Cross-reactivity of NRASQ61R antibody in a subset of Spitz nevi with 11p gain: a potential confounding factor in the era of pathway-based diagnostic approach

Summary

The most recent World Health Organization classification for skin tumors (2018) categorizes melanomas and their precursor lesions, benign or intermediate, into nine pathways based not only on their clinical and histomorphologic characteristics but also on their molecular profile and genetic fingerprint. In an index case of a partially sampled atypical spitzoid lesion, which proved to be an 11p-amplified Spitz nevus with HRASQ61R mutation, we observed cross-reactivity with the NRASQ61R antibody (clone SP174). Overall, we assessed the status of HRAS and NRAS genes and their immunoreaction to NRASQ61R antibody in 16 cases of 11p-amplified Spitz nevi/atypical Spitz tumors. We also assessed the immunoexpression of NRASQ61R antibody in various malignancies with proven BRAFV600E, NRASQ61R, L or K, KRASQ61R and HRASQ61R, and HRASQ61R mutations and ALK Spitz lesions. Finally, we assessed the expression of PReferentially expressed Antigen in MElanoma (PRAME) immunohistochemistry in our 11p Spitz cohort. Three of 16 cases (3/16) harbored the HRASQ61R mutation and exhibited diffuse immunoreaction with the NRASQ61R antibody. All the cases in our cohort were negative for the NRASQ61R mutation. All NRASQ61R-, KRASQ61R-, and HRASQ61R-mutated neoplasms were positive for the antibody, further supporting the cross-reactivity between the RAS proteins. All the cases of our cohort were essentially negative for PRAME immunohistochemistry. In the era of pathway-based approach in the diagnosis of melanocytic neoplasms, the cross-reactivity between the NRASQ61R- and HRASQ61R-mutated proteins can lead to a diagnostic pitfall in the assessment of lesions with spitzoid characteristics.

1. Introduction

Spitz nevi fall into the category of Spitz tumors, that also include atypical Spitz tumors (ASTs) and Spitz melanoma [1,2]. More than 20% of Spitz nevi genetically harbor an increase in copy number of 11p chromosome, where the HRAS gene is located [3,4]. The amplification of HRAS in approximately 67% of cases is also associated with gain-of-function mutations that commonly involve exon 2 (codon 61). This results in the replacement of glutamine with arginine (HRASQ61R) or less frequently with leucine (HRASQ61L) [5]. Exons 1 [4] and 3 [6] can also be involved. Increased copy numbers of wild-type HRAS [5] with or without accompanying mutations lead to aberrant activation of the MAPK/ERK and the PI3K/AKT/mechanistic Target Of Rapamycin pathway.

NRAS belongs to the Ras gene family, which is implicated in cell proliferation, migration, and survival through activation of Raf/Mek/Erk (MAPK) and PI3K/mTOR pathways as well [7]. BRAF and NRAS mutations are the driver genetic events toward the formation of many melanocytic lesions ranging from banal nevi to melanoma [1]. In the era of personalized medicine, the detection of these mutations is critical for predictive purposes. NRASQ61R antibody is an immunohistochemical marker commonly used as a surrogate marker for an underlying NRASQ61R mutation in malignant melanoma as well as colorectal and thyroid carcinomas [8,9]. However, its specificity is hampered owing to cross-reactivity with HRASQ61R- and KRASQ61R-mutated gene products in a subset of medullary thyroid carcinomas, adenomyoepitheliomas of the breast, colorectal carcinomas, and melanomas [8,10–12].

As we enter the era of pathway-based approach for the diagnosis of melanocytic lesions, our the morphological threshold, in an effort to accurately predict biologic behavior, may differ depending on which molecular category the tumor belongs to. For example, true Spitz melanocytic proliferations may exhibit significant cytologic atypia and still behave in an indolent fashion. On the other hand, a melanocytic lesion that resembles a Spitz lesion but carries BRAF/NRAS mutations (therefore spitzoid rather than true Spitz or spitzian) may behave aggressively despite similar cytologic features.

In an index case of a partially sampled atypical spitzoid lesion, proved to be an 11p-amplified Spitz nevus with HRASQ61R mutation, we observed cross-reactivity with the NRASQ61R antibody (clone SP174), which initially created a diagnostic dilemma. We subsequently revisited all the 11p-amplified Spitz nevi in the databases of our academic institutions and assessed the status of HRAS and NRAS genes and their immunoreaction to NRASQ61R antibody. We also assessed the immunoexpression of NRASQ61R antibody in various malignancies from our database with proven BRAFV600E, NRASQ61R, L or K, KRASQ61R and HRASQ61R mutations and ALK+ Spitz lesions. Finally, as a secondary branch, we assessed the expression of PReferentially expressed Antigen in MElanoma (PRAME) immunohistochemistry (IHC) in our 11p Spitz cohort. The latter is a relatively new immunohistochemical marker whose expression is supposedly useful in the diagnosis of suspected melanomas. However, the data with regard to its expression in Spitz lesions are relatively scarce.

2. Materials and methods

2.1. Case selection

On the basis of an index case, a retrospective search (2015–2020) through the Chromosomal Microarray database of Dartmouth-Hitchcock Medical Center (NH, USA) identified 15 additional melanocytic cases with 11p gain as monoaberration or oligoaberration from 15 patients out of approximately 300 cases. One additional case was contributed from the University of Virginia (after additional comprehensive review of their microarray database). All cases had been diagnosed as either Spitz nevus or AST. All slides were evaluated for diagnostic confirmation, and patients’ demographics as well as information on the location of the lesion, initial and subsequent procedure, and follow-up were gathered from medical records.

This study received approval from the Dartmouth-Hitchcock Health Institutional Review Board (DHH IRB STUDY02000523).

2.2. Control groups

A retrospective search through the molecular database of Dartmouth-Hitchcock Medical Center (NH, USA) from 2014 to 2020 was conducted to identify various neoplasms that were analyzed during routine clinical workup and harbored the following mutations: NRASQ61R, NRASQ61K, NRASQ61L, HRASQ61R, KRASQ61R, and BRAFV600E. Finally, a group of 5 ALK+ ASTs, detected either by IHC, fluorescence in situ hybridization, next-generation sequencing (NGS), or a combination thereof, was also included.

2.3. Immunohistochemistry

NRASQ61R antibody (dilution 1:50, clone SP174; Abcam, Cambridge, MA) and PRAME (dilution 1:50, #219650, Mab EPR20330; Abcam) IHC were performed on all our Spitz nevi/AST cases with 11p gain. NRASQ61R IHC was also performed on the control cases when they had not already been part of the diagnostic workup. PRAME IHC was assessed as per the recent article by Lezcano et al. [13], who defined it as positive when more than 75% of the lesional nuclei exhibited immunoreactivity. The studies were performed using the Leica Bond polymer visualization system (product #DS9800) on the Leica Bond III autostainer (Leica Biosystems, Buffalo Grove, IL).

2.4. Chromosome microarray analysis

Tumor tissue samples from unstained formalin-fixed paraffin-embedded (FFPE) tissue sections were macro-dissected using a scalpel or razor blade based on tumor regions marked by a pathologist on a corresponding hematoxylin and eosin (H&E)–stained slide. DNA was extracted from these tissue samples using the QIAamp DNA FFPE tissue kit (Qiagen, Hilden, Germany) and subjected to chromosome microarray analysis (CMA) testing using the OncoScan FFPE Assay or Oncoscan CNV Assay (Thermo Fisher Scientific, Waltham, MA) as previously described by Hedayat et al [14]. CMA analysis and review was performed using Chromosome Analysis Suite software (Thermo Fisher Scientific, Waltham, MA).

2.5. Next Generation Sequencing

All cases were reviewed, circled, and quantified by a pathologist to assess for adequacy and enrichment for tumor content. After the pathologist’s review, macrodissection was performed to ensure ≥10% tumor cellularity.

2.5.1. CHPv2 Hotspot Panel (50 genes) methods

DNA was extracted from eight 4-μm sections of FFPE tumor tissue using the Gentra PureGene Blood Kit Plus or Qiacube (Qiagen, Hilden, Germany) and quantified using PicoGreen (Promega, Madison, WI). Bar-coded libraries were prepared from 10 ng of DNA using the Ion AmpliSeq Library Kit 2.0 with Ion AmpliSeq Cancer Hotspot Panel version 2 according to the manufacturer’s protocol (Life Technologies, Rockville, MD). Libraries were sequenced using the Ion Torrent Personal Genome Machine (PGM) System and Ion 318 Chips (Life Technologies, Rockville, MD). Sequencing reads were aligned to hg19, and variants were called using Torrent Suite Variant Caller Plugin version 4.0 (Thermo Fisher Scientific, Waltham, MA). Variant annotation was performed using Golden Helix SNP and Variation Suite software version 8.2.1 (Golden Helix, Bozeman, MT). Filters were applied to remove benign polymorphisms, as well as noncoding and synonymous variants. A report detailing variants detected in the tumor and the resultant amino acid changes was included in each patient’s medical record.

2.5.2. TruSight tumor 170

DNA and RNA were isolated from eight 4-μm FFPE tissues using the AllPrep DNA/RNA FFPE Kit Protocol on the QIAcube instrument (Qiagen). DNA was sheared using the Covaris^® ME220 Focused-ultrasonicator^™ (Covaris, Woburn, MA). Libraries were prepared using the Biomek NX^p (Beckman Coulter, Brea, CA Illumina, San Diego, CA) using the Illumina TruSight Tumor 170 kit and sequenced using the NextSeq^® 500 System (Illumina). After sequencing, TST170 Local App version 1.0.1 (Illumina), housed in Clinical Genomics Workspace (CGW) (Pierian DX, Creve Coeur, MO), was used to perform alignment and variant calling. Variants detected from paired DNA and RNA samples were combined into a single sample output and reported in CGW. Variants were classified as IA to III as per the AMP (Association of Molecular Pathology)/ASCO (American Society of Clinical Oncology)/CAP (College of American Pathologists) guidelines for interpretation and reporting of sequence variants in cancer.

2.6. Droplet digital polymerase chain reaction

DNA from FFPE tissue was extracted on the QIAcube, using the QIAamp DNA FFPE Tissue kit (QIAGEN, Germany), and quantified on the Qubit 3.0 fluorometer (Thermo Fisher Scientific, USA). The samples were tested using two droplet digital polymerase chain reaction (ddPCR) mutation detection assays, HRAS p.Q61R c.182 A>G and NRAS p.Q61R c.182 A>G (Bio-Rad Laboratories, Hercules, CA); standard conditions were followed. Each reaction mix was prepared using 2× ddPCR supermix for probes (Bio-Rad, USA), 20× ddPCR mutation detection assay (Bio-Rad, USA), HaeIII (10 IU/μl) (New England BioLabs, Ipswich, MA), nuclease-free water, and sample DNA. Droplets were generated on the Bio-Rad Automated Droplet Generator, and the plate was sealed on a PX1 PCR Plate Sealer (Bio-Rad, USA). The plate was then subjected to the following conditions in a C1000 Touch Thermal Cycler (Bio-Rad, USA): 95°C for 10 min and 40 cycles at 94°C for 30 s and 53°C for 1 min, 98°C for 10 min, and 4°C indefinitely (at a ramp rate of 2°C/sec). The plate was then transferred to the Bio-Rad QX200 Droplet Reader to be read.

2.7. NRAS/BRAF rapid

The FFPE tissue specimens were tested using the Idylla^™ system (Biocartis, Mechelen, Belgium) using the NRAS-BRAF cartridges. For testing, tumor tissue was enriched by macrodissection after the tumor area and neoplastic cellularity were determined by a pathologist to ensure the specimens meet assay requirements (≥10% neoplastic cellularity and 50–600 mm² tissue area). A total of 23 common mutations located in the BRAF (5) and NRAS (18) genes are detected by this cartridge: NRAS codon 12 (exon 2): G12C (c.34G>T), G12S (c.34G>A), G12D (c.35G>A), G12A (c.35G>C), G12V (c.35G>T); codon 13 (exon 2): G13D (c.38G>A), G13V (c.38G>T), G13R (c.37G>C); codon 59 (exon 3): A59T (c.175G>A); codon 61 (exon 3): Q61K (c.181C>A), Q61L (c.182A>T), Q61R (c.182A>G), Q61H (c.183A>C; c.183A>T); codon 117 (exon 4): K117N (c.351G>C; c.351G>T); codon 146 (exon 4): A146T, A146V (c.436G>A), (c.437C>T). BRAF codon 600: V600E (c.1799T>A; c.1799_1800delinsAA), V600D (c.1799_1800delinsAC), V600K (c.1798_1799delinsAA), V600R (c.1798_1799delinsAG).

2.8. 23-gene expression signature (myPath melanoma)

An H&E-stained glass slide with the lesion marked and several unstained slides were submitted to Myriad A pathologist from the assay’s manufacturer macro-dissected the corresponding lesion on unstained slides for extraction and analysis using the gene expression signature based on published criteria [15]. Using a proprietary scoring system, melanocytic lesions were labeled as benign (−16.7 to −2.1), indeterminate (−2.0 to −0.1), or malignant (0.0 to +11.1).

3. Results

3.1. HRAS Spitz nevi/tumors

3.1.1. Clinical and histologic findings

Sixteen patients were identified (including our index case), 10 women and 6 men, with a mean age of 33 years (range = 12–72 years). After rereview, 13 cases were diagnosed as Spitz nevus and 3 cases were diagnosed as AST, favor indolent biologic behavior. Five cases involved the thigh, 5 the upper extremities, 3 the trunk, and 2 the head and neck. The initial procedure was shave biopsy or removal in 9 cases, punch biopsy in 3 cases, and excision or excisional biopsy in 4 cases. Margins in 3 of the excisional specimens and 2 shave removals were reported as negative; thus, re-excision was not performed. In 10 of 11 remaining cases, patients underwent re-excision, whereas in one recent case, it is pending. In 4 cases, no residual lesion was identified, whereas in 5 cases, focal residuum was reported with negative margins. In one case, the residuum extended focally to the peripheral margin, but no further action was taken. The patients’ clinical characteristics are summarized in Table 1.

Table 1.

Patients’ clinical characteristics.

Case no.	Age (years)	Sex	Location	Initial procedure	Re-excision	Diagnosis
1	37	M	Neck	Shave biopsy	NR	Compound Spitz nevus
2	40	F	Thigh	Shave biopsy	R, MF	Compound Spitz nevus
3	23	F	Thigh	Excision (MF)	NA	Predominantly intradermal Spitz nevus
4	69	M	Trunk	Shave biopsy	R, MF	Dysplastic nevus-like Spitz
5	49	M	Arm	Excision (MF)	NA	AST
6	21	M	Arm	Punch biopsy	NR	Compound Spitz nevus
7	39	M	Chest	Shave biopsy	NA	Compound Spitz nevus
8	53	F	Thigh	Punch biopsy	R, MF	Compound Spitz nevus
9	25	F	Shoulder	Shave biopsy	R, MF	Predominantly intradermal AST
10	14	F	Jaw	Shave biopsy	R, MF	AST
11	12	F	Thigh	Excision (MF)	NA	Predominantly intradermal Spitz nevus
12	46	F	Arm	Shave removal	NR	Compound Spitz nevus
14	72	F	Back	Excision (MF)	NR	Predominantly intradermal Spitz nevus
15	35	F	Arm	Shave biopsy	NA	Predominantly intradermal Spitz nevus
16	42	M	Arm	Shave biopsy	NR	Predominantly intradermal Spitz nevus

Abbreviations: M, male; F, female; NR, no residuum; R, residuum; MF, margins free; NA, not available; AST, atypical Spitz tumor.

^aResiduum focally extending to the peripheral specimen margin.

Thirteen of 16 cases were compound, intradermal, or predominantly intradermal and showed classic 11p Spitz nevus morphology, with large epithelioid melanocytes, which infiltrated the dermis toward the base in a single-cell pattern of growth, associated with desmoplastic stroma (Fig. 1A–C and 2A–C). Three lesions were diagnosed as ASTs owing to their higher-grade cytological atypia and/or additional oligoaberrations other than 11p amplification and/or changing clinical appearance; they were still favored to be biologically indolent (Fig. 3A–C). One lesion shared common architectural features with dysplastic nevi. Eight of 16 cases were mitotically active, with 1 or 2 superficial or deep mitoses per mm².

Fig. 1.

Case 1. A, A melanocytic proliferation involving predominantly the superficial and deep reticular dermis is evident (H&E, ×10). B, The melanocytes exhibit voluminous cytoplasm and fill the papillary dermis (H&E, ×100). C, The cells appear epithelioid with nucleomegaly, irregular nuclear contours, amphophilic cytoplasm, and occasional prominent nucleoli (H&E, ×400). D, Strong membranous and weak cytoplasmic immunoreaction with the NRASQ61R antibody (×200). E, Whole-genome view of CMA data that exhibit the 11p gain supporting its classification within the Spitz category. The top panel shows the weighted log₂ ratio (y-axis on the left) and smooth signal or copy number plot (y-axis on the right) with B-allele frequency plots showing allelic balance or imbalance in the bottom panels. Chromosomes 1–22, X, and Y are indicated along the x-axis with gain of 11p indicated by the black arrow. A CN-LOH (Copy number neutral loss of heterozygosity) within the short arm of chromosome 3 can be seen in the allelic imbalance (purple arrow) in the BAF plot. F, Molecular detection of HRAS p.Q61R by NGS. NGS, next-generation sequencing; H&E, hematoxylin and eosin; CMA, chromosome microarray analysis.

Fig. 2.

Case 6. A, A predominantly intradermal proliferation is appreciated (H&E, ×10). B, Spindled and epithelioid melanocytes are arranged in nests and cords infiltrating the dermis in a single-cell fashion at the deeper aspect of the lesion (H&E, ×100). C, The lesional cells are large with epithelioid features and amphophilic cytoplasm, nucleomegaly, relative pleomorphism, and occasional prominent nucleoli. A mitotic figure is also seen (black arrowhead). (H&E, ×400). D, Strong membranous and weak cytoplasmic immunoreaction with the NRASQ61R antibody (×200). E, Whole-genome view of CMA data (similar to Fig. 1E) that exhibit the 11 p gain, indicated by the black arrow, supporting its classification within the Spitz category. F, Molecular detection of HRAS p.Q61R by ddPCR. The green droplets along the xc-axis (channel 2) represent the wild-type HRAS p.Q61 allele, and the blue droplets along the y-axis represent the mutated HRAS p.Q61R allele. ddPCR, droplet digital polymerase chain reaction; CMA, chromosome microarray analysis; H&E, hematoxylin and eosin.

Fig. 3.

Case 9. A, A hypercellular proliferation involving the superficial and deep dermis is evident (H&E, ×14). B, Atypical epithelioid melanocytes arranged in nests of varying size and surrounded by collagen stroma (H&E, ×100). C, The epithelioid melanocytes exhibit abundant amphophilic to gray cytoplasm with nucleomegaly and pleomorphism. A few melanophages are also noted (×400). D, Strong membranous and weak cytoplasmic immunoreaction with NRASQ61R antibody (×200). E, Whole-genome view of CMA data that exhibit the 11 p gain, indicated by the black arrow, supporting its classification within the Spitz category. F, Molecular detection of HRAS p.Q61R by ddPCR. The green droplets along the x-axis (channel 2) represent the wild-type HRAS p.Q61 allele, and the blue droplets along the y-axis represent the mutated HRAS p.Q61R allele. CMA, chromosome microarray analysis; ddPCR, droplet digital polymerase chain reaction; H&E, hematoxylin and eosin.

3.1.2. NRASQ61R/PRAME immunohistochemical results

IHC with NRASQ61R antibody (clone SP174) exhibited strong membranous and weak cytoplasmic immunoreaction in 3 of 16 cases (Figs. 1D, 2D and 3D). The rest of the cases were negative. PRAME was also essentially negative in all cases. One case showed rare weak nuclear positivity for PRAME, whereas 2 cases showed rare weak to moderate nuclear positivity in intraepidermal and dermal lesional melanocytes (Table 2). Various additional immunohistochemical studies performed during clinical sign-out are presented in Table 2.

Table 2.

Summary of immunohistochemical features.

Case no.	PRAME	Melan-A/Ki67 (proliferation index)	HMB-45	P16	BAP1	BRAFV600E	ALK
1	–	NP	NP	Retained	Retained	–	NP
2	–	Low	+a	Retained	NP	–	NP
3	–	Low	NP	Retained	Retained	NP	NP
4	–	Low	NP	NP	Retained	NP	NP
5	Rare cells	NP	NP	NP	NP	NP	NP
6	Few cells	Low	NP	Retained	NP	NP	NP
7	–	Low	+b	Retained	NP	NP	NP
8	–	Low	NP	Retained	NP	NP	NP
9	–	Low	+b	Retained	Retained	NP	NP
10	–	Low	NP	Retained	NP	NP	NP
11	–	Increased	+	NP	NP	NP	NP
12	–	NP	NP	Retained	Retained	NP	–
13	Rare cells	NP	NP	NP	NP	NP	NP
14	–	NP	NP	NP	NP	NP	NP
15	–	Low	NP	Retained	Retained	–	–
16		NP	NP	NP	NP	NP	NP

Abbreviations: NP, not performed; NGS, next-generation sequencing; ddPCR, droplet digital polymerase chain reaction.

^aIn the intraepidermal component of the lesion.

^bIn the superficial component of the lesion.

3.1.3. CMA results

CMA identified an 11p chromosome gain in 16 of 16 cases either as a monoaberration (11/16) or in combination with other chromosomal oligoaberrations (5/16). A copy neutral loss of heterozygosity involving the distal 41.5 Mb of 3p (not including BAP1) was present in the index case (Fig. 1E). A low-level gain of 46 Mb from the terminal end of 7q was detected in another case. In two ASTs, low-level loss of all or most of chromosome 9 and low-level loss of chromosome 8p were detected, respectively. A low-level gain of all or most of chromosome 7 was present in one case, but the intensity of this finding was below than the normal threshold, which was suggestive of trisomy 7 in only a small percentage of cells in the tissue tested. One of 16 cases displayed multiple regions of copy number gain throughout the entire short arm of chromosome 11.

3.1.4. NGS and ddPCR results

NGS identified the HRASQ61R mutation in our index case (Fig. 1F). The additional two cases that were positive for the NRASQ61 antibody also harbored the HRASQ61R mutation, identified by ddPCR (Figs. 2F and 3F). The remaining 13 cases were negative for HRASQ61R mutation; the entire cohort was negative for the NRASQ61R mutation. The results of the CMA, NGS, and ddPCR are presented in Table 3.

Table 3.

Summary of molecular features compared with the NRASQ61R immunohistochemistry results.

Case no.	SNP array	HRASQ61R mutation	NRASQ61R mutation	NRASQ61R IHC
1	11p gain copy neutral LOH involving the distal 41.5 Mb of 3p	+ by NGS	–	+
2	11p gain low-level gain of 46 Mb from the terminal end of 7q	–	–	–
3	11p gain	–	–	–
4	11p gain	–	–	–
5	11p gain low-level loss of all or most of Ch. 9	–	–	–
6	11p gain	+ by ddPCR	–	+
7	11p gain	–	–	–
8	11p gain	–	–	–
9	11p gain	+ by ddPCR	–	+
10	11p gain low-level loss of 8p	–	–	–
11	11p gain	–	–	–
12	Multiple regions of copy number gain throughout the short arm of Ch. 11	–	–	–
13	11p gain	–	–	–
14	11p gain	–	–	–
15	11p gain	–	–	–
16	11p gain	–	–	–

Abbreviations: LOH, loss of heterozygosity; NGS, next-generation sequencing; IHC, immunohistochemistry; ddPCR, droplet digital polymerase chain reaction; Ch., Chromosome; SNP, Single Nucleotide Polymorphism.

3.1.5. 23-gene expression signature (Myriad myPath)

Two compound Spitz nevi with 11p gain as a monoaberration were tested for the 23-gene expression signature and were scored indeterminate (−0.7) and benign (−5.7). The case that scored indeterminate displayed weak to moderate PRAME immunoreaction in a few dermal lesional cells, and intraepidermal melanocytes and was positive for the NRASQ61R antibody. The case that was scored benign was completely negative for PRAME.

3.2. Control cases

3.2.1. Cases harboring NRASQ61R, L, or K mutations

All six cases harboring NRASQ61R mutation (one case of primary cutaneous melanoma, 4 cases of metastatic melanoma, and 1 case of colonic adenocarcinoma) were immunoreactive with the antibody as expected. Five of 6 cases showed strong membranous and weak cytoplasmic immunoreactivity, whereas one case (primary cutaneous melanoma) showed moderate cytoplasmic and focal weak to moderate membranous positivity.

All five cases harboring NRASQ61K mutation (2 cases of primary cutaneous melanoma, 2 cases of metastatic melanomas, and 1 case of metastatic non–small-cell malignant neoplasm) were negative for the NRASQ61R antibody. Two cases of melanoma that harbored the NRASQ61L mutation were also negative.

3.2.2. Cases harboring HRASQ61R or KRASQ61R mutations

This control group included 5 cases of colorectal adenocarcinoma (2 of them being metastatic), 2 cases of poorly differentiated thyroid carcinoma, and one case of lung pleomorphic carcinoma with squamous differentiation, all harboring either the HRASQ61R or the KRASQ61R mutation. All cases exhibited immunoreaction to NRASQ61R IHC that was either diffuse (7 of 8 cases) or focal (1 of 8 cases). Five of 8 cases showed both membranous and cytoplasmic immunoreaction, ranging from weak to strong; 2 of 8 cases showed only cytoplasmic immunoreaction; and 1 case showed only focal weak membranous.

3.2.3. Cases harboring BRAFV600E V600E mutation

All cases harboring BRAFV600E mutation (2 cases of primary cutaneous melanoma, 2 cases of metastatic melanoma, and 2 cases of primary colorectal adenocarcinomas) were negative for NRASQ61R IHC.

3.2.4. ALK+ Spitz tumors

All ALK-positive ASTs were negative for NRASQ61R IHC.

The representative H&E and IHC images of 6 control cases are presented in Fig. 4.

Fig. 4.

Representative H&E and NRASQ61R immunohistochemistry images of 6 control cases. A and B, NRASQ61R-mutated metastatic melanoma (panel A, H&E, ×200) that is immunoreactive for the NRASQ1R antibody (panel B, ×200). C and D, HRASQ61R-mutated lung pleomorphic adenocarcinoma (panel C, H&E, ×200) that is immunoreactive for the NRASQ61R antibody (panel D, ×200). E and F, KRASQ61R-mutated colonic adenocarcinoma (panel E, H&E, ×200) that is immunoreactive for the NRASQ61R antibody (panel F, ×200). G and H, NRASQ61K-mutated metastatic melanoma (panel G, H&E, ×200) that is negative for the NRASQ61R antibody (panel H, ×200). I and J, NRASQ61L-mutated primary undifferentiated melanoma (panel I, H&E, ×200) that is negative for the NRASQ61R antibody (panel J, ×200). K and L, BRAFV600E-mutated metastatic melanoma (panel K, H&E, ×200) that is negative for the NRASQ61R antibody (panel L, ×200). H&E, hematoxylin and eosin.

4. Discussion

The past few years have been instrumental in understanding the pathophysiology and genetic profile of cutaneous melanocytic tumors. The most recent World Health Organization classification for skin tumors (2018) [2] categorizes melanomas and their precursor lesions, benign or intermediate, into nine pathways based not only on their clinical and histomorphologic characteristics but also on their molecular profile and genetic fingerprint. The majority of congenital nevi carry NRAS mutations [1], whereas acquired nevi usually carry BRAF mutations [16]. Spitz nevi/tumors are defined by 11p copy number gain, which are usually accompanied by HRAS mutations, and fusions in the kinase domains of ALK, ROS, NTRK, MET, RET, BRAF, and others [17]. Intermediate lesions include entities such as atypical deep penetrating nevus/melanocytoma, BAP1-inactivated nevus/melanocytoma, and pigmented epithelioid melanocytoma. The most common oncogene mutations encountered in primary cutaneous melanomas are those of BRAF and NRAS, particularly BRAFV600E and NRASQ61R [18,19]. Until recently, evidence showed that the HRAS mutations and/or amplification do not promote malignant transformation [20]. However, Raghavan et al. [21], who studied the genetic profile of 25 spitzoid melanomas, identified two lesions harboring HRAS amplifications with accompanying hot spot mutations, characterized as Spitz melanomas. These also harbored several other genetic aberrancies including CDKN2A loss or inactivation.

Historically, whether a melanocytic tumor belongs to the Spitz category or simply resembles one (spitzoid ) has suffered from interobserver variability. Indeed, the original description by Dr. Sophie Spitz of Spitz nevi used the term Melanomas of Childhood [22]. The unraveling of the molecular underpinnings of various melanocytic neoplasms has provided us with tools for more accurate classification. This can have significant predictive consequences as shown by Raghavan et al. [21], who reported low sensitivity of histomorphology in differentiating true Spitz from spitzoid melanomas. In fact, only 36% of the Spitzoid melanomas were placed in the Spitz category. The rest of them harbored either MAPK pathway–activating mutations (other than HRAS and BRAV-600E ) or non-MAPK–activating mutations. From a diagnostic point of view, we cannot overemphasize the dilemma a diagnostician may face between a true AST and a spitzoid melanocytic tumor with BRAF/NRAS mutations; the former one can behave in an indolent fashion, whereas similar atypical features in the latter may be adequate for the diagnosis of melanoma. This is nicely illustrated by the study of Quan et al. [23], who reported a melanocytic tumor with spitzoid features that was initially diagnosed on morphologic grounds as an indolent Spitz tumor. This lesion resulted in distant metastasis, and metachronous molecular analysis revealed BRAFV600E and TERT promoter mutations as well as TP53 loss, which are commonly detected in low and high cumulative solar damage melanoma (pathway I and II) [1].

Although molecular analysis on diagnostically challenging melanocytic lesions is becoming more common, it is still costly, time-consuming, and, in many instances, unavailable. Conversely, various immunohistochemical markers such as BRAFV600E, NRASQ61R, K or L, ALK, ROS1, PanTrk, PRK1R1A, and others [24] can act as rapid, cost-effective surrogate markers for underlying genetic events [25,26]. Nevertheless, as every ancillary study, there are limitations in terms of sensitivity and specificity. For example, although a positive result can be valuable in the appropriate morphologic context, a negative result is less helpful and may indicate other mutation variants [23]. In regard to the NRASQ61R antibody specifically, its immunoreaction in neoplasms that harbor either HRASQ61R or KRASQ61R mutations is well described in the literature. Reagh et al. [10] reported immunoreaction for the NRASQ61R antibody in all cases of medullary thyroid carcinomas that harbored the HRASQ61R (c.182A>G) mutation. Recently, Pareja et al. [11] reported moderate sensitivity and high specificity of the SP174 clone in the detection of the HRASQ61R mutation in a series of 26 cases of adenomyoepithelioma of the breast. NRASQ61R immunoreaction has also been described in cases of KRASQ61R-mutated colorectal carcinomas [27]. The marker’s specificity for identifying the NRASQ61R mutation in melanomas was also questioned by Felisiak-Golabek et al. [12], who described clone SP174 positivity in HRAS- and KRAS-mutated melanomas. All published studies on cross-reactivity of the NRASQ61R antibody used Abcam’s or Spring Bioscience’s clone SP174 (rabbit monoclonal). To the best of our knowledge, none of the commercially available clones are entirely specific for the NRASQ61R mutation. The cross-reactivity between the different RAS gene products can be explained by the fact that the NRAS, HRAS, and KRAS proteins share homologous amino-terminal catalytic domains (amino acid sequence 1e166: … LLDILDTAGQEEYSAMRDQY …). The NRASQ61R antibody [14], originally manufactured to detect the NRAS gene mutation in melanomas, identifies the region that contains arginine (R) instead of glutamine (Q) in position 61. Although there is a slight difference in the corresponding base pair sequences at codon 61 between NRAS/KRAS and HRAS (CAA>CGA vs CAG>CGG), the amino acid substitution is identical for both mutations [10].

Based on our index case, an NRASQ61R-positive atypical spitzoid lesion that proved to be an 11p-amplified Spitz nevus harboring the HRASQ61R mutation, we revisited all 11p Spitz nevi/tumors in our institutions. As expected, all 11p-amplified Spitz nevi/tumors harboring the HRASQ61R mutation exhibited strong membranous immunoreaction for the NRASQ61R antibody. Furthermore, the rest of the Spitz nevi/tumors that were negative for the antibody were also negative for the mutation, rendering the marker 100% sensitive for detection of the HRASQ61R mutation in 11p Spitz nevi. The marker was also 100% sensitive in identifying the NRASQ61R-, HRASQ61R-, and KRASQ61R-mutated proteins in all the control cases that were known to harbor the corresponding gene mutations. Thus, the term RASQ61R could be more accurate in describing the marker’s detection utility.

Although the clinical and histomorphologic appearance of 11p Spitz nevi are usually distinct, their diagnosis can sometimes be challenging especially in superficial partial samples [28]. The large epithelioid cells with nuclear pleomorphism can raise the differential of melanoma with spitzoid features that are most commonly characterized by BRAF and NRAS mutations [1,4,6,29,30]. In addition, very recently, Lezcano et al. [31] published their experience on Spitz tumors with 11p gains, expanding their morphologic spectrum. In their study, 7 of 40 cases had no significant desmoplasia, whereas two tumors exhibited atypical findings with a bulbous nodular pushing border raising concern for melanoma. We hope to stress with our cohort that owing to the cross-reaction of the NRASQ61R antibody with the HRASQ61R-mutated protein, its use in such cases can lead to a misdiagnosis of melanoma harboring NRASQ61R mutation instead of an 11p Spitz nevus/tumor. In challenging cases that are either superficially sampled or do not exhibit the classic stromal desmoplasia, molecular analysis may be necessary to categorize the proliferation in the appropriate pathway and render an accurate diagnosis.

PRAME is an antigen that is present in cutaneous, mucosal, and uveal melanomas [3–34]. Its diffuse expression in melanocytic neoplasms is very sensitive for supporting the diagnosis of melanoma, with the exception of desmoplastic and spindle cell melanoma. Absence of expression is commonly encountered in benign lesions—including nodal nevi [35]—and normal melanocytes. However, focal and weak positivity in junctional nests and/or intradermal melanocytes of acquired and dysplastic nevi has been described [33]. Immunoreaction has also been reported in rare junctional melanocytes of sun-damaged skin [33]. PRAME is also part of the 23-gene expression signature (Myriad myPath Melanoma), designed to aid the differentiation between melanoma and indolent melanocytic lesions [36].

Data of PRAME IHC on Spitz/Spitzoid lesions are limited. Diffuse PRAME immunoreaction in Spitz nevi and ASTs has been reported in the literature [13,33,37], which suggests that PRAME might be less reliable as an ancillary tool in the evaluation of Spitz neoplasms [37]. Nevertheless, Lezcano et al. [13] described high concordance between molecular findings and final diagnosis with PRAME immunoexpression in challenging spitzoid melanocytic lesions. In the very recent study on 11p Spitz lesions by Lezcano et al. [31], one particularly challenging case with several genomic abnormalities and atypical morphologic features raised significant concern for Spitz melanoma; nevertheless, the lack of PRAME expression in conjunction with the absence of TERT promoter mutations pointed toward a melanocytoma. All the cases in our cohort were essentially negative for PRAME, which can be potentially very helpful in 11p Spitz tumors with NRASQ61R immunoreactivity against the differential diagnosis of melanoma. Intriguingly, one Spitz nevus with 11p gain, which showed focal weak to moderate immunoreaction for PRAME, was scored indeterminate by Myriad myPath. This was in contrast with the second case, which showed no immunoreactivity and scored in the benign category. The algorithm of the test is not publicly available, but our observation raises the question if protein expression of one of the main genes even by a minority of cells skews the final categorization. Ostensibly, more data are needed in this regard.

In conclusion, we present 3 of 16 cases of 11p-amplified Spitz nevi with HRASQ61R mutation exhibiting immunoreaction for the NRASQ61R antibody (clone SP174). This cross-reaction between NRAS- and HRAS-mutated gene products can lead to a diagnostic pitfall in the assessment of melanocytic lesions with spitzoid characteristics.