Original Research

Otolaryngology–
Head and Neck Surgery

Impact of Biopsy Modality on the 1–6

Ó American Academy of
Otolaryngology—Head and Neck
Management of Cutaneous Melanoma of Surgery Foundation 2017
Reprints and permission:
the Head and Neck sagepub.com/journalsPermissions.nav
DOI: 10.1177/0194599817740568
http://otojournal.org

Arya W. Namin, MD1 and Robert P. Zitsch III, MD1

No sponsorships or competing interests have been disclosed for this article. Received June 27, 2017; revised August 23, 2017; accepted October
13, 2017.

Abstract
Objective. The purpose of this study was to examine how

A
n estimated 2.2% of men and women in the United
biopsy modality affects the treatment course and outcomes States will be diagnosed with melanoma of the skin
of patients with cutaneous melanoma of the head and neck. at some point during their lifetimes, and the inci-
Specifically, we investigated if partial biopsy techniques are dence continues to rise.1 Melanoma of the head and neck
associated with positive margins on definitive wide local exci- region accounts for approximately 20% of all cutaneous
sion (DWLE), the need for early reoperation to obtain ade- melanoma cases.2,3 It poses unique challenges in diagnosis
quate margins or sentinel lymph node biopsy, and survival. and management given the complex form and function of
Study Design. Retrospective case series. the region. One challenge is making the decision of what
biopsy modality to utilize. Classic teaching holds that all
Setting. Tertiary care academic center. lesions suspicious for melanoma be biopsied through an
Subjects and Methods. Subjects (N = 170) included all patients excisional technique with narrow margins.4 According to
who were surgically treated for primary cutaneous melanoma current National Comprehensive Cancer Network guide-
of the head and neck at the University of Missouri–Columbia lines, lesions suspicious for melanoma should receive an
between January 1, 2000, and December 31, 2015. For analy- excisional biopsy with 1- to 3-mm margins, including deep
sis, patients were divided into 4 groups based on biopsy mod- shave/saucerization but excluding superficial shave.5 An
ality: shave (n = 61), excisional (n = 62), punch (n = 33), and excisional biopsy is defined as excising the macroscopic
incisional (n = 14). extent of the lesion in the subcutaneous tissue plane so that
the entire dermis deep to the lesion is excised. A shave
Results. The shave biopsy group (P = .0324) and the punch biopsy utilizes a sharp blade, and the lesion is shaved off.
biopsy group (P = .0479) were significantly more likely to The depth of the shave biopsy is thus poorly controlled. The
have positive margins on DWLE. The shave biopsy group (P = poor control in the depth of a shave biopsy thus portends
.0042) and the punch biopsy group (P = .0479) were also sig- great variability in this technique. An incisional biopsy is
nificantly more likely to need early reoperation. The mean defined as an excision of a portion of the lesion in the sub-
number of sentinel nodes and incidence of positive sentinel cutaneous tissue plane. The depth of the biopsy is therefore
nodes detected on pathologic examination did not differ sig- better controlled; however, the portion of the lesion that is
nificantly across biopsy modality (P = .3600). Overall survival biopsied is at the discretion of the provider. The punch
(P = .4605) and disease-free survival (P = .5011) did not differ biopsy technique uses a circular blade of variable diameter
significantly among the groups. to remove either a portion or the entirety of a lesion. The
Conclusions. Patients diagnosed with shave and punch biopsy lesion is usually removed in the subcutaneous tissue plane
techniques are significantly more likely to have positive mar- with a punch biopsy technique. However, owing the high
gins after DWLE and more frequently require early reopera-
tion. Biopsy modality does not appear to influence the
1
number of sentinel nodes detected, the incidence of detect- Department of Otolaryngology–Head and Neck Surgery, University of
Missouri, Columbia, Missouri, USA
ing regional metastases in sentinel nodes, the overall sur-
vival, or the disease-free survival. This article was presented at the 2017 AAO-HNSF Annual Meeting & OTO
Experience; September 10-13, 2017; Chicago, Illinois.
Corresponding Author:
Keywords Robert P. Zitsch III, MD, Department of Otolaryngology–Head and Neck
Surgery, University of Missouri, One Hospital Dr, MA314, Columbia, MO
melanoma, head and neck, biopsy, shave, excisional, margin, 65212, USA.
sentinel lymph node biopsy, prognosis Email: zitschr@health.missouri.edu
2 Otolaryngology–Head and Neck Surgery

Diagnosis Definive Surgical Neck dissecon +/- parodectomy

Management aer posive SLNB
Excisional
Shave Wide Local Need for early unexpected
Excision +/- SLNB reoperaon
Punch based on biopsy
1. Re-excision aer posive
Incisional findings
margins
2. SLNB aer T upstaging
Figure 1. Surgical management of primary cutaneous melanoma. SLNB, sentinel lymph node biopsy.

number of skin lesions seen by primary care physicians and Note that this study was not designed to measure the
dermatologists, compounded with the difficulty in accu- appropriateness of different biopsy modalities in assessing
rately predicting melanoma based on clinical appearance pigmented cutaneous lesions of the head and neck. To do so
alone, melanoma is not infrequently diagnosed by partial properly, the study population would have had to include a
biopsy techniques (shave, incisional, and punch). group of patients with pigmented lesions prior to pathologic
Past evidence suggested that patients diagnosed with diagnosis. This study was designed to provide the head and
cutaneous melanoma of the head and neck by partial biopsy neck surgeon with information regarding whether biopsy
techniques had worse disease-specific survival when com- modality is associated with treatment obstacles and out-
pared with those diagnosed by excisional biopsy.6 More comes among patients diagnosed with cutaneous melanoma
recently, multiple studies showed no difference in prognosis of the head and neck.
based on biopsy modality; however, all these studies
included cutaneous melanoma from various regions of the Methods
body and did not specifically address the head and neck.2,7,8
The University of Missouri Institutional Review Board
No recent data have been published regarding the effect of
approved this work. The study is a retrospective review of
biopsy modality on the prognosis of cutaneous melanoma of
all patients who were treated surgically for primary cuta-
the head and neck. The current literature suggests that
neous melanoma of the head and neck at the University of
patients who are diagnosed with partial biopsy techniques
Missouri–Columbia between January 1, 2000, and December
are much more likely to be T-upstaged after definitive wide
31, 2015. Patients with previously diagnosed melanoma and
local excision (DWLE), but the data have not demonstrated
patients who were referred for treatment of recurrent disease
a negative impact on survival.2,7-10 Cutaneous melanoma of were excluded from the study. Patients were excluded if the
the head and neck is more likely to be T-upstaged after
pathology report from the initial biopsy was unavailable.
DWLE when compared with cutaneous melanoma of other
Demographic, clinical, and pathologic data were recorded.
body locations.11
The biopsy modality, stage of melanoma based on biopsy,
The typical treatment of primary cutaneous melanoma
and stage of melanoma after DWLE 6 SLNB were also
often involves 2 procedures (Figure 1): the first is a biopsy
recorded.
to obtain a diagnosis; the second is definitive surgical man-
The final depth assigned to each lesion was based on the
agement (DWLE 6 sentinel lymph node biopsy [SLNB])
deeper of the measurements made on (1) the initial biopsy
selected per the staging information obtained from the or (2) the pathologic analysis of the DWLE specimen. The
biopsy. Some patients will need to return to the operating
incidence of T-upstaging after the DWLE was then com-
room after definitive surgical management for a third proce-
pared across biopsy modalities (excisional, shave, punch,
dure, which we define as an early unexpected reoperation so
and incisional). The incidence of positive margins on the
that surgeons can (1) perform an SLNB based on T-upstaging
DWLE specimen was also recorded and compared across
and patient inclusion criteria or (2) obtain adequate margins
biopsy modalities. Need for early unexpected reoperation
after finding positive margins on the DWLE specimen (ie,
was defined as follows: (1) reexcision to obtain adequate
per the diagnostic details of the final pathology report). An
margins after positive margins were found on the DWLE
important outcome to measure for patients undergoing treat- specimen or (2) SLNB after T-upstaging made the patient
ment for melanoma is how biopsy modality affects the need
eligible for it. Neck dissection 6 parotidectomy after posi-
for early unexpected reoperation. No data have been pub-
tive sentinel lymph node was not included in the ‘‘need for
lished on the association between biopsy modality and the
early unexpected reoperation’’ group, given that this could
need for such additional operations after DWLE among
not be predicted per the initial biopsy alone (Figure 1). The
patients with cutaneous melanoma of the head and neck.
number of sentinel lymph nodes found on pathologic analy-
The purpose of this study was to examine whether biopsy
sis and the incidence of positive SLNB were compared
modality (1) correlates with finding positive margins on the
across biopsy modalities. Last, overall survival and disease-
DWLE specimen, (2) is predictive of early unexpected reo- free survival were compared across biopsy modalities.
peration, and (3) influences overall and disease-free survival.
Namin and Zitsch 3

Table 1. Tumor Characteristics across Biopsy Modalities.a

Shave (n = 61) Excisional (n = 62) Punch (n = 33) Incisional (n = 14) P Value

Mean age, y 64 59 66 64 .3160

Stage .6781
1a 20 (33) 12 (19) 6 (19) 6 (44)
1b 16 (26) 14 (23) 9 (27) 3 (21)
2a 9 (15) 12 (19) 11 (33) 3 (21)
2b 5 (8) 11 (18) 4 (12) 0 (0)
2c 5 (8) 6 (10) 2 (6) 1 (7)
3a 3 (5) 4 (6) 1 (3) 1 (7)
3b 2 (3) 3 (5) 0 (0) 0 (0)
3c 1 (2) 0 (0) 0 (0) 0 (0)
Histology .6144
Superficial spreading 12 (20) 17 (27) 9 (27) 4 (28)
Lentigo maligna melanoma 21 (34) 16 (26) 12 (37) 5 (36)
Nodular 12 (20) 12 (19) 2 (6) 0 (0)
Desmoplastic 2 (3) 3 (5) 3 (9) 0 (0)
Granular 0 (0) 2 (3) 0 (0) 0 (0)
Spindle cell 1 (2) 0 (0) 0 (0) 0 (0)
Blue nevus like 0 (0) 1 (2) 0 (0) 0 (0)
Not specified 13 (21) 11 (18) 7 (21) 5 (36)
Subsite .6897
Scalp 24 (39) 19 (31) 7 (21) 5 (37)
Forehead 5 (8) 7 (11) 4 (12) 1 (7)
Cheek 13 (21) 18 (29) 9 (28) 1 (7)
Nose 1 (2) 0 (0) 3 (9) 1 (7)
Ear 7 (11) 9 (14) 5 (15) 3 (21)
Cutaneous lip 1 (2) 1 (2) 1 (3) 0 (0)
Eyelid 1 (2) 0 (0) 1 (3) 0 (0)
Neck 9 (15) 8 (13) 3 (9) 3 (21)
a
Values are presented as n (%) unless noted otherwise.

Statistical Analysis and incisional group (P = .0319). No other pairs of groups dif-
Analysis was done with SAS 9 (SAS Institute Inc, Cary, fered significantly (Table 2).
North Carolina). For categorical data, groups (biopsy modal- Twelve patients had positive margins based on the final
ities) were compared with exact chi-square to test the null pathology report of the DWLE specimen (11 with mela-
hypothesis of homogeneity of proportions among the 4 noma in situ at the margin and 1 with malignant melanoma
groups. When there was evidence of an overall difference at the margin). The shave biopsy group (P = .0324) and the
among groups, they were compared pairwise with Fisher’s punch biopsy group (P = .0479) were significantly more
exact test. The numeric variables tended to have skewed likely than the excisional biopsy group to have positive mar-
distributions. For these variables, groups were compared gins on the DWLE specimen (Table 3). No other pairs of
with Kruskal-Wallis tests. Survival curves were estimated groups differed significantly. Positive margins were found
with the Kaplan-Meier product limit estimates. Groups were in 14.3% (6 of 42) of women and 4.7% (6 of 128) of men,
compared with the log-rank test. Results with P values \.05 which was not statistically significant (P = .0743). Among
were considered statistically significant. the 12 patients with positive margins on DWLE, the scalp,
cutaneous lip, cheek, forehead, nose, and ear subsites were
Results represented. Among patients who had positive margins on
One hundred seventy patients met inclusion criteria: 42 DWLE, subsite breakdown was as follows: scalp, 9% (5 of
women and 128 men, with a mean age of 63. There was no 55); lip, 67% (2 of 3); cheek, 5% (2 of 41); forehead, 6% (1
significant difference across biopsy modalities according to of 17); nose, 20% (1 of 5); and ear, 4% (1 of 24).
patient age, final pathologic stage, melanoma subtype, or Fifteen patients needed early unexpected reoperation
location subsite (Table 1). The proportion of T-upstaging (Table 4). Twelve of these patients needed reoperation due
after DWLE was significantly different in the excisional group to positive margins, and 3 needed reoperation for SLNB
than the shave group (P \ .0001), punch group (P = .0013), after T-upstaging. Compared with the excisional biopsy
4 Otolaryngology–Head and Neck Surgery

Table 2. Upstaged Based on T Stage.

Upstaged

Biopsy Modality Total, n n % 95% CI, % P Valuea

All patients 170 24 14.1

Excisional 62 0 0 0-5.8
Shave 61 16 26.2 15.8-39.1 \.0001
Punch 33 6 18.2 7.0-35.5 .0013
Incisional 14 2 14.3 1.8-42.8 .0319
a
Pairwise comparison with Fisher’s exact test between the respective sub-
groups and the excisional biopsy group.

Table 3. Positive Margins.

Figure 2. Overall survival across biopsy modalities. No significant
Positive Margins difference among the biopsy modality groups (log-rank test, P =
.4605).
Biopsy Modality Total, n n % 95% CI, % P Valuea

All patients 170 12 7.1

detecting a positive sentinel node was 13% (4 of 31), 13%
Excisional 62 1 1.6 0.04-8.7
(5 of 38), 11% (1 of 9), and 5% (1 of 19) for shave, exci-
Shave 61 7 11.5 4.7-22.2 .0324
sional, incisional, and punch, respectively. The incidence of
Punch 33 4 12.1 3.4-28.2 .0479
detecting positive nodes did not differ across biopsy modal-
Incisional 14 0 0 0.0-23.2
ities (Fisher’s exact test, P = .9001).
a
Pairwise comparison with Fisher’s exact test between the respective sub- Thirty-five patients (20.6%) developed a recurrence of
groups and the excisional biopsy group. melanoma during the study period. The incidence of recur-
rence was 14.3% (6 of 42) among women and 22.7% (29 of
128) among men, which was not statistically significant (P =
Table 4. Need for Early Unexpected Reoperation. .2797). Eleven patients developed a local recurrence (6.5%).
The incidence of local recurrence for shave, excisional, inci-
Reoperation
sional, and punch biopsy groups was 4.9%, 6.5%, 14.3%,
Biopsy Modality Total, n n % 95% CI, % P Valuea and 6.1% respectively. This was not statistically significant
(P = .56).
All patients 170 15 8.8 For overall survival (Figure 2), we did not find a signifi-
Excisional 62 1 1.6 0.0-8.7 cant difference among groups (log-rank test, P = .4605). For
Shave 61 10 16.4 8.2-28.1 .0042 disease-free survival (Figure 3), we did not find a signifi-
Punch 33 4 12.1 3.4-28.2 .0479 cant difference among groups (log-rank test, P = .5011).
Incisional 14 0 0 0.0-23.2
a
Discussion
Pairwise comparison with Fisher’s exact test between the respective sub-
groups and the excisional biopsy group. Biopsy modality is an important topic in the management of
cutaneous melanoma because depth of melanoma, mitoses,
and ulceration are used to stage and guide initial treatment.
Excisional biopsy has been advocated for the diagnosis of
group, the shave biopsy group (P = .0042) and the punch melanoma because it samples the entirety of the lesion and
biopsy group (P = .0479) were significantly more likely to provides the most accurate pathologic information.8,9 We
need early unexpected reoperation. No other pairs of groups emphasize that we are not drawing any conclusions regard-
differed significantly. Eleven underwent the additional oper- ing the appropriateness of different biopsy techniques in
ation, while 3 patients with positive margins and 1 for assessing cutaneous pigmented lesions of the head and
whom an SLNB became indicated elected for observation. neck. Dermatologists and primary care physicians see pig-
Ninety-seven patients underwent SLNB. The mean mented lesions on an ongoing basis. Shave and punch
number of sentinel nodes detected on pathologic examina- biopsy techniques involve significantly less time, making
tion was 3.5, 2.6, 2.9, and 2.5 for shave, excisional, inci- them frequently utilized methods. The purpose of this study
sional, and punch, respectively. The mean number of was to determine if biopsy modality is associated with treat-
sentinel nodes detected did not differ significantly across ment obstacles and outcomes among patients diagnosed
biopsy modalities (Kruskal-Wallis test, P = .3600). Eleven with cutaneous melanoma of the head and neck. The data
patients had positive sentinel lymph nodes. The incidence of from this study support prior published evidence that partial
Namin and Zitsch 5

cutaneous melanoma of the head and neck versus other body

regions.11
Nevertheless, the data from this study show that it is
more commonly the positive margins that necessitate early
unexpected reoperation rather than SLNB. Specifically,
patients in the shave biopsy and punch biopsy groups were
significantly more likely to have positive margins after
DWLE and consequently needed early unexpected reopera-
tion. These data provide insight into recent studies that
found that cutaneous melanoma of the head and neck is
more likely to be upstaged and more likely to have positive
margins on the DWLE specimen when compared with cuta-
neous melanoma of other regions of the body.11,13
Last, there is suspicion that partial biopsy techniques
may negatively affect SLNB and prognosis. This study con-
Figure 3. Disease-free survival across biopsy modalities. No signif- firms that biopsy modality does not influence lymph node
icant difference among the biopsy modality groups (log-rank test, identification for the SLNB procedure, as evidenced by the
P = .5011). number of sentinel nodes detected. More important, the inci-
dence of detecting regional metastases in the sentinel nodes,
the overall survival, and the disease-free survival seem to be
biopsy techniques are more prone to T-upstaging after unaffected by biopsy modality.2,7,8
pathologic examination of the subsequent DWLE speci-
men.8,9 The implications of T-upstaging after DWLE are Conclusions
twofold: (1) patients may become candidates for SLNB,
Biopsy techniques other than narrow excision for suspected
and/or (2) the patient must come to terms with a changed
cutaneous melanoma of the head and neck region have been
prognosis, which can be substantially worse in some cases.
subject of considerable controversy. They have been purported
The recommended excision margins may also change after
to result in broader dissemination of melanoma to regional
upstaging; however, the data regarding appropriate clinical
lymph nodes as well as compromised survival. This study
resection margins at the time of wide local excision are very
shows that patients diagnosed with the shave biopsy and punch
limited with respect to cutaneous melanoma of the head and
neck.5,12 Current recommendations suggest that an excision biopsy techniques are significantly more likely to be T-upstaged
after DWLE, are more likely to have positive margins after
margin of 1 to 2 cm is sufficient for primary cutaneous mel-
DWLE, and more frequently require early unexpected reopera-
anoma of the head and neck.5,12 The high incidence of posi-
tion to obtain adequate margins or SLNB. Biopsy modality
tive margins after DWLE among those in the shave biopsy
does not appear to influence the number of sentinel nodes
and punch biopsy groups was an unexpected finding in this
detected, the incidence of detecting regional metastases in senti-
study. To be sure, determining the lesion borders can be
nel nodes, the overall survival, or the disease-free survival.
more difficult after shave biopsy, which could be a contri-
buting factor to the high incidence of positive margins in Acknowledgments
this group. Excisional biopsy, by definition, removes the We acknowledge Rebecca Schneider, MS, for assistance in obtain-
macroscopic extent of the lesion, which helps explain the ing approval from the Institutional Review Board at the University
low incidence of positive margins on the formal DWLE spe- of Missouri, Peggy Canoy for assistance in performing the billing
cimen in this subgroup. The data from this study provide search, and Richard W. Madsen, PhD, for assistance with the sta-
additional insights into the relatively high incidence of posi- tistical analysis.
tive margins in head and neck melanoma versus melanoma
in other areas of the body. This observation was tradition- Author Contributions
ally thought to be due to adjacent vital structures that could
Arya W. Namin, study design, Institutional Review Board
not be violated. approval, data acquisition, interpretation of data, drafting of manu-
An additional indication for early unexpected reoperation script, critical revision of manuscript, final approval of the version
after DWLE is SLNB after T-upstaging based on the to be published, agreement to be accountable for all aspects of the
DWLE specimen. The data from this study show that 1.8% work; Robert P. Zitsch III, study design, data interpretation, criti-
(3 of 170) of patients qualified for SLNB due to T-upstaging, cal revision of manuscript, final approval of version to be pub-
which is slightly higher when compared with a study by lished, agreement to be accountable for all aspects of the work.
Kaiser et al, which found that 0.5% (4 of 853) qualified for
Disclosures
SLNB for T-upstaging.3 The need to return to the operating
Competing interests: None.
room for SLNB when T-upstaging resulted is slightly higher
in the head and neck than in other regions of the body and is Sponsorships: None.
very possibly due to the higher incidence of T-upstaging in Funding source: None.
6 Otolaryngology–Head and Neck Surgery

References 7. Martin RCG, Scoggins CR, Ross MI, et al. Is incisional biopsy
of melanoma harmful? Am J Surg. 2005;190:913-917.
1. Surveillance, Epidemiology, and End Results Program. Cancer
8. Egnatios GL, Dueck AC, Macdonald JB, et al. The impact of
stat facts: melanoma of the skin. https://seer.cancer.gov/stat
biopsy technique on upstaging, residual disease, and outcome
facts/html/melan.html. Accessed June 13, 2017.
in cutaneous melanoma. Am J Surg. 2011;202:771-777.
2. Mills JK, White I, Diggs B, Fortino J, Vetto JT. Effect of
9. Ng JC, Swain S, Dowling JP, Wolfe R, Simpson P, Kelly JW.
biopsy type on outcomes in the treatment of primary cutaneous
The impact of partial biopsy on histopathologic diagnosis of
melanoma. Am J Surg. 2013;205:585-590.
cutaneous melanoma. Arch Dermatol. 2010;146:234-239.
3. Kaiser S, Vassell R, Pinckney RG, Holmes TE, James TA.
10. Karimipour DJ, Schwartz JL, Wang TS, et al. Microstaging
Clinical impact of biopsy method on the quality of surgical
accuracy after subtotal incisional biopsy of cutaneous mela-
management in melanoma. J Surg Oncol. 2014;109:775-779.
noma. J Am Acad Dermatol. 2005;52:798-802.
4. Ricotti C, Cather J, Cockerell CJ. Pathology of melanoma:
11. Etzkorn JR, Sharkey JM, Grunyk JW, Shin TM, Sobanko JF,
interpretation and new concepts. In: Rigel DS, Robinson JK,
Miller CJ. Frequency of and risk factors for tumor upstaging
Ross M, et al, eds. Cancer of the Skin. 2nd ed. Philadelphia,
after wide local excision of primary cutaneous melanoma. J
PA: Elsevier Inc; 2011.
Am Acad Dermatol. 2017;77:341-348.
5. National Comprehensive Cancer Network. NCCN Clinical
12. Haigh PI, DiFronzo LA, McCready DR. Optimal excision mar-
Practice Guidelines in Oncology: Melanoma (Version 1.2017).
gins for primary cutaneous melanoma: a systematic review
Fort Washington, PA: National Comprehensive Cancer
and meta-analysis. Can J Surg. 2003;46:419-426.
Network.
13. Al-Qurayshi Z, Hassan M, Srivastav S, et al. Risk and survival
6. Austin JR, Byers RM, Brown WD, Wolf P. Influence of
of patients with head and neck cutaneous melanoma: national
biopsy on the prognosis of cutaneous melanoma of the head
perspective. Oncology. 2017;93:18-28.
and neck. Head Neck. 1996;18:107-117.