Outcomes of nodal metastatic cutaneous squamous cell carcinoma of the head and neck treated in a regional center

ORIGINAL ARTICLE Outcomes of nodal metastatic cutaneous squamous cell carcinoma of the head and neck treated in a regional center Campbell Schmidt, MBBS(Hons), 1 * Jarad M. Martin, DMed(Res), FRANZCR, 2 Eric Khoo, MBChB, FRANZCR, 3 Ashley Plank, PhD, 4 Roger Grigg, MBBS, FRACS 5 1 School of Medicine, University of Queensland, Queensland, Australia, 2 Department of Radiation Oncology, Calvary Mater Newcastle, Newcastle, Australia, 3 Radiation Oncology Queensland, St Andrew s Cancer Care Center, St Andrew s Hospital, Toowoomba, Queensland, Australia, 4 Oncology Research Australia, Australia, 5 Department of Otolaryngology Head and Neck Surgery, Toowoomba Rural Clinical School, University of Queensland, Queensland, Australia. Accepted 30 June 2014 Published online 27 January 2015 in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/hed.23843 ABSTRACT: Background. Given its metastatic potential and high incidence, cutaneous squamous cell carcinoma of the head and neck (SCCHN) has significant morbidity and mortality. Methods. We conducted a retrospective review of prospectively collected data for 113 consecutive patients with nodal metastatic cutaneous SCCHN treated surgically with curative intent in a regional center. Survival curves were generated by the Kaplan Meier method. Results. Five-year overall survival (OS), disease-specific survival (DSS), and disease-free survival (DFS) were 80%, 83%, and 75%, respectively. Twenty-six patients (23%) relapsed, with 92% of relapses occurring within 2 years of surgery. Immunosuppression (p 5.008) and N classification (p 5.043) predicted decreased DFS on univariate analysis. On multivariate analysis, only immunosuppression independently predicted DFS (p 5.034). Conclusion. This study validates the current N classification system, supports the adverse effect of immunosuppression, and suggests that intense follow-up for 2 years postsurgery is warranted. Survival at this regional center is comparable to that achieved at metropolitan tertiary cancer centers. VC 2015 Wiley Periodicals, Inc. Head Neck 37: 1808 1815, 2015 KEY WORDS: carcinoma, squamous cell, head and neck neoplasms, lymphatic metastasis, survival analysis, rural health services INTRODUCTION Cutaneous squamous cell carcinoma (SCC) is a malignant proliferation of epidermal keratinocytes, caused predominantly by the mutagenic effects of solar ultraviolet radiation. 1 Australia has the highest incidence of cutaneous SCC in the world with an age-standardized rate of 387 in 100,000 and the rate exceeding 1300 in 100,000 men in north Queensland. 2 4 The vast majority of cutaneous SCC occurs on the sun-exposed head and neck, with the disease exhibiting a male predominance associated with occupational sun exposure. 1 Other risk factors include age, fair skin type, and immunodeficiency. 2 With an aging population, ongoing ozone depletion, and increasing numbers and survival of immunosuppressed patients, the incidence of cutaneous SCC is rising. 2 In 95% of cutaneous SCC of the head and neck (SCCHN), complete surgical excision is curative. 5 However, cutaneous SCCHN has metastatic potential, with a recent systematic analysis reporting rates of regional and distant metastasis to be 5% and 1%, respectively, with a subsequent 1% disease-specific death rate. 6 Regional nodal spread is equally distributed between parotid and/or *Corresponding author: C. Schmidt, School of Medicine, University of Queensland, Mayne Medical School, 288 Herston Rd, Herston, Queensland, Australia 4006. E-mail: campbell.schmidt@uqconnect.edu.au cervical nodes, with metastatic cutaneous SCCHN recognized as the most common parotid malignancy in Australia. 7 The high absolute numbers of cutaneous SCCHN mean there is paradoxically significant morbidity and mortality associated with this disease. 8 In Australia, on average, 382 deaths/year are attributed to nonmelanoma skin cancer, with the majority of these likely to be nodal metastatic cutaneous SCCHN. 2 The standard of care treatment for cutaneous SCCHN metastatic to regional lymph nodes consists of comprehensive surgical dissection with or without adjuvant radiotherapy. 7 Research efforts thus far have stratified high-risk primary lesions and nodal disease based on unfavorable clinicopathological factors, some of which have been incorporated in the seventh edition of the American Joint Committee on Cancer (AJCC) staging manual. 7,9,10 The morbidity associated with treatment of nodal metastatic cutaneous SCCHN justifies attempts to risk-stratify patients and facilitate individualized treatment. 7 There is a paucity of patient outcome data for nodal metastatic cutaneous SCCHN treated in Queensland and, to our knowledge, no published data from regional Australia. It is well established that oncology patients living in regional and rural areas, including those with cutaneous SCC, have poorer outcomes than those living in major cities. 2,11 This is due, in part, to reduced access to and availability of diagnostic and treatment services. 11 1808 HEAD & NECK DOI 10.1002/HED DECEMBER 2015

OUTCOMES OF NODAL METASTATIC CUTANEOUS SCCHN Moreover, undertaking treatment at metropolitan tertiary cancer centers adds a significant personal and financial burden for patients. 11 Conceivably, if an equivalent standard of care can be provided in a regional setting, this could positively impact patient survival and quality of life. The Darling Downs region accounts for 16% of the estimated incidence of nonmelanoma skin cancer in Queensland, while containing only 8% of the state population. 12 In Toowoomba, the provincial capital, patients with nodal metastatic cutaneous SCCHN are treated by a multidisciplinary head and neck cancer service comprising head and neck surgery, plastic surgery, radiation oncology, medical oncology, pathology, radiology, and allied health. The primary purpose of this study was to report patient outcome data for this regional center and benchmark survival to that achieved at metropolitan tertiary cancer centers. A secondary purpose was to evaluate the impact of relevant clinicopathological factors on survival. MATERIALS AND METHODS This is a retrospective review of prospectively collected data for patients with nodal metastatic cutaneous SCCHN treated surgically by one of the co-authors in Toowoomba, Queensland, Australia, between 1998 and 2011. Ethical approval was obtained from the Darling Downs Human Research Ethics Committee and the University of Queensland School of Medicine internal low-risk ethics review panel. Only patients with histologically proven parotid and/or cervical nodal metastatic cutaneous SCCHN treated initially with surgery with curative intent were included. All patients had imaging studies confirming absence of distant disease before treatment. Demographic and clinicopathological data were extracted from a prospectively managed database. Data were crosschecked with existing patient medical records and updated where necessary. Fields included sex, age, immunosuppression, dates of primary and secondary diagnosis, extent of nodal involvement, surgical and adjuvant radiotherapy treatments, relapse status, and status at last follow-up. Relapse was defined as the first instance of confirmed disease recurrence. Locoregional relapse was defined as recurrence in the head and neck. Patients were staged using the seventh edition of the AJCC staging manual. 9 Data were deidentified post-collection and statistical analysis conducted independently using SPSS version 21.0 and R version 3.0.2, incorporating the Survival package version 2.37-2 for Kaplan Meier estimation. 13 15 End points for analysis included disease-free survival (DFS), disease-specific survival (DSS), and overall survival (OS). Survival was calculated from the date of surgery to the date of relapse, death, or last follow-up, as appropriate. Survival curves were generated by the Kaplan Meier method and compared using the log-rank test. Cox univariate analysis was performed for all relevant clinicopathological factors. Multivariate analysis was conducted using a Cox proportional hazards model to control for statistically and clinically significant covariates. A p value <.05 was considered statistically significant and all tests are 2- TABLE 1. Treatment. Treatment No. of patients (%)* Surgery Parotidectomy and neck dissection 73 (65) Neck dissection alone 38 (34) Parotidectomy alone 2 (2) Adjuvant radiotherapy Yes 100 (88) No 13 (12) * The total number of patients in the study (113) is the denominator for percentages. sided. Hazard ratios (HRs) with 95% confidence intervals (CIs) are reported to illustrate trends. RESULTS Of the 113 patients who met inclusion criteria, 93 (82%) were men and 20 (18%) were women, with a median age at diagnosis of 74 years (range, 41 93 years). Twelve patients (11%) were chronically immunosuppressed; 1 patient had a history of renal transplant and 11 patients had chronic leukemia. The majority of patients (91 of 113; 81%) presented with nodal metastasis within 5 years of an identifiable likely primary lesion. Within this group, 36 of 91 patients (40%) presented with concurrent primary and nodal disease. Regarding the 55 of 91 patients (60%) with nodal metastatic disease diagnosed over the subsequent 5 years after a likely primary lesion, 45 of 55 (82%) presented within 2 years and 51 of 55 (93%) within 3 years. In 22 of 113 patients (19%), the index lesion was unidentifiable on a background of multiple head and neck SCCs. In the 91 patients with an identifiable primary, 15 (16%) and 11 (12%) were located in the high-risk regions of the ear and non-glabrous lip, respectively. 9 Treatment All patients underwent individualized adequate surgical dissection based on the site of the primary and location of clinical metastatic involvement (Table 1). Patients with clinically positive parotid node disease received a parotidectomy with preservation of the facial nerve, unless obvious nerve involvement was visualized intraoperatively. Five patients had facial nerve resection where, intraoperatively, the facial nerve was obviously involved and/or encased, despite having normal facial nerve function preoperatively. On frozen section there was extensive microscopic invasion in each of these cases, with clear nerve margins confirmed by intraoperative frozen section. Lateral temporal bone resection was required in 2 patients to achieve this. Patients with cranial parotid disease had thorough examination of the deep lobe, as this area seems to be at risk with increasingly cranial parotid lymph node metastases. Patients with clinically positive cervical node disease received a selective level I to IV neck dissection for isolated level I and IIa metastases or a selective level I to V neck dissection for level IIb or multilevel disease. Surgical intent was to preserve the accessory nerve unless macroscopically involved. The majority of patients with clinically positive parotid disease but a clinically negative neckalsoreceivedanelective HEAD & NECK DOI 10.1002/HED DECEMBER 2015 1809

SCHMIDT ET AL. ipsilateral selective neck dissection, given the risk of occult neck disease. The site of parotid metastasis determined the extent of neck dissection. For high superficial disease, a level I to III dissection was performed. For lower and deep lobe involvement, a level I to IV dissection was performed. For posterior and parotid tail lesions, a level II to V dissection was performed. The 2 patients who received parotidectomy alone were treated for local parotid invasion with incidentally detected parotid node involvement on pathology. They declined subsequent elective neck dissection but received elective radiotherapy to the neck. The majority of patients (100 of 113; 88%) received adjuvant radiotherapy to pathologically involved fields (Table 1). Radiotherapy typically consisted of 60 Gy in 2 Gy fractions over 30 days, although other fractionated schedules were used. Six of these patients received concomitant chemoradiotherapy under enrollment in an international randomized clinical trial. 16 Of the 13 of 113 patients (12%) who did not receive radiotherapy, 4 were ineligible because of previous radiotherapy, 1 declined, and 8 had small (<2 cm), completely excised, single node disease. Pathology of nodal metastatic disease Of the 113 patients, on pathology, 54 (48%) had parotid node disease alone, 42 (37%) had cervical node disease alone, whereas 17 (15%) had both parotid and cervical nodes involved. Thus, of the 71 patients with pathological parotid disease, 17 (24%) had synchronous pathological cervical disease. Of these 17 patients, only 6 presented with clinical evidence of neck disease on preoperative examination/imaging. Therefore, the incidence of occult neck disease in patients presenting with parotid involvement was 15% (11 of 71) in our cohort. The median number of involved lymph nodes in positive parotid specimens was 2 (range, 1 10) and in positive neck specimens was 1 (range, 1 32). The median size of an involved lymph node was 24.5 mm (range, 7 75 mm). Patient outcomes Median duration to death or last follow-up was 40 months (range, 4 149 months). At last follow-up, 88 of 113 patients (78%) were alive and 25 of 113 (22%) were dead. Of those who died, the large majority died of disease (19 of 25; 76%). Of the total 113 patients, 26 (23%) experienced relapse, with 18 having locoregional recurrence and 8 developing distant or contralateral metastases. Only 4 patients relapsed in the parotid/temporal region with no local recurrence in the parotid occurring on the facial nerve. Median time to relapse was 6.5 months (range, 1 39 months) with 92% (24 of 26) of relapses occurring within 24 months of surgery. The majority of patients who relapsed died of disease (19 of 26; 73%). OS was 84% and 80% at 2 and 5 years, respectively (Figure 1A). DSS was 86% and 83%, whereas DFS was 78% and 75% at 2 and 5 years, respectively (Figure 1B and 1C). Univariate analysis Results of Cox univariate analysis of clinicopathological factors on DFS are reported in Table 2. Immunosuppression was associated with significantly worse DFS (p 5.008). Patients who received adjuvant radiotherapy had a nonsignificantly increased DFS (p 5.597). Increasing maximum node size (p 5.014) and number of involved nodes (p 5.001) both correlated with reduced DFS on univariate analysis. For the purpose of further analysis, the 104 patients with nodal size data were divided into 2 groups based on the revised AJCC N classification: maximum node size 30 mm (n 5 78; 75%) versus >30 mm (n 5 26; 25%). 9 DFS was not significantly different on log-rank test (p 5.115), with 5-year DFS of 79% and 63% for maximum node size 30 mm and >30 mm, respectively. The 104 patients with nodal number data were similarly divided according to AJCC staging: 50 (48%) had only 1 involved node and 54 (52%) had >1 involved node. 9 DFS was not significantly different on log-rank test (p 5.134) with 5-year DFS 82% and 70% for 1 and >1 nodes involved, respectively. When the same analysis was performed independently for parotid and cervical node data, >1 cervical node involved predicted significantly worse DFS in those with cervical node disease (n 5 55; p 5.020), whereas >1 parotid node involved did not in those with parotid node disease (n 5 66; p 5.456). Under revised AJCC N classification of the 102 patients with both nodal size and nodal number data available, 37 (36%) had N1 disease, 63 (62%) had N2 disease, and 2 (2%) had N3 disease. 9 Given only 2 patients met N3 criteria, the latter 2 groups were combined to facilitate analysis. Patients with N2 to N3 disease had significantly worse DFS compared with patients with N1 disease (p 5.043), with 5-year DFS 68% and 88%, respectively (Figure 2A). Of the total 113 patients, 96 (85%) had either parotid or cervical nodes that were positive, whereas 17 (15%) had both nodal sites involved. The presence of synchronous cervical and parotid disease trended toward worse DFS compared to disease in either site alone on log-rank test (p 5.089). In a subgroup analysis of the 71 patients with pathological parotid disease, 54 (76%) had parotid disease alone, whereas 17 (24%) had synchronous cervical disease. The presence of cervical disease was associated with borderline worse DFS in this group on log-rank test (p 5.078) with 5-year DFS 80% for parotid disease alone versus 54% for synchronous parotid and cervical disease (Figure 2B). Multivariate analysis To control for the effect of multiple variables, a backward conditional Cox proportional hazards model for DFS was created. Immunosuppression, adjuvant radiotherapy, N classification, and presence of synchronous cervical and parotid node disease were included as initial covariates. Maximum node size and number of involved nodes were not included in the model, as these variables are taken into account by the N classification. In the final model (Table 3), only immunosuppression was an independent predictor of DFS (p 5.034) with N classification a borderline independent predictor of worse DFS (p 5.061). DISCUSSION In this study, we report patient outcome data for nodal metastatic cutaneous SCCHN treated by a 1810 HEAD & NECK DOI 10.1002/HED DECEMBER 2015

OUTCOMES OF NODAL METASTATIC CUTANEOUS SCCHN FIGURE 1. (A) Overall survival (OS) for 113 patients with nodal metastatic cutaneous squamous cell carcinoma of the head and neck (SCCHN). Dotted lines define 95% point-wise confidence bounds. Crosses denote censored observations. Five-year OS was 80% with 95% confidence interval (CI) 72% to 88%. (B) Disease-specific survival (DSS) for 113 patients with nodal metastatic cutaneous SCCHN. Dotted lines define 95% point-wise confidence bounds. Crosses denote censored observations. Five-year DSS was 83% with 95% CI 76% to 91%. (C) Disease-free survival (DFS) for 113 patients with nodal metastatic cutaneous SCCHN. Dotted lines define 95% point-wise confidence bounds. Crosses denote censored observations. Five-year DFS was 75% with 95% CI 67% to 84%. TABLE 2. Cox univariate analysis of clinicopathological factors on disease-free survival. Factor HR (95% CI) p value (sample size) Immunosuppression, yes vs no 3.46 (1.38 8.68).008 (n 5 113) Adjuvant radiotherapy, yes vs no 0.75 (0.26 2.18).597 (n 5 113) Maximum node size 1.03 (1.01 1.06).014 (n 5 104) Number involved nodes 1.06 (1.03 1.10).001 (n 5 104) Maximum node size, >30 mm vs 30 mm 2.06 (0.89 4.76).092 (n 5 104) Number involved nodes, >1 vs 1 1.90 (0.81 4.48).143 (n 5 104) Cervical nodes, >1 vs 1 3.26 (1.13 9.41).028 (n 5 55) Parotid nodes, >1 vs 1 1.51 (0.51 4.51).462 (n 5 66) N classification, N2 3 vs N1 2.88 (0.98 8.47).054 (n 5 102) Site of nodes, P and C vs P only 2.31 (0.88 6.09).088 (n 5 71) Site of nodes, P and C vs P or C only 2.08 (0.87 4.95).098 (n 5 113) Abbreviations: HR, hazard ratio; CI, confidence interval; P, parotid; C, cervical. HEAD & NECK DOI 10.1002/HED DECEMBER 2015 1811

SCHMIDT ET AL. FIGURE 2. (A) Comparison of disease-free survival (DFS) for 102 patients with nodal metastatic cutaneous squamous cell carcinoma of the head and neck (SCCHN) with N2 to N3 (n 5 65) versus N1 (n 5 37) disease (p 5.043). Crosses denote censored observations. Five-year DFS was 68% and 88% for patients with N2 to N3 and N1 disease, respectively. (B) Comparison of DFS for 71 patients with nodal metastatic cutaneous SCCHN with parotid and cervical nodes (n 5 17) versus parotid nodes alone (n 5 54) involved (p 5.078). Crosses denote censored observations. Five-year DFS was 54% and 80% for patients with parotid and cervical node disease versus parotid node disease alone, respectively. multidisciplinary head and neck cancer service in Toowoomba, Queensland, Australia. There is a paucity of outcome data for Queensland despite this Australian state having the highest incidence of cutaneous SCCHN worldwide and the significant morbidity and mortality associated with nodal metastatic disease. 2,4 In addition, to our knowledge, this is the first published outcome data from a regional cancer center in Australia. A review of the literature shows the vast majority of survival data is from metropolitan tertiary cancer centers in Australia, New Zealand, and North America. Reported survival rates vary between institutions, with 5-year survival data from selected recent studies summarized in Table 4. 17 21 The variation in outcome measures and cohort composition make comparisons difficult. In particular, different numbers of patients with immunosuppression, recurrence, and receiving single modality therapy are likely to have a significant impact on cohort survival rates. However, it is evident that our 5-year survival results (DFS, 75%; DSS, 83%; and OS, 80%) are comparable to those achieved at metropolitan tertiary cancer centers (Table 4). Although the question of a selected population at this regional center may be raised, it is important to note that this cohort represents 113 consecutive patients presenting with nodal metastatic disease to a single surgeon and that no patients were referred on for management in a tertiary center. These excellent survival figures reflect our individualized treatment of nodal metastatic cutaneous SCCHN in the context of a multidisciplinary head and neck cancer team. A volume-outcome paradigm has favored centralized cancer care in metropolitan centers. Our results support the treatment of nodal metastatic cutaneous SCCHN in regional centers when appropriate specialized services are available and a sufficient caseload is maintained, as at this single surgeon center. Provision of such a service in regional centers has the potential to reduce the burden on the regional/rural patient and the health care system alike. 11 These survival data also reflect our deliberately comprehensive treatment philosophy. All patients underwent adequate surgical dissection. No patients had node sampling or isolated nodal biopsy. Almost all patients with clinical parotid disease and a clinically negative neck had an elective ipsilateral selective neck dissection, given the risk of occult neck disease, which has been reported at up to 44% in other series. 22 27 In our cohort, 15% of patients with parotid disease had occult neck disease, which is consistent with the rates of 15% and 16% reported in 2 other Australian series and 1 Canadian study. 22,23,25 Given the incidence of occult neck disease seems to be at least 15%, we would continue to recommend elective neck dissection for patients with clinical parotid disease and a clinically negative neck. If pathology is negative, we believe radiotherapy to the neck can be safely avoided. The majority of patients treated at our center received adjuvant radiotherapy, in accordance with the consensus in the literature that combined modality therapy is best practice. 28,29 However, in this study, we were unable to demonstrate a survival benefit of adjuvant radiotherapy TABLE 3. Final multivariate Cox proportional hazards model after backward conditional elimination. Factor HR (95% CI) p value Immunosuppression, yes vs no 2.94 (1.08 7.97).034 N classification, N2 N3 vs N1 2.80 (0.95 8.24).061 Abbreviations: HR, hazard ratio; CI, confidence interval. 1812 HEAD & NECK DOI 10.1002/HED DECEMBER 2015

OUTCOMES OF NODAL METASTATIC CUTANEOUS SCCHN TABLE 4. Selected recent studies reporting 5-year survival data for nodal metastatic cutaneous squamous cell carcinoma of the head and neck. 17 21 Authors (year) Cohort size Center DFS DSS OS Current study (2013) 113 Toowoomba, AUS 75% 83% 80% Ebrahimi et al 17 (2013) 229 Sydney, AUS* NR 77% NR Forest et al 18 (2010) 215 Sydney, AUS 73% 77% 69% Ch ng et al 19 (2008) 170 Multicentre, NZ 59% 69% 48% Hinerman et al 20 (2008) 121 Florida, USA 70% NR 54% Andruchow et al 21 (2006) 322 Multicenter, USA/AUS NR 74% NR Abbreviations: DFS, disease-free survival; DSS, disease-specific survival; OS, overall survival; AUS, Australia; NR, not reported; NZ, New Zealand; USA, United States of America. * Westmead Cancer Centre. Royal Prince Alfred Hospital. (p 5.597). This was probably because of the small number of patients receiving surgery alone and a significant patient selection bias, with the majority of these patients having minimal nodal disease burden. Given the accepted benefit of adjuvant radiotherapy, the role of cytotoxic and biological therapies is now at the frontline of research efforts. 7 Six of the patients in our cohort are currently enrolled in an international randomized controlled trial investigating the impact of adding weekly carboplatin to adjuvant radiotherapy. 16 In mucosal SCCHN, the addition of radiosensitizing platinum-based chemotherapy provides a well-characterized improvement in locoregional control and it is hypothesized a similar advantage will be demonstrated in nodal metastatic cutaneous SCCHN. 16 A not insignificant number of patients in our cohort were chronically immunosuppressed (12 of 113; 11%) and our results are consistent with earlier data in showing that immunosuppression predicts worse survival for nodal metastatic disease on univariate (p 5.008) and multivariate (p 5.034) analysis. 30 Immunosuppression is a wellestablished risk factor in the pathogenesis of nodal metastatic cutaneous SCCHN. 7 Organ-transplant recipients and patients with chronic lymphocytic leukemia have a much higher incidence of cutaneous SCC, with tumors more aggressive in nature. 31 34 In 2009, Oddone et al 35 demonstrated the utility of incorporating immunosuppression into a prognostic score for nodal metastatic cutaneous SCCHN. Although the revised AJCC staging precluded the inclusion of immunosuppression on account of its status as a risk factor, the authors recommend qualifying staging in the presence of immunosuppression (I) and our study supports the value of this recommendation. 9 We also rigorously assessed the impact of nodal disease burden on DFS. The seventh edition of the AJCC staging manual incorporated several evidence-based changes in the stratification of lymph node status. 9 Previously, nodal (N) disease was defined as either present (N1) or absent (N0). 10 The revised TNM staging adopted the wellestablished N0 to N3 classification used for mucosal SCCHN based on recent evidence that both increasing number and size of pathologically involved lymph nodes reduces survival. 10,18,36 Under this system, a single involved node 30 mm in size represents N1 disease, maximum node size >30 mm and/or multiple nodes distinguishes N2 disease, and maximum node size >60 mm denotes N3 disease. N1 disease represents stage III disease, whereas both N2 and N3 represent stage IV disease. Although 2 of our patients met N3 criteria based on maximum node size, they were grouped with N2 for the purposes of statistical analysis in evaluating the prognostic value of N classification. Initially, we assessed maximum node size and number of involved nodes independently. On univariate analysis, both factors predicted significantly worse DFS, in concordance with recent evidence (Table 2). 18,21,36 We then demonstrated that maximum node size >30 mm and >1 involved nodes both trended toward worse DFS without statistical significance, lending some support to the new N classification thresholds adopted by the AJCC. 9 A seminal study by O Brien et al, 37 in 2002, separated parotid from cervical involvement and found that extent of parotid disease correlated with local control but not survival, whereas increasing cervical nodal involvement heralded a significantly worse prognosis. Similarly, in our study, on independent analysis of extent of parotid and cervical nodal involvement, multiple cervical nodes correlated with worse DFS in patients with neck disease (p 5.020), whereas multiple parotid nodes did not worsen DFS in those with parotid disease (p 5.456). We also found that patients with neck and parotid involvement had significantly worse outcomes than those with parotid disease alone (p 5.078), again in agreement with O Brien et al. 37 In our cohort, the presence of parotid and neck disease compared to disease in either site alone was also associated with worse survival (p 5.089). Other studies have demonstrated the independent prognostic significance of parotid disease and the adverse impact of synchronous parotid and neck involvement, supporting a role for separate parotid and neck staging. 21,38 However, a later study by Hinerman et al, 20 in 2008, failed to validate separation of parotid and neck stage with Forest et al, 18 in 2010, demonstrating the utility of instead incorporating the parotid as a regional level for the purposes of staging. For practical purposes, parotid and neck staging remained combined in the revised AJCC staging system. 10 Our data supports this collective approach to nodal burden, but again highlights the prognostic significance of multisite disease. On restaging our cohort using the revised AJCC criteria, patients with N2 to N3 disease had significantly worse DFS than N1 patients on univariate (p 5.043) analysis and borderline worse survival on multivariate (p 5.061) analysis. 9 A recent study by Brunner et al, 39 in 2012, found N3 to be significantly worse than N1 but no significant difference between N1 and N2 disease. Our results similarly demonstrate broad survival correlation HEAD & NECK DOI 10.1002/HED DECEMBER 2015 1813

SCHMIDT ET AL. with N classification and affirm the prognostic value of the revised nodal stratification system. However, 2 recent articles have questioned the heterogenic composition of AJCC stage IV disease, which now contains all cancers meeting T4, N2, N3, or M1 criteria. 36,39 Our study excluded patients presenting with distant metastasis (M1) as they are not usually treated with curative intent. However, the prognosis of this group, particularly in comparison to regional nodal metastatic cutaneous SCCHN, is evidently an area for future research. Of the 26 patients who experienced recurrence in our cohort, 50% (13 of 26) relapsed within 6.5 months and 92% (24 of 26) relapsed within 24 months of surgery. Other studies have reported a median time to recurrence of 8 to 15 months. 29,40 Our data therefore suggest that intense follow-up for at least 2 years postsurgery is warranted. Beyond 24 months, relapse is unlikely and possibly discharge to a general practitioner would be appropriate at this time. At our center, patients alternate 3-month follow-up visits with the surgeon and radiation oncologist for 2 years. This was then extended to 4- month then 6-month reviews up to 5 years, with some patients receiving ongoing annual follow-up beyond 5 years. Patients in our cohort received 6-month chest radiograph and neck ultrasound, with CT of the neck, chest, and abdomen performed at 18 months. As in other series, the majority of recurrences were locoregional (18 of 26). 21,29 However, because of the aggressive multimodality primary therapy received by our patients, treatment options for recurrence were limited, with only 7 of 26 patients (27%) successfully salvaged. The majority of patients (19 of 26; 73%) who relapsed died of disease. Patients who recurred with distant metastasis (8 of 26) presumably had undetectable seeding before primary treatment and inevitably failed therapy. We acknowledge the inherent limitations of this retrospective study of a relatively small number of patients treated over a significant period of time. However, we believe retrospective review of a prospectively collected dataset has allowed us to robustly assess outcomes at this center and benchmark survival against other centers. Despite the high incidence of cutaneous SCCHN, nodal metastatic spread is still relatively uncommon, which necessitates a protracted time course for evaluation of a reasonably sized cohort. Although this has the potential to introduce treatment heterogeneity, our study is distinctive in that it represents the experience of a single surgeon who has maintained a consistently aggressive approach to this disease. CONCLUSION In conclusion, this study validates the current N classification system, supports the adverse effect of immunosuppression, and suggests that intense follow-up for 2 years is warranted postsurgery for nodal metastatic cutaneous SCCHN. Most importantly, we have demonstrated that survival at this regional center is comparable to that achieved at metropolitan tertiary cancer centers. This is a unique prospectively maintained data set with only 1 surgeon involved. An aggressive treatment philosophy and close multidisciplinary collaboration has contributed to good patient outcomes. This study supports the role of expert regional head and neck cancer centers in managing nodal metastatic cutaneous SCCHN. REFERENCES 1. Ramirez CC, Federman DG, Kirsner RS. Skin cancer as an occupational disease: the effect of ultraviolet and other forms of radiation. Int J Dermatol 2005;44:95 100. 2. Australian Institute of Health and Welfare, Cancer Australia. Nonmelanoma skin cancer: general practice consultations, hospitalisation and mortality. Cancer series: no. 43. Cat. no. 39. Canberra: Australian Institute of Health and Welfare; 2008. 3. Buettner PG, Raasch BA. Incidence rates of skin cancer in Townsville, Australia. Int J Cancer 1998;78:587 593. 4. Staples MP, Elwood M, Burton RC, Williams JL, Marks R, Giles GG. Non-melanoma skin cancer in Australia: the 2002 national survey and trends since 1985. Med J Aust 2006;184:6 10. 5. Clayman GL, Lee JJ, Holsinger FC, et al. Mortality risk from squamous cell skin cancer. J Clin Oncol 2005;23:759 765. 6. Jambusaria Pahlajani A, Miller CJ, Quon H, Smith N, Klein RQ, Schmults CD. Surgical monotherapy versus surgery plus adjuvant radiotherapy in high-risk cutaneous squamous cell carcinoma: a systematic review of outcomes. Dermatol Surg 2009;35:574 585. 7. Veness MJ, Porceddu S, Palme CE, Morgan GJ. Cutaneous head and neck squamous cell carcinoma metastatic to parotid and cervical lymph nodes. Head Neck 2007;29:621 631. 8. Givi B, Andersen PE, Diggs BS, Wax MK, Gross ND. Outcome of patients treated surgically for lymph node metastases from cutaneous squamous cell carcinoma of the head and neck. Head Neck 2011;33:999 1004. 9. Edge SB, Byrd DR, Compton CC, Fritz AG, Greene FL, Trotti A, editors. Cutaneous squamous cell carcinoma and other cutaneous carcinomas. AJCC Cancer Staging Manual. New York, NY: Springer; 2010. 10. Warner CL, Cockerell CJ. The new seventh edition American Joint Committee on Cancer staging of cutaneous non-melanoma skin cancer: a critical review. Am J Clin Dermatol 2011;12:147 154. 11. Jong KE, Smith DP, Yu XQ, O Connell DL, Goldstein D, Armstrong BK. Remoteness of residence and survival from cancer in New South Wales. Med J Aust 2004;180:618 622. 12. Giles GG, Marks R, Foley P. Incidence of non-melanocytic skin cancer treated in Australia. Br Med J (Clin Res Ed) 1988;296:13 17. 13. SPSS Statistics for Windows, version 21.0. Armonk, NY: IBM Corp; 2012. 14. R Development Core Team. R: a language and environment for statistical computing, version 3.0.2. Vienna, Austria: R Foundation for Statistical Computing; 2013. 15. Therneau T. A package for survival analysis in S, R package, version 2.37 2, 2013. 16. Trans-Tasman Radiation Oncology Group (TROG). Postoperative concurrent chemo-radiotherapy versus postoperative radiotherapy in high-risk cutaneous squamous cell carcinoma of the head and neck. ClinicalTrials.- gov, [updated 5 August 2013]. Available at: http://clinicaltrials.gov/show/ NCT00193895. Accessed August 5, 2013. 17. Ebrahimi A, Clark JR, Ahmadi N, Palme CE, Morgan GJ, Veness MJ. Prognostic significance of disease-free interval in head and neck cutaneous squamous cell carcinoma with nodal metastases. Head Neck 2013;35: 1138 1143. 18. Forest VI, Clark JJ, Veness MJ, Milross C. N1S3: a revised staging system for head and neck cutaneous squamous cell carcinoma with lymph node metastases: results of 2 Australian Cancer Centers. Cancer 2010;116:1298 1304. 19. Ch ng S, Maitra A, Allison RS, et al. Parotid and cervical nodal status predict prognosis for patients with head and neck metastatic cutaneous squamous cell carcinoma. J Surg Oncol 2008;98:101 105. 20. Hinerman RW, Indelicato DJ, Amdur RJ, et al. Cutaneous squamous cell carcinoma metastatic to parotid-area lymph nodes. Laryngoscope 2008; 118:1989 1996. 21. Andruchow JL, Veness MJ, Morgan GJ, et al. Implications for clinical staging of metastatic cutaneous squamous carcinoma of the head and neck based on a multicenter study of treatment outcomes. Cancer 2006;106: 1078 1083. 22. Kirke DN, Porceddu S, Wallwork BD, Panizza B, Coman WB. Pathologic occult neck disease in patients with metastatic cutaneous squamous cell carcinoma to the parotid. Otolaryngol Head Neck Surg 2011;144:549 551. 23. Audet N, Palme CE, Gullane PJ, et al. Cutaneous metastatic squamous cell carcinoma to the parotid gland: analysis and outcome. Head Neck 2004;26: 727 732. 24. O Brien CJ, McNeil EB, McMahon JD, Pathak I, Lauer CS. Incidence of cervical node involvement in metastatic cutaneous malignancy involving the parotid gland. Head Neck 2001;23:744 748. 25. Dona E, Veness MJ, Cakir B, Morgan GJ. Metastatic cutaneous squamous cell carcinoma to the parotid: the role of surgery and adjuvant radiotherapy to achieve best outcome. ANZ J Surg 2003;73:692 696. 26. Moore BA, Weber RS, Prieto V, et al. Lymph node metastases from cutaneous squamous cell carcinoma of the head and neck. Laryngoscope 2005; 115:1561 1567. 27. Ying YL, Johnson JT, Myers EN. Squamous cell carcinoma of the parotid gland. Head Neck 2006;28:626 632. 1814 HEAD & NECK DOI 10.1002/HED DECEMBER 2015

OUTCOMES OF NODAL METASTATIC CUTANEOUS SCCHN 28. Veness MJ, Palme CE, Smith M, Cakir B, Morgan GJ, Kalnins I. Cutaneous head and neck squamous cell carcinoma metastatic to cervical lymph nodes (nonparotid): a better outcome with surgery and adjuvant radiotherapy. Laryngoscope 2003;113:1827 1833. 29. Wang JT, Palme CE, Morgan GJ, Gebski V, Wang AY, Veness MJ. Predictors of outcome in patients with metastatic cutaneous head and neck squamous cell carcinoma involving cervical lymph nodes: Improved survival with the addition of adjuvant radiotherapy. Head Neck 2012;34:1524 1528. 30. Southwell KE, Chaplin JM, Eisenberg RL, McIvor NP, Morton RP. Effect of immunocompromise on metastatic cutaneous squamous cell carcinoma in the parotid and neck. Head Neck 2006;28:244 248. 31. Martinez JC, Otley CC, Stasko T, et al. Defining the clinical course of metastatic skin cancer in organ transplant recipients: a multicenter collaborative study. Arch Dermatol 2003;139:301 306. 32. Mihalov ML, Gattuso P, Abraham K, Holmes EW, Reddy V. Incidence of post-transplant malignancy among 674 solid-organ-transplant recipients at a single center. Clin Transplant 1996;10:248 255. 33. Veness MJ, Quinn DI, Ong CS, et al. Aggressive cutaneous malignancies following cardiothoracic transplantation: the Australian experience. Cancer 1999;85:1758 1764. 34. Mehrany K, Weenig RH, Lee KK, Pittelkow MR, Otley CC. Increased metastasis and mortality from cutaneous squamous cell carcinoma in patients with chronic lymphocytic leukemia. J Am Acad Dermatol 2005;53:1067 1071. 35. Oddone N, Morgan GJ, Palme CE, et al. Metastatic cutaneous squamous cell carcinoma of the head and neck: the Immunosuppression, Treatment, Extranodal spread, and Margin status (ITEM) prognostic score to predict outcome and the need to improve survival. Cancer 2009;115:1883 1891. 36. Buethe D, Warner C, Miedler J, Cockerell CJ. Focus issue on squamous cell carcinoma: practical concerns regarding the 7th edition AJCC Staging Guidelines. J Skin Cancer 2011;2011:156391. 37. O Brien CJ, McNeil EB, McMahon JD, Pathak I, Lauer CS, Jackson MA. Significance of clinical stage, extent of surgery, and pathologic findings in metastatic cutaneous squamous carcinoma of the parotid gland. Head Neck 2002;24:417 422. 38. Palme CE, O Brien CJ, Veness MJ, McNeil EB, Bron LP, Morgan GJ. Extent of parotid disease influences outcome in patients with metastatic cutaneous squamous cell carcinoma. Arch Otolaryngol Head Neck Surg 2003;129:750 753. 39. Brunner M, Veness MJ, Ch ng S, Elliott M, Clark JR. Distant metastases from cutaneous squamous cell carcinoma-analysis of AJCC stage IV. Head Neck 2013;35:72 75. 40. Veness MJ, Morgan GJ, Palme CE, Gebski V. Surgery and adjuvant radiotherapy in patients with cutaneous head and neck squamous cell carcinoma metastatic to lymph nodes: combined treatment should be considered best practice. Laryngoscope 2005;115:870 875. HEAD & NECK DOI 10.1002/HED DECEMBER 2015 1815