POSTOPERATIVE IRRADIATION FOR SQUAMOUS CELL CARCINOMA OF THE ORAL CAVITY: 35-YEAR EXPERIENCE

Transcription

1 POSTOPERATIVE IRRADIATION FOR SQUAMOUS CELL CARCINOMA OF THE ORAL CAVITY: 35-YEAR EXPERIENCE Russell W. Hinerman, MD, 1 William M. Mendenhall, MD, 1 Christopher G. Morris, MS, 1 Robert J. Amdur, MD, 1 John W. Werning, MD, 2 Douglas B. Villaret, MD 2 1 Department of Radiation Oncology, University of Florida Shands Cancer Center, PO Box , Gainesville, FL hinerrw@shands.ufl.edu 2 Department of Otolaryngology, University of Florida, Gainesville, Florida Accepted 19 April 2004 Published online 30 September 2004 in Wiley InterScience ( DOI: /hed Abstract: Background. The purpose of this study was to analyze factors influencing outcome in patients who received postoperative irradiation for advanced squamous cell carcinoma of the oral cavity. Methods. Between October 1964 and November 2000, 226 patients with 230 previously untreated primary invasive squamous cell carcinomas of the oral cavity were treated postoperatively with continuous-course external beam irradiation. All patients had a minimum follow-up of 2 years (analysis, November 2002). No patient was lost to follow-up. Results. The 5-year actuarial rates of locoregional control by pathologic American Joint Committee on Cancer stage were: stage I, 100%; stage II, 84%; stage III, 78%; and stage IV, 66%. Recurrence of cancer above the clavicles developed in 55 patients (24%). In multivariate analysis of locoregional control, positive margins, vascular invasion, perineural invasion, extracapsular extension, and T classification remained significant. Conclusions. This article provides additional data defining relatively favorable and unfavorable groups of patients in the postoperative setting. Dose recommendations are re-examined and selectively increased for high-risk patients. A 2004 Wiley Periodicals, Inc. Head Neck 26: , 2004 Correspondence to: R. W. Hinerman Presented at the 88th Scientific Assembly and Annual Meeting of the Radiological Society of North America, Chicago, IL, Dec 1 6, Reprints are not available from the authors. B 2004 Wiley Periodicals, Inc. Keywords: carcinoma; squamous cell; mouth; radiotherapy; mouth neoplasms; postoperative care The rationale for postoperative radiotherapy (RT) is that it is most likely to be effective against microscopic deposits of cancer cells, which would progress and lead to locoregional recurrence if left unchecked. 1 In 1989, our results were published on 134 patients who underwent postoperative irradiation for tumors of oral, pharyngeal, or laryngeal sites at the University of Florida. Included were 38 patients with oral cavity cancers. During the period covered in our initial review, the institutional policy was to use primary irradiation for most oral cavity cancers. In the early 1980s, the approach to oral cavity cancer changed at the University of Florida such that almost all patients were treated initially by operation with postoperative irradiation added as indicated by pathologic findings. The main reason for the change was to avoid the relatively high incidence of bone exposure and soft tissue necrosis observed in patients treated predominantly for oral tongue and floor of mouth cancers with interstitial irradiation. This series updates 984 Postoperative Radiotherapy for Squamous Cell Carcinoma

2 Table 1. Patient population (n = 226). Parameter No. patients (%) Sex Male 167 (74) Female 59 (26) Primary site Retromolar trigone 53 (23) Alveolar ridge 27 (12) Floor of mouth 71 (31) Oral tongue 58 (25) Buccal mucosa 4 (2) Gingivobuccal sulcus 9 (4) Hard palate 4 (2) Neck dissection Unilateral 154 (68) Bilateral 57 (25) None 15 (7) Clinical N status N0 124 (55) N1 47 (21) N2 N3 55 (24) Pathologic N status N0 99 (44) N1 42 (19) N2 N3 85 (38) Clinical T classification T1 21 (9) T2 78 (35) T3 48 (21) T4 79 (35) Pathologic T classification T1 18 (8) T2 75 (33) T3 44 (19) T4 89 (39) No. of pathologically positive nodes No neck dissection 15 (7) 0 86 (38) 1 44 (19) (19) z 4 37 (16) Margin status Positive 68 (30) Negative 83 (37) Close/equivocal 75 (33) Extracapsular extension (ECE) ECE+ 43 (19) ECE 42 (19) No data 141 (62) Perineural invasion (PNI) PNI+ 83 (37) PNI 40 (18) No data 103 (46) Clinical AJCC stage I 16 (7) II 47 (21) III 51 (23) IVA 110 (49) IVB 2 (1) Pathologic AJCC stage I 9 (4) II 35 (15) III 39 (17) Table 1. (continued). Parameter No. patients (%) IVA 142 (63) IVB 1 (1) No. of risk factors per patient 0 12 (5) 1 60 (27) 2 52 (23) 3 53 (23) 4 30 (13) 5 14 (6) 6 or 7 5 (2) Dose received (Gy) V (21) > (28) > (25) our experience to include 226 patients with oral cavity malignancies. METHODS AND MATERIALS Between October 1964 and November 2000, 226 patients with 230 previously untreated primary invasive squamous cell carcinomas of the oral cavity (excluding the lip) were treated postoperatively with continuous-course external beam irradiation in the University of Florida Department of Radiation Oncology. All living patients had a minimum follow-up of 2 years (analysis, November 2002). No patient was lost to follow-up. The mean and median follow-up times were 4.8 and 3.2 years, respectively. Patients ranged from 15 to 85 years of age (mean, 60; median, 61). There were 167 men and 59 women. Disease was staged according to the guidelines of the 1997 American Joint Committee on Cancer (AJCC) 2 (Table 1). Stages IVA and IVB were grouped together for statistical analyses, because only two patients were initially seen with clinical stage IVB disease. Four patients had two synchronous oral cavity cancers. In each of these cases, only one TNM stage was assigned, corresponding to the most advanced lesion. Statistical analysis was done for 226 patients rather than 230 separate tumors. All patients underwent a major cancer operation (eg, partial glossectomy, maxillectomy, tongue-jaw-neck dissection) with or without neck dissection before irradiation. Patients referred for irradiation after excisional biopsy of early cancers were excluded. Most patients were treated in the 1980s and 1990s. Before that time, most early and moderately advanced oral cancers received pri- Postoperative Radiotherapy for Squamous Cell Carcinoma 985

3 mary irradiation. Irradiation was started as soon as feasible after surgery, usually within 4 to 6 weeks. No patient received chemotherapy except for palliation after tumor recurrence. Indications for postoperative RT are listed in Table 2. Twelve patients had none of our usual indications for postoperative irradiation but were irradiated because the surgeon believed the patient was at high risk of local recurrence. Margins of resection at the primary site were negative in 83 patients (37%), positive in 68 patients (30%), and close/equivocal in 75 patients (33%) (close [invasive or carcinoma in situ {CIS} V5 mm] in 41 [18%], initially positive but ultimately negative in 26 [12%], contained CIS or dysplasia in seven [3%], and uncertain in one [0.4%]). The latter grouping into the close/equivocal category was done because of the small number of patients in each category and because no significant difference was seen in local or locoregional control rates between the different subgroups. All patients were treated with megavoltage photon beams, usually 60 Co or 6 MV. The mean doses according to margin status in 218 patients who were treated by external irradiation alone (eight additional patients received an interstitial implant) were as follows: negative margins, 64.8 Gy; positive margins, 72 Gy; and close/ equivocal, 66 Gy. No patient in this series was treated with intensity-modulated radiation therapy (IMRT). All patients were treated with planned continuous-course external beam RT. Sixty-nine patients received twice-a-day irradiation at 1.2 Gy per fraction to a median dose of 74.4 Gy (range, 64 Gy 76.8 Gy). One hundred forty-nine patients were treated once daily to a median dose of 65 Gy (range, 21.6 Gy 70.5 Gy). Eight additional patients received an interstitial implant after external beam irradiation. The posterior cervical and low-neck regions (unilateral or bilateral) were irradiated in most patients. All scars, needle tracks, and drain sites Table 2. Indications for postoperative radiation.. T4 (Bone invasion/soft tissue of neck). Multiple positive nodes. Extracapsular spread. Perineural invasion. Vascular invasion. Margins, close (V 5 mm), positive, CIS/dysplasia, or initially positive but negative after re-excision Abbreviation: CIS, carcinoma in situ. were bolused during irradiation with 0.5-cm thick petrolatum gauze. Statistical Methods. All statistical analyses were performed with SAS software (SAS Institute, Cary, NC). 3 Univariate survival estimates were obtained by the Kaplan-Meier method. 4 The logrank statistic was used to detect differences between strata, and corresponding two-sided p values of less than.05 were interpreted as statistically significant. Cox regression was implemented for multivariate analyses. 5 Specifically, backward selection was performed to attain the most parsimonious model for each endpoint of interest. Multiple imputation was implemented to allow the most unbiased estimates in the presence of explanatory variables with randomly missing information. Those patients not receiving a neck dissection were excluded from multivariate analyses. The presence of severe complications was analyzed with logistic regression; those patients receiving an implant rather than external beam RT were excluded, because hyperfractionation was one of the covariates in this analysis. RESULTS Local (Primary) Control. The 5-year actuarial rate of primary site control was 83%. There was no significant difference ( p =.16) in the 5-year local control rates according to tumor location within the oral cavity: oral tongue, 84%; retromolar trigone, 76%; floor of mouth, 85%; and other, 70%. There was a significant difference in local control according to margin status ( p =.0004): positive, 63% (n = 68); negative, 82%; and close/ equivocal, 92%. The difference in local control rates according to preoperative clinical T classification was also significant ( p =.04): T1, 90%; T2, 88%; T3, 79%; and T4, 70%. Local control rates did not correlate with clinical N status or with the number of lymph nodes involved in the pathologic neck dissection specimen. Five-year local control rates according to clinical N status were as follows ( p =.96): N0, 79%; N1, 78%; and N2 N3, 84%. Local control according to the number of pathologically positive lymph nodes were as follows: none, 81%; one, 82%; two to three, 86%; four or more, 78% ( p =.73). Regional (Neck) Control. The 5-year rates of neck control by the number of positive lymph nodes at 986 Postoperative Radiotherapy for Squamous Cell Carcinoma

4 neck dissection were as follows: none, 94%; one, 93%; two to three, 84%; four or more, 76% ( p =.02). There was a significantly reduced probability of neck control in patients with extracapsular extension (ECE) compared with those without ECE (the only patients included were those whose pathology report specifically commented on the presence or absence of ECE): 78% versus 94%, respectively ( p =.02). Locoregional Control According to Primary Site. Five-year actuarial locoregional control rates according to primary site were as follows: oral tongue, 72%; floor of mouth, 78%; retromolar trigone, 75%; and other sites, 61% ( p =.23). Time to Locoregional Failure. Locoregional failure developed in 55 patients (24%). Fifty (91%) of 55 recurrences were detected within 2 years of the date of surgery. The median time to recurrence was 5.4 months. Locoregional Control According to AJCC Stage. Five-year actuarial locoregional control rates according to clinical AJCC stage were as follows: stage I, 87%; stage II, 74%; stage III, 68%; and stage IV, 72% (p =.6). Five-year actuarial locoregional control rates according to pathologic AJCC stage are shown in Figure 1. Locoregional Control According to Margin Status. Five-year actuarial locoregional control rates according to margin status were as follows: negative margins, 77%; positive margins, 52%; and close/ equivocal, 86% ( p <.0001). Locoregional control results for patients with positive margins who received twice-a-day irradiation to 74.4 Gy or more were higher than for FIGURE 2. Locoregional control for positive margins versus dose. those treated once a day and/or to less than 74.4 Gy (Figure 2). Locoregional Control According to Other Pathologic Variables. Five-year actuarial locoregional control rates were significantly worse in patients with ECE, perineural invasion, or pathologic T3 or T4 tumors. There were no significant differences in 5-year locoregional control rates in patients with vascular invasion, bone invasion, or poor differentiation (Table 3). In addition, the number of lymph nodes involved did not significantly affect 5-year locoregional control rates ( p =.30): none, 78%; one, 75%; two to three, 74%; and four or more, 65%. Disease in 10 of the 12 patients with none of the usual indications for postoperative radiation was locoregionally controlled. Locoregional Control According to Number of Indications for Irradiation. Five-year actuarial locoregional control rates were significantly worse in patients with three or more indications for ad- Table 3. Univariate analysis of 5-year locoregional control. Risk factor % of locoregional control with listed factor (no. patients) % of locoregional control without listed factor (no. patients) p value T3/T4 primary* 63 (133) 86 (93).0003 Positive margins 52 (68) 77 (83).0018 Perineural invasion 62 (83) 88 (40).003 ECE 61 (43) 87 (42).007 z 3 risk factors 63 (49) 79 (177).007 Bone invasion 67 (80) 76 (146).24 Vascular invasion 70 (49) 76 (46).40 Poor differentiation 76 (47) 72 (179).57 FIGURE 1. Locoregional control versus pathologic AJCC stage. Abbreviation: ECE, extracapsular nodal spread. *pt1, 88%; pt2, 86%; pt3, 64%; pt4, 63%. Postoperative Radiotherapy for Squamous Cell Carcinoma 987

5 ministering postoperative irradiation (Figure 3). A steady decline was observed in the rates of locoregional control with an increase in the number of indications for administering postoperative irradiation (Table 4). On the basis of these findings, patients were divided into relatively favorable and unfavorable groups. Patients in the favorable group had fewer than three indications to receive irradiation; those in the unfavorable group had three or more indications. Multivariate Analysis of Pathologic Variables Affecting Locoregional Control. A multivariate analysis of pathologic factors that might influence locoregional control is shown in Table 5. T classification, margin status, and the presence of ECE, vascular or perineural invasion were significant. Time Analyses. Treatment time was analyzed in three ways to determine its influence on locoregional control. First, the interval (in days) between surgery and the beginning of irradiation was analyzed (Table 6). Data were analyzed by four time periods, which were chosen for statistical reasons to give near equal numbers of patients in each group. Patients were separated into low (< three indications for RT) and high (z three indications for RT) risk groups. There was a trend toward higher failure rates in highrisk patients whose irradiation was initiated more than 51 days after surgery (Figure 4). No trend was seen in low-risk patients (0 35 days, 23%; days, 4%; days, 30%; and >51 days, 17%). Second, the length of the overall treatment package, calculated from the date of surgery until the last day of irradiation, was analyzed for its prognostic impact (Table 6). In high-risk patients, there was a trend toward higher failure rates when the overall treatment time exceeded 101 days (Figure 5). For low-risk patients, the failure rates were not significantly Table 4. Locoregional control by indications for postoperative radiation therapy. No. of indications No. of recurrences/no. of patients % locoregional control 0 2/ / / / / / /5 40 different according to length of the overall treatment package (0 80 days, 19%; days, 16%; days, 28%; and >101 days, 17%). The length of the radiation course (in days) was analyzed for its potential impact on locoregional control. No trend was identified for either lowrisk or high-risk patient groups. Dose Analysis. Five-year locoregional control rates for patients who received 60 Gy or less (74%), > 60 to 65 Gy (84%), > 65 to 70 Gy (69%), or more than 70 Gy (63%) were similar because of patient selection ( p =.07). Patients with positive margins received higher mean doses (72 Gy) than did those with negative margins (64.8 Gy), yet still had lower locoregional control rates (63% vs 82%) at 5 years. Conversely, patients with close/ equivocal margins had the highest control rate (92%), their less favorable prognosis being offset by higher mean doses (66 Gy) ( p =.0004). Salvage of Locoregional Recurrence. Of 55 patients with locoregional failures, surgical salvage with or without additional irradiation was attempted in 17 (31%). Nine of 17 patients (53%) Table 5. Multivariate analysis of pathologic variables affecting locoregional control. FIGURE 3. Locoregional control versus number of risk factors. Risk factor p value T classification.0002 ECE.0037 Margins.0102 Perineural invasion.0103 Vascular invasion.0258 N status.1108 Overall stage.1541 Bone invasion.3045 Differentiation.5916 Abbreviation: ECE, extracapsular nodal spread. 988 Postoperative Radiotherapy for Squamous Cell Carcinoma

6 Table 6. Locoregional failure by surgery radiation therapy interval and overall treatment time. By no. of risk factors No. of failures/total no. Overall of patients (% failure) No. of failures/total no. patients (% failure) 0 2 z 3 Surgery RT interval in days /56 (25) 8/35 (23) 6/21 (29) /57 (16) 1/23 (4) 8/34 (24) /57 (28) 9/30 (30) 7/27 (26) >51 16/56 (29) 6/36 (17) 10/20 (50) Overall treatment time in days /60 (23) 6/32 (19) 8/28 (29) /56 (20) 5/32 (16) 6/24 (25) /54 (26) 7/25 (28) 7/29 (24) z101 16/56 (29) 6/35 (17) 10/21 (48) Abbreviation: RT, radiation therapy. were successfully salvaged, for an overall salvage rate of 16% (nine of 55 patients). Distant Metastasis. Five-year actuarial distant metastasis-free survival by pathologic AJCC stage was as follows: stage I, 100%; stage II, 97%; stage III, 90%; and stage IV, 81% (p =.07). In patients with positive nodes, those with ECE had significantly lower distant metastasis free survival rates (64%) than did those without ECE (95%, p =.001). Survival. Actuarial 5-year absolute and causespecific survival rates for the entire group were 47% and 67%, respectively. Absolute 5-year survival rates by pathologic AJCC stage were as follows: stage I, 63%; stage II, FIGURE 5. Locoregional control versus overall treatment time (patients with z three risk factors). 70%; stage III, 48%; and stage IV, 40% ( p =.03). Node-positive patients with ECE had a significantly lower absolute survival rate (21%) than did those without ECE (58%; p =.0003). Cause-specific survival rates at 5 years by clinical AJCC stage were as follows: stage I, 81%; stage II, 77%; stage III, 62%; and stage IV, 63% (p =.30). Cause-specific survival rates at 5 years by pathologic AJCC stage are shown in Figure 6. Node-positive patients with ECE had a significantly lower 5-year cause-specific survival (38%) than those without ECE had (81%, p <.0001). There was a significant improvement in 5-year cause-specific survival for patients without perineural invasion (85%) compared with those in whom it was present (56%, p =.002). A multivariate analysis of pathologic factors that might influence cause-specific survival is shown in Table 7. The presence of ECE or perineural invasion was significant. Postirradiation, Late Complications. Complications of treatment were graded in severity as mild, moderate, or severe. Mild complications were those that healed spontaneously in 1 to 5 months and were most often small (V0.5 cm) FIGURE 4. Locoregional control versus surgery radiation therapy interval (patients with z three risk factors). FIGURE 6. Cause-specific survival versus pathologic AJCC stage. Postoperative Radiotherapy for Squamous Cell Carcinoma 989

7 Table 7. Multivariate analysis of pathologic variables affecting cause-specific survival. Risk factor p value ECE.0001 Perineural invasion.0205 T classification.0612 Vascular invasion.2040 Differentiation.2599 Bone invasion.5086 Overall stage.4521 N status.6982 Margins.8837 Abbreviation: ECE, extracapsular nodal spread. areas of soft tissue necrosis or bone exposure. Moderate complications were of longer duration and produced more disability, yet resolved after conservation measures to promote healing such as discontinuation of denture use, oral rinses, antibiotics, and, in some instances, hyperbaric oxygen, without requiring surgery. Severe complications required hospitalization, surgery, or both. Patients requiring a permanent gastrostomy tube were also coded as having a severe complication. There were 58 mild and moderate complications (26%): 22 bone exposures, 21 soft tissue necroses, one small wound dehiscence, seven fistulas, and seven wound infections. There were 33 severe injuries (15%): 14 osteoradionecrosis, seven permanent gastrostomies, four severe dehiscences, three fistulas, two soft tissue necrosis, one wound infection, one mandibular fracture, and one plate extrusion. The median dose was 66 Gy in patients with severe complications. Logistics regression showed no correlation with respect to total dose (p = 0.81) or fractionation ( p = 0.25) (twice daily vs once daily) in patients experiencing severe complications. DISCUSSION The usual indications to give postoperative irradiation at the University of Florida for carcinoma of the oral cavity include one or more of the following pathologic findings related to the primary site: bone invasion or extension of tumor into the soft tissues of the neck (T4 primary); margins that are positive, close (V5 mm), contain CIS or dysplasia, or that were initially positive (but ultimately negative after re-resection 6 9 ;) perineural invasion; vascular space invasion; 13 or multicentricity (scattered, discontinuous foci or islands of tumor adjacent to the primary site, making margins of resection uncertain). Factors in the neck specimen for which postoperative irradiation is generally added include multiple positive lymph nodes or ECE. 13,16,18 21 High risk of occult tumor in an undissected, clinically negative N0 neck is not an adequate reason to add adjuvant RT unless additional indications for treatment are present. The high-risk neck should be electively dissected. Some of these factors are more ominous than others. For example, early bone erosion of the mandible is not likely to have the same negative prognostic influence as positive margins. Factors thought to be particularly ominous are the presence of positive margins, 7,22,23 multiple indications for administering postoperative irradiation, 22,24,25 and/or ECE. 17,19 21,23,25 27 Although we have never used poor differentiation as the sole indication for postoperative irradiation, there are data to suggest it may significantly affect locoregional control, 13,17,24 and in otherwise borderline cases, poor differentiation might be a reason to add irradiation postoperatively. This series shows statistically significant correlations on univariate analysis between locoregional control and T classification (T1 and T2 vs T3 and T4), pathologic AJCC stage, ECE, perineural invasion, margin status, and the number of indications for postoperative radiation (Table 3). Multivariate analysis shows T classification, margin status, and the presence of ECE, vascular or perineural invasion to be independent predictors of locoregional control (Table 5). Does Postoperative Radiotherapy Improve Locoregional Control and Survival? No randomized trials have addressed the efficacy of postoperative adjuvant RT. Some of the best available data are from the Medical College of Virginia, where two groups of surgeons operated on patients with head and neck cancer: general surgical oncologists who used surgery alone and reserved RT for treatment of recurrent disease, and otolaryngologists who routinely sent patients with locally advanced disease for postoperative RT. 23 One hundred twenty-five of 441 patients treated surgically between 1982 and 1988 had ECE and/ or positive margins; 71 were treated with surgery alone, and 54 received postoperative RT. Patients were irradiated at 1.8 to 2.0 Gy per fraction with 60 Co or 4-MV x-rays to doses of 50 to Gy in 26 patients and to 60 Gy or more in the remainder. Local control rates at 3 years after surgery alone compared with surgery and RT were as 990 Postoperative Radiotherapy for Squamous Cell Carcinoma

8 follows: ECE, 31% and 66% ( p =.03); positive margins, 41% and 49% ( p =.04); and ECE and positive margins, 0% and 68% ( p =.001). A multivariate analysis of local control was performed evaluating the impact of T classification, N status, use of postoperative RT, the number of positive nodes, the number of nodes with ECE, primary site, microscopic and macroscopic ECE, and margin status. For the local control endpoint, use of postoperative RT ( p =.0001), macroscopic ECE ( p =.0001), and margin status ( p =.09) were of independent significance. Disease-free survival at 3 years was 25% after surgery alone and 45% after combined-modality treatment ( p =.0001). Cause-specific survival rates at 3 years were 41% for surgery alone and 72% for surgery and postoperative RT ( p =.0003). Multivariate analysis of cause-specific survival showed postoperative RT ( p =.0001) and the number of nodes with ECE (p =.0001) significantly influenced this endpoint. Two irradiated patients experienced mandibular necrosis; one was treated with hyperbaric oxygen treatments and the other with conservative management. Lundahl et al 28 reported a series of 95 patients with node-positive squamous cell carcinoma who were treated with a neck dissection and postoperative RT at the Mayo Clinic. A matched-pair analysis was performed with a series of patients treated with surgery alone; 56 matched pairs of patients were identified. The rates of recurrence in the dissected neck (relative risk [RR] = 5.82; p =.0002), recurrence in either side of the neck (RR = 2.21; p =.0052), and death from any cause (RR = 1.67; p =.0182) were significantly higher for patients treated with neck dissection alone. Thus, it seems for patients who are at high risk for locoregional failure after surgery, postoperative RT significantly improves disease control above the clavicles as well as survival. Factors Affecting Locoregional Control. Data from the University of Florida, 22 the Radiation Therapy Oncology Group (RTOG) trial, 29 and The University of Texas M. D. Anderson Cancer 25,30 33 Center all suggest locoregional control rates after postoperative irradiation are lower for oral cavity primary tumors than for other head and neck primary sites. Zelefsky et al 34 noted that among patients with oral cavity primary tumors, those with oral tongue lesions fared the worst (p =.03). In this series, there were no significant differences according to primary site within the oral cavity. Another treatment-related issue that may influence the likelihood of cure is beam energy. Data exist in the literature suggesting better control in the neck with 60 Co and 4-MV than with 6-MV photons, although local and locoregional control is just as good or better with 6-MV beams. 35,36 Thus, most patients can be adequately treated with 6-MV photons. A beam spoiler may be added to increase the surface dose for patients believed to be at particularly high risk for tumor in the subcutaneous tissues. Salvage of Patients with Locoregional Recurrence. In this series, 55 patients (25%) who were treated had recurrence above the clavicles develop. Seventeen underwent salvage surgery with or without additional RT. Nine of 17 (53%) were successfully salvaged, for an overall salvage rate of 16% (nine of 55). Fifty-eight of 240 patients (24%) treated with surgery and postoperative RT at M. D. Anderson Cancer Center had a locoregional recurrence develop. 25 Thirty-six patients underwent salvage therapy with surgery alone or combined with chemotherapy (11 patients), RT alone or combined with chemotherapy (three patients), or chemotherapy alone (22 patients). Only one of 58 patients (2%) was a long-term survivor 22 months after salvage therapy. Analysis of Time and Dose Factors. Intuitively, it seems residual microscopic nests of tumor would be able to regrow in proportion to the length of time between surgery and the beginning of RT. However, the available data are mixed on this issue. 22,37 40 This study demonstrates a decrease in locoregional control for patients with three or more risk factors with surgery RT intervals greater than 51 days (Figure 4). Several authors have also suggested a correlation between tumor control and the length of RT, the latter being altered by the dose per fraction and/or number of fractions per day. 22,25,40 42 No such trend was detected in this series. These two indices (ie, the surgery RT interval and the length of the radiotherapy course) combine to give the overall treatment time, defined as the time from surgery to the last day of irradiation. Data exist in the literature correlating overall treatment time with local control, at least for high-risk patients. 40,43 This study demonstrates a decrease in locoregional control for patients having three or more risk factors with overall treatment times greater Postoperative Radiotherapy for Squamous Cell Carcinoma 991

9 than 101 days (Figure 5), although the difference is not statistically significant. The previously mentioned data and common sense indicate irradiation should begin as soon as healing is complete; it can often be started within 3 to 4 weeks of surgery. If irradiation is delayed until gross recurrence has occurred, the chance for successful salvage by irradiation is only 5% to 10%. 7,44,45 If healing is not complete by 6 weeks, data from earlier University of Florida series 43,46 show irradiation can often be safely initiated, and approximately two thirds of patients will heal spontaneously during or after irradiation. Patients at high risk (z three risk factors) for recurrence and/or those whose surgery RT interval has been prolonged (>7 weeks) should be considered for altered dose/fractionation schemes that shorten the length of the RT course, thereby decreasing the overall treatment time. Adjuvant Chemotherapy and Future Directions. The issue of whether chemotherapy is beneficial concomitantly with postoperative irradiation for head and neck cancer was recently resolved with the publication of two randomized trials (RTOG 9501 and EORTC 22391). 47,48 Each of these showed an improvement in locoregional control and disease-free survival with cisplatin (100 mg/m 2 ) given on days 1, 22, and 43 of the radiotherapy regimen. 47,48 Prior to that, two publications showed a similar benefit using concomitant chemoradiation, although an earlier review of the RTOG 9501 data that appeared in abstract form was inconclusive Severe acute effects are seen more frequently with concomitant treatment compared with postoperative radiation alone. Although none of the patients in this series received chemotherapy, our current policy is to consider concomitant cisplatin in selected highrisk cases. IMRT has been used postoperatively at other institutions We currently use IMRT mainly to treat de novo head and neck tumors definitively with irradiation. It is used sparingly in the postoperative setting, mainly because of concerns about the dose distribution to skin and subcutaneous tissues, which may be at more risk after surgery. As the role of IMRT evolves, its use may increase in postoperative situations. We continue to favor more conventional beam arrangements when feasible, to optimize dose distributions to the skin and subcutaneous tissues at risk. The data presented herein suggest a strikingly negative effect on locoregional control when perineural spread, ECE, and/or positive margins are present. More than one third of patients continue to have treatment failure above the clavicles when any of these three factors exist. An improvement in locoregional control was seen in patients with positive margins treated twice daily to doses of 74.4 Gy or higher. We will likely consider the same aggressive treatment for patients with oral cavity lesions with perineural and/or ECE, as well. Conversely, consideration should be given to re-resection in cases in which the final margins contain invasive cancer if the site of positivity can be accurately determined and another operation performed without adding excessively to the morbidity. Our current recommendation for patients with oral cavity primary tumors who require postoperative irradiation is to deliver 64 to 66 Gy at 2 Gy per fraction for relatively favorable situations and 74.4 to 76.8 Gy in 62 to 64 fractions, twice a day, for unfavorable scenarios. Concomitant cisplatin should be considered in high-risk cases. We recognize that these dosefractionation recommendations for high-risk patients are not strictly supported by our data or by randomized data published elsewhere in the literature. 25,40 However, we believe these guidelines are justified in view of the unsatisfactory rates of locoregional control seen in this unfavorable subset of patients. REFERENCES 1. MacComb WS, Fletcher GH. Planned combination of surgery and radiation in treatment of advanced primary head and neck cancers. Am J Roentgenol Radium Ther Nucl Med 1957;77: American Joint Committee on Cancer, American Cancer Society, American College of Surgeons. Fleming ID, et al, editors. AJCC cancer staging handbook from the AJCC Cancer Staging Manual, ed 5. Philadelphia: Lippincott- Raven, SAS Institute Inc. SAS OnlineDoc, Version 8. Cary, NC: SAS Institute Inc., Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958;53: Cox DR. Regression models and life tables series B (methodological). J R Stat Soc 1972;34: Chen TY, Emrich LJ, Driscoll DL. The clinical significance of pathological findings in surgically resected margins of the primary tumor in head and neck carcinoma. Int J Radiat Oncol Biol Phys 1987;13: Huang D, Johnson CR, Schmidt-Ullrich RK, Sismanis A, Neifeld JP, Weber J. Incompletely resected advanced squamous cell carcinoma of the head and neck: the effectiveness of adjuvant vs. salvage radiotherapy. Radiother Oncol 1992;24: Looser KG, Shah JP, Strong EW. The significance of positive margins in surgically resected epidermoid carcinomas. Head Neck Surg 1978;1: Scholl P, Byers RM, Batsakis JG, Wolf P, Santini H. Microscopic cut-through of cancer in the surgical treat- 992 Postoperative Radiotherapy for Squamous Cell Carcinoma

10 ment of squamous carcinoma of the tongue. Prognostic and therapeutic implications. Am J Surg 1986;152: Lydiatt DD, Robbins KT, Byers RM, Wolf PF. Treatment of stage I and II oral tongue cancer. Head Neck 1993;15: O Brien CJ, Lahr CJ, Soong SJ, et al. Surgical treatment of early-stage carcinoma of the oral tongue would adjuvant treatment be beneficial? Head Neck Surg 1986; 8: Soo KC, Carter RL, O Brien CJ, Barr L, Bliss JM, Shaw HJ. Prognostic implications of perineural spread in squamous carcinomas of the head and neck. Laryngoscope 1986;96: Close LG, Brown PM, Vutich MF, Reisch J, Schaefer SD. Microvascular invasion and survival in cancer of the oral cavity and oropharynx. Arch Otolaryngol Head Neck Surg 1989;115: Farr HW, Arthur K. Epidermoid carcinoma of the mouth and pharynx J Laryngol Otol 1972; 86: Francheschi D, Gupta R, Spiro RH, Shah JP. Improved survival in the treatment of squamous carcinoma of the oral tongue. Am J Surg 1993;166: O Brien CJ, Smith JW, Soong SJ, Urist MM, Maddox WA. Neck dissection with and without radiotherapy: prognostic factors, patterns of recurrence, and survival. Am J Surg 1986;152: Olsen KD, Caruso M, Foote RL, et al. Primary head and neck cancer. Histopathologic predictors of recurrence after neck dissection in patients with lymph node involvement. Arch Otolaryngol Head Neck Surg 1994;120: Cachin Y, Eschwege F. Combination of radiotherapy and surgery in the treatment of head and neck cancers. Cancer Treat Rev 1975;2: Johnson JT, Barnes EL, Myers EN, Schramm VL Jr, Borochovitz D, Sigler BA. The extracapsular spread of tumors in cervical node metastasis. Otolaryngology 1981;107: Johnson JT, Myers EN, Bedetti CD, Barnes EL, Schramm VL, Thearle PB. Cervical lymph node metastases. Incidence and implications of extracapsular carcinoma. Arch Otolaryngol 1985;111: Synderman NL, Johnson JT, Schramm VL, Myers EN, Bedetti CD, Thearle PB. Extracapsular spread of carcinoma in cervical lymph nodes. Impact upon survival in patients with carcinoma of the supraglottic larynx. Cancer 1985;56: Amdur RJ, Parsons JT, Mendenhall WM, Million RR, Stringer SP, Cassisi NJ. Postoperative irradiation for squamous cell carcinoma of the head and neck: an analysis of treatment results and complications. Int J Radiat Oncol Biol Phys 1989;16: Huang DT, Johnson CR, Schmidt-Ullrich R, Grimes M. Postoperative radiotherapy in head and neck carcinoma with extracapsular lymph node extension and/or positive resection margins: a comparative study. Int J Radiat Oncol Biol Phys 1992;23: Foote RL, Buskirk SJ, Stanley RJ, et al. Patterns of failure after total laryngectomy for glottic carcinoma. Cancer 1989;64: Peters LJ, Goepfert H, Ang KK, et al. Evaluation of the dose for postoperative radiation therapy of head and neck cancer: first report of a prospective randomized trial. Int J Radiat Oncol Biol Phys 1993;26: Arriagada R, Eschwege F, Cachin Y, Richard JM. The value of combining radiotherapy alone with surgery in the treatment of hypopharyngeal and laryngeal cancers. Cancer 1983;51: Bartelink H, Breur K, Hart G, Annyas B, van Slooten E, Snow G. The value of postoperative radiotherapy as an adjuvant to radical neck dissection. Cancer 1983;52: Lundahl RE, Foote RL, Bonner JA, et al. Combined neck dissection and postoperative radiation therapy in the management of the high-risk neck: a matched-pair analysis. Int J Radiat Oncol Biol Phys 1998;40: Kramer S, Gelber RD, Snow JB, et al. Combined radiation therapy and surgery in the management of advanced head and neck cancer: final report of study of the Radiation Therapy Oncology Group. Head Neck Surg 1987; 10: El Badawi SA, Goepfert H, Fletcher GH, Herson J, Oswald MJ. Squamous cell carcinoma of the pyriform sinus. Laryngoscope 1982;92: Feldman M, Fletcher GH. Analysis for the parameters relating to failures above the clavicles in patients treated by postoperative irradiation for squamous cell carcinoma of the oral cavity or oropharynx. Int J Radiat Oncol Biol Phys 1982;8: Goepfert H, Jesse RH, Fletcher GH, Hamberger A. Optimal treatment for the technically resectable squamous cell carcinoma of the supraglottic larynx. Laryngoscope 1975;85: Hamberger AD, Fletcher GH, Guillamondegui OM, Byers RM. Advanced squamous cell carcinoma of the oral cavity and oropharynx treated with irradiation and surgery. Radiology 1976;119: Zelefsky MJ, Harrison LB, Fass DE, et al. Postoperative radiotherapy for oral cavity cancers: impact of anatomic subsite on treatment outcome. Head Neck 1990; 12: Aref A, Berkey BA, Schwade JG, et al. The influence of beam energy on the outcome of postoperative radiotherapy in head and neck cancer patients: secondary analysis of RTOG Int J Radiat Oncol Biol Phys 2000;47: Fortin A, Allard J, Albert M, Roy J. Outcome of patients treated with cobalt and 6 MV in head and neck cancers. Head Neck 2001;23: Vikram B. Importance of the time interval between surgery and postoperative radiation therapy in the combined management of head and neck cancer. Int J Radiat Oncol Biol Phys 1979;5: Schiff PB, Harrison LB, Strong EW, et al. Impact of the time interval between surgery and postoperative radiation therapy on locoregional control in advanced head and neck cancer. J Surg Oncol 1990;43: Bastit L, Blot E, Debourdeau P, Menard J, Bastit P, Le Fur R. Influence of the delay of adjuvant postoperative radiation therapy on relapse and survival in oropharyngeal and hypopharyngeal cancers. Int J Radiat Oncol Biol Phys 2001;49: Ang KK, Trotti A, Brown BW, et al. Randomized trial addressing risk features and time factors of surgery plus radiotherapy in advanced head-and-neck cancer. Int J Radiat Oncol Biol Phys 2001;51: Zelefsky MJ, Harrison LB, Fass DE, et al. Postoperative radiation therapy for squamous cell carcinomas of the oral cavity and oropharynx: Impact of therapy on patients with positive surgical margins. Int J Radiat Oncol Biol Phys 1992;25: Pfreundner L, Willner J, Marx A, Hoppe F, Beckmann G, Flentje M. The influence of the radicality of resection and dose of postoperative radiation therapy on local control and survival in carcinomas of the upper aerodigestive tract. Int J Radiat Oncol Biol Phys 2000;47: Parsons JT, Mendenhall WM, Stringer SP, Cassisi NJ, Million RR. An analysis of factors influencing the outcome of postoperative irradiation for squamous cell carcinoma of the oral cavity. Int J Radiat Oncol Biol Phys 1997;39: Ampil RL. Surgery and postoperative external irradiation for head and neck cancer recurrent after surgery alone. Laryngoscope 1998;98: Postoperative Radiotherapy for Squamous Cell Carcinoma 993

11 45. Fletcher GH, Evers WT. Radiotherapeutic management of surgical recurrences and postoperative residuals in tumors of the head and neck. Radiology 1970;95: Isaacs JH Jr, Stiles WA, Cassisi NJ, Million RR, Parsons JT. Postoperative radiation of open head and neck wounds updated. Head Neck 1997;19: Bernier J, Domenge C, Ozsahin M, et al. Postoperative irradiation with or without concomitant chemotherapy for locally advanced head and neck cancer. N Engl J Med 2004;350: Cooper JS, Pajak TF, Forastiere AA, et al. Postoperative concurrent radiotherapy and chemotherapy for high-risk squamous cell carcinoma of the head and neck. N Engl J Med 2004;350: Bachaud J-M, Cohen-Jonathan E, Alzieu C, et al. Combined postoperative radiotherapy and weekly cisplatin infusion for locally advanced head and neck carcinoma: Final report of a randomized trial. Int J Radiat Oncol Biol Phys 1996;36: Bernier J, Domenge C, Eschwege F, Ozsahin M, Matuszewska K, Moncho V, Greiner RH, Giralt J, Kirkpatrick A, van Glabbeke M. Chemo-radiotherapy, as compared to radiotherapy alone, significantly increases disease-free and overall survival in head and neck cancer patients after surgery: Results of EORTC phase III trial [Abstr.] Int J Radiat Oncol Biol Phys 2001;51(3 Suppl. 1): Cooper JS, Pajak TF, Forastiere AA, Jacobs J, Saxman SB, Kish JA, Kim HE, Cmelak AJ, Rotman M, Machtay M, Ensley JF, Chao KS, et al. Postoperative concurrent radiochemotherapy in high-risk SCCA of the head and neck: Initial report of RTOG 9501/intergroup phase III trial. [Abstr.] Presented at the 38 th Annual Meeting of the Am Soc Clin Oncol 2002, Orlando, FL. 52. Chao KS, Ozyigit G, Tran BN, et al. Patterns of failure in patients receiving definitive and postoperative IMRT for head and neck cancer. Int J Radiat Oncol Biol Phys 2003;55: Chao KS, Wippold FJ, Ozyigit G, et al. Determination and delineation of nodal target volumes for head and neck cancer based on patterns of failure in patients receiving definitive and postoperative IMRT. Int J Radiat Oncol Biol Phys 2002;53: Chao KS, Majhail N, Huang CJ, et al. Intensity-modulated radiation therapy reduces late salivary toxicity without compromising tumor control in patients with oropharyngeal carcinoma: A comparison with conventional techniques. Radiother Oncol 2001;61: Dawson LA, Anzai Y, Marsh L, et al. Patterns of localregional recurrence following parotid-sparing conformal and segmental intensity-modulated radiotherapy for head and neck cancer. Int J Radiat Oncol Biol Phys 2000;46: Claus F, Vakaet L, De Gersem W, et al. Postoperative radiotherapy of paranasal sinus tumours: A challenge for intensity-modulated radiotherapy. Acta Otorhinolaryngol Belg 1999;53: Postoperative Radiotherapy for Squamous Cell Carcinoma