KANK1-NTRK3 fusions define a subset of BRAF mutation negative renal metanephric adenomas

Background Metanephric adenoma (MA) is a rare benign renal neoplasm. On occasion, MA can be difficult to differentiate from renal malignancies such as papillary renal cell carcinoma in adults and Wilms̕ tumor in children. Despite recent advancements in tumor genomics, there is limited data available regarding the genetic alterations characteristic of MA. The purpose of this study is to determine the frequency of metanephric adenoma cases exhibiting cytogenetic aberration t (9;15)(p24;q24), and to investigate the association between t (9,15) and BRAF mutation in metanephric adenoma. Methods This study was conducted on 28 archival formalin fixed paraffin-embedded (FFPE) specimens from patients with pathologically confirmed MA. Tissue blocks were selected for BRAF sequencing and fluorescent in situ hybridization (FISH) analysis for chromosomal rearrangement between KANK1 on chromosome 9 (9p24.3) and NTRK3 on chromosome 15 (15q25.3), which was previously characterized and described in two MA cases. Results BRAFV600E mutation was identified in 62% of our cases, 9 (38%) cases were BRAFWT, and 4 cases were uninformative. Of the 20 tumors with FISH results, two (10%) were positive for KANK1-NTRK3 fusion. Both cases were BRAFWT suggesting mutual exclusivity of BRAFV600E and KANK1-NTRK3 fusion, the first such observation in the literature. Conclusions Our data shows that BRAF mutation in MA may not be as frequent as suggested in the literature and KANK-NTRK3 fusions may account for a subset of BRAFWT cases in younger patients. FISH analysis for KANK1-NTRK3 fusion or conventional cytogenetic analysis may be warranted to establish the diagnosis of MA in morphologically and immunohistochemically ambiguous MA cases lacking BRAF mutations.


Background
Metanephric adenoma (MA) is a rare benign renal tumor classified under the rubric of metanephric tumors, which also include metanephric stromal tumor and metanephric adenofibroma [1]. BRAF mutations have been identified in metanephric stromal tumor and metanephric adenofibroma in addition to metanephric adenoma, which justifies their grouping as family of metanephric tumors by the World Health Organization (WHO) [2]. MA is uncommon and generally occurs in adults between the fourth and sixth decades of life and occasionally in children [3]. The male-to-female ratio is between 1:2 to 1:3 with a mean age of approximately 41 years [3,4]. In adults, MA accounts for approximately 0.2% of adult renal epithelial neoplasms [5]. Despite the fact that fewer than 25 cases have been reported in children, it is considered to be the most common benign pediatric renal epithelial tumor [6,7]. Including both pediatric and adult cases, fewer than 200 cases of MA have been reported in the literature, thus, illustrating its rarity and the scarcity of available data [5][6][7][8].
The majority of MA cases can be diagnosed on routine hematoxylin and eosin (H&E) stained slides. However, in some challenging cases MA can be difficult to morphologically differentiate from malignant renal neoplasms [9]; particularly the solid variant of papillary renal cell carcinoma (PRCC) and epithelial-predominant Wilms' tumor (WT) [10]. The distinction between these renal tumor subtypes can be aided by the use of diagnostic modalities such as immunohistochemistry, cytogenetic studies, and advanced molecular analyses [3]. The correct classification of a renal tumor is not only critical from a diagnostic standpoint, but also from a prognostic and therapeutic standpoint [11,12].
Immunohistochemistry may be helpful in distinguishing metanephric adenoma from solid variant of papillary renal cell carcinoma (s-pRCC) and epithelial predominant Wilm's tumor (e-WT). Specifically, MAs are generally diffusely CD57 and WT1 positive, only focally CK7 positive, and CD56 and AMACR negative (rarely weakly positive), s-pRCCs are generally diffusely CK7 and AMACR positive, often CD57 positive (in contrast to conventional [non solid-variant] pRCC, which is usually CD57 negative) and WT1 and CD56 negative, and e-WTs are generally diffusely CD56 and WT1 positive and variably CD57, CK7, and AMACR positive.
The genetic alterations underlying MA tumorigenesis have only been defined relatively recently [11]. Previous cytogenetic studies have revealed a paucity of genetic alterations in MA. The molecular and cytogenetics data reported in the literature in regard to MA is sparse and often consists of literature reviews of previously published cases or isolated case reports. We and another group each reported a case of MA showing a t (9;15) [11][12][13]. A study by Choueiri et al. demonstrated that approximately 90% of MAs harbor BRAF V600E mutations; the genetics of the remaining 10% in their study is unclear [14]. BRAF V600E gene mutations are frequently detected in a wide range of benign and malignant human tumors, however, BRAF mutations in renal tumors such as renal cell carcinoma (RCC), oncocytoma, and WT are essentially absent [4,[15][16][17][18][19][20][21]. This data coupled with Choueiri's data suggests that BRAF mutations are specific for MA amongst renal tumors.
The present study was undertaken to determine the frequency of MA cases exhibiting cytogenetic aberration t (9;15)(p24;q24) as previously reported in the literature, and to investigate the association between t (9;15) and BRAF mutation in MAs [11][12][13]. Two cases included in this study have previously been reported in literature (case # 1 and #12- Table 1.). Catic et al. have demonstrated the specific gene fusion that results from the chromosomal translocation t (9,15)(p24;q24) [11]. Rakheja et al. reported chromosomal translocation mentioned above with only a karyotype [13]. We examined 28 cases of MA at the genetic and molecular level, using a combination of BRAF sequencing and fluorescent in situ hybridization (FISH) to detect chromosomal rearrangement between KANK1 on chromosome 9 (9p24.3) and NTRK3 on chromosome 15 (15q25.3).

Patients and samples
This study was conducted on 28 archival formalin fixed paraffin-embedded (FFPE) specimens from renal metanephric adenomas. FFPE blocks and H&E stained slides were obtained from the departments of pathology at four participating institutions, including Advocate Lutheran General Hospital (Park Ridge, IL, USA), Northwestern Memorial Hospital (Chicago, IL, USA), Children's Medical Center of Dallas (Dallas, TX, USA), and Charles University Hospital (Plzen, Czech Republic). Fourteen cases were of American origin and 14 of European origin. All samples received for this study and data reported have been de-identified. Because this was a retrospective study, ethics committee ruled that no formal ethics approval was required in this particular study.
The MA specimens and hematoxylin and eosinstained slides were retrieved and reviewed by expert pathologists at each institution. All pathologic specimens were acquired after partial or complete nephrectomy, and none were diagnosed by needle biopsy. The diagnosis of MA was then re-confirmed by two genitourinary pathologists ( Fig. 1) and de-identified representative tissue blocks were further selected for BRAF V600E exon 15 sequencing and FISH analysis. Histologically, the tumors are composed of epithelial cells arranged in tubules and papillary configurations. The relatively small tumor cells have a high nuclear: cytoplasmic ratio, ovoid nuclei, uniformly dispersed chromatin, inconspicuous nucleoli, scant eosinophilic cytoplasm, and ill-defined cell borders with nuclear overlap, Fig. 1, a and b. Mitoses are not conspicuous. Occasional psammomatoid calcifications are seen, Fig. 2. Patient demographics and clinicopathologic characteristics such as: age, gender, tumor size, laterality, and chromosomal analysis results were provided by pathologists at each institution (Table 1).

Fluorescence in situ hybridization (FISH)
Metaphase chromosome spreads and interphase nuclei were prepared on a glass microscope slide in accordance with standard cytogenetic procedure and according to the manufacturer's instructions. Paraffin embedded tissue slides were cut at 2-μm thickness using a microtome and floated in a protein free waterbath at 40°C. A concurrent H&E slide was stained and marked by a pathologist to delineate the area of tumor for analysis. Fluorescent in situ hybridization probes were purchased from BlueGnome (Illumina, Cambridge, United Kingdom) and Empire Genomics (Buffalo, NY, USA). Briefly, prepared slides from the tumor were placed in a Coplin jar with 40 mL of 2XSSC pH 7.0 at 37°C for 15 min. Treatment of slides in 2XSSC was used to artificially age the chromosomes, making them less sensitive to overdenaturation. Next, the slides were dehydrated in 70, 85, and 100% ethanol at room temperature for 2 min each, followed by drying on a 50°C warmer for 15 min. Paraffin embedded tissue slides were baked in a 60°C oven for a minimum of 1 h. These slides were then placed on the VP processor (Abbott Molecular, Des Plaines, IL, USA) for de-paraffinization, pretreatment, and protease digestion. A mixture of 3.5 μl of locus specific identifier (LSI) hybridization buffer (Abbott Molecular, Des Plaines, IL, USA), 1 μl of sterile water, and 0.5 μl of probe was prepared for the BlueGnome probes and 4 μl of Empire Genomics buffer was used with 1 μl of probe for the Empire probes for each hybridization area. 5 μl of probe mixture was applied to each hybridization area of patient and control slides. Prepared slides and probes were then co-denatured using the ThermoBrite machine at a denature temperature of 76°C for 5 min and then hybridized overnight. After hybridization, slides were washed using 40 ml of 0.4X SSC/0.3%NP40 for 2 min, followed by 40 ml of 2X SSC/0.1%NP40 for 1 min to remove any excess or unbound probe. After slides were air-dried, 10 μl of DAPI II counterstain on a 22 × 22 coverslip was applied to the targeted area of the slide.
FISH analysis was performed following standard techniques using a fluorescent microscope with appropriate  [22].
BRAF mutation analysis FFPE sections were evaluated for the BRAF mutation on a Roche LightCycler 2.0 instrument (Mannheim, Germany) utilizing the allelic discrimination by real time In short, the appropriate FFPE tissue block was selected by the pathologist. Four-micrometer-thick sections from FFPE tissue blocks were enriched by manual microdissection and DNA was isolated by the ZymoPinPiont method (Irvine, CA, USA), according to the manufacturer's protocol. Exon 15 of the BRAF gene was amplified from 50 ng of genomic DNA by real time PCR using sequence specific primers ordered from Invitrogen-Life Technologies (Carlsbad, CA, USA) (forward primer: 5′-CTCTTCATAATGCTTGCTCTGATAGG-3′, and reverse primer: 5′-TAGTAACTCAGCAGCATCTCAGG-3′). Melting curve analysis was performed by optimized fluorescent probes 5′-FL-TGGAGTGGGTCCCATCAG TTTGAACAGTTGTCTGGATCCATT SpacerC35'-TGGTCTAGCTACAGTGAAATCTC-LC640. The PCR products were amplified in the following conditions: initial denaturation at 95°C for 10 min; amplification 45 cycles of 95°C for 5 s, 60°C for 10 s, and 72°C for 20 s; melting curves 1 cycle of 95°C for 2 min, 40°C for 2 min, and 85°C at 5 s; and cooling period of 1 cycle at 40°C for 30 s. The method of BRAF exon 15-mutation analysis interpretation has been previously described [4,11,14]. Specifically, our assay detects 28 nucleotide changes involving the following codons: V600E, M, L, R, Q, D, K, A and G, L597V, S, Q, R and L, K601E, del, and N, A598V, A598_T599insV, T599I, T599_V600insT, T599_V600insTT, and V600_K601 > E.
Fisher's exact analysis was performed for comparison between the differential prevalence of BRAF V600E mutation in patients under the age of 30 and those patients over 30 years of age. A p value <.05 was used to indicate statistical significance.

Results
We analyzed 28 MAs. Among those, there were 17 women, 10 men and 1 unknown (F:M, 1.7:1). Patient age ranged from 9.8 to 73 years with median age of 52.5 years (52 among women; and 53.5 among men). Twenty patients were over the age of 30, while 7 were under the age of 30 and 1 patient was of unknown age (Table 1 and Table 2). Tumor size ranged from 0.5 cm to 12 cm with a median size of 3 cm. Patient cohort and clinical characteristics are summarized in Table 1.
Out of 28 cases studied, cytogenetic analysis was available in only six (21%) cases, Table 2. Three (50%) cases out of 6 exhibited chromosomal aberration. All three patients were under the age of 30. Two cases exhibited translocation involving chromosome 9 (9p24) and chromosome 15 (15q24), or t (9;15)(p24;q24) [11,13]. To rule out constitutional abnormality in these two patients, peripheral blood samples revealed normal karyotypes [11,13]. One additional case that was translocation positive exhibited a translocation involving chromosome 6 (6q26) and chromosome 22 (q11.2) with a karyotype of 46,XX,t (6;22)(q26;q11.2). For this case, it is unknown if any genes are affected and if constitutional chromosome analysis was performed to rule out constitutional abnormality. The remaining three tumor cases demonstrated normal karyotypes of 46,XX or 46,XY and were over the age of 30. There were no differences noted among genders. Cytogenetic results are further summarized in Table 2.  The focus of the study was to determine the frequency of t (9;15) or KANK1-NTRK3 gene fusion using FISH, Fig. 1, C and D. Cases 1 through 9 were tested using FISH probes RP11-130C19 on 9p24.3 (green signals) and RP11-62D2 on 15q25.3 (orange signals) from BlueGnome (Illumina, Cambridge, United Kingdom). Due to unavailability of previously used probes from BlueGnome, the following replacement probes were purchased from Empire Genomics (Buffalo, NY, USA): RP11-1107A23 on 9p24.3 (green signals) and RP11-  Fig. 4 a Schematic representation of the BRAF status wild type (WT) is colored green and mutation is colored red. Translocation presence of either t(9;15) or t (6;22) is colored red and no translocation is colored green. Gray represents uninformative results. b Schematic representation of the results demonstrating the presence and the absence of translocation, and BRAF V600E mutation or wild type, suggesting mutual exclusivity between BRAF V600E and KANK1-NTRK3 fusion 608H9 on 15q25.3 (orange signals). A clinical cytogeneticist picked the best replacement probes based on available data from the vendor as well as University of California Santa Cruz (UCSC) Genome Browser (Genome Build 38), making sure probes covered the genes of interest. FISH analysis of the remaining cases (10 through 28) was carried out using the Empire Genomics probes. Clear signals and reportable FISH results were obtained in 20 (71%) cases. Of the 20 tumors with FISH results, two tumors (10%) were positive for KANK1-NTRK3 fusion (Fig. 1, D). These two cases were also found to harbor the same translocation t (9;15)(p24;q24) by conventional cytogenetics analysis [11,13]. Thus, concordance between FISH testing and chromosome analysis evaluation for the detection of the rearrangement and involvement of KANK1-NTRK3 genes was 100% (2/2 cases). The remaining 18 (90%) tumors were negative for KANK1-NTRK3 fusion, Fig. 1d. FISH in eight tumors (29%) was technically unsuccessful due to lack of hybridization signals and therefore uninformative. This was most likely due to DNA degradation in archival material, a conclusion supported by the fact that many of the same cases were uninformative for BRAF by PCR. FISH results are summarized in Table 2.
In the present study, the t (9;15) resulted in a KANK1-NTRK3 fusion transcript in which the first seven exons of KANK1 are fused to exon fourteen of the NTRK3, Fig. 3.
Out of the 9 cases lacking BRAF mutation, 3 cases (33%) exhibited chromosomal translocation. Four out of six BRAF WT cases did not have cytogenetic results available, but showed normal FISH pattern for t (9;15). The remaining 2 cases were uninformative by FISH. Overall, 13 of 20 (65%) cases lacking t (9;15) harbored BRAF V600E mutations. There were no cases demonstrating both the translocation and BRAF V600E mutation, suggesting mutual exclusivity between BRAF V600E and KANK1-NTRK3 fusion, Fig. 4 (a and b).
In this study, we confirm that a subset of biologically distinct MAs in younger patients (< 30 yrs. of age) exhibit aberrant chromosomal translocations and do not harbor BRAF mutations. The novel findings of our study are that the typical MAs which do not harbor BRAF mutations can demonstrate cytogenetic aberrations. Additional larger cohort studies are necessary to confirm and further elucidate the frequency of the cytogenetic aberrations found in this subset of MAs.
Unlike chromosome analysis data, FISH analysis and immunohistochemistry analyses on MA are more frequently performed and reported in the literature. Most reported FISH studies have been primarily focused on testing for trisomies of chromosome 7 and 17, and loss of Y chromosome. Brown et al., reported trisomies of chromosomes 7 and 17 and loss of Y chromosome in eight of 11 cases classified as MA [31]. However, their eight cases almost certainly represented examples of solid variant of PRCC, an entity that was described in the literature subsequent to their publication [32][33][34]. Recent studies strongly recommend utilization of FISH analysis to aid in differentiating those cases that deviate from the expected immunohistochemical staining pattern [3,4].
Recent studies of MA have emphasized the importance of advanced molecular testing. Previous studies have demonstrated that the vast majority of MAs harbor BRAF V600E mutations, and that epithelial WTs lack BRAF V600E mutations [19,20,35]. Choueiri et al. published the largest series of MA demonstrating that 90% of MAs harbor a BRAF V600E mutation [14]. This was the first large study to shed light on the molecular underpinnings of MA and to investigate BRAF V600E mutation in this indolent tumor. Additionally, they tested 129 non-MA renal neoplasms and detected BRAF V600E mutation in only one PRCC. Cytogenetic analysis of this PRCC case revealed the presence of trisomy of chromosomes 7 and 17. Padilha et al. and Choueiri et al. suggest that molecular BRAF V600E mutation analysis is a valuable diagnostic tool in the differential diagnosis of this rare kidney tumor that may be diagnostically challenging [8,14].
More recently, Chami et al. studied pediatric MA cases for BRAF V600E mutations and found three out of four cases to be positive for BRAF V600E mutation; 10 cases of pediatric renal cell carcinoma and 10 cases of Wilms' tumor did not exhibit BRAF V600E mutation [6]. To [35].
Our assay, which is capable of detecting 28 different BRAF mutations including BRAF V600D and BRAF K601L only detected BRAF V600E mutations. Our results differ significantly from those in the literature in the overall lower frequency of BRAF mutations (62% vs. 90%). This may be due to the fact that in our study 25% of our patients were under the age of 30. We observed a possible trend towards patients younger than 30 years-old showing BRAF WT tumors. This data suggests that MA in younger patients may be genetically distinct from its counterpart in older patients, but a larger patient cohort is needed to confirm this observation.
Lastly, we compared BRAF V600E mutation analysis results with FISH for KANK1-NTRK3 gene fusion. Of the nine cases that exhibited BRAF WT , 3 cases were those that demonstrated chromosomal aberrations by conventional karyotyping. Of the three cases that showed chromosomal translocations, two were positive for t (9; 15) and KANK1-NTRK3 gene fusion by FISH. The remaining six BRAF WT cases did not have conventional cytogenetic analyses. Four of these six cases were negative for KANK1-NTRK3 fusion by FISH, and 2 cases were uninformative. Overall, 13 of 20 (65%) cases lacking t (9;15) harbored BRAF mutations. There were no cases with both t (9;15) and BRAF V600E mutation suggesting exclusivity between BRAF V600E and t (9;15) and that the latter may be the genetic event behind a subset of BRAF WT MAs.
A significant limitation of our study is the retrospective nature of case series with inability to test the BRAFwild type cohort for additional mutations. Unfortunately, many cases had very little or no additional sample material to perform NTRK3, pan-Trk and BRAF immunohistochemistry testing.

Conclusion
In conclusion, we report KANK1-NTRK3 fusion without BRAF V600E mutation in two MA cases. This finding supports the initial suggestion that KANK1-NTRK3 is the pathogenetically significant fusion transcript in tumors carrying a t (9,15)(p24;q24) and lacking BRAF V600E mutation in younger patient cohort. In this study, we have provided additional evidence that metanephric adenomas have relatively noncomplex karyotypes and have distinctive cytogenetic profiles. The cytogenetic profile can be useful in resolving a differential diagnosis of metanephric adenoma.
Classic histopathological features of MA coupled with a documented BRAF V600E mutation are diagnostic of MA; however, we and others have demonstrated that the absence of BRAF V600E mutation does not exclude a diagnosis of MA. For those case lacking BRAF mutations, alternative testing such as FISH analysis for KANK1-NTRK3 fusion and/or cytogenetic chromosome analysis to look for t(9;15)(p24;q24) may be warranted.