Metastatic head and neck cutaneous squamous cell carcinoma: Defining a low‐risk patient

ORIGINAL ARTICLE

METASTATIC HEAD ANDNECKCUTANEOUS SQUAMOUS CELLCARCINOMA: DEFINING A LOW-RISK PATIENT

Ardalan Ebrahimi, MBBS (Hons), FRACS,1 Jonathan R. Clark, MBBS (Hons), FRACS,1

Balazs B. Lorincz, MD,1 Christopher G. Milross, MD, FRANZCR,1

Michael J. Veness, MMed, MD, FRANZCR2

1 Sydney Head and Neck Cancer Institute, The Sydney Cancer Centre, Royal Prince Alfred Hospital, Missenden Road,Camperdown, New South Wales, 2050, Sydney, Australia. E-mail: [email protected]

2Head and Neck Cancer Service, Westmead Hospital, University of Sydney, Sydney, Australia

Accepted 4 January 2011Published online 5 April 2011 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/hed.21743

Abstract: Background. The purpose of this study was to

determine whether there is a ‘‘low-risk’’ subset of patients with

regional metastatic head and neck cutaneous squamous cell

carcinoma (SCC) suitable for treatment with surgery alone and

omission of adjuvant radiotherapy.

Methods. We conducted a retrospective analysis of 168

patients with a single parotid gland or neck nodal metastasis

�3 cm in size from cutaneous SCC treated with curative intent

by surgery � adjuvant radiotherapy.

Results. Disease-specific survival for the 33 patients

treated with surgery alone was 97% at 5 years. In the subset

of 19 patients without extracapsular nodal spread (ECS), there

was 1 regional recurrence which was successfully salvaged

yielding a 5-year disease-specific survival of 100%.

Conclusion. In head and neck cutaneous SCC, the subset

with a single node �3 cm in size without ECS are at low risk of re-

gional failure and death from cutaneous cancer. These patients

may be suitable for single-modality therapy with surgery alone.VVC 2011 Wiley Periodicals, Inc. Head Neck 34: 365–370, 2012

Keywords: head and neck; cutaneous squamous cell

carcinoma; lymph nodes; metastases; radiotherapy

Cutaneous squamous cell carcinoma (cutaneous SCC)is a major public health issue in countries with largewhite populations and high solar ultraviolet exposure,such as Australia and New Zealand.1 Approximately5% of head and neck cutaneous SCCs will metastasizeto regional lymph nodes in the parotid gland orneck,2,3 and there is evidence to support surgery andpostoperative radiotherapy as best practice in thesepatients.4–8 At present, the results of a Trans-TasmanRadiation Oncology Group trial (Postoperative SkinTrial 05.01; available at www.trog.com.au) are awaitedto determine whether adjuvant chemoradiation usingcarboplatin is beneficial in high-risk metastatic cuta-neous SCC of the head and neck. Hence, it seems that

there is a trend toward more aggressive treatmentwith inherent greater toxicity, although there is noevidence that this is required in all patients with met-astatic nodal cutaneous SCC.

One factor limiting risk stratification has been thelack of a suitable staging system to discriminatebetween low-risk and high-risk patients; however,staging systems developed by O’Brien et al,9 andmore recently Forest et al,10 define subsets of patientsbased on the size and number of involved lymphnodes akin to the widely accepted American JointCommittee on Cancer staging system for mucosalSCC.11 Oddone et al12 proposed an alternative prog-nostic model using the ‘‘ITEM’’ score based on thepresence of immunosuppression, treatment modality,extracapsular nodal spread (ECS) and margin status.This model emphasized the relative importanceof ECS in relation to the other variables. The primarypurpose of this study was to determine whetherthere is a ‘‘low-risk’’ group of patients with metastaticcutaneous SCC to the parotid gland and/or neck whomay be suitable for treatment with surgery alone.

MATERIALS AND METHODS

Study Population. Clinicopathological data frompatients treated for metastatic cutaneous SCC at theSydney Head and Neck Cancer Institute (SHNCI),the Royal Prince Alfred Hospital, the Sydney Hospi-tal, and the Westmead Hospital, Sydney, Australia,were extracted from their respective databases andcombined. These databases prospectively record de-mographic data, clinical and pathological staging,treatment details, histopathological features, theduration of follow-up, and status at last visit.

The N1S3 staging system developed by Forestet al10 (Table 1) was used to select a low-risk group ofpatients with metastatic cutaneous SCC to the parotidgland or neck. All patients with stage I (single nodalmetastasis �3 cm in maximal diameter) disease

Correspondence to: A. Ebrahimi

This work was presented at the Australian Society of OtolaryngologyHead & Neck Surgery, Sydney, Australia, March 30, 2010.VVC 2011 Wiley Periodicals, Inc.

Low-Risk Nodal Cutaneous SCC HEAD & NECK—DOI 10.1002/hed March 2012 365

treated with curative intent between 1980 and 2007were included in this study. The pathology detailsrecorded in the database were crosschecked with theoriginal pathology reports, and the database wasupdated where necessary. All patients had undergoneparotidectomy and/or neck dissection � adjuvantradiotherapy and clinicopathological variables weresimilar for both SHNCI and Westmead cohorts, unlessspecified.

Statistical Analysis. Data were collated and filteredusing Excel (Microsoft, Redmond, WA), and the statis-tical analysis was performed using SPSS version 12.0(SPSS, Chicago, IL). The endpoints for analysisincluded regional failure and disease-specific survival(DSS). Regional failure was defined as pathologicallyproven tumor relapse in the parotid gland or neck.DSS was calculated from the date of surgery to thedate of death from head and neck cutaneous SCC orlast follow-up. Patients that died from other causeswere censored from the study at the time of death.Differences in recurrence and survival were calcu-lated using the Kaplan–Meier method and curveswere compared using the log rank test. Significancelevels were set at p < .05, and all tests were 2-sided.

RESULTS

Study Group. Our study population consisted of 168patients with metastatic cutaneous SCC of the headand neck classified as N1S3 stage I, after detailedevaluation of their medical records for completenessof pathological information. There were 142 men and26 women with a median age of 72 years (range, 37–89 years). The median and mean follow-up was 3.1years and 3.7 years, respectively. The distribution ofprimary sites is summarized in Table 2.

Treatment. All patients had surgery as their pri-mary treatment modality (Table 3). This included 93

patients undergoing parotidectomy with neck dissec-tion, 43 having parotidectomy alone, and 33 patientsundergoing neck dissection only. One patient had abilateral neck dissection for a metastatic cutaneousSCC of the scalp.

One hundred thirty-five patients (80%) receivedadjuvant radiotherapy. Over the course of the studyperiod, there has been a move toward routine admin-istration of adjuvant radiotherapy to the parotidgland and neck. Specifically, radiotherapy was recom-mended to all patients with involved or close margins,extracapsular spread, multiple metastatic nodes,large nodal deposits, immunosuppression, and con-cern regarding possible tumor contamination of theoperative field. The median dose of adjuvant radiother-apy given to the parotid bed was 60 Gy (range, 46–70Gy) at Westmead Hospital and 54 Gy (range, 20–66Gy) at the SHNCI while the neck received a mediandose of 50 Gy in both institutions (range, 32–66 Gy).Table 4 summarizes the adverse histopathological fea-tures based on single versus combined modality treat-ment and institution. In the surgery alone cohort, 16patients (48.5%) had at least 1 adverse feature. The av-erage number of adverse features in the surgery alonecohort was 1.1 compared with 1.3 in those receiving ad-juvant radiotherapy. Of the 33 patients who did notreceive adjuvant radiotherapy, reasons includedpatient refusal, poor performance status, physicianpreference based on favorable histopathological fea-tures, and previous radiotherapy to the neck.

Pathology of Regional Metastatic Disease. Themedian size of the metastatic node was 15 mm(range, 5–29 mm). ECS was present in 102 patients(57%). The excision margins were involved or close in45% and 57% of patients from the SHNCI and West-mead cohorts, respectively. Of the 33 patients whowere treated with surgery alone, 19 did not haveECS.

Regional Control. Regional recurrence occurred in19 of 168 patients (11%) overall. The regional controlrate (parotid gland and neck) was 93% and 89% at 2and 5 years, respectively. As shown in Figure 1A, the5-year regional control rate based on the presence ofECS was equivalent at 88% for those with and with-out ECS (p ¼ .94). Similarly, there was no significant

Table 2. Distribution of primary sites by institution.

Location SHNCI Westmead Total

Face 36 30 66

External ear 11 11 22

Scalp or neck 16 7 23

Skin, NOS 35 22 57

Total 98 70 168

Abbreviations: SHNCI, Sydney Head and Neck Cancer Institute; NOS, not otherwisespecified.

Table 3. Types of surgery by institution.

Type of surgery SHNCI Westmead Total

Parotidectomy 33 10 43

Parotidectomy þ SND 25 32 57

Parotidectomy þ CND 30 6 36

SND 6 15 21

CND 4 7 11

Total 98 70 168

Abbreviations: SHNCI, Sydney Head and Neck Cancer Institute; SND, selectiveneck dissection; CND, comprehensive neck dissection.

Table 1. N1S3 staging system.10

Stage I: single lymph node �3 cm

Stage II: single lymph node measuring >3 cm or multiple

lymph nodes measuring �3 cm

Stage III: multiple lymph nodes measuring >3 cm

366 Low-Risk Nodal Cutaneous SCC HEAD & NECK—DOI 10.1002/hed March 2012

difference in regional control based on whetherpatients did or did not receive adjuvant radiotherapy(87% vs 91%, respectively, p ¼ .54; Figure 1B).

Disease-Specific Survival. There were 42 deaths(25%) overall, of which 11 (7%) were due to head andneck cutaneous SCC. The cumulative DSS rate for the168 patients with N1S3 stage I disease was 94% and92% at 2 years and 5 years, respectively. As shown inFigure 2, the 5-year DSS was 90% in the presence ofECS and 94% in those without ECS (p ¼ .24). Therewas also no significant difference in DSS based onwhether patients received adjuvant radiotherapy ornot, with 5-year DSS rates of 90% versus 97% (p ¼ .40).

Surgery Alone Cohort. There were 33 patientstreated with surgery alone. Relevant pathologicaldata for this group are summarized in Table 4. Basedon the ITEM prognostic model, 19 of these patients(57%) were low risk (ITEM score �2.6) and 14 werehigh risk (43%) due to the presence of ECS �involved margins (ITEM score >3.0). There was 1 dis-ease-related death and 3 regional recurrences, 1 ofwhich occurred in a patient with an ITEM score of�2.6. The regional control rate (parotid gland and

neck) was 91% at 2 years and 5 years, respectively.The DSS rate for the 33 patients with N1S3 stage Idisease treated with surgery alone was 97% at 2years and 5 years, respectively.

The 5-year regional control rate based on thepresence of ECS was 92% (ECS) versus 92% (no ECS;p ¼ .36). In this subset of patients treated with sur-gery alone, only 1 patient without ECS developedregional failure, and was successfully salvaged withfurther surgery and adjuvant radiotherapy. Thispatient did not have any other adverse pathologicalfeatures. The 5-year DSS was 100% in patients with-out ECS and 93% in the presence of ECS (p ¼ .29).The Kaplan–Meier curves comparing DSS based ontreatment modality and ECS are shown in Figure 3.

DISCUSSION

Head and neck cutaneous SCC is a major public healthproblem around the world, and although regionalmetastases occur in less than 5% of patients, the dis-ease and necessary treatment are associated with sig-nificant morbidity and mortality.13 In recent years,there has been a trend toward more aggressive com-bined-modality therapy utilizing surgery and adjuvantradiotherapy with many authors recommending this

FIGURE 1. (A) Kaplan–Meier curves of regional failure based on the presence of extracapsular nodal spread in the parotid gland or

neck. (B) Kaplan–Meier curves comparing regional failure rates based on whether patients received adjuvant radiotherapy. ECS,

extracapsular spread.

Table 4. Adverse clinicopathological data by treatment modality and institution.

Surgery alone Surgery þ RT

Adverse pathology SHNCI, 21 patients n (%) Westmead, 12 patients n (%) SHNCI, 77 patient n (%) Westmead, 58 patients n (%)

Involved margins 3 (14.3) 4 (33.3) 10 (13.0) 35 (60.3)

Extracapsular spread 8 (44.4) 6 (50) 32 (54.2) 44 (75.9)

Poorly differentiated 2 (9.5) 8 (66.7) 18 (23.4) 19 (32.8)

Perineural invasion 5 (23.8) 0 (0.0) 8 (10.4) 2 (3.4)

Lymphovascular invasion 1 (4.8) 0 (0.0) 1 (1.3) 2 (3.4)

Immunosuppression 0 (0.0) 0 (0.0) 0 (0.0) 2 (3.4)

Abbreviations: RT, adjuvant radiotherapy; SHNCI, Sydney Head and Neck Cancer Institute.

Low-Risk Nodal Cutaneous SCC HEAD & NECK—DOI 10.1002/hed March 2012 367

as standard treatment.4–8 Although excellent locore-gional control rates can be expected with such anapproach, the acute and long-term side effects, psycho-social impact, and economic burden of adjuvant radio-therapy warrant consideration.14 With this in mind,the current study was designed to determine whetherthere is a ‘‘low-risk’’ group of patients with metastaticcutaneous SCC to the parotid gland and/or neck thatmay be suitable for treatment with surgery alone.

In this retrospective multi-institution analysis,we used the staging system recently developed byForest et al10 to identify 168 patients with metastaticcutaneous SCC to the parotid gland or neck, classifiedas N1S3 stage I (single nodal metastasis �3 cm insize). The data suggest that within this group ofpatients with a single small node, there is a low-risksubset defined by the lack of ECS who do well when

treated with surgery alone. This was reflected in a 5-year DSS rate of 100%. Although 1 of 19 patients(5%) in this low-risk subset developed regional recur-rence, this was successfully salvaged with surgeryand adjuvant radiotherapy.

The ITEM prognostic model developed by Oddoneet al12 does not take into account the number andsize of involved nodes which are integral to definingthe low-risk patient in the current study. Despitethis, the ITEM staging system defines other criterionthat makes patients ineligible for single modalitytherapy, in particular ECS. Interestingly, despite 43%of the surgery alone cohort having a high-risk ITEMscore in the current study, the 5-year DSS rate of97% was excellent in contrast to the 44% 5-year DSSreported by Oddone et al.12 This probably reflects dif-ferences in tumor biology observed in patients with

FIGURE 3. (A) Kaplan–Meier curves of disease-specific survival in patients without extracapsular spread (n ¼ 66), based on whether

patients had surgery alone or surgery and adjuvant radiotherapy. (B) Kaplan–Meier curves of disease-specific survival in patients with.

Extracapsular spread (ECS; n ¼ 102), based on whether patients underwent surgery alone or surgery and adjuvant radiotherapy.

FIGURE 2. (A) Kaplan–Meier curves comparing disease-specific survival based on the presence or absence of extracapsular nodal

spread. (B) Kaplan–Meier curves comparing disease-specific survival based on whether patients received adjuvant radiotherapy. ECS,

extracapsular spread.

368 Low-Risk Nodal Cutaneous SCC HEAD & NECK—DOI 10.1002/hed March 2012

small solitary nodal metastases and difficulties inapplying a prognostic scoring system in a cohort ofpatients in which 1 criteria is, by definition, alwayshigh risk (ie, treatment modality).

The low-risk subset (single node �3 cm with noECS) represents a minority, comprising only 35% ofpatients with head and neck metastatic cutaneousSCC in our series. This subset is further reduced bythe presence of adverse factors making the additionof adjuvant radiotherapy preferable, such as perineu-ral invasion, involved or close surgical margins,immunosuppression, and concern regarding tumorcontamination of the surgical field.

Ultimately, only 19 patients (9%) who fulfilledthe low-risk criteria did not receive adjuvant radio-therapy. It is acknowledged that the reasons dictat-ing omission of adjuvant radiotherapy were notalways disease-related in this retrospective series,and factors such as physician preference and patientrefusal must be considered. This is reflected in thefact that 2 of these patients had involved marginsand 4 had perineural invasion. On the other hand,many of the patients that did receive adjuvant radio-therapy did so based on institutional protocol ratherthan tailoring treatment to the individual after riskstratification. We estimate that approximately 15%of patients may be suitable for single-modality ther-apy with surgery alone after consideration of all rel-evant clinicopathological information. Clearly, thecurrent study has significant limitations due to itsretrospective design and our hypothesis needs to bevalidated by future prospective trials before accep-tance into routine clinical practice.

With only 11 disease-related deaths and 19 re-gional recurrences, there were insufficient events inour study population to include all potentially rele-vant prognostic factors in a multivariable model.However, given that the need for adjuvant radiother-apy is determined by the presence of adverse histolo-gical features, it may be considered a surrogatemarker for their presence. This is reflected in thefact that we found no significant difference in re-gional failure (p ¼ .54) or DSS (p ¼ .40) when com-paring patients based on whether they had receivedadjuvant radiotherapy, suggesting that radiotherapyis compensating for adverse features not measuredin this study.

Whereas adjuvant radiotherapy has an integralrole in the management of metastatic cutaneous SCCof the head and neck, it may be associated with con-siderable morbidity including the acute toxicities ofskin erythema, ulceration, mucositis, xerostomia,anorexia, lethargy, and temporary loss of taste, aswell as late effects which may include dental caries,hearing loss, neck fibrosis, hypothyroidism, acceler-ated atherosclerosis, osteoradionecrosis, and rarelyradiation-induced cancers.14 In addition, postopera-tive radiotherapy is a major inconvenience forpatients and frequently involves multiple dental

extractions and is associated with significant cost andresource demands on overburdened healthcare sys-tems with escalating medical costs. Of course, thisneeds to be balanced against the additional morbidityof recurrence in the parotid gland bed or neck, and assuch, a low threshold for adjuvant radiotherapy isappropriate at present.

All these factors must be synthesized before rec-ommendations for parotid gland and neck irradiationare given to patients. Based on the results of thisstudy, patients with a single node �3 cm in size andno ECS may be suitable for single-modality therapywith surgery alone if no other adverse clinicopatho-logical features are present. However, if omission ofadjuvant radiotherapy is being considered, it is im-portant that patients have undergone a parotidec-tomy and appropriate neck dissection based on thelocation of the primary tumor,15 because there is asignificant risk of occult cervical node16,17 and parotidgland18,19 involvement in the presence of clinical pa-rotid gland and cervical node metastases, respectively.This allows accurate pathological staging and ensuresthat there is only a single small node involved.

CONCLUSION

In this multi-institutional analysis, we have identi-fied a subset of patients with metastatic nodal cuta-neous SCC of the head and neck who are at lowrisk of regional recurrence and could be treated withsurgery alone. This group was defined by a singlenode �3 cm in size without ECS and no otheradverse clinicopathological features. We believe theregional failure rate of 5% in our series is accepta-ble, particularly in view of successful salvage withsurgery and adjuvant radiotherapy yielding 100% 5-year DSS rates. Approximately 15% of all patientswith regional metastases from head and neck cuta-neous SCC may be suitable for such an approach,which aims to avoid the morbidity, inconvenience tothe patient, and resource demands associated withadjuvant radiotherapy.

REFERENCES

1. Staples MP, Elwood M, Burton RC, Williams JL, Marks R, GilesGG. Non-melanoma skin cancer in Australia: the 2002 nationalsurvey and trends since 1985. Med J Aust 2006;184:6–10.

2. Moore BA, Weber RS, Prieto V, et al. Lymph node metastasesfrom cutaneous squamous cell carcinoma of the head and neck.Laryngoscope 2005;115:1561–1567.

3. Joseph MG, Zulueta WP, Kennedy PJ. Squamous cell carcinomaof the skin of the trunk and limbs: the incidence of metastasesand their outcome. Aust N Z J Surg 1992;62:697–701.

4. Taylor BW Jr, Brant TA, Mendenhall NP, et al. Carcinoma of theskin metastatic to parotid area lymph nodes. Head Neck1991;13:427–433.

5. delCharco JO, Mendenhall WM, Parsons JT, Stringer SP, CassisiNJ, Mendenhall NP. Carcinoma of the skin metastatic to the pa-rotid area lymph nodes. Head Neck 1998;20:369–373.

6. Shimm DS, Wilder RB. Radiation therapy for squamous cell car-cinoma of the skin. Am J Clin Oncol 1991;14:383–386.

7. Veness MJ, Palme CE, Smith M, Cakir B, Morgan GJ, Kalnins I.Cutaneous head and neck squamous cell carcinoma metastatic to

Low-Risk Nodal Cutaneous SCC HEAD & NECK—DOI 10.1002/hed March 2012 369

cervical lymph nodes (nonparotid): a better outcome with surgeryand adjuvant radiotherapy. Laryngoscope 2003;113:1827–1833.

8. Veness MJ, Morgan GJ, Palme CE, Gebski V. Surgery and adju-vant radiotherapy in patients with cutaneous head and necksquamous cell carcinoma metastatic to lymph nodes: combinedtreatment should be considered best practice. Laryngoscope2005;115:870–875.

9. O’Brien CJ, McNeil EB, McMahon JD, Pathak I, Lauer CS,Jackson MA. Significance of clinical stage, extent of surgery, andpathologic findings in metastatic cutaneous squamous carcinomaof the parotid gland. Head Neck 2002;24:417–422.

10. Forest VI, Clark JJ, Veness MJ, Milross C. N1S3: a revised stag-ing system for head and neck cutaneous squamous cell carci-noma with lymph node metastases: results of 2 AustralianCancer Centers. Cancer 2010;116:1298–1304.

11. American Joint Committee on Cancer. AJCC Cancer StagingHandbook, 7th ed. New York: Springer Verlag; 2010.

12. Oddone N, Morgan GJ, Palme CE, et al. Metastatic cutaneoussquamous cell carcinoma of the head and neck: the Immunosup-pression, Treatment, Extranodal spread, and Margin status(ITEM) prognostic score to predict outcome and the need toimprove survival. Cancer 2009;115:1883–1891.

13. Clayman GL, Lee JJ, Holsinger FC, et al. Mortality risk fromsquamous cell skin cancer. J Clin Oncol 2005;23:759–765.

14. August M, Wang J, Plante D, Wang CC. Complications associ-ated with therapeutic neck radiation. J Oral Maxillofac Surg1996;54:1409–1415; discussion 1415–1416.

15. Ebrahimi A, Moncrieff MD, Clark JR, et al. Predicting the pat-tern of regional metastases from cutaneous squamous cell carci-noma of the head and neck based on location of the primary.Head Neck 2010;32:1288–1294.

16. O’Brien CJ, McNeil EB, McMahon JD, Pathak I, Lauer CS. Inci-dence of cervical node involvement in metastatic cutaneous malig-nancy involving the parotid gland. Head Neck 2001;23:744–748.

17. Dona E, Veness MJ, Cakir B, Morgan GJ. Metastatic cutaneous squa-mous cell carcinoma to the parotid: the role of surgery and adjuvantradiotherapy to achieve best outcome. ANZ J Surg 2003;73:692–696.

18. Vauterin TJ, Veness MJ, Morgan GJ, Poulsen MG, O’Brien CJ.Patterns of lymph node spread of cutaneous squamous cell carci-noma of the head and neck. Head Neck 2006;28:785–791.

19. Ch’ng S, Maitra A, Lea R, Brasch H, Tan ST. Parotid metasta-ses—an independent prognostic factor for head and neck cutane-ous squamous cell carcinoma. J Plast Reconstr Aesthet Surg2006;59:1288–1293.

370 Low-Risk Nodal Cutaneous SCC HEAD & NECK—DOI 10.1002/hed March 2012