University Hospital Ulm Institute of Pathology

Chairman: Prof. Dr. med. Peter Möller

A Diagnostic Algorithm to Distinguish Desmoplastic from Spindle Cell Melanoma

Thesis for the Degree Doctor of Medicine (Dr. med.)

Presented to

the Medical Faculty of the Ulm University

by

Stephanie Ellen Weißinger

born in Nürtingen

2013

Published online 20th of September 2013: Modern Pathology (2014) 27,

524–534; doi:10.1038/modpathol.2013.162;

http://www.nature.com/modpathol/journal/v27/n4/full/modpathol2013162a.html

Dean of the Faculty: Prof. Dr. T. Wirth

Reviewer 1: Prof. Dr. P. Möller

Reviewer 2: PD Dr. A. Kleger

Date of Examination: 18th of June 2015

I

Index

Index I

Abbreviations II

1. Introduction 1

2. Material and Methods 3 2.1. Case Selection 3 2.2. Immunohistochemistry 3 2.3. Molecular Genetic Analysis 4 2.4. Literature Review and Biomarkerselection 4 2.5. Statistical Analysis and Data Processing 5

3. Results 6

4. Discussion 144.1. The Terminology remains confusing 15 4.2. The Prevalence of Spindle Cell and Desmoplastic Melanoma Maybe Higher Than Previously Assumed 15 4.3. The diagnostic categories are part of an extended biological spectrum 15 4.4. There is a Dearth of Specific Desmoplastic Melanoma Markers 16

5. Summary 18

6. References 19 6.1 Software Packages and Internet Resources 23

Appendix 24

Acknowledgements 36

Curriculum vitae 37

II

Abbreviations

Abbreviation Explanation(s)

aa amino acid + sum of positively staining cases Amp amplification probe BA/RA break-apart/rearrangement probe Bx biopsy c Cytoplasmic or centromeric probe cat catalogue CEP centromere enumeration probe conc concentration EDTA ethylenediaminetetraacetic Ex excsision f female G:CN gene to copy-number ratio GIST gastrointestinal stromal tumor HN head and neck ID case identification number IHC immunohistochemistry LE lower extremity m monoclonal M Male or mouse MEM membranous MW microwave N Nuclear or no pigment present N/A not applicable No number NPV negative predictive value NT non-tumor ON oncology probeset p polyclonal PC pressure cooker PMID pubmed identification number PPV positive predictive value

III

Abbreviation Explanation(s)

R rabbit Retrieval antigen retrieval S steamer T tumor t telomeric probe TRS target retrieval solution UE upper extremity V vessels Y pigment containing

Introduction

1

1. Introduction

Spindle cell- and desmoplastic melanoma are two distinct subtypes of

malignant melanoma that differ by clinical, histopathological, and prognostic

features (McCarthy et al., 2006). Histologically, both spindle cell and

desmoplastic melanoma are characterized by atypical, spindled, malignant

melanocytes. In desmoplastic melanoma, these malignant cells are intimately

admixed with ropy and dense collagen fibrils. Anatomically, desmoplastic

melanoma usually occurs in the head and neck region and presents clinically

with a flat or nodular, scar-like lesion. Desmoplastic melanoma is usually not

well-circumscribed and malignant cells track along pre-existing structures

such as skin appendages or nerves (e.g., neurotropic melanoma). This

specific growth pattern of desmoplastic melanoma is likely responsible for the

higher rate of local recurrences, when compared to other melanoma subtypes

(Chen et al., 2008; Hawkins et al., 2005; Jaroszewski et al., 2001; Maurichi et

al.). There is an ongoing debate whether sentinel lymph node biopsy should

be performed in desmoplastic melanoma because its metastatic rate is

somewhat lower when compared to conventional melanoma and spindle cell

melanoma (Broer et al.; Gyorki et al., 2003; Lens et al., 2005; Livestro et al.,

2005; Mohebati et al.; Murali et al.; Shaw et al., 2006). The overall prognosis

in desmoplastic melanoma tends to be better when compared stage-by-stage

to conventional melanoma (de Almeida et al., 2008; Livestro et al., 2005). In

contrast spindle cell melanoma is frequently amelanotic, can occur essentially

anywhere on the body, and presents typically with widespread metastatic

disease. Histologically spindle cell melanoma is often confused with other

(e.g. mesenchymal or neural) tumors. This is why knowledge of specific

immunophenotypes in spindle cell and desmoplastic melanoma can be helpful

(de Almeida et al., 2008; Livestro et al., 2005; Ohsie et al., 2008).

Due to these differences, diagnostic distinction of spindle cell and

desmoplastic melanoma is clinically relevant; however, currently mainly rests

on the amount of scar-like tissue (i.e., collagen content) (McCarthy et al.,

2006). Diagnostic distinction of spindle cell from desmoplastic melanoma is

not always straightforward because of partially overlapping features and

Introduction

2

heterogeneity in collagen content. One approach to resolve this diagnostic

challenge has been the introduction of an intermediate category (termed

‘mixed’ or ‘combined’ spindle/desmoplastic melanoma) that suggests a

seemingly continuous biological spectrum (Busam et al., 2004; Hawkins et al.,

2005; McCarthy et al., 2004; Mohebati et al.; Scolyer and Thompson, 2005;

Skelton et al., 1997). Formally, the diagnostic categories in these overtly

spindled melanomas are: spindle cell melanoma when the collagen content is

<10%, mixed spindle/desmoplastic melanoma when collagen content is 10-

90%, and desmoplastic melanoma when the collagen content is >90%

(Coupelon et al., 2012; Longacre et al., 1996; McCarthy et al., 2006;

McCarthy et al., 2004). From a practical perspective these formal definitions

are problematic. First, superficial or partial biopsies may preclude

assessment of the entire lesion. Second, lesional variability of the collagen

content makes the assignment of one quantitative value difficult. Third, the

immunoprofiles of spindle cell and desmoplastic melanoma differ not only

from conventional melanoma but also between spindle cell and desmoplastic

melanoma (Longacre et al., 1996). These immunophenotypic differences

indicate that collagen amount is not the only difference between desmoplastic

and spindle cell melanoma. Collectively, this situation argues for a simple,

reliable and practical assay for diagnostic distinction of desmoplastic from

spindle cell melanoma. It is surprising that despite the substantial body of

work on immunophenotypes in spindle cell and desmoplastic melanoma, a

direct comparative study of multiple diagnostic biomarkers has not been

performed.

Therefore, a comprehensive biomarker panel for the distinction of

desmoplastic from spindle cell melanoma and included biomarkers with

potential therapeutic implications was performed.

Material and Methods

3

2. Material and Methods

2.1 Case selection

After regulatory approval by the local Human Studies Committee, computer-

based free-text queries were combined with ICD-based searches to identify

patients with the diagnosis of spindle cell-, mixed-, or desmoplastic

melanoma. After review and confirmation of the primary diagnosis according

to established criteria (Longacre et al., 1996; McCarthy et al., 2006) (H&E-

based Ulm: BK, JKL and myself; New York: BAH, DNS) cases with sufficient

material were included. To minimize false diagnostic classification due to

tumor heterogeneity, primary excision specimen (35 of 38 cases) rather than

biopsies or re-excisions were selected whenever possible. The three included

core-punch biopsies were also primary diagnoses and demonstrated classical

features of spindle cell and desmoplastic melanoma (3 of 38 cases). Stringent

selection criteria were chosen because morphology-based diagnoses served

as the benchmark (‘gold-standard’) for subsequent biomarker assessments.

All samples had been formalin-fixed and paraffin-embedded. Masson

trichrome-staining was scored as positive when >10% of the lesional area

consisted of blue collagen fibrils. The New York samples were freshly cut and

unstained sections were combined into arrays using a slide-to-slide transfer

technique (Weissinger et al., 2013) as follows. After deparaffinization, tissues

were epoxy-resin-covered, lifted, microdissected into >2mm2 tiles and arrayed

onto separate recipient slides. The staining performance using STS-arrays

and regular histological sections was confirmed in the one cohort (Ulm) before

sections of the other cohort (New York) were arrayed, stained and analyzed. For the Ulm cases combinations of large-core (9 mm diameter) tissue

microarrays were used, traditional 2µm sections as well as slide-to-slide

transfer arrays, capturing the superficial aspect including the epidermis

whenever possible. Each array included control tissues composed of

epithelial, neuronal, basement membrane, tonsillar-, lymphatic-, and

collageneous tissue as well as conventional melanoma.

2.2 Immunohistochemistry

Immunohistochemistry was performed according to established protocols and

details are provided (Supplemental Table 1). Specific staining was

Material and Methods

4

ascertained by using built-in controls (see above), negative controls

(secondary antibody used without the primary antibody), and comparison of

staining with prior reports and data from publicly available resources

(www.proteinatlas.org; last accessioned 18th, April, 2013). All images were

captured using an Olympus BX51 microscope connected to a digital camera

or a whole-slide scanning system (.slide, both Olympus, Hamburg, Germany).

Images were captured at 1440 x 900 pixel and cropped using Photoshop CS3

(Adobe Systems, San Jose, CA).

2.3. Molecular Genetic Analysis

Fluorescent in situ hybridization (FISH) was performed to identify gene copy

number changes on formalin-fixed paraffin-embedded tissue by using a total

of 7 separate hybridizations. Probe details are provided in Supplemental

Table 2. Hybridizations followed routine protocols and samples were counted

and photographed using an Olympus BX60 Fluorescent Microscope

(Olympus, Melville, NY) with appropriate filter settings and software packages

(MetaSystems, Altlussheim, Germany). Cutoffs for amplification probes

followed definitions established for HER2 testing (gene copy number ratio >2

scored in at least 50 non-overlapping nuclei). Rearrangement was defined as

splitting of otherwise fused yellow signals into separate green and red signals

whereby the distance was apart more than 1 signal width. More than 15% of

nuclei had to have split signals for a case to be scored as positive for a

rearrangement. The 15% cutoff was based on counts of non-neoplastic

controls, and represents the mean±3 standard deviations of nuclei with split

signals per 100 nuclei in normal tissue. For BRAF V600 genotyping tumor

regions were either sectioned and microdissected, or cored using a 2mm

dermal punch needle. After deparaffinization, DNA extraction and PCR-

reactions (primers: -5’-TGC TTG CTC TGA TAG GAA AAT G-3’; R-5’-AGC

ATC TCA GGG CCA AAA AT-3’), pyrosequencing (R-5’- GAC CCA CTC CAT

CGA G-3’) was performed on a PyroMarkQ24 sequencer (Qiagen, Hilden,

Germany).

2.4. Literature Review and Biomarker Selection Two authors (JKL and myself) performed a meta-review of the literature.

Specific questions were: anatomic location by histotype, age at diagnosis,

sex, and biomarker expression status. A total of 193 published articles were

Material and Methods

5

reviewed and independently extracted using the pre-determined criteria. After

resolving differences via discussions, final tabulations were composed (see

supplement) and formed the starting point for the biomarker screening. A set

of therapeutically relevant markers from other tumor settings was also

included.

2.5. Statistical Analysis and Data Processing For contingency testing the Fisher’s exact test (association of histotype with

dichotomous factors), chi-square (association of histotype with >2 categories),

or student’s t-test (comparison of means) were applied. Data analysis was

performed with Microsoft Excel 2008 (version 12.1.9; Microsoft Corp.,

Redmond, WA, USA) or Prism 5.0b (GraphPad Softare, San Diego, CA,

USA). Heatmaps were generated under the R programming language

environment (www.r-project.org; version 2.13.2) and the pheatmap library. To

sort all cases according to their overall biomarker-phenotype, we performed

unsupervised clustering analysis using the hclust function. Based on the

maximum fraction of positive cases, each biomarker was assigned to one of

two categories: “positive in spindle cell melanoma” or “positive in

desmoplastic melanoma” (referred to as labeling direction). Differentially

expressed biomarkers were defined as markers that showed significant

differences in at least one cross validation. Cross validations compared

differences within each cohort (e.g., spindle cell melanoma-Ulm vs.

desmoplastic melanoma-Ulm), between cohorts (e.g., spindle cell melanoma-

Ulm vs. spindle cell melanoma-New York) and/or by histotype (spindle cell all

vs. desmoplastic melanoma all). To estimate the standard deviations of

sensitivity and specificity for all differentially expressed biomarkers, a leave

one out analysis was performed. The final decision tree for biomarker-based

classification of spindle cell vs. desmoplastic melanoma followed prior

approaches (Choi et al., 2009) and was based on considering the top three

markers “positive in spindle cell melanoma” and “positive in desmoplastic

melanoma”, selecting for the highest sensitivity. Statistical performance

measures were determined using the Hutchon toolkit

(http://www.hutchon.net/EPRval.htm) and 2-way contingency table analysis

(www.statpages.org, last accessed April, 18th, 2013). Statistical significance

was defined as P<0.05.

Results

6

3. Results

Archival searches, performed at two institutions, identified 38 cases

composed of 16 spindle cell, 18 desmoplastic melanoma and 4 mixed

spindle/desmoplastic melanoma. Key clinicopathological parameters of each

case are provided in Supplemental Table 3 and the cohorts were combined

for contingency testing (Table 1) because characteristics did not differ

between cohorts

(Supplemental Table 4).

In the current series,

desmoplastic melanoma

was non-pigmented and

more common in the

head and neck region.

To ensure the

representativeness of

the cohort a meta-review

of 86 publications was

performed, containing a

total of 5955 spindle cell and desmoplastic melanoma cases (Supplemental

Table 5). A summary along with the cases, including the anatomic

distribution, is provided in Figure 1. Briefly, there was a male predominance

(m-f-ratio: here 1.9:1 vs. literature 1.8:1 P-range: 0.18-1.0), and the highest

ratio of head and neck tumors in desmoplastic melanoma (here 52.6% vs.

literature 53.5%; P=1.0). By histopathological, epidemiologic, and

clinicopatholgical parameters, the cohort was considered representative.

For biomarker screening, a meta-review of 94 publications containing a total

of 10,340 data points describing the reported immunophenotypes in spindle

cell vs. desmoplastic melanoma vs. conventional melanoma (Supplemental

Tables 6, 7, 8; respectively) was performed. Analysis of the positive fraction

for each marker (Supplemental Table 9) revealed that desmoplastic

melanoma, while being essentially negative for all classic melanocytic

markers (e.g. HMB45, MelanA, Tyrosinase), is S100 positive (96.4%), and

Results

7

expresses SOX10 (92.2%) and p75 (75%). These markers also label a

significant fraction of spindle cell melanoma (S100 in 91.3%, SOX10 in 100%

and p75 in 79.5%) and can therefore be employed in reaching or confirming

the diagnosis of melanoma. These literature data underscore a lack of

markers that specifically label desmoplastic melanoma, which precludes

reliable positive diagnostic distinction of desmoplastic from spindle cell

melanoma (Figure 1).

Results

8

A limitation of this literature-

based approach is that the

percentage of positive cases is

based on a theoretical summary

of separate studies rather than a

direct comparison of multiple

markers in one study. The

results of the direct comparative

study of 50 routinely available

biomarkers (38 IHC features, 10

FISH probes, one trichrome

stain and the BRAF V600 status)

are provided in Figure 2A.

There were two key findings

when examining how spindle

cell, desmoplastic and mixed

cases relate to the labeling

pattern in a cluster analysis

(Figure 3).

First, three cases (one spindle

cell, two desmoplastic) had a

signature that was more similar

to the opposite diagnostic

category; however, re-review of

the histopathology in these

‘outliers’ did not confirm the

signature-based classification.

Although no re-classification was

performed, these ‘outliers’ were included in further analysis to take the

apparent variation in immunophenotypes into account. Second, as a result

the four mixed spindle/desmoplastic cases did not cluster together and, at

least by signature-based similarity, had some phenotypic variability that

covers the entire spindle cell-to-desmoplastic melanoma spectrum.

Results

9

To derive an algorithm for diagnostic distinction of desmoplastic from spindle

cell melanoma, it was argued that analysis of the marker pattern in the two

extremes (spindle cell and desmoplastic melanoma only) would identify the

most accurate markers and instead of re-assignment of the mixed cases to

spindle cell or desmoplastic categories, and the 4 mixed cases were excluded

from further analysis.

Systematic marker comparison in the 34 spindle cell and desmoplastic

melanoma cases showed significant differences in 10 markers (Figure 2A,

box and Supplemental tables 10 and 11). Representative images comparing

labeling patterns in spindle cell and desmoplastic melanoma are provided in

Figure 4.

Results

10

Notably, the labeling-direction of each marker has to be taken into account.

Specifically, laminin (tumor and non-tumor), p75, HMB45, c-kit, MelanA were

significantly more often positive in spindle cell melanoma whereas collagen

IV, CD68, MDM2 and trichrome were significantly more often positive in

desmoplastic than in spindle cell melanoma (Figure 2A). Briefly, collagen IV

showed strong immunolabeling around individual tumor cells (Figure 4D),

trichrome labeled the interspersed collagen fibrils (Figure 4H), CD68

demonstrated immunoreactivity in tumor cells (Figure 4L) and MDM2 showed

nuclear labeling (Figure 4P). Based on test performance comparisons (Figure

2B), the two markers with the highest sensitivity in spindle cell (i.e. MelanA)

and desmoplastic melanoma (i.e. Trichrome) were chosen for the design of a

Results

11

diagnostic decision tree (see discussion). Assessment in both cohorts

revealed robust diagnostic distinction (Supplemental Table 11), and the

overall test characteristics are shown in Table 2.

Results

12

As part of the biomarker screen, a set of therapeutically relevant molecules

(Figure 2A; red capsules) was included. FISH assessed ten markers with

potential therapeutic relevance HER2, EGFR, MET, MDM2, TP53, ALK, and

MYC. A total of 266 hybridizations were performed; however, only a single

case with MDM2 amplification was identified (Figure 5A). Thus, from a

practical standpoint, FISH-based genotyping for the included markers in

spindle cell and desmoplastic melanoma has no value. The potential targets

by immunolabeling were also assessed, and representative examples are

provided in Figure 5.

By fraction of positive cases, KIT and MYC were most frequently expressed in

spindle cell melanoma whereas TP53 and MDM2 were the most prevalent

markers in desmoplastic melanoma. While positive labeling for each marker

was individually rare, the distribution of positive cases showed that several

markers were expressed in a mutual exclusive fashion (e.g., EGFR, MET,

HER2; Figure 2A), which adds up to a significant number immunopositive

cases (Table 3).

Results

13

Comparison of marker positivity in spindle cell and desmoplastic melanoma

revealed that only KIT was significantly more commonly expressed in spindle

cell melanoma. While these findings are in line with an overall dearth of

positive markers in desmoplastic melanoma, there were only 3 cases (2

desmoplastic and 1 mixed) that were completely negative for any of the ten

potentially targetable molecules. Therefore, a total of 35 of 38 tumors in this

cohort (92%) were positive for at least one biomarker with therapeutic

significance in other tumor settings. Obviously, the practical significance of

this finding requires validation; however, in conjunction with the FISH-based

genotyping, it is save to assume that the (ectopic) expression of these

targetable molecules (Figure 5E, 5G, 5I, 5K) is the result of molecular

mechanisms other than gene amplification (Figure 5D, 5F, 5H, 5J). Finally,

BRAF genotyping (Figure 5C) was included and V600 mutations in 31% of

spindle cell and 5% of desmoplastic melanoma (Figure 2A) were found.

These numbers surmount to an overall mutation frequency of 16% (n=6/38) in

spindle cell and desmoplastic melanoma, which –given the treatment

implications in melanoma– argues for BRAF testing in this setting, when

clinically indicated.

Discussion

14

4. Discussion

Here the results of a comprehensive screening study of diagnostic and

therapeutic biomarkers in spindle cell and desmoplastic melanoma. Based on

the results in two independent cohorts, an algorithm was build that allows

diagnostic distinction of desmoplastic from spindle cell melanoma when

histological analysis is not decisive (Figure 6).

In diagnostic practice, spindle cell and desmoplastic melanoma may reveal

their identity as melanoma only after a first set of adjunct tests, typically

including immunophenotyping. Importantly, spindle cell and desmoplastic

melanoma may feature an atypical immunophenotype and even be negative

for classic melanocytic markers. Thus, when confronted with a spindled

lesion, demonstrating an atypical immunophenotype, the differential-diagnosis

should include melanocytic lesions and in particular spindle cell and

desmoplastic melanoma. The meta-review delineated that S100 and the less

widely established markers SOX10 and p75 have the highest fraction of

positivity in both spindle cell and desmoplastic melanoma (Supplemental

Table 9). Thus, S100 remains as a reliable routine marker to reach the

diagnosis of melanoma in spindle cell and desmoplastic melanoma (Ohsie et

al., 2008). The prognostic differences in spindle cell and desmoplastic

melanoma make their diagnostic distinction relevant; however when lesions

are partially sampled or heterogeneous, diagnosis may become difficult. In

this setting – namely when a spindled lesion has been confirmed as a

melanoma – the algorithm can be applied as an adjunct diagnostic tool. The

diagnostic dilemma of spindle cell and desmoplastic melanoma (Ohsie et al.,

2008) extends, however, beyond atypical morphological- and

immunohistochemical features and at least four additional points require

consideration.

Discussion

15

4.1. The terminology remains confusing Desmoplastic melanoma is currently defined as a subtype of spindle cell

melanoma (McCarthy et al., 2006); however, the term spindle cell melanoma

can either serve as an umbrella term for spindle cell and desmoplastic

melanoma category, might be applied as a descriptive diagnosis for a

conventional melanoma with prominent spindle cells, or – as applied here –

may refer to a specific entity that is typically amelanotic, contains less than

10% collagen fibrils, shows immunophenotypic differences to conventional

melanoma (Mohamed et al.; Pages et al., 2008; Skelton et al., 1997; Xu et al.,

2002) and typically present as a thicker lesion or in advanced stages.

Furthermore, BRAF mutations occur with different prevalence in spindle cell

melanoma (Kim et al., 2012). These findings collectively suggest that the

terminologic imprecision of spindle cell melanoma may explain the variable

incidence of 1-14% reported in the literature (Piao et al., 2008).

4.2. The prevalence of spindle cell and desmoplastic melanoma may be higher than previously assumed. The meta-review from the first description in 1971 until April 2013 lists nearly

6000 cases, arguing that spindle cell and desmoplastic melanoma cannot be

considered orphan diseases anymore.

4.3. The diagnostic categories are part of an extended biological spectrum

The partly overlapping morphological features along with clinicopathological

differences have led to the introduction of a mixed spindle and desmoplastic

melanoma category (Busam et al., 2004; Hawkins et al., 2005; McCarthy et

al., 2004; Scolyer and Thompson, 2005). Briefly, mixed spindle cell and

desmoplastic melanoma is defined as composed of spindled melanocytes

admixed with 10-90% collagen fibrils.

While the distinct immunoprofile for Ki-67, CD117, and Nestin in mixed

spindle cell and desmoplastic melanoma has been reported (Miller et al.,

2011), mixed cases were not included into the analysis due to the small

number (n=4 in the cohorts), the phenotypic variability (see results) and to

establish the decision tree based on findings in the extremes (i.e., ‘pure’

Discussion

16

spindle cell vs. desmoplastic melanoma). This should, however, not imply

that the clinically distinct mixed subtype ought to be neglected (Busam et al.,

2004; Hawkins et al., 2005; McCarthy et al., 2004; Scolyer and Thompson,

2005). Rather it has to be emphasized that the algorithm is not a one size-

fits-all approach but allows inclusion of existing or additional categories. The

concept of an extended spectrum is also supported by comparative genome

hybridization data demonstrating that 67% of desmoplastic melanoma share

allelic losses of NF1, which matches the immunophenotypic overlap of

desmoplastic melanoma with malignant peripheral nerve sheet tumors (Boyle

et al., 2002; Kanik et al., 1996; Prieto and Woodruff, 1998; Pryor et al.; Xu et

al., 2002).

Collectively these data suggest that spindle cell and desmoplastic melanoma

represent discrete diagnostic categories within an extended biological

spectrum.

4.4 There is a dearth of specific desmoplastic melanoma markers Gene expression profiling in desmoplastic melanoma revealed at least 629

differentially expressed genes (Busam et al., 2005) that have been partially

validated by immunohistochemistry. However, the overall rarity of

desmoplastic melanoma, may not justify implementation of novel antibodies in

every laboratory. Thus, it is obvious to focus on routinely available and

established markers and extend the list of specific dsmoplastic melanoma

markers. MDM2 was identified as positive in desmoplastic melanoma (albeit

with less optimal test performance characteristics), and a subset of

desmoplastic melanoma that is CD68 positive. The latter is important because

one key differential diagnosis of desmoplastic melanoma is a scar that

typically contains CD68-positive histiocytes (Ramos-Herberth et al.). Also,

CD68 may not be helpful in differentiating histiocytic tumors from

desmoplastic melanoma (Pernick et al., 1999). Whether CD68 can become

important in terms of classification remains to be determined. Currently the

reason for the ectopic CD68 expression remains unclear (Longacre et al.,

1996; Morgan et al., 2008; Shah et al., 1997) and this histiocytic feature of

desmoplastic melanoma was interpreted as part of the “great imitator”

malignant melanoma. Collagen IV was also identified around tumor cells as a

Discussion

17

relatively sensitive and specific phenomenon in desmoplastic melanoma

(Figure 4D). It seems unlikely that this staining pattern has not been

previously noted in desmoplastic melanoma (Prieto and Woodruff, 1998);

however, peritumoral collagen IV labeling is in accord with the overall

histomorphology and indicates that each tumor cells synthesized at least one

principle component of the surrounding basement membrane, essentially

providing its own structural encasement. Since addition of collagen IV did not

increase test performance when trichrome staining was included, the special

stain was chosen as a cost-effective and reliable approach. Thus, two readily

available markers (MelanA and trichrome staining) achieved the best

diagnostic performance; however, the presented algorithm should not replace

careful histomorphological examination and clinicopathological correlation.

In sum, this is the first literature-based, comprehensive screening study that

compares the profile of diagnostic and potentially targetable molecules in

spindle cell vs. desmoplastic melanoma with respect to commonly used

tissue-biomarkers. A panel approach identifies a significant fraction of

patients that may benefit from molecularly targeted therapy and the presented

diagnostic algorithm may guide classification when histopathological

classification is not straightforward.

Summary

18

5. Summary

Spindle cell melanoma and desmoplastic melanoma differ clinically in

prognosis and therapeutic implications; however, due to partially overlapping

histopathological features, diagnostic distinction of spindle cell from

desmoplastic melanoma is not always straightforward. A direct comparison of

diagnostic and therapeutic biomarkers has not been performed.

Meta-review of the literature discloses key clinicopathological differences

between spindle cell and desmoplastic melanoma, including

immunophenotypes. Using 50 biomarkers available in routine diagnostics 38

archival cases (n=16 spindle, 18 desmoplastic, 4 mixed spindle/desmoplastic

melanoma) were examined. S100 remains as the most reliable routine

marker to reach the diagnosis of melanoma in spindle cell and desmoplastic

melanoma. There were 9 distinctly labeling markers with spindle cell

melanoma showing positivity for laminin, p75, HMB45, c-kit, and MelanA, and

desmoplastic melanoma preferentially labeling with collagen IV, trichrome,

CD68, and MDM2. Based on comparisons of test performance measures,

MelanA and trichrome were used to devise a 94% sensitive diagnostic

algorithm for the distinction of desmoplastic from spindle cell melanoma.

Gene amplification and expression status was assessed for a set of

potentially drugable targets (HER2, EGFR, MET, MDM2, TP53, ALK, MYC,

FLI-1, KIT). Fluorescent in situ hybridizations did not reveal a significant

number of gene abberations/rearrangements; however, protein

overexpression for at least one of these markers was identified in 35 of 38

cases (92%). In addition, we found BRAF mutations in 31% of spindle cell

and 5% of desmoplastic melanoma with an overall mutation frequency of 16%

(n=6/38).

This is the first comprehensive screening study of diagnostic and therapeutic

biomarkers in spindle cell and desmoplastic melanoma. The devised

algorithm allows diagnostic distinction of desmoplastic from spindle cell

melanoma when routine histology is not decisive.

References

19

6. References

1. Boyle JL, Haupt HM, Stern JB, Multhaupt HA. 2002. Tyrosinase

expression in malignant melanoma, desmoplastic melanoma, and

peripheral nerve tumors. Arch Pathol Lab Med 126(7):816-822.

2. Broer PN, Walker ME, Goldberg C, Buonocore S, Braddock DT,

Lazova R, Narayan D, Ariyan S. 2013. Desmoplastic melanoma: A 12-

year experience with sentinel lymph node biopsy. Eur J Surg Oncol

39((7)):681-685.

3. Busam KJ, Mujumdar U, Hummer AJ, Nobrega J, Hawkins WG, Coit

DG, Brady MS. 2004. Cutaneous desmoplastic melanoma: reappraisal

of morphologic heterogeneity and prognostic factors. Am J Surg Pathol

28(11):1518-1525.

4. Busam KJ, Zhao H, Coit DG, Kucukgol D, Jungbluth AA, Nobrega J,

Viale A. 2005. Distinction of desmoplastic melanoma from non-

desmoplastic melanoma by gene expression profiling. J Invest

Dermatol 124(2):412-418.

5. Chen JY, Hruby G, Scolyer RA, Murali R, Hong A, Fitzgerald P, Pham

TT, Quinn MJ, Thompson JF. 2008. Desmoplastic neurotropic

melanoma: a clinicopathologic analysis of 128 cases. Cancer

113(10):2770-2778.

6. Choi WW, Weisenburger DD, Greiner TC, Piris MA, Banham AH,

Delabie J, Braziel RM, Geng H, Iqbal J, Lenz G, Vose JM, Hans CP,

Fu K, Smith LM, Li M, Liu Z, Gascoyne RD, Rosenwald A, Ott G,

Rimsza LM, Campo E, Jaffe ES, Jaye DL, Staudt LM, Chan WC. 2009.

A new immunostain algorithm classifies diffuse large B-cell lymphoma

into molecular subtypes with high accuracy. Clin Cancer Res

15(17):5494-5502.

7. Coupelon S, Franck F, Jarrousse AS, Dechelotte P, Souteyrand P,

D'Incan M. 2012. Desmoplastic malignant melanoma: a study of ten

References

20

cases and status of BRAF mutation. Dermatology 225(2):168-171.

8. de Almeida LS, Requena L, Rutten A, Kutzner H, Garbe C, Pestana D,

Gomes MM. 2008. Desmoplastic malignant melanoma: a

clinicopathologic analysis of 113 cases. Am J Dermatopathol

30(3):207-215.

9. Gyorki DE, Busam K, Panageas K, Brady MS, Coit DG. 2003. Sentinel

lymph node biopsy for patients with cutaneous desmoplastic

melanoma. Ann Surg Oncol 10(4):403-407.

10. Hawkins WG, Busam KJ, Ben-Porat L, Panageas KS, Coit DG, Gyorki

DE, Linehan DC, Brady MS. 2005. Desmoplastic melanoma: a

pathologically and clinically distinct form of cutaneous melanoma. Ann

Surg Oncol 12(3):207-213.

11. Jaroszewski DE, Pockaj BA, DiCaudo DJ, Bite U. 2001. The clinical

behavior of desmoplastic melanoma. Am J Surg 182(6):590-595.

12. Kanik AB, Yaar M, Bhawan J. 1996. p75 nerve growth factor receptor

staining helps identify desmoplastic and neurotropic melanoma. J

Cutan Pathol 23(3):205-210.

13. Kim J, Lazar AJ, Davies MA, Homsi J, Papadopoulos NE, Hwu WJ,

Bedikian AY, Woodman SE, Patel SP, Hwu P, Kim KB. 2012. BRAF,

NRAS and KIT sequencing analysis of spindle cell melanoma. J Cutan

Pathol 39(9):821-825.

14. Lens MB, Newton-Bishop JA, Boon AP. 2005. Desmoplastic malignant

melanoma: a systematic review. Br J Dermatol 152(4):673-678.

15. Livestro DP, Muzikansky A, Kaine EM, Flotte TJ, Sober AJ, Mihm MC,

Jr., Michaelson JS, Cosimi AB, Tanabe KK. 2005. Biology of

desmoplastic melanoma: a case-control comparison with other

melanomas. J Clin Oncol 23(27):6739-6746.

16. Longacre TA, Egbert BM, Rouse RV. 1996. Desmoplastic and spindle-

References

21

cell malignant melanoma. An immunohistochemical study. Am J Surg

Pathol 20(12):1489-1500.

17. Maurichi A, Miceli R, Camerini T, Contiero P, Patuzzo R, Tragni G,

Crippa F, Romanidis K, Ruggeri R, Carbone A, Santinami M. 2010.

Pure desmoplastic melanoma: a melanoma with distinctive clinical

behavior. Ann Surg 252(6):1052-1057.

18. McCarthy SW, Crotty KA, Scolyer RA. 2006. Desmoplastic melanoma

and desmoplastic neurotropic melanoma. In: LeBoit PE, editor.

Pathology and genetics of skin tumours. Lyon: IARC Press.

19. McCarthy SW, Scolyer RA, Palmer AA. 2004. Desmoplastic

melanoma: a diagnostic trap for the unwary. Pathology 36(5):445-451.

20. Miller DD, Emley A, Yang S, Richards JE, Lee JE, Deng A, Hoang MP,

Mahalingam M. 2011. Mixed versus pure variants of desmoplastic

melanoma: a genetic and immunohistochemical appraisal. Mod Pathol

25(4):505-515.

21. Mohamed A, Gonzalez RS, Lawson D, Wang J, Cohen C. 2012.

SOX10 Expression in Malignant Melanoma, Carcinoma, and Normal

Tissues. Appl Immunohistochem Mol Morphol:in press.

22. Mohebati A, Ganly I, Busam KJ, Coit D, Kraus DH, Shah JP, Patel SG.

2012. The role of sentinel lymph node biopsy in the management of

head and neck desmoplastic melanoma. Ann Surg Oncol 19(13):4307-

4313.

23. Morgan MB, Purohit C, Anglin TR. 2008. Immunohistochemical

distinction of cutaneous spindle cell carcinoma. Am J Dermatopathol

30(3):228-232.

24. Murali R, Shaw HM, Lai K, McCarthy SW, Quinn MJ, Stretch JR,

Thompson JF, Scolyer RA. 2010. Prognostic factors in cutaneous

desmoplastic melanoma: a study of 252 patients. Cancer

116(17):4130-4138.

References

22

25. Ohsie SJ, Sarantopoulos GP, Cochran AJ, Binder SW. 2008.

Immunohistochemical characteristics of melanoma. J Cutan Pathol

35(5):433-444.

26. Pages C, Rochaix P, al Saati T, Valmary-Degano S, Boulinguez S,

Launay F, Carle P, Lauwers F, Payoux P, Le Guellec S, Brousset P,

Lamant L. 2008. KBA.62: a useful marker for primary and metastatic

melanomas. Hum Pathol 39(8):1136-1142.

27. Pernick NL, DaSilva M, Gangi MD, Crissman J, Adsay V. 1999.

"Histiocytic markers" in melanoma. Mod Pathol 12(11):1072-1077.

28. Piao Y, Guo M, Gong Y. 2008. Diagnostic challenges of metastatic

spindle cell melanoma on fine-needle aspiration specimens. Cancer

114(2):94-101.

29. Prieto VG, Woodruff JM. 1998. Expression of basement membrane

antigens in spindle cell melanoma. J Cutan Pathol 25(6):297-300.

30. Pryor JG, Brown-Kipphut BA, Iqbal A, Scott GA. 2011. Microarray

comparative genomic hybridization detection of copy number changes

in desmoplastic melanoma and malignant peripheral nerve sheath

tumor. Am J Dermatopathol 33(8):780-785.

31. Ramos-Herberth FI, Karamchandani J, Kim J, Dadras SS. 2010.

SOX10 immunostaining distinguishes desmoplastic melanoma from

excision scar. J Cutan Pathol 37(9):944-952.

32. Scolyer RA, Thompson JF. 2005. Desmoplastic melanoma: a

heterogeneous entity in which subclassification as "pure" or "mixed"

may have important prognostic significance. Ann Surg Oncol

12(3):197-199.

33. Shah IA, Gani OS, Wheler L. 1997. Comparative immunoreactivity of

CD-68 and HMB-45 in malignant melanoma, neural tumors and nevi.

Pathol Res Pract 193(7):497-502.

References

23

34. Shaw HM, Quinn MJ, Scolyer RA, Thompson JF. 2006. Survival in

patients with desmoplastic melanoma. J Clin Oncol 24(8):e12; author

reply e13.

35. Skelton HG, Maceira J, Smith KJ, McCarthy WF, Lupton GP, Graham

JH. 1997. HMB45 negative spindle cell malignant melanoma. Am J

Dermatopathol 19(6):580-584.

36. Weissinger SE, Moeller P, Lennerz JK. 2013. Slide-to-slide arrays for

high-throughput molecular profiling or rare tumor specimens. Mod

Pathol 26(52 (s2)):503A.

37. Xu X, Chu AY, Pasha TL, Elder DE, Zhang PJ. 2002. Immunoprofile of

MITF, tyrosinase, melan-A, and MAGE-1 in HMB45-negative

melanomas. Am J Surg Pathol 26(1):82-87.

6.1 Software Packages and Internet Resources

38. R-Project, statistical computing and graphics software environment

http://www.r-project.org; version 2.13.2; download October, 4th, 2011)

39. Hutchon toolkit, online test performance calculator

http://www.hutchon.net/EPRval.htm; last accessed April, 18th, 2013

40. Statpages, statistical webpage used for contingency calculations

http://www.statpages.org, last accessed April, 18th, 2013

24

Appendix to

“A Diagnostic Algorithm to Distinguish Desmoplastic from Spindle Cell Melanoma” Weissinger SE, Keil P, Silvers DN, Klaus BM, Möller P, Horst BA, Lennerz JK.

Modern Pathology (2013) ePub ahead of print.

Supplemental Table 1. Antibodies used for Immunohistochemistry

Name Host/Clone Antigen Characteristics Clone, source (cat. no.) Conc. Retrieval

Staining pattern; Control tissue; Staining assessed in

References PMID

Laminin M/m GST-laminin-gamma2-III fusion protein (aa 391-567)

4G, Dako Braunschweig Germany (M7262)

1:25 S:TRSpH6,1

C; epithelial cells; T, NT

11299788

S100 R/p S100 Protein (cow brain); antibody reacts strongly with S100A and S100B

Dako (Z0311) 1:1000 Pronase

C/N; histiocytes; C, N*

3415443

NSE M/m Gamma gamma Enolase from human brain

BBS/NC/VI-H14, Dako (M0873) 1:500 MW

C; neuroendocrine cells T

7812919

Vimentin M/m Vimentin from bovine eye lens (Epitope on coil 2 from rod domain)

VIM3B4, Dako (M7020) 1:300 MW

C; mesenchymal cells T

7812919

GFAP M/m Human glial fibrillary acidic protein MIG-G2G, Linaris Dossenheim, Germany (E008)

1:5 MW

C/MEM; Schwann cells T

11481518

CD34 M/m Endothelial cell membranes obtainded as vesicles from human placenta

QBEnd10, Dako (M7165) 1:100 S:TRSpH6,1

MEM; endothelial cells T, NT

22456515

CD56 (N-CAM) M/m CD56 glycoprotein, neural cell adhesion molecule

BC56C04, Sanova, Vienna, Austria (CM164, ABC)

1:50 MW

MEM; neurons T

10644945

CD57 M/m Human peripheral blood mononuclear cells

NK-1, Novocastra/Leica Berlin Germany (NCL-NK1)

1:20 Pronase

MEM; neurons T

22050593

HER2 R/p Synthetic human c-erbB-1 peptide from the intracytoplasmatic portion

N/A, Dako (A0485) 1:600 PC

MEM; breast carcinoma T

19434003

FLI-1 R/p Synthetic peptide from C-terminus N/A, Medac Wedel Germany (RB-9295)

1:100 MW

N; keratinocytes, melanocytes T

18785112

NF M/m NF normal human brain NR4, Dako (M0762) 1:500 none

C; neurones central & periphere T

6190849

MelanA/MART-1 M/m Recombinant Melan-A protein expressed in E.coli

A103, Dako (M7196) 1:200 MW

C; melanocytes T

11474288

Collagene IV (Col IV)

M/m Human glomeruli PHM-12, Novocastra/Leica (NCL-Coll-IV)

1:200 Citrate

C; basement membrane T, NT

15936594

HMB45 (gp100) M/m Metastasing melanoma cells (Anti human melanosome)

HMB45, Zytomed Berlin, Germany (MSK007-05)

1:200 MW

C; melanocytes, nevus cells; T

11756773

c-kit (CD117) R/p Peptide corresp. to cytoplasmatic part of c-kit (aa 963-976)

N/A, Dako (A4502) 1:200 MW

C/MEM; mast cells T

21897303

CD99 (MIC2 gene product)

M/m ALL T-cells 12E7, Dako (M3601) 1:100 none

MEM; Ewing sarkoma T

17026517

CD68 M/m Human mononuclear spleen cells containing > 80% Gaucher's cells

PG-M1, Dako (M0876) 1:100 S:TRSpH6,1

C; histiocytes T, NT

16078259

Desmin M/m Intermediate filament proteins DE-R-11, Menarini Berlin, Germany (NCL-L-Des-DerII)

1:50 Prot K

C; skleletal and cardiac muscle T

7512798

SMA M/m N-terminal synthetic decapeptide of α-SMA

1A4, Dako (M0851) 1:400 MW

MEM; smooth muscle cells T, V

7812919

CK (AE1.3) R/p Bovine muzzle epidermal keratin AE1+AE3, Dako (M3515) 1:100 Pronase

C; epithelial differentiation T

7812919

CAM5.2 (K7) M/m OT11- ovarian carcinoma cell line OV-TL 12/39, Dako (M7018) 1:200 Pronase

C; CK8>CK7+ epithelium; T

7812919

p75 R/p Third exon of mouse p75 (aa 43-161)

Millipore Schwalbach, Germany (AB1554)

1:100 PC

MEM; Schwann cells and neurons C, N*

21178583

DOG1 R/p Synthetic human DOG1 protein BV10, Zytomed (RMAB031) 1:200 PC

C/MEM; GIST T

21685034

p53 M/m Recombinant human wild-type p53 protein

DO-7, Dako (M7001) 1:500 MW

N; colorectal carcinoma T

18780655

MUM1 M/m Recombinant GST-MUM1 fusion protein

MUM1p, Dako (M7259) 1:100 MW

N; germinal center B cells T

11455001

MDM2 M/m Peptide epitope around Thr-216 of murine MDM2

SMP14, Abcam (ab3110) 1:100 PC

N/C; breast carcinoma C, N*

17895748

EGFR M/m Extracellular domain of human recombinant EGFR protein

2.1E1, Zytomed Systems (MSK014-05)

1:100 Pronase

C/MEM; carcinoma T

10098284

c-MET R/m Synthetic peptide corresp. to the C-terminus of the human Met protein

Met(D1C2), Cell Signaling, Frankfurt Germany; (8198S)

1:200 Citrate

C/MEM; carcinoma T

8180811

ALK M/m DHFR-ALK recombinant protein comprising mouse DHF reductase

ALK1, Dako (M7195) 1:100 MW

C/N ALK-rearranged NSCLC T

22100693

c-myc R/p Synthetic peptide corresp. to residues in N-terminus of human c-myc protein

EP121, Epitomics 1:100 S:EDTApH8

C/N; Burkitt lymphoma T

9621133

Abbreviations: aa, amino acid; cat, catalogue; Conc., concentration; C: cytoplasmic; EDTA, ethylenediaminetetraacetic acid; GIST, gastro intestinal stromal tumor; MEM: membranous; m: monoclonal; M, mouse; MW, microwave; No, number; N, nuclear; NT, non-tumor; p, polyclonal; PC, pressure cooker; PMID, pubmed identification number; R, rabbit; Retrieval, antigen retrieval; S, steamer; T, tumor; TRS, target retrieval solution; V, vessels (see note); *, indicates that cytoplasmic (C) and nuclear (N) staining was assessed in tumor cells.

Note: In the CD34 (V) assessment, presence of at least 3 arterioles (≥ 25µm diameter) in association with arborizing capillaries in 2 high power fields (x400) was required to assign positivity. For all other markers, a case was scored positive when ≥10% of the assessed cells showed immunoreactivity.

Supplemental Table 2. Fluorescent in situ Hybridization (FISH) Probe Sets

Gene Type of Probe (diagnostic cut-off)

Chromosomal Position

Control Probe Source (Order number)

References PMID

HER2 Amp (G:CN>2)

17q11.2-q12 17q11.1-q11 Abbott, Pathvysion (02J01-030)

16184453

EGFR Amp (G:CN>2)

7p11 CEP7 Kreatech, ON (KBI-10702)

11597399

c-MET Amp (G:CN>2)

7q31 CEP7 Kreatech, ON (KBI-10719)

20107422

MDM2 Amp (G:CN>2)

12q15 CEP12 Kreatech, ON (KBI-10717)

20431481

TP53 Del (<2) Amp (>2)

17p13.1 N/A Abbott, Vysis (08L64-020)

21713341

ALK BA/RA (cut off:15%)

2p23 c[488];t[575]

N/A Abbott, Vysis (05J89-001)

9490693

MYC BA/RA (cut off:15%)

8q24 c[575];t[488]

N/A Abbott, Vysis (05J91-001)

8392399

Abbreviations: Abbott, Wiesbaden, Germany; Amp, amplification probe; BA/RA, break-apart/rearrangement probe; c, centromeric probe; CEP, centromere enumeration probe; G:CN, gene to copy-number ratio; Kreatech Diagnostics, Amsterdam, Netherlands; N/A, not applicable; ON, Oncology probeset; PMID, www.pubmed.com identification number; t, telomeric probe; [ ], filter setting in µm.

Supplemental Table 3. Clinicopathological Characteristics of the Study Cases

Abbreviations: Bx, biopsy; Ex, excision; F, female; ID, case identification number (Figure 1); HN, head and neck; LE, lower extremity; M, male; N, no pigment present; UE, upper extremity; Y, pigment containing.

Case ID Age Sex Site Specimen Histotype Pigment Stage

1 S.U1 52 M Trunk Ex Spindle Cell Melanoma Y IIB 2 S.U2 53 M Trunk Ex Spindle Cell Melanoma Y IB 3 S.U3 65 F HN Ex Spindle Cell Melanoma Y IB 4 S.U4 48 M Trunk Bx Spindle Cell Melanoma N IIB 5 S.U5 82 M Trunk Ex Spindle Cell Melanoma N IV 6 S.U6 65 M Trunk Ex Spindle Cell Melanoma N IV 7 S.U7 66 F Trunk Ex Spindle Cell Melanoma N IIA 8 S.U8 76 M Trunk Ex Spindle Cell Melanoma N IIB 9 S.U9 79 M Trunk Ex Spindle Cell Melanoma Y IA

10 S.N1 96 M UE Ex Spindle Cell Melanoma N IIA 11 S.N2 61 M HN Ex Spindle Cell Melanoma N IB 12 S.N3 70 M UE Ex Spindle Cell Melanoma N IIB 13 S.N4 82 M Trunk Ex Spindle Cell Melanoma N IIB 14 S.N5 69 M Trunk Ex Spindle Cell Melanoma N IIA 15 S.N6 48 F UE Ex Spindle Cell Melanoma N IIA 16 S.N7 72 M HN Ex Spindle Cell Melanoma N IIB

17 M.N1 78 M HN Ex Mixed Spindle/Desmoplastic N IIB 18 M.N2 81 M HN Ex Mixed Spindle/Desmoplastic N IIB 19 M.N3 74 M UE Ex Mixed Spindle/Desmoplastic N IIA 20 M.N4 61 F HN Ex Mixed Spindle/Desmoplastic N IB

21 D.U1 94 F HN Ex Desmoplastic Melanoma N IV 22 D.U2 77 M HN Ex Desmoplastic Melanoma N IIA 23 D.U3 87 M HN Ex Desmoplastic Melanoma N IIB 24 D.U4 54 F HN Ex Desmoplastic Melanoma N IA 25 D.U5 78 M HN Ex Desmoplastic Melanoma N IB 26 D.U6 96 F HN Ex Desmoplastic Melanoma N IA 27 D.U7 89 F HN Ex Desmoplastic Melanoma N IIB 28 D.U8 92 F HN Ex Desmoplastic Melanoma N III 29 D.U9 53 F Trunk Ex Desmoplastic Melanoma N IV 30 D.N1 88 M HN Bx Desmoplastic Melanoma N IIB 31 D.N2 75 M LE Bx Desmoplastic Melanoma N IB 32 D.N3 79 F HN Ex Desmoplastic Melanoma N IIA 33 D.N4 84 M HN Ex Desmoplastic Melanoma N IIA 34 D.N5 78 M UE Ex Desmoplastic Melanoma N IIA 35 D.N6 87 F HN Ex Desmoplastic Melanoma N IIA 36 D.N7 79 M UE Ex Desmoplastic Melanoma N IIA 37 D.N8 55 M HN Ex Desmoplastic Melanoma N IB 38 D.N9 28 F HN Ex Desmoplastic Melanoma N IIA

Supplemental Table 4. Comparison of Clinicopathological Characteristics in Both Study Cohorts by Histological Subtype

Ulm New York

Spindle Cell Melanoma

(N=9)

Desmoplastic Melanoma

(N=9)

P Spindle vs.

Desmoplastic

Melanoma

Spindle Cell Melanoma

(N=7)

Desmoplastic Melanoma

(N=9)

P Spindle vs.

Desmoplastic

Melanoma

P Spindle Ulm vs.

Spindle New York

P Desmoplastic Ulm vs.

Desmoplastic New York

Characteristic No. % No. % No. % No. %

Age Median 65 87 0.04 70 79 0.88 0.39 0.39 Range 48-82 53-96 48-96 28-88

Sex Female 2 22 6 67 0.15 1 14 3 33 0.58 1.0 0.35 Male 7 78 3 33 6 86 6 67

Location 0.15 0.09 0.12 Head&Neck 1 11 8 89 0.003 2 29 6 67 0.31 Trunk 8 89 1 11 2 29 0 - Extremities 0 - 0 - 3 43 3 33

Pigmentation Yes 4 44 0 0 0.08 0 0 0 0 1 0.08 1.0 No 5 56 9 100 7 100 9 100

Stage 0.96 0.35 0.39 0.18 IA 1 11 1 11 0 - 0 - IB 2 22 1 22 1 14 2 22 IIA 1 11 2 11 3 43 6 67 IIB 3 33 3 33 3 43 1 11 III 0 - 0 - 0 - 0 - IV 2 22 2 22 0 - 0 -

Abbreviations: No., number of cases; P-values derived from student t-test (age) Fisher’s exact (sex, pigmentation) or Chi-square test (location, stage).

Supplemental Table 5. Meta-review of Epidemiology and Anatomic Location by Histological Subtype Location

Histotype Reference PMID No. Male Female Age (Median) Head & Neck Trunk Upper Extr

Lower Extr. Spindle cell 1 Longacre et al., 1996 8944042 2 N/A N/A N/A 0 0 2 0

Spindle cell 2 Skelton et al., 1997 9415614 56 28 27 N/A 25 7 23 1 Spindle cell 3 Prieto et al., 1998 9694618 20 N/A N/A 60.1 9 5 3 3 Spindle cell 4 Thelmo et al., 2001 11209114 13 13 0 70 5 2 2 4 Spindle cell 5 Granter et al., 2001 11391097 11 6 5 72 6 1 1 3 Spindle cell 6 Winnepenninckx et al., 2003 14663357 9 6 3 69 7 0 2 0 Spindle cell 7 Morgan et al., 2008 18496422 3 3 0 59 2 0 1 0 Spindle cell 8 Lazova et al., 2009 20950740 8 7 1 79 3 2 3 0 Spindle cell 9 Diaz et al., 2011 21997694 24 N/A N/A 62 5 7 8 3

Spindle cell 10 Kim et al., 2012 22809251 24 15 9 57 5 10 1 2 Spindle cell 11 Buonaccorsi et al., 2012 22257901 12 7 6 52 5 6 0 2 Spindle cell 12 Sigal et al., 2012 22207443 11 8 3 83 7 1 3 0 Sum

Spindle cell total 191 93 54 65.5 79 41 49 18 187 Spindle cell % 63.3 36.7 Age range (23-99) 42.2 21.9 26.2 9.6 100%

Histotype Reference PMID No. Male Female Age (Median) Head & Neck Trunk Upper Extr

Lower Extr. Mixed 1 Almeida et al., 2008 18496419 27 27 35 70 46 9 3 2

Mixed 2 Busam et al., 2004 15489657 37 26 11 61.1 24 5 5 3 Mixed 3 Hawkins et al., 2005 15827812 39 35 4 60 19 9 6 5 Mixed 4 Longacre et al., 1996 8944042 10 N/A N/A N/A 5 1 2 2 Mixed 5 Matuichi et al., 2010 211071162

1107116 124 71 53 63 68 33 15 8

Mixed 6 Miller et al., 2012 22157936 18 8 10 75 9 3 4 2 Mixed 7 Murali et al., 2010 20564101 129 79 50 59.9 24 27 41 37 Mixed 8 Broer et al., 2013 23522951 8 N/A N/A 54.5 N/A N/A N/A N/A Sum

Mixed total 392 246 163 61.1 195 87 76 59 417 Mixed % 60.1 39.9 Age range (16-92) 46.8 20.9 18.2 14.1 100%

Histotype Reference PMID No. Cases

Male Female Age (Median) Head & Neck Trunk Upper Extr

Lower Extr. Desmoplastic 1 Conley et al., 1971 5286448 7 5 2 50 5 1 0 1

Desmoplastic 2 Frolow et al., 1975 1137419 1 1 0 68 0 0 1 0 Desmoplastic 3 Labrecque et al., 1976 953965 4 2 2 56 3 0 0 1 Desmoplastic 4 Reed et al., 1979 539614 22 13 9 60 19 1 2 0 Desmoplastic 5 Valensi 1979 427719 2 1 1 67.5 2 0 0 0 Desmoplastic 6 Man et al., 1981 7459530 1 1 0 50 1 0 0 0 Desmoplastic 7 Bryant et al., 1982 6293333 1 1 0 61 1 0 0 0 Desmoplastic 8 DiMaio et al., 1982 7139530 7 3 3 62.5 3 1 0 3 Desmoplastic 9 From et al., 1983 6357991 4 2 2 64.5 3 0 1 0

Desmoplastic 10 Kossard et al., 1987 3606170 10 6 4 67 5 3 2 0 Desmoplastic 11 Reiman et al., 1987 3440236 9 5 4 66 9 0 0 0 Desmoplastic 12 Egbert et al., 1988 3167815 25 11 14 61.2 21 1 3 0 Desmoplastic 13 Noodlemann et al., (from Egbert_1988) 3167815 1 1 0 59 1 0 0 0 Desmoplastic 14 Yannopoulos et al., (from Egbert_1988) 3167815 8 5 2 68.5 5 2 0 1 Desmoplastic 15 Di Maio et al., (from Egbert_1988) 3167815 6 3 3 62.5 3 1 0 2 Desmoplastic 16 Bryant et al., (from Egbert_1988) 3167815 1 N/A N/A 61 1 0 0 0 Desmoplastic 17 Man et al., (from Egbert_1988) 3167815 1 1 0 50 0 1 0 0 Desmoplastic 18 Valensi et al., (from Egbert_1988) 3167815 1 1 0 77 1 0 0 0 Desmoplastic 19 Frolow et al., (from Egbert_1988) 3167815 1 1 0 68 0 0 1 0 Desmoplastic 20 Valensi et al., (from Egbert_1988) 3167815 2 0 2 58.5 1 0 0 1 Desmoplastic 21 Labreque et al., (from Egbert_1988) 3167815 4 2 2 56 3 0 0 1 Desmoplastic 22 Walsh et al., 1988 3415443 14 10 4 58 4 5 3 2 Desmoplastic 23 Jain et al., 1989 2712188 45 31 14 64 35 5 3 2 Desmoplastic 24 Beenken et al., 1989 2917074 16 16 0 62.7 13 2 1 0 Desmoplastic 25 Whitaker et al., 1992 1583168 3 1 2 65 3 0 0 0 Desmoplastic 26 Smithers et al., 1992 1373257 44 N/A N/A 63 24 20 7 7 Desmoplastic 27 Anstey et al., 1993 8217747 25 14 11 65 18 3 1 3 Desmoplastic 28 Carlson et al., 1995 7812919 28 15 13 59 21 5 2 0 Desmoplastic 29 Skelton et al., 1995 7722014 121 N/A N/A 63 61 21 30 9 Desmoplastic 30 Weinzweig et al., 1995 7870781 11 7 4 68.4 4 4 1 2 Desmoplastic 31 Longacre et al., 1996 8944042 22 N/A N/A N/A 12 5 5 0 Desmoplastic 32 Al-Alouis et al., 1996 8721445 23 12 11 62 N/A N/A N/A N/A Desmoplastic 33 Jennings et al., 1996 8872151 5 3 2 67 5 0 0 0 Desmoplastic 34 Kilpatrick et al., 1996 8697380 8 6 2 59 8 0 0 0 Desmoplastic 35 Papadopoulos et al., 1997 9240143 10 N/A N/A N/A 10 0 0 0 Desmoplastic 36 Quinn et al., 1998 9740077 280 178 102 59.6 106 67 69 32 Desmoplastic 37 Wharton et al., 1999 10333224 18 8 10 64.1 12 3 3 0 Desmoplastic 38 Thelmo et al., 2001 11209114 15 11 5 57.5 7 4 4 1 Desmoplastic 39 Gerami et al., 2001 21323721 15 13 2 71 7 6 1 1 Desmoplastic 40 Granter et al., 2001 11391097 10 8 2 62 9 1 0 0 Desmoplastic 41 Hoang et al., 2001 11737519 1 1 0 72 0 0 1 0 Desmoplastic 42 Payne et al., 2001 11603540 30 21 9 63 12 10 5 3 Desmoplastic 43 Jaroszewski et al., 2001 11839322 59 37 22 62.8 36 11 6 6 Desmoplastic 44 Gyorki et al., 2003 12734089 27 20 7 64 14 8 3 2 Desmoplastic 45 Vongtama et al., 2003 12784232 44 N/A N/A 66 30 6 5 3 Desmoplastic 46 Busam et al., 2004 15489657 55 32 23 64.5 29 12 11 3 Desmoplastic 47 Kay et al., 2004 14765268 26 17 11 64.5 26 0 0 0 Desmoplastic 48 Hawinks et al., 2005 15827812 92 56 36 63 47 21 12 12 Desmoplastic 49 Livestro et al., 2005 16170181 89 N/A N/A 63.9 59 12 12 6 Desmoplastic 50 Posther et al., 2006 16538415 129 82 47 55 61 27 17 17 Desmoplastic 51 Almeida et al., 2008 18496419 27 27 24 72.2 38 7 1 0 Desmoplastic 52 Chen et al., 2008 18823042 128 94 34 65.5 65 27 27 9 Desmoplastic 53 George et al., 2009 19278427 87 52 35 68 46 15 13 13 Desmoplastic 54 Lazova et al., 2009 20950740 4 2 2 81.3 4 0 0 0 Desmoplastic 55 Leinweber et al., 2009 19318805 37 27 10 79 23 5 8 1 Desmoplastic 57 Matuichi et al., 2010 21107116 118 68 50 65 66 30 15 7 Desmoplastic 58 Murali et al., 2010 (incl. 26 from 21438911) 20564101 123 88 35 61.2 40 36 40 7 Desmoplastic 59 Ivette et al., 2010 20653825 9 5 7 73 7 2 0 0 Desmoplastic 60 Maurichi et al., 2010 21107116 242 139 103 64 134 63 30 15 Desmoplastic 61 Feng et al., 2011 21518379 1129 717 412 66.4 578 195 269 74 Desmoplastic 62 Wasif et al., 2011 21259250 1735 1126 609 69 882 307 260 260 Desmoplastic 63 Bernaba et al., 2011 21610456 12 11 1 74.5 8 3 1 0 Desmoplastic 64 Pryor et al., 2011 21785329 5 5 0 77 4 1 0 0 Desmoplastic 65 Miller et al., 2012 22157936 23 14 9 70 11 6 5 1 Desmoplastic 66 Mohebati et al., 2012 22766985 47 27 20 N/A 47 0 0 0 Desmoplastic 67 Coupelon et al., 2013 23095503 10 7 3 N/A 3 4 2 1 Desmoplastic 68 Broer et al., 2013 23522951 24 16 6 65 16 6 1 1 Desmoplastic 69 Jaimes et al., 2013 23325288 16 9 7 67 8 5 2 1 Desmoplastic 70 Carrera et al., 2013 23506474 8 4 4 73.5 8 0 0 0 Desmoplastic 71 Han et al., 2013 23389470 205 141 64 66.3 105 50 25 25 Sum

Desmoplastic total 5372 3243 1824 64 2853 1031 911 536 5331 Desmoplastic % 64.0 35.9 Age range (13-92) 53.5 19.3 17.1 10.1 100%

Histotype Reference PMID No. Cases

Male Female Age (Median) Head & Neck Trunk Upper Extr

Lower Extr. Conventional 1 Newell et al., 1988 3409232 13255 6636 6619 N/A 2686 4097 2914 2821

Conventional 2 Osterlind et al., 1988 3179192 1533 949 1427 N/A 418 809 306 787 Conventional 3 Cress et al., 1995 8634653 242 N/A N/A N/A 16 71 66 89 Conventional 4 Garbe et al., 1995 736393 5093 N/A N/A N/A 982 1490 850 1771 Conventional 5 Naldi et al., 2005 16132799 531 N/A N/A N/A 81 219 48 183 Conventional 6 Hawinks et al., 2005 15827812 3976 2227 1749 53 684 1375 1917 0 Conventional 7 Livestro et al., 2005 16170181 178 N/A N/A 63.5 36 65 37 40 Conventional 8 Lasithiotakis et al., 2006 16909413 1911 N/A N/A 53.1 332 684 297 598 Conventional 9 Clark et al., 2007 17337091 549 N/A N/A N/A 94 237 94 124 Sum

Conventional total 27268 9812 9795 53.1 5329 9047 6529 6413 27318 Conventional % 50.0 49.9 19.5 33.1 23.9 23.5 100%

Supplemental Table 6. Meta-review of the Immunoprofile in Spindle Cell Melanoma

Reference PubmedID NSE

SMA

S100

Col

l IV

Lam

inin

Vim

entin

NK

I/C-3

HM

B45

Mel

anA

Tyr

MIT

F

MU

M1

p75

GFA

P

L1 (N

AM

)

CD

68

Des

min

CD

34

cKIT

CK

Clu

ster

in

MSA

SOX1

0

HM

B50

KB

A.6

2

NF

MA

GE1

CD

10

Podo

plan

in

p63

HH

V8N

A1

Fasc

in

34be

taE1

2

Buonaccorsi 2012 22257901 0/13 7/13 2/13 0/13 Daniels 1994 8164154 3/3 2/3 Granter 2001 11391097 10/11 3/11 1/3 4/11 Grogg 2005 15467709 2/6 5/6 Hoang 2001 11737519 0/1 0/1 0/1 0/1 1/1 0/1 0/1 Iwamoto 2001 11481518 17/17 6/16 7/15 6/17 16/16 9/17 9/10 4/15 Iwamoto 2002 11934320 3/5 5/5 5/5 5/5 5/5 0/5 5/5 Jo 2011 22031315 1/20 Jungbluth 1998 9591730 0/5 1/5 Kanik 1996 8793654 3/3 3/3 0/3 Lazova 2009 20950740 2/3 0/3 3/3 Longacre 1996 8944042 0/1 2/2 2/2 1/2 0/2 Mohamed 2012 23197006 19/19 Morgan 2008 18496422 0/3 3/3 0/3 0/3 0/3 2/3 0/3 Nonaka 2008 18636017 7/7 7/7 Pages 2008 18495211 4/4 4/4 4/4 Patel 2004 14990970 0/3 Piao 2008 18260089 15/21 3/8 5/20 Prieto 1998 9694618 21/22 17/22 8/20 13/20 Ringens 1989 2661335 18/19 Sigal 2012 22207443 10/11 0/11 11/11 Skelton 1997 9415614 22/22 11/22 Sundram 2003 12920225 3/3 2/3 1/3 2/3 Winnepenninckx 2003 14663357 0/1 0/1 1/1 1/1 1/1 0/1 0/1 0/1 1/1 0/1 0/1 0/1 1/1 Xu 2002 11756773 9/14 6/14 9/14 11/14

Sum 0/1 0/6 126/139 17/22 8/20 1/1 19/20 54/118 29/66 17/37 18/30 2/3 31/39 9/17 9/10 1/5 0/7 1/2 0/1 0/17 2/6 0/1 26/26 9/20 4/4 1/1 11/14 7/13 2/13 3/36 0/3 5/6 0/3 % Positive 0 0 90.6 77.3 40 100 95 45.8 43.9 45.9 60 66.7 79.5 52.9 90 20 0 50 0 0 33.3 0 100 45 100 100 78.6 53.8 15.4 8.3 0 83.3 0

Supplemental Table 7. Meta-review of the Immunoprofile in Desmoplastic Melanoma

Reference PubmedID

NSE

SMA

S100

Vim

entin

NK

I/C-3

HM

B45

Mel

anA

Tyr

MIT

F

MU

M1

PNL2

p75

CD

68

Des

min

Cal

desm

on

Cal

pind

in

CD

34

c-ki

t

bFG

F

FXIII

a

CK

Clu

ster

in

Perip

herin

SOX1

0

CD

57

EMA

CA

M5.

2

KB

A.6

2

MA

GE1

Nes

tin

Proc

olla

gen

Cal

cycl

in (S

100A

6)

DO

G1

Fasc

in

VEG

FR2

Leu7

A1a

CT

FVIII

a

Al Alousi 1996 8721445 22/23 Anstey 1994 7512798 21/21 24/24 15/18 13/23 6/24 0/17 0/22 Aung 2012 22067329 2/34 28/34 Bahrami 2010 20829680 4/4 Boyle 2002 12088451 4/5 4/5 4/5 Busam 2001 11176068 20/20 1/20 6/20 3/20 Busam 1998 9706977 4/14 Busam 2005 15725810 1/13 1/13 0/13 1/13 1/13 Busam 2005 15675962 4/4 0/4 0/4 0/4 0/4 4/4 Carlson 1994 9420906 25/47 13/23 198/204 48/49 9/19 11/117 7/27 22/38 1/42 10/24 1/42 1/42 0/12 0/1 Clarkson 2000 11253130 2/2 0/20 0/2 0/20 Coupelon 2012 23095503 10/10 1/10 Egbert 1988 3167815 8/8 Espinosa 2008 18223323 1/10 1/10 1/10 Fetsch 1999 10328095 4/4 George 2008 19278427 54/55 4/34 3/8 5/16 Granter 2001 11391097 10/10 1/10 1/10 2/10 Grogg 2005 15467709 0/10 10/10 Harris 1999 10403301 9/9 4/9 1/7 2/8 0/7 Ivette 2010 20653825 4/9 0/9 1/9 2/9 4/9 Jain 1989 2712188 19/22 Jennings 1996 8872151 Kanik 1996 8793654 10/10 0/10 10/10 Kay 2004 14765268 6/28 24/28 28/28 2/28 4/28 4/28 0/28 0/28 3/28 0/28 0/28 King 2001 11145251 1/14 Kossard 1987 3606170 7/10 Kucher 2004 15618925 44/44 3/42 3/43 Lazova 2009 20950740 10/10 3/10 10/10 Leinweber 2009 19318805 37/37 19/37 Longacre 1996 8944042 8/19 22/22 0/22 2/22 0/15 Miettinen 2001 11176069 27/27 0/27 0/27 1/30 1/30 Miettinen 2012 22314185 0/17 Miller 2012 22157936 26/41 20/41 2/41 37/41 35/41 13/37 Mohamed 2012 23197006 22/22 Morgan 2008 18496422 3/3 0/3 0/3 0/3 Nonaka 2008 1863017 27/28 28/28 North 2011 21645038 2/14 Orchard 1999 11095072 6/6 3/6 1/6 2/6 3/6 Orosz 1999 10383696 2/2 0/2 1/2 Pages 2008 18495211 5/5 4/5 5/5 Palla 2013 23291581 15/15 Prasad 2001 11395556 5/5 1/5 1/5 4/5 0/5 Reimann 1987 3440236 8/9 7/8 Ribe 2003 12748257 0/3 Riccioni 1999 10608246 0/2 3/3 3/3 0/3 0/2 0/3 0/2 2/2 0/2 0/3 Shaw 2006 16525175 Skelton 1991 1725245 0/30 Skelton 1994 7722014 5/10 67/69 0/64 0/16 Sundram 2003 12920225 5/5 0/5 0/5 0/5 Warner 1984 6360335 3/4 4/4 Wharton 1999 10333224 16/16 1/9 Whitaker 1992 1583168 3/3 1/3 1/3 Winnepenninckx 2003 14663357 6/6 1/6 8/8 8/8 0/8 0/7 2/6 0/8 0/6 2/6 Xu 2002 11756773 0/8 0/8 0/8 3/8

Sum 55/80 36/89 718/745 105/111 25/48 49/559 44/234 29/161 10/113 0/5 3/47 46/61 2/25 14/109 0/2 2/2 4/39 23/57 22/23 24/68 1/78 6/55 0/7 106/115 3/44 10/52 1/73 33/39 3/8 35/41 13/37 0/3 1/10 10/10 0/17 1/42 0/12 0/1 % Positive 68,8 40,4 96,4 94,6 52,1 8,8 18,8 18 8,8 0 6,4 75,4 8 12,8 0 100 10,3 40,4 95,7 35,3 1,3 10,9 0 92,2 6,8 19,2 1,4 84,6 37,5 85,4 35,1 0 10 100 0 2,4 0 0

Supplemental Table 8. Meta-review of the Immunoprofile in Conventional Melanoma

Reference PubmedID NSE

SMA

S100

Vim

entin

NK

I/C-3

HM

B45

Mel

anA

Tyro

sina

se

MIT

F

MU

M1

PNL2

p75

GFA

P

L1 (N

AM

)

CD

68

Des

min

Cal

desm

on

Cal

poni

n

CD

34

c-ki

t

CK

Clu

ster

in

CD

56

GA

P43

Perip

herin

SOX1

0

HM

B50

EMA

KB

A.6

2

NF

MA

GE1

Cal

cycl

in

(S10

0A6)

VEG

FR2

CD

10

Aung 2012 22067329 204/214 186/214 187/214 198/214 183/214 25/30 Bertucci 2007 17970093 38/67Blessing 1998 9543670 37/37 25/37 35/37 Boyle 2002 12088451 10/10 10/10 10/10 6/10 0/10 Busam 1998 9706977 26/26 Busam 2005 15725810 29/38 31/38 35/38 32/38 33/38 Busam 2005 15675962 5/5 5/5 5/5 5/5 5/5 1/5 Clarkson 2000 11253130 21/23 19/23 20/23 22/23 Daniels 1994 8164154 5/5 5/5 Espinosa 2008 18223323 8/21 Gajjar 2004 15223957 55/53 42/53 23/53 49/53 23/53 Huttenbach 2002 12358815 5/30 5/30 6/30 14/14 Ivette 2010 20653825 2/3 2/3 2/3 2/3 2/3 Iwamoto 2001 11481518 11/11 10/10 10/10 9/9 6/10 2/10 7/8 9/10 Iwamoto 2002 11934320 9/10 10/10 10/10 10/10 10/10 0/10 10/10 Jungbluth 1998 9591730 56/75 61/75 Kanik 1996 8793654 8/8 0/8 5/8 King 1999 10487831 71/76 69/76 76/76 King 2001 11145251 44/44 Koganehira 2003 12786828 53/53 22/53 35/53 Miettinen 2001 11176069 250/266 230/266 234/266 247/266 235/266 Miettinen 2012 22314185 0/107 Mohamed 2012 23197006 49/49 Morris 2007 18228523 8/8 8/9 7/8 3/5 7/9 Nonaka 2008 18636017 37/43 41/43 Orchard 1999 11095072 54/55 47/55 38/55 43/55 47/55 Orosz 1999 10383696 7/10 7/10 7/10 Pages 2008 18495211 84/85 75/85 78/85 Piao 2008 18260089 6/8 2/4 1/4 Prasad 2001 11395556 76/79 78/79 74/79 77/79 66/79 Reimann 1987 3440236 10/10 10/10 Ribe 2003 12748257 32/81 Ringens 1989 2661335 281/303 Rode 1984 6526386 35/39 24/28 Shah 1997 9342756 6/6 0/6 6/6 0/6 Skelton 1991 1725245 28/30 Sundram 2005 12920225 36/36 28/36 27/36 33/36 Tsuchihashi 2011 22189238 0/1 1/1 1/1 1/1 1/1 1/1 0/1 1/1 0/1 Wick 1988 3053811 62/62 62/62 62/62 0/62 0/62 Xu 2002 11756773 8/8 8/8 8/8 3/8

Sum 35/39 53/54 1099/1152 73/73 328/358 1025/1210 812/961 715/770 664/748 33/36 40/47 16/58 2/10 7/8 6/6 10/10 22/53 35/53 0/11 9/22 0/63 1/5 5/30 6/30 14/14 92/95 19/20 0/62 103/115 0/6 26/61 32/81 0/107 38/67% Positive 89.7 98.1 95,4 100 91.6 84.7 84.5 93 88.8 91.7 85.1 27.6 20 87.5 100 100 41.5 66 0 40.9 0 20 16.7 20 100 96.8 95 0 89.6 0 42.6 39.5 0 56.7

Supplemental Table 9. Summary of the Published Immunoprofile.

Spindle Cell Melanoma

Desmoplastic Melanoma

Conventional Melanoma*

Marker + Tested % + Tested % + Tested % PSpindle vs.

Desmoplastic PSpindle vs.

Conventional PDesmoplastic

vs. Conventional

34bE12 0 3 0 - - - - - - - - - A1aCT - - - 0 12 0 - - - - - - bFGF - - - 22 23 95.7 - - - - - -

Calcyclin (S100A6) - - - 0 3 0 32 81 39.5 - - 0.28 Caldesmon - - - 0 2 0 22 53 41.5 - - 0.51

Calponin - - - 2 2 100 35 53 66 - - 1 CAM5.2 - - - 1 73 1.4 - - - - - -

CD10 7 13 53.8 - - - 38 67 56.7 - 1 - CD146 - - - - - - - - - - - - CD31 - - - - - - - - - - - - CD34 1 2 50 4 39 10.3 0 11 0 0.23 0.15 0.56

CD56/N-CAM - - - - - - 5 30 16.7 - - - CD57 - - - 3 44 6.8 - - - - - - CD68 1 5 20 2 25 8 6 6 100 0.43 0.02 <0.001

CK (AE1.3) 0 17 0 1 78 1.3 0 63 0 1 1 1 c-kit 0 1 0 23 57 40.4 9 22 40.9 1 1 1

Clusterin 2 6 33.3 6 55 10.9 1 5 20 0.17 1 0.47 Collagen IV 17 22 77.3 - - - - - - - - -

Desmin 0 7 0 14 109 12.8 10 10 100 1 <0.001 <0.001 DOG1 - - - 1 10 10 - - - - - - EMA - - - 10 52 19.2 0 62 0 - - <0.001

Fascin 5 6 83.3 10 10 100 - - - 0.38 - - Factor-VIIIa - - - 0 1 0 - - - - - - Factor-XIIIa - - - 24 68 35.3 - - - - - -

GAP43 - - - - - - 6 30 20 - - - GFAP 9 17 52.9 - - - 2 10 20 - 0.12 -

HHV8NA1 0 3 0 - - - - - - - - - HMB45 54 118 45.8 49 559 8.8 1025 1210 84.7 <0.001 <0.001 <0.001 HMB50 9 20 45 - - - 19 20 95 - 0.001 - KBA.62 4 4 100 33 39 84.6 103 115 89.6 1 1 0.39

L1 (NAM) 9 10 90 - - - 7 8 87.5 - 1 - Laminin 8 20 40 - - - - - - - - -

Leu7 - - - 1 42 2.4 - - - - - - MAGE1 11 14 78.6 3 8 37.5 26 61 42.6 0.08 0.02 1

MelanA/MART-1 29 66 43.9 44 234 18.8 812 961 84.5 <0.001 <0.001 <0.001 MITF 18 30 60 10 113 8.8 664 748 88.8 <0.001 <0.001 <0.001 MSA 0 1 0 - - - - - - - - -

MUM1 2 3 66.7 0 5 0 33 36 91.7 0.11 0.28 <0.001 Nestin - - - 35 41 85.4 - - - - - -

NF 1 1 100 - - - 0 6 0 - 0.14 - NKI/C-3 19 20 95 25 48 52.1 328 358 91.6 <0.001 1 <0.001

NSE 0 1 0 55 80 68.8 35 39 89.7 0.32 0.13 0.01 P63 3 36 8.3 - - - - - - - - -

p75/NGFR/CD271 31 39 79.5 46 61 75.4 16 58 27.6 0.81 <0.001 <0.001 Peripherin - - - 0 7 0 14 14 100 - - <0.001

PNL2 - - - 3 47 6.4 40 47 85.1 - - <0.001 Podoplanin 2 13 15.4 - - - - - - - - - Procollagen - - - 13 37 35.1 - - - - - -

S100 126 139 90.6 718 745 96.4 1099 1152 95.4 0.006 0.024 0.35 SMA 0 6 0 36 89 40.4 53 54 98.1 0.15 <0.001 <0.001

SOX10 26 26 100 106 115 92.2 92 95 96.8 0.21 1 0.23 Tyrosinase 17 37 45.9 29 161 18 715 770 92.9 <0.001 <0.001 <0.001

VEGFR2 - - - 0 17 0 0 107 0 - - 1 Vimentin 1 1 100 105 111 94.6 73 73 100 1 1 0.08

Abbreviations: IHC, immunohistochemistry; +, sum of positively staining cases; tested, sum of cases tested in the literature; P values derived from Fisher’s exact test. *Note: numbers for conventional melanoma are derived from those cases included in the reviewed spindle and desmoplastic melanoma publications (see supplemental tables 6-8).

Supplemental Table 10. Marker Comparison by Cohort

P P P P Biomarker Spindle cell

Ulm (n=9) Spindle cell

New York (n=7) Spindle cell

Ulm vs. New York Desmoplastic

Ulm (n=9) Desmoplastic

New York (n=9) Desmoplastic

Ulm vs. New York Ulm

Spindle vs. Desmoplastic New York

Spindle vs. Desmoplastic Positive % Positive % Fisher Positive % Positive % Fisher Fisher Fisher

Vimentin 9 100

7 100

1.00 9 100 9 100 1.00 1.00 1.00 Coll IV (NT) 9 10

0 7 10

0 1.00 9 100 9 100 1.00 1.00 1.00

S100 (C) 9 100

5 71.42857143

0.18 8 88.88888889

9 100 1.00 1.00 0.18 NSE 8 88

.88888889

7 100

1.00 9 100 6 67 0.21 1.00 0.21 CD68 (NT) 8 88

.88888889

7 100

1.00 6 66.66666667

9 100 0.21 0.58 1.00 S100 (N) 7 77

.77777778

5 71.42857143

1.00 8 88.88888889

9 100 1.00 1.00 0.18 p75 (N) 7 77

.77777778

5 71.42857143

1.00 8 88.88888889

5 56 0.29 1.00 0.63 CD34 (V) 6 66

.66666667

4 57.14285714

1.00 8 88.88888889

4 44 0.13 0.58 1.00 p53 3 33

.33333333

5 71.42857143

0.31 7 77.77777778

7 78 1.00 0.15 1.00 c-myc 4 44

.44444444

5 71.42857143

0.36 0 0 4 44 0.08 0.08 0.36 Laminin (NT) 6 66

.66666667

0 0 0.01 1 11.11111111

0 0 1.00 0.05 1.00 p75 (C) 7 77

.77777778

2 28.57142857

0.13 1 11.11111111

1 11 1.00 0.02 0.55 Laminin (T) 2 22

.22222222

4 57.14285714

0.30 0 0 0 0 1.00 0.47 0.02 HMB45 8 88

.88888889

2 28.57142857

0.04 1 11.11111111

1 11 1.00 0.00 0.55 c-kit 5 55

.55555556

4 57.14285714

1.00 1 11.11111111

0 0 1.00 0.13 0.02 MelanA/MART-1 8 88

.88888889

4 57.14285714

0.26 0 0 0 0 1.00 0.00 0.02 Coll IV (T) 1 11

.11111111

0 0 1.00 7 77.77777778

2 22 0.06 0.02 0.48 Trichrome 1 11

.11111111

1 14.28571429

1.00 7 77.77777778

4 44 0.33 0.02 0.31 CD68 (T) 0 0 0 0 1.00 7 77.7

7777778

2 22 0.06 0.00 0.48 MDM2 (N) 3 33

.33333333

2 28.57142857

1.00 9 100 1 11 0.00 0.01 0.55 MDM2 (C) 4 44

.44444444

3 42.85714286

1.00 5 55.55555556

2 22 0.33 1.00 0.60 SMA (V) 5 55

.55555556

1 14.28571429

0.15 3 33.33333333

4 44 1.00 0.64 0.31 CD56 1 11

.11111111

0 0 1.00 1 11.11111111

2 22 1.00 1.00 0.48 MUM1 2 22

.22222222

0 0 0.48 0 0 0 0 1.00 0.47 1.00 CD34 (T) 1 11

.11111111

0 0 1.00 0 0 0 0 1.00 1.00 1.00 CD99 0 0 0 0 1.00 0 0 1 11 1.00 1.00 1.00 HER2 0 0 0 0 1.00 1 11.1

1111111

0 0 1.00 1.00 1.00 EGFR 0 0 1 14

.28571429

0.44 0 0 0 0 1.00 1.00 0.44 c-MET 4 44

.44444444

0 0 0.09 1 11.11111111

1 11 1.00 0.29 1.00 ALK 0 0 0 0 1.00 0 0 1 11 1.00 1.00 1.00 FLI-1 1 11

.11111111

1 14.28571429

1.00 0 0 0 0 1.00 1.00 0.44 BRAF 4 44

.44444444

1 14.28571429

0.31 1 11.11111111

0 0 1.00 0.29 0.44 Desmin 0 0 0 0 1.00 0 0 0 0 1.00 1.00 1.00 CD57 0 0 0 0 1.00 0 0 0 0 1.00 1.00 1.00

NF 0 0 0 0 1.00 0 0 0 0 1.00 1.00 1.00 GFAP 0 0 0 0 1.00 0 0 0 0 1.00 1.00 1.00

CK 0 0 0 0 1.00 0 0 0 0 1.00 1.00 1.00 CAM5.2 0 0 0 0 1.00 0 0 0 0 1.00 1.00 1.00 DOG1 0 0 0 0 1.00 0 0 0 0 1.00 1.00 1.00

SMA (T) 0 0 0 0 1.00 0 0 0 0 1.00 1.00 1.00

HER2 0(9) 0 0(5) 0 1.00 0(9) 0 0(6) 0 1.00 1.00 1.00 EGFR 0(7) 0 0(6) 0 1.00 0(9) 0 0(8) 0 1.00 1.00 1.00 c-MET 0(8) 0 0(6) 0 1.00 0(9) 0 0(8) 0 1.00 1.00 1.00 MDM2 1(8) 13 0(5) 0 1.00 0(9) 0 0(9) 0 1.00 0.47 1.00 TP53 0(7) 0 0(5) 0 1.00 0(9) 0 0(8) 0 1.00 1.00 1.00 ALK 0(6) 0 0(6) 0 1.00 0(9) 0 0(7) 0 1.00 1.00 1.00

c-myc 0(9) 0 0(6) 0 1.00 0(9) 0 0(7) 0 1.00 1.00 1.00

Abbreviations: C, cytoplasmic; N, nuclear stain; NT, non-tumor cells; T, tumor cells; V, vascular density; Color code: IHC, FISH, special stain

Supplemental Table 11. Marker Comparison by Histological Subtype

Test Performance Characteristics

Biomarker Spindle cell melanoma all (n=16)

Desmoplastic melanoma all (n=18)

P PPV NPV Sensitivity Specificity Accuracy Youden Index

Selection

Positive % Positive % Fisher Positive in

Vimentin 16 100 18 100 1.000 SM 47 N/A 100 0 47 0Coll IV (NT) 16 100 18 100 1.000 SM 47 N/A 100 0 47 0

S100 (C) 14 87.5 17 94 0.591 DM 55 67 94 13 56 7NSE 15 93.75 15 83 0.604 SM 50 75 94 17 53 10

CD68 (NT) 15 93.75 15 83 0.604 SM 50 75 94 17 53 10S100 (N) 12 75 17 94 0.164 DM 59 80 94 25 62 19p75 (N) 12 75 13 72 1.000 SM 48 56 75 28 50 3

CD34 (V) 10 62.5 12 67 1.000 DM 55 50 67 38 53 5p53 8 50 14 78 0.151 DM 64 67 78 50 65 28

c-myc 9 56.25 4 22 0.076 SM 69 67 56 78 68 34Laminin (NT) 6 37.5 1 6 0.035 SM 86 63 38 94 68 32 SM6

p75 (C) 9 56.25 2 11 0.009 SM 82 70 56 89 74 45 SM4Laminin (T) 6 37.5 0 0 0.006 SM 100 64 38 100 71 38 SM5

HMB45 10 62.5 2 11 0.003 SM 83 73 63 89 76 51 SM2c-kit 9 56.25 1 6 0.002 SM 90 71 56 94 76 51 SM3

MelanA/MART-1 12 75 0 0 0.000 SM 100 82 75 100 88 75 SM1Coll IV (T) 1 6.25 9 50 0.008 DM 90 63 50 94 71 44 DM3Trichrome 2 12.5 11 61 0.005 DM 85 67 61 88 74 49 DM1CD68 (T) 0 0 9 50 0.001 DM 100 64 50 100 74 50 DM2MDM2 (N) 5 31.25 10 56 0.185 DM 67 58 56 69 62 25 DM4MDM2 (C) 7 43.75 7 39 0.052 SM 50 55 44 61 53 5SMA (V) 6 37.5 7 39 1.000 DM 54 48 39 63 50 2

CD56 1 6.25 3 17 0.604 DM 75 50 17 94 53 11MUM1 2 12.5 0 0 0.214 SM 100 56 13 100 59 13

CD34 (T) 1 6.25 0 0 0.471 SM 100 55 6 100 56 6CD99 0 0 1 6 1.000 DM 100 49 6 100 50 6HER2 0 0 1 6 1.000 DM 100 49 6 100 50 6EGFR 1 6.25 0 0 0.471 SM 100 55 6 100 56 6c-MET 4 25 2 11 0.387 SM 67 57 25 89 59 14ALK 0 0 1 6 1.000 DM 100 49 6 100 50 6FLI-1 2 12.5 0 0 0.214 SM 100 56 13 100 59 13BRAF 5 31.25 1 6 0.078 SM 83 61 31 94 65 26

Desmin 0 0 0 0 1.000 N/A N/A N/A N/A N/A N/A N/A CD57 0 0 0 0 1.000 N/A N/A N/A N/A N/A N/A N/A

NF 0 0 0 0 1.000 N/A N/A N/A N/A N/A N/A N/A GFAP 0 0 0 0 1.000 N/A N/A N/A N/A N/A N/A N/A AE1.3 0 0 0 0 1.000 N/A N/A N/A N/A N/A N/A N/A

CAM5.2 0 0 0 0 1.000 N/A N/A N/A N/A N/A N/A N/A DOG1 0 0 0 0 1.000 N/A N/A N/A N/A N/A N/A N/A

SMA (T) 0 0 0 0 1.000 N/A N/A N/A N/A N/A N/A N/A

HER2 0(14) 0 0(15) 0 1.000 N/A N/A N/A N/A N/A N/A N/A EGFR 0(13) 0 0(17) 0 1.000 N/A N/A N/A N/A N/A N/A N/A c-MET 0(14) 0 0(17) 0 1.000 N/A N/A N/A N/A N/A N/A N/A MDM2 1(13) 8 0(18) 0 0.420 SM 100 60 8 100 59 8TP53 0(12) 0 0(17) 0 1.000 N/A N/A N/A N/A N/A N/A N/A ALK 0(12) 0 0(16) 0 1.000 N/A N/A N/A N/A N/A N/A N/A

c-myc 0(15) 0 0(16) 0 1.000 N/A N/A N/A N/A N/A N/A N/A

Abbreviations: C, cytoplasmic; N, nuclear stain; NPV, negative predictive value; NT, assessed in non-tumor cells; PPV, positive predictive value, T, assessed in tumor cells; V, vascular density; Color code: IHC, FISH, special stain

36

Acknowledgements

First, I would like to thank the chairman of the Institute of Pathology at University

of Ulm, Prof. Dr. med. Peter Möller, who served as my thesis advisor. He

provided me the unique opportunity and support to complete my thesis and

publication. Prof. Möller was always willing to share his extensive knowledge.

Next, I would like to thank my supervisor Dr. Jochen Lennerz for his continuous

support and guidance. His patience and enthusiasm is contagious.

I also want to thank all of my coworkers at the Institute of Pathology. Specifically,

I would like to thank Michaela Buck, Julia Kiedaisch, Iwona Nerbas and Dr. Ulrike

Kostezka for their expert technical assistance and their patience in early times of

my laboratory work. In addition I would like to thank my co-doctoral students

Philipp Nagel, Fenja Feld and Anne-Sophie K. Meyer.

In addition, I would like to thank Dr. Basil A. Horst and Prof. David N. Silvers from

the Columbia University New York, NY, USA, for sharing samples and their

expertise.

Lastly, I want to thank my parents and family for never loosing faith.

Curriculum Vitae

Cand. Med. Stephanie E. Weißinger

Date of Birth: 6th of August 1982 Place of Birth: Nürtingen, Baden-Württemberg, Germany Nationality: German

Content removed for reasons of data protection.

Curriculum Vitae Cand. Med. Stephanie E. Weißinger

Content removed for reasons of data protection.