal lymph node groups. This is also true in the case of the patient with a clinically negative neck.

Our preferred strategy entails EBRT followed by an interstitial 192Ir implant as a boost to the base of tongue. We believe that this offers the most consistent excellent local control and best function results compared with either EBRT alone or to primary surgery. External beam radiation is delivered to a dose of 5400 cGy in 30 fractions (180 cGy/fraction). A brachytherapy boost to the tongue base (2000 to 3000 cGy) is performed approximately 3 weeks after EBRT (5000 to 5400 cGy). The implant consists of 192Ir after-loading catheters placed with a looping technique,^66,67 which involves the percutaneous introduction of curved trocars by means of a submental approach through the base of tongue. The catheter is threaded through the trocar and looped back through an adjacent trocar, creating a loop in the tongue. For patients with disease extension toward the pharyngoepiglottic fold, lateral loops are added to encompass this region.⁶⁶ Both ends exit the skin of the neck. An array of loops is fashioned to encompass the target volume plus a 1.0- to 1.5-cm margin. The spacing between each end of the loop is 1 cm and between each plane is 1.0 to 1.5 cm (see Radiation Therapy Techniques). As a safety precaution, a temporary tracheostomy is performed in patients immediately before the implantation. A temporary nasogastric feeding tube is placed at the completion of the procedure. The neck is managed by RT alone in the patient with N0 or N1 disease. For patients with more advanced neck disease, chemotherapy is given concomitantly with RT. When a neck dissection is planned, the implantation and dissection are performed during the same anesthetic period. Patients are loaded several days after the procedure, after the majority of swelling has subsided and patients are able to suction themselves and perform tracheostomy care. Efforts are made to protect the palate by use of a customized shield that is under patient control.

Surgical Therapy

Selected patients with small base-of-tongue lesions and a negative neck can be managed with primary resection and elective neck dissection. A significant percentage will be found to harbor positive microscopic nodes, thereby necessitating postoperative RT. Therefore, for purposes of maximizing functional outcome and using only one modality, primary RT is the preferred strategy. Surgery generally is reserved for patients who cannot receive RT or for those with particularly endophytic, locally advanced lesions that are difficult to control with RT alone.

By contrast, patients with nodal metastases routinely will undergo a surgical procedure in the neck if the primary tumor is being treated by radiation alone. The neck dissection preferably is performed at the completion of RT. If the primary tumor is being managed surgically, surgery followed by RT is preferred. In our practice, patients generally undergo a combination of EBRT and a brachytherapy implant (discussed later in this chapter). In this setting, the implant and the neck dissection are carried out during the same anesthetic period, approximately 3 weeks after the EBRT is completed. Andersen et al.⁶⁸ have shown that preservation of the spinal accessory nerve in a modified neck dissection in patients with clinically evident nodal metastases was not associated with increased risk for treatment failure in the dissected neck.

Base-of-tongue tumors may be resected transorally or through a mandibulotomy and transhyoid pharyngotomy. The transoral approach is indicated only for small, well-circumscribed lesions that are located superficially. The mandibulotomy technique allows superior access and frequently is combined with a graft.⁶⁹ When a patient has evidence of bone invasion or close encroachment of tumor to bone, frequently a mandibulectomy is performed. The use of a flap or plate with this procedure depends on the age of the patient, the amount of tissue resected, and the anticipated functional or cosmetic outcome.

Patients with advanced tumors usually will require a major or total glossectomy. In addition, some patients may require a total glossectomy owing to local extension of their tumor. The need for laryngeal resection (either subtotal supraglottic resection or total laryngectomy) depends on the extent of disease and the risk for aspiration. In most patients in whom a laryngectomy is deemed necessary for base-of-tongue cancer, the indication is for prevention of chronic aspiration, not for oncologic purposes.

After surgery, patients will spend a great deal of time learning to swallow and avoiding aspiration. If the dysfunction is significant, the patient may require placement of a percutaneous endoscopic gastrostomy (PEG), which can be performed at the time of surgery.⁷⁰ A prosthesis may have to be constructed to aid speech and swallowing. Speech rehabilitation is started as soon as possible while the patient is still in the hospital.

Postoperative Adjuvant Radiation Therapy

After careful examination of the resected specimen by the pathologist, attention should be paid to the size of the primary tumor, histology, surgical margins, evidence of perineural extension, and lymph nodes. Patients with small tumors (stage T2), clear negative margins, no evidence of perineural extension, and histologically negative lymph nodes do not require postoperative RT. RT generally is used when the surgical wound is seeded, when close or positive margins are present, and in the presence of extranodal tumor spread (extracapsular extension), multiple positive nodes, or extensive vascular or perineural invasion.⁷¹ Because larger primary tumors are poorly controlled locally with surgery alone, postoperative RT is suggested in this group of patients. The retropharyngeal lymph nodes are included in the treatment of the primary tumor.⁷² Owing to the high probability for bilateral neck disease, RT is used electively to control the area of unoperated contralateral neck.

The interval between surgery and RT was examined by Vikram,⁷³ who reports that a delay of 7 weeks or more was associated with increased locoregional failure and decreased survival. The neck disease recurrence rate was 2% if the postoperative RT was started within 6 weeks compared with 22% at later than 6 weeks. A reanalysis by Schiff et al.⁷⁴ suggests that a prolonged delay in postoperative RT does not have a negative impact on LRC as long as appropriate tumoricidal doses of more than 6000 cGy are employed. Seventy-three percent in the delayed treatment group received doses less than 56 Gy, accounting for their increased risk for locoregional failure. Bastit et al. reported no difference in LRC or survival in a retrospective study of 420 oropharyngeal and hypopharyngeal carcinomas who were treated with similar doses (59 Gy/6 weeks) either within (n = 219) or after 30 days (n = 201) from surgery.⁷⁵ However, Ang et al. reported that among high-risk patients (positive margins, extracapsular nodal extension, or multiple lymph nodes), a duration of less than 11 weeks from the end of surgery to the end of radiation resulted in better LRC and survival compared with those treated during a longer period.⁷⁶ We recommend starting radiation as soon as possible.⁷⁴

Surgery Alone versus Radiotherapy

Surgical results for early base-of-tongue tumors reflect relatively high local control rates—from 75% to 85%.^75,76 Primary EBRT, alone or followed by a planned neck dissection, also has a high local control rate of approximately 80% to 90% for T1 to T2 lesions and 70% to 85% for T3 tumors (Table 28-5). Survival for RT is similar to that after surgery, but with a lower risk for severe complications than surgery.⁷⁷

Data from numerous authors have shown that the local control for early stage tumors is in the range of 80% to 100% when treated with combined EBRT and brachytherapy implant (Table 28-6). Pernot et al. reported a 5-year respective local control of 93% and 72% for T1 (n = 14) and T2 (n = 27) base-of-tongue tumors treated with a combined EBRT and brachytherapy approach. Corresponding 5-year disease-specific survivals were 76% and 62%, respectively.⁴⁶

Housset et al.⁶⁷ have evaluated a comparison among surgery plus postoperative RT, EBRT plus implantation, and EBRT alone for a series of patients with T1 to T2 base-of-tongue lesions. This series is unique in that it attempts to address the issue of which management strategy is optimal. Demographic and oncologic characteristics of the patients were well balanced except that, among the EBRT-alone group, there were significantly more patients with exophytic (more favorable) lesions. Despite this imbalance, the patients who received EBRT alone had approximately twice the local failure rate as those in the other two groups (40% versus 20%). This study suggests that EBRT plus brachytherapy is the preferred technique.

Surgery Alone

The use of surgical resection alone for pharyngeal wall tumors has been reported by Guillamondegui et al.⁷⁸ Twenty-eight percent of patients had recurrent tumors above the clavicles. Salvage in the form of RT or surgery was successful in less than one third of those patients. Patients with positive retropharyngeal nodes had an increased rate of distant metastases (22%) compared with those who did not have positive nodes (15%).

Definitive Radiation Therapy Alone

Hull et al.⁷⁹ from the University of Florida retrospectively reviewed their experience with definitive radiation for pharyngeal wall tumors treated between 1964 and 2000. Thirty-seven percent of patients had oropharyngeal wall tumors, while the remainder was of hypopharyngeal origin. Sixty-four percent of patients were treated with a hyperfractionated technique, and 11 patients also received chemotherapy. Five-year local control rates for T1, T2, T3, and T4 tumors were 93%, 82%, 59%, and 50%, respectively. On multivariate analysis, factors associated with improved LRC included twice-daily fractionation (P = .0009), Stage I to II disease (P = .0051), and oropharyngeal primary site (P = .0193). The same group previously reported the effect of the use of once-daily to twice-daily fractionation on local control in 99 patients with carcinomas of the hypopharyngeal or oropharyngeal wall or both. The local control rates for patients treated with once-daily versus twice-daily fractionation were, for T1 lesions, 100% versus 100%; for T2 lesions, 67% versus 92%; for T3 lesions, 43% versus 80%; and for T4 lesions, 17% versus 50%. These investigators also examined their former technique (posterior border placed at middle of the vertebral body when the portals were reduced off the spinal cord) compared with their current, modified technique (posterior border placed at posterior edge of the vertebral body). The local control rates were 100% each for T1 lesions; 57% versus 100% for T2 lesions; 46% versus 73% for T3 lesions; and 29% versus 75% for T4 lesions.⁸⁰

Meoz-Mendez et al.⁸¹ reported on 164 patients treated at the M.D. Anderson Cancer Center for carcinomas of the hypopharynx and pharyngeal walls. Local control rates for T1 through T4 lesions were 71%, 73%, 61%, and 37%, respectively. For patients with T3 and T4 tumors, the combination of surgery and RT was superior to RT alone (75% local control versus 51%).

Surgery and Radiation Therapy

Marks et al.⁸² retrospectively compared the results of treatment in 51 patients after low-dose (2500 to 3000 cGy) preoperative RT and surgery to the results in those who had definitive RT. No difference was found in local control or survival (17%); however, the surgery plus RT group experienced a significant number of complications (fistula, 31%; carotid rupture, 14%; operative mortality, 14%). The same authors updated their experience with a group consisting of 89 patients.^83,84 These patients were treated with high doses of radiation (5000 to 7200 cGy) either for preoperative intent or for definitive therapy. Treatment outcome, survival, and tumor and nodal control were better for the RT-plus-surgery group than for the RT-alone group. The patterns of relapse differed for the two groups, with low-dose preoperative irradiation and surgery offering greater control of the primary tumor and high-dose irradiation achieving better control of nodal disease. To draw any conclusions from this article is difficult, as the reported study spanned almost 20 years, during which time many technologic advances were made in radiation oncology, anesthesia, and surgery.

Spiro et al.⁸⁵ reviewed a 12-year experience with 295 patients treated for SCC of the pharynx at the Memorial Sloan-Kettering Cancer Center. Of these patients, 78 patients had lesions in the posterior wall. Surgery was the definitive therapy for the primary tumor in 73%. A second group of 21 patients with more extensive tumors required a laryngectomy and complex reconstruction, often with postoperative RT, and five had lesions that were implanted after access was provided by a mandibulotomy. The cumulative 5-year survival for the entire group was 32% and ranged from 44% in those with favorable lesions to 15% in those with extensive tumors. The overall complication rate was 50%. For the group that received an implant, local control was excellent.

Julieron et al.⁸⁶ from the Institut Gustav-Roussy reported their results among 77 patients treated surgically between 1984 and 1995. Twenty-three of these patients had been previously irradiated, and all patients who were previously untreated underwent postoperative radiation. Of the 54 patients who were previously untreated, 89% were locally controlled and 78% were locoregionally controlled, with a 5-year survival rate of 35%.

The use of a brachytherapy implant (either 192Ir or iodine-125 [125I]) and EBRT was reported by Son and Kacinski⁸⁷ for a small group of patients. The local control rate was 86% and actuarial survival was 82% at 5 years. These results are impressive but must be confirmed. In general, results in patients with extensive lesions still leave much to be desired, despite radical surgery and aggressive RT, and creative brachytherapy techniques warrant further investigation.

Phase II Studies

Harrison reported the results of a phase II trial that treated 82 patients with unresectable head and neck cancer using the DCB technique (70 Gy/6 weeks, b.i.d. RT last 2 weeks) with concurrent cisplatin (100 mg/m²) given every 3 weeks for 2 cycles.¹²⁵ Adjuvant chemotherapy was given in 68%. Although 40% had initial skull base invasion, the response rate was 94% with 60% complete responses. Oropharynx cancers comprised 23% of all tumors treated. For all patients at a minimal follow-up of 3 years, the 3-year LC, OS, and distant-metastasis-free survival was 58%, 36%, and 58%, respectively. Seventy percent of patients with base of skull invasion were locally controlled. The treatment was reasonably well tolerated. The 3-year LC for oropharynx cancers was 64%. All patients experienced grade 3 acute mucositis usually during the concomitant boost phase. Twenty-four percent required a treatment break, the majority requiring less than one week. Two deaths due to sepsis occurred during treatment. Severe chronic toxicity occurred in three patients with one osteoradionecrosis, one frontal lobe necrosis, and one case of lung toxicity secondary to adjuvant chemotherapy. Given the especially poor prognosis of this group, the LC was good and survival better than expected with such an aggressive chemoradiation program. Our experience in treating advanced oropharyngeal disease was recently updated.¹²⁶ Forty patients were treated with the above regimen, with either two cycles of cisplatin (100 mg/m² on weeks 1 and 4) or daily carboplatin (10 to 17.5 mg/m²). Seventy-five percent of patients had T3 or T4 disease, and 57.5% of patients were stage N2 or N3. The 2-year LC, regional control, LRC, freedom from distant metastasis, DFS and OS rates were 81%, 97%, 81%, 74%, 76% and 80%, respectively. Grade 3 toxicity rates were as follows: mucositis, 67.5%; dysphagia, 62.5%, odynophagia, 62.5%. Patients required a median 2.24 days of treatment break. Late toxicities were mainly due to grade 3 to 4 xerostomia (36%, 2.2%, respectively) and grade 3 dysphagia (11.1%). One patient developed osteonecrosis and one patient developed soft tissue necrosis. Of the 80% of patients who had a gastrostomy tube placed before or during treatment, two patients were still dependent on parenteral feeds 5 and 17 months after treatment completion.

A modified version of DCB radiation was reported by Bieri et al. in which a planned total dose of 69.6 Gy was shortened to total of 5.5 weeks and one third of the patients received primarily concurrent cisplatin-based chemotherapy.¹²⁷ Among the 55 patients with oropharynx carcinoma (76% stage III/IV), LRC at 3 years was 69.5% at a median follow-up of 32 months. Eighty-two percent experienced grade 3 or 4 mucositis. Patients receiving chemotherapy had a greater rate of grade 3 dysphagia (68% versus 25%, P = .003), hospitalization (37% versus 14%, P = .08), and need for nasogastric tube (68% versus 22%, P = .001). Late RTOG grade 3/4 complications occurred in 12% of patients and consisted of laryngeal edema, mucosal necrosis, and mandibular osteoradionecrosis.

Induction chemotherapy followed by RT was used in the treatment of patients with T4 (AJCC, 2nd ed.) oropharyngeal carcinomas at the University of Florida.¹²⁸ Chemotherapy consisted of cisplatinum (100 mg/m²) and 5-FU (1000 mg/m²/day × 5 days) every 3 to 4 weeks for several cycles (63% received 3 cycles) followed by definitive RT (83% received hyperfractionated RT to 74.4 to 81.75 Gy). Twenty-six patients were able to complete the chemotherapy and high-dose radiation. The major response rate after induction chemotherapy was 97% (35% complete response and 62% partial response). These rates were compared with a similar group of 34 patients with T4 oropharyngeal tumors treated with a comparable radiation regimen without chemotherapy during the same time period and also to 83 patients with T4 tumors treated with RT alone during a three-decade time span. All patients were followed for a minimum of 1 year and none were lost to follow-up until death from 1964 to 1996. On multivariate analysis, no difference in local failure (37% versus 62% versus 52%, respectively) or regional or distant failure (13% versus 27% versus 24%, respectively) was detected in those receiving induction chemotherapy compared with patients treated with RT alone. However, disease-specific survival and OS were improved in those who received induction chemotherapy (58% versus 27% versus 37% and 42% versus 17% versus 23%, respectively). Because of the nonrandomized nature of the study and the lack of statistically significant improvement in parameters of tumor control, the authors cautioned against any conclusions regarding the benefit of induction chemotherapy.

A phase II study using induction chemotherapy followed by concurrent chemoradiation was reported by Vokes et al.¹²⁹ Sixty-one patients with advanced oropharyngeal carcinomas (97% IV, 82% T3-4, and 85% N2 to N3) were treated with three cycles of cisplatinum (100 mg/m²), 5-FU (640 mg/m²/days × 5 days), leucovorin (300-600 mg/m²/day × 6 days), and interferon-α (2 million U/m²/day × 6 day) for three cycles then proceeded to concurrent chemoradiotherapy if at least a near complete response was obtained. This consisted of split courses of weekly concurrent hydroxyurea (2000 mg/day × 5 days) and 5-FU (800 mg/m² × 5 days) with 900 to 1000 cGy per 5 days given every other week to a total dose of 68 to 75 Gy for gross disease. Neck dissections (n = 35) were performed for N2 to N3 disease. After induction chemotherapy, 65% obtained a complete response and 34% a partial response. Sixteen patients underwent resection of the primary tumor (11 organ-sparing and 13 after induction chemotherapy). At a median follow-up of 39 months (68 months among survivors), LRC was 70%, distant-metastasis-free survival 89%, DFS 64%, and OS 51%. LRC was 100% in patients with T1 to T2 (n = 11) or N0 (n = 1), 81% for T3 (n = 16), 53% for T4 (n = 34), and 67% to 69% N1 to N3 (n = 60). No difference was noted in LC among those undergoing surgery of the primary site versus those treated with chemoradiation alone (5-year LRC 66% versus 72%). Regional control was obtained in 98%. An update with 93 patients showed similar results.¹³⁰ Acute toxicity was substantial with severe or life-threatening mucositis in 57% and leukopenia in 65% during the induction phase, whereas 81% had grade 3 or 4 mucositis during the concurrent chemoradiation. The authors concluded that the treatment sequence of induction chemotherapy followed by concurrent chemoradiation and optional organ-preservation surgery is promising but that less toxic regimens need to be identified.¹³⁰

Bensadoun et al. reported on 54 patients with unresectable oropharynx and hypopharynx carcinoma treated with concomitant hyperfractionated radiation (75.6 to 80.4 Gy) and three cycles of 5-FU/cisplatin (CDDP) (750 mg/m²/day or 750 mg/day × 5 days and 100 mg/m², respectively on weeks 1, 4, and 7).¹³¹ The regimen was acutely toxic (4% mortality from treatment-related septicemia, 86% grade 3/4 mucositis by cycle 2 of chemotherapy, grade 3 or 4 neutropenia in 43% by cycle 2) but no patient required a treatment break greater than 4 days due to mucositis. No grade ≥3 late toxicities were reported; grade 2 xerostomia was reported in 70% and grade 2 cervical fibrosis in 45%. At a median follow-up of 16 months, LRC was 67% at 6 months and disease-specific survival 72% at 2 years. Other chemoradiation regimens for oropharyngeal carcinomas with encouraging LRC rates but short-term follow-up or small patient numbers include those using concurrent weekly paclitaxel (30-50 mg/m²/week)¹³² and the use of induction chemotherapy with preoperative concurrent chemoradiation followed by organ-preserving surgery.¹³³

Randomized Trials

Concurrent chemotherapy with hyperfractionated radiation was explored by Brizel et al. in a phase III randomized trial.¹³⁴ One hundred and sixteen patients with advanced head and neck cancer were randomized to hyperfractionated radiation alone treated with 1.25 Gy b.i.d. (6-hour interfraction interval) 5 days per week to 75 Gy during a 6-week period versus a concurrent chemoradiation arm consisting of 5-FU/CDDP given on weeks 1 and 6 of split-course hyperfractionated radiation. In contrast to the control arm, an intentional 1-week treatment break was required in the experimental arm and the total dose was slightly lower (70 Gy). Both groups received two adjuvant courses of 5-FU/CDDP after completion of radiation. At a median follow-up of 41 months, the chemoradiation showed improved LRC (70% versus 44%, P = .01) and a trend toward improved 3-year OS (55% versus 34%, P = .07) and relapse-free survival (61% versus 41%, P = .08). However, patients in the chemoradiation arm developed more acute toxicity including the requirement for more feeding tubes (44% versus 29%) and worse hematologic suppression. Three fourths of patients in both arms experienced confluent mucositis. Chronic toxicity was no different, with about a 10% incidence of necrosis of the skin or bone in both arms. The trial has been criticized, not only for the added toxicity, but also because of the imbalance in the proportion of advanced neck disease (44% versus 63%) treated in the concurrent chemoradiation arm, which may have accounted for the difference in LRC.

Jeremic et al. reported a phase III randomized study testing whether daily low-dose cisplatin improved outcome for patients undergoing hyperfractionation radiation compared with those treated with the same hyperfractionated radiation alone in locally advanced head and neck SCCs (37% were oropharynx).¹³⁵ One hundred and thirty patients with stage III or IV disease were randomized to 1.1 Gy b.i.d. to 77 Gy per 7 weeks alone or with cisplatin (6 mg/m²/day). At a median follow-up of 79 months, the investigational arm showed improved LRC (50% versus 36% at 5 years, P = .041), progression-free survival (46% versus 25% at 5 years, P = .0068 and OS (46% versus 25% at 5 years, P = .0075), and fewer distant metastases (14% versus 43% at 5 years, P = .0013). The latter result was unexpected and raised the possibility that concurrent daily cisplatin may impact on incidence of distant metastases by a direct systemic effect or secondarily through improved LRC. Daily concurrent chemotherapy was well tolerated with no increase in acute grade 3 mucositis and esophagitis. There were no increases in late skin or severe effects to bone or salivary gland.

A German multicenter randomized trial tested whether the combination of hyperfractionated accelerated radiation (69.6 Gy/5 5.5 weeks) with carboplatin (70 mg/m²) and 5-FU (600 mg/m²/day × 5 days) on weeks 1 and 5 of RT improved outcome compared with the same radiation regimen alone.¹³⁶ Two hundred and forty patients with stage III (4%) to IV (96%) oropharyngeal (n = 178) and hypopharynx (n = 62) tumors were entered in a 2 × 2 study randomizing patients to receive radiation with or without chemotherapy and then again to receive granulocyte colony-stimulating factor (G-CSF) or not. G-CSF was administered to determine its effect on mucositis. Treatment was tolerable in both study arms with higher mucosal and hematologic toxicity in the chemotherapy/RT arm but no difference in duration of treatment (41 versus 42 days) or total radiation dose delivered (68.5 Gy versus 69.9 Gy) in the chemotherapy/RT versus RT arms, respectively. Response rates at 6 weeks after treatment were similar 92% (complete response = 40%) after chemotherapy/RT and 88% (complete response = 34%) after RT alone. At a median follow-up of 22 months, the 1- and 2-year respective rates of LRC were 69% and 51% after chemotherapy/RT compared with 58% and 45% after RT (P = .14). On subset analysis, patients with oropharyngeal carcinomas had l-year improved survival with LC after chemoradiation (60% versus 40%, P = .01) and trend toward improved LRC (70% versus 58%, l-year and 51% versus 42% 2-year, P = .07) compared with RT alone. Patients receiving chemotherapy/RT had increased grade 3 and 4 toxicities including mucositis (68% versus 52%, P = .01), vomiting (8% versus 2%, P = .02), and hematologic toxicity (neutropenia, 18% versus 0%) with similar rates of dermatitis (30% versus 28%). Patients receiving chemotherapy/RT had more swallowing problems and continuous need for feeding tube (51% versus 25%, P = .02). Interestingly, patients receiving G-CSF had reduced LRC (55% versus 38%, P = .0072) and decreased mucositis (P = .06). Unfortunately, this finding was not supported by RTOG 99-01, a large, multicenter randomized placebo-controlled study of GM-CSF that failed to show any improvement in radiation-induced mucositis rates.¹³⁷

The GORTEC group conducted a study (GORTEC 96-01), randomizing 109 patients to receive either very accelerated RT (62 to 64 Gy/3 weeks) or moderately accelerated RT (62 to 64 Gy/5 weeks) with concurrent cisplatin (100 mg/m² days 1, 16, and 32) and 5-FU (1 g/m² days 1 to 5 and 31 to 35),¹³⁸ followed by 2 additional cycles of chemotherapy 28 days and 56 days after RT. This trial was stopped early because of an increased incidence of treatment-related deaths in the chemoradiation arm. At preliminary report, no statistical difference on LRC, event-free, or OS could be detected between the two arms.

One question that remains to be answered is whether the employment of altered fractionation schedules further enhances the efficacy of concurrent chemotherapy and radiotherapy using a standard fractionation scheme. In order to address this issue, the RTOG¹³⁹ has randomized patients with stage III and IV head and neck SCC to receive either standard RT (70 Gy / 7 weeks or accelerated RT (72 Gy / 6 weeks via DCB), with both groups receiving concomitant cisplatin. The accelerated RT group will receive 100 mg/m² on days 1 and 22, while the standard RT group will receive an additional cycle on day 43. All patients with N2 to 3 disease, and N1 with residual disease at the completion of therapy, will undergo a planned neck dissection. This trial completed accrual in 2005, but preliminary results have not yet been reported.

Surgery and Postoperative Radiation

In the past, definitive resection was recommended in patients with advanced disease. Hicks reported on the Roswell Park Cancer Institute experience of 76 patients with tonsillar cancer treated with single modality therapy—surgery (56 patients) and radiation (20 patients).⁴⁴ Among stage III to IV patients (n = 52), surgery resulted in better 5-year DFS (47% versus 27%, P < .05) compared with radiation alone, although a greater percentage of patients were stage IV in the RT cohort (75% versus 44%, P < .05) and split-course treatment was delivered to about half those receiving radiation.

Surgery frequently results in close or positive margins and multiple positive neck nodes. Two reports highlight the importance of postoperative adjuvant RT in advanced cases. The first, by Zelefsky et al.,⁷² reported the long-term treatment results for advanced oropharyngeal carcinomas treated with surgery and postoperative RT at the Memorial Sloan-Kettering Cancer Center. Twenty patients with SCC of the tonsillar fossa were treated with surgery plus RT. The 7-year actuarial LC rate for tonsillar fossa lesions was 83%. LC was achieved in 94% of patients with T3 lesions and in 75% of patients with T4 lesions. Among patients with positive or close margins who received postoperative doses of 6000 cGy or more, the long-term control rate was 93%. At 7 years, the OS for all patients was 52%, and the DFS was 64%. The actuarial incidence of neck failure was 18%. For all patients, the likelihood of having distant metastasis at 7 years was 30%.

The second study, by Foote et al., evaluated 72 patients who had surgery either with or without postoperative adjuvant RT for advanced disease.³⁸ These investigators note that the main pattern of treatment failure was above the clavicles. This occurred in 39% of patients treated with surgery alone compared with 31% of patients undergoing surgery and postoperative adjuvant RT (despite the more advanced neck disease of the surgery and RT group) and was significantly related (P = .002) to the overall clinical tumor, node, metastasis (TNM) stage. Five-year OS rates for patients with clinical stage III and IV disease who were treated with surgery and postoperative adjuvant RT were 100% and 78%, compared with 56% and 43%, respectively.

Perez et al.¹⁶⁴ analyzed 296 patients with histologically proved epidermoid carcinoma of the tonsillar fossa: 127 were treated with irradiation alone (5500 to 7000 cGy), 133 received preoperative RT (2000 to 3000 cGy) or were planned initially for preoperative irradiation but were treated with RT alone, and 36 received postoperative irradiation (5000 to 6000 cGy). The primary tumor recurrence rate in the T1 to T2 groups was approximately 20% for patients treated with irradiation and surgery and 30% for those treated with irradiation alone (difference not statistically significant), 30% in patients with stage T3 lesions in all treatment groups, and 33% in patients with T4 disease treated with surgery and postoperative irradiation, compared with 52% for patients treated with irradiation alone (P = .03). Dasmahapatra et al.¹⁶³ likewise reported for stage III disease that the 5-year survival after RT and surgery was 31%, compared with 11% for radiation alone; and for stage IV disease, the respective 3- and 5-year survival rates for RT and surgery were 24% and 15%, compared with 6% and 0%, respectively, for RT alone.

By contrast, Spiro and Spiro¹⁶⁵ reviewed 162 patients with carcinoma of the tonsillar fossa treated between 1969 and 1983. Combined surgery and RT were used in 29% of patients with stage II disease, 40% of patients with stage III disease, and 67% of patients with stage IV disease. The 3-year determinate LC rates were 89%, 83%, 58%, and 49% for stages I through IV, respectively. The overall 2-year crude survival is 58%. A 60% 2-year survival for T3 lesions compares favorably with other series for treatment of T3 lesions. No survival benefit was seen in advanced stage disease, but this might reflect a selection bias against combined modality treatment compared with those receiving RT alone.

A number of retrospective reports have examined preoperative RT (4500 to 5000 cGy during 5 weeks) versus definitive RT or neck dissection.^52,53,166 No advantage to preoperative RT was seen. Thus, preoperative RT for resectable lesions for which surgery is planned is no longer used as a strategy for the primary site. Patients with early stage primary lesions but advanced neck disease are often treated with definitive RT to the primary with preoperative radiation to the neck followed by planned neck dissection.

External Beam Radiation Therapy and Brachytherapy

The use of a combination of external megavoltage irradiation and interstitial 192Ir implants for T3 lesions has been reported by Puthawala et al.¹⁶⁷ LC of disease in patients with T3 and T4 lesions was 79%, compared with 95% for T1 and T2 lesions. Treatment-related complications such as soft tissue necrosis or osteoradionecrosis occurred in 6% of patients in the primary group and in 23% in the recurrent group. No significant functional or aesthetic impairments were reported.

A number of investigators have reported that patients with tumor that extends to the tongue have an inferior outcome compared with patients with no tongue extension. However, when an implant is used, this difference is negated. In the retrospective report of Leborgne et al.,¹⁶⁸ local relapse was 64% and the 3-year DFS rate was 23% in 39 patients with tongue extension who were treated with EBRT alone, compared with 33% and 43%, respectively, for 90 patients with no tongue extension. However, in those treated with EBRT plus brachytherapy, the local relapse rate was 40% and the 3-year survival rate 60%.

Recurrent Disease and Salvage

Surgical salvage of recurrent disease after RT has a greater chance of success for tumors of the anterior tonsillar pillar than for those in the tonsillar fossa. Gehanno et al.⁴² reported on salvage surgery for 50 patients with tumors of the tonsillar region. The actuarial survival rates at 3 and 5 years after salvage surgery were 38% and 24%, respectively. Compared with primary surgery, a higher postoperative mortality (8% versus 1.4%) also was seen. Tumor extension that required resection into the tongue base was found to be a negative prognostic factor; survival declined dramatically in such cases.

Peiffert et al.¹⁶⁹ reported on using brachytherapy salvage (6000 cGy) in 73 patients who presented with velotonsillar SCC in a previously irradiated area. The 5-year actuarial LC rates for T1N0 and T2N0 disease were 80% and 67%, respectively. The regional relapse rate was 10% in both groups. Grade 2 complications occurred in 13% of patients, and these were not related to the volume treated or to the dose rate. No grade 3 or 4 complications occurred. The 5-year specific survival rate is 64%. Of note, 42% of the patients in this series died from another carcinoma.

Puthawala et al.¹⁶⁷ using implantation alone, obtained a 75% LC rate in patients with recurrent disease, with a 2-year absolute DFS rate of 42%. Treatment-related complications such as soft tissue necrosis or osteoradionecrosis occurred in 23% of these patients who had received previous RT.

Surgery and Postoperative Adjuvant Radiotherapy

Recognizing that advanced carcinomas treated surgically will need postoperative RT, Zelefsky et al.,⁷² reported the long-term treatment results for base-of-tongue and tonsillar fossa carcinomas treated at the Memorial Sloan-Kettering Cancer Center. Between 1973 and 1986, 51 patients were treated with surgery plus RT. Indications included advanced disease (stage T3 or T4, 66%), close or positive margins (64%), and multiple positive neck nodes (84%). The 7-year actuarial LC rates are impressive: 81% for carcinomas of the base of tongue. LC was achieved in 94% and 75% of patients with T3 and T4 lesions, respectively. For patients with positive or close margins who received postoperative doses of 6000 cGy or more, the long-term control rate was 93%. The authors also examined the influence of treatment interruptions. The actuarial control rate among patients who required a treatment break was 64%, in contrast to those who did not require interruption of their treatment, in whom the actuarial control rate was 93% (P = .05). At 7 years, the OS for all patients was 52%, and the DFS was 64%. For all patients, the likelihood of having distant metastasis at 7 years was 30%.

Two randomized trials have examined preoperative versus postoperative RT, and the results appear to be better with postoperative therapy. Although no trials address tumors of the oropharynx specifically, it is reasonable to infer from data on other head and neck sites. The Radiation Therapy Oncology Group (RTOG) trial 73-03 randomized patients with advanced operable tumors of the supraglottic larynx or hypopharynx to 5000 cGy before surgery or 6000 cGy after surgery, and patients with tumors of the oral cavity or oropharynx were assigned to preoperative, postoperative, or definitive RT.⁸⁴ The LRC rate at 4 years was superior in the postoperative groups (58% versus 48%, respectively; P = .04). The Gustav-Roussy trial¹⁷⁹ randomized patients with primary tumors of the hypopharynx to preoperative or postoperative RT. A statistically significant difference (P < .01) in survival rates, complications, and QOL existed in favor of postoperative RT. The postoperative RT group had a superior 5-year survival rate of 56% compared with 20% in the preoperative group.¹⁷⁹ In the modern era, preoperative radiation no longer is used for lesions that are resectable and for which resection will be part of the planned management strategy.

In an attempt to reduce the probability of tumor repopulation during a long course of treatment, investigators have used the concomitant radiation boost schedule in a postoperative setting. The regimen is characterized by delivering the boost (10 fractions) as second daily treatments at a time when accelerated repopulation is theorized to occur.^76,88 Ang et al. reported a multi-institutional randomized trial prospectively randomizing 213 patients who underwent surgical treatment to adjuvant radiation based on previously defined pathologic risk factors.⁷⁶ Low-risk patients (n = 31) were observed, intermediate-risk patients (n = 31) received 57.6 Gy in 6.5 weeks, and high-risk patients were randomized to 63 Gy in 7 weeks (n = 75) or in 5 weeks (n = 76) using the DCB schedule. LRC was similar between low-risk patients who received no adjuvant radiation and intermediate-risk patients who received conventionally fractionated radiation to 57.6 Gy (5-year LRC 90% versus 94%, respectively). High-risk patients treated with DCB showed a trend toward improved LRC and survival compared with those treated with CF (P = .11, P = .08, respectively). This benefit was postulated to be due to the shortening of treatment duration between surgery and the end of radiation, thereby overcoming tumor repopulation. Among high-risk patients, shorter treatment duration significantly improved LRC (5-year LRC was as follows: 76% for <11 weeks, 62% for 11 to 13 weeks, and 38% for >13 weeks). Moreover, among high-risk patients treated with conventionally fractionated radiation, those treated below the median treatment duration had a higher 5-year LRC (71% versus 50%, P = .03, estimated from graph) and survival (45% versus 12%, P = .01, estimated from graph) compared with those whose treatment duration exceeded the median time. Acute toxicity was increased in the DCB arm compared with conventionally fractionated (confluent mucositis 62% versus 36%) but late grade 3 to 4 chronic toxicity was no different (38% versus 42%, respectively, at 5 years).

Bernier et al., reported results of the European Organization for the Research and Treatment of Cancer 22931 phase III multicenter randomized trial demonstrating the benefit of concurrent chemoradiation compared with conventional radiation as postoperative adjuvant treatment.^180,181 Three hundred thirty-four patients with resected advanced head and neck cancers were randomized to conventional fractionated radiation (2 Gy up to 66 Gy) or to cisplatinum (100 mg/m²/cycle on weeks 1, 4, and 7) concurrent with the same radiation regimen. At a median follow-up of 60 months, patients receiving postoperative chemoradiation had better 5-year PFS (47% versus 36%, P = .04), 5-year OS (53% versus 40%, P = .02), and 5-year local-regional relapse (18% versus 31%, P = .007). Five-year distant metastatic (21% versus 25%) and secondary malignancy rates (12% versus 13%) were not significantly different between the two groups. Patients receiving concurrent chemoradiation had increased grade 3 or higher acute toxicity, the most common being mucositis (41% versus 21%, P = .001). Other severe toxicities in the group receiving chemoradiation included leukopenia (16%), granulocytopenia (13%), nausea (12%) and vomiting (11%). However, the cumulative incidence of late complications was not significantly different between the two groups.

Cooper et al¹⁸² reported findings from a similar trial from the RTOG. RTOG 95-01 randomized 459 patients with high-risk squamous-cell carcinoma of the head and neck after total resection to receive conventional radiotherapy alone (60 to 66 Gy in 30 to 33 fractions) or concurrent chemoradiation consisting of the same radiation treatment with three cycles of cisplatin (100 mg/m² on days 1, 22, and 43). After a median follow-up of 45.9 months, patients in the chemoradiation arm had an improvement in 2-year LRC (82% versus 72%, P = .01) and 2-year DFS (70% versus 60%, P = .04). Two-year OS did not differ significantly between the two groups. The incidence of acute grade 3 toxicity was 77% in the combined-therapy group, compared with 34% in the radiotherapy alone group (P < .001).¹⁸² With longer follow-up, however, the improvements in LRC (and DFS rates were no longer significant. At five years, the LRF rate was 20.4% and 28.7% (P = .083) and the DFS rate was 37.4% versus 29.1% (P = .098).¹⁸³ Nevertheless, based on the earlier results of the trial in conjunction with the findings from EORTC 22931, concomitant chemoradiation has become standard for postoperative patients with high-risk features.