BRAF Exon 15 Mutations in the Evaluation of Well-Differentiated Epithelial Nephroblastic Neoplasms in Children: A Report From the Children’s Oncology Group Study AREN03B2

All patients registered in the Children’s Oncology Group AREN03B2 study from 2006 to 2017 with tumors classified as disease stage I epithelial-predominant favorable-histology WT on central pathology review were identified, yielding 174 patients; of these, 36 were exclusively epithelial (EFHWT).⁷  Consent was provided by all patients or their legally authorized guardians for participation in the AREN03B2 study. This study was approved by the National Cancer Institute Central Institutional Review Board (IRB) or by local IRBs, and by the IRB of the Anne & Robert H. Lurie Children’s Hospital of Chicago, Illinois. A complete set of hematoxylin-eosin–stained slides was available for all. These 36 EFHWTs were separated into 2 groups. One group of 13 tumors were noted at the time of central review to have morphologic overlap with MA, including tubular and acinar features with bland nuclei. Such tumors were classified at the time as EFHWT rather than MA because of the presence of at least a partial fibrous pseudocapsule and/or occasional mitoses; these 13 tumors are considered EFHWT with MA features in the current study. The remaining group of 23 tumors lacked these morphologic features and were classified as EFHWT without MA features. The hematoxylin-eosin slides of these 36 cases were rereviewed and the presence and completeness of a layer of collagen separating the normal kidney from the lesion (fibrous pseudocapsule) and the presence of mitotic activity were assessed. Clinical and outcome features were extracted from the Children’s Oncology Group database. It should be noted that typical MAs lacking encapsulation and mitotic activity were seldom registered in AREN03B2; if they were, they were classified from the start as MA, were not included in therapeutic protocols, and were not followed. These were excluded from the current study.

DNA was extracted from unstained paraffin-embedded slides using the Qiagen DNeasy Blood and Tissue Kit (catalog #69504), per manufacturer’s protocol, with modifications as described below. Tumor areas were macrodissected as needed using a razor blade. The samples were incubated with xylene at 55°C for 15 minutes before centrifuging, followed by overnight incubation twice in proteinase K at 56°C. Samples were then eluted in 50 μL of Buffer AE (Qiagen, Germantown, Pennsylvania) and incubated at room temperature for 5 minutes before centrifugation. The DNA concentrations were read using a Nanodrop Lite Spectrophotometer (ThermoFischer Scientific, Waltham, Massachussetts).

Human DNA sequences for BRAF exon 15 were downloaded from the NCBI Nucleotide database (http://www.ncbi.nlm.nih.gov/nucleotide/) to generate primer sets. The primer sequences were forward 5′-CTA CTG TTT TCC TTT ACT TAC TAC ACC TCA GA-3′; reverse 5′-ATC CAG ACA ACT GTT CAA ACT GAT G-3′. The resulting amplicon consists of base pairs 140753289–140753424 on chromosome 7 (GRCh38/hg38 assembly), corresponding to the 24 terminal base pairs of BRAF intron 14 and the first 122 base pairs of exon 15. An M13 sequencing tail was added to the 5′ end of each primer, allowing direct sequencing of the amplicon generated during the polymerase chain reaction (PCR). PCR amplicon cleanup was performed following the ExoSAP-IT manufacturer instructions (Affymetrix, Santa Clara, California) or the QIAquick PCR Purification Kit (Qiagen, Valencia, California), and 2 μL of the cleaned product was used for sequencing reactions with the Big Dye Terminator v1.1 Cycle Sequencing kit (ThermoFisher Scientific) using the M13 forward primers on an ABI Prism 3130 xl automatic sequencer (ThermoFisher Scientific) in both directions. Data were analyzed by Sequencing Analysis 5.1.1 (ThermoFisher Scientific).

Selected demographics and clinical characteristics of the stage I EFHWT cohorts with versus without MA features were reported as descriptive statistics and compared using Fisher exact tests for categorical variables and 2-sample t tests for continuous variables (or Wilcoxon rank-sum test if normality of the continuous data could not reasonably be assumed). Observed qualitative differences between the cohorts were also described by the reviewing pathologist. The rates of BRAF exon 15 mutation and incidence of relapse were also reported from each cohort, along with median duration of follow-up for relapse events.

There were 174 stage I epithelial-predominant favorable-histology WT patients registered in AREN03B2 at the time of case identification; and 36 of 174 were exclusively epithelial (EFHWT). The 13 patients with EFHWTs with MA features (8 male, 5 female) had a mean age of 8.4 years (range, 2.2–16.4 years); the 23 patients with EFHWTs without MA features (13 male, 10 female) were significantly younger, with a mean age of 1.9 years (range, 0.4–10.0 years; P < .001).

On histologic examination, the 13 EFHWT patients with MA features all showed similar histology and had tubular and acinar patterns and bland cytologic features. The tumor directly interfaced with the adjacent normal parenchyma in areas (Figure, A); however, an incomplete, variably thickened fibrous pseudocapsule was present at least focally in all 13 cases (Figure, B and C). EFHWTs without MA features showed a wider range of epithelial patterns, from more primitive to more well differentiated. All slides submitted that contained the interface between the tumor and the adjacent kidney demonstrated a thick fibrous pseudocapsule (Figure, D). Among EFHWTs with MA features, mitotic activity ranged from 0 to 5 (mean, 1) mitotic figures per 10 high-power fields (hpfs). Conversely, mitotic activity in patients with EFHWT without MA features was significantly higher, ranging from 19 to 104 (mean, 48) mitoses/10 hpfs (P < .001). Average tumor diameter in EFHWT patients with MA features was significantly smaller, at 6.0 cm (range, 1.6–9.0 cm) versus 9.7 cm (range, 2.7–14.0 cm) for EFHWT patients without MA features (P < .001) (Table). There have been no recurrences or deaths among EFHWT patients with MA features with a median follow-up of 5.9 years.

Of the 13 EFHWTs with MA features, 1 had been analyzed for BRAF mutation by the submitting institution subsequent to the initial evaluation and was determined to have BRAF V600E mutation. Formalin-fixed, paraffin-embedded tissue was available in 11 of the 12 remaining EFHWT with MA features for BRAF exon 15 sequencing. Combined, 7 of 12 BRAF-evaluated FHWT patients with MA features had the valine 600 glutamate (V600E) amino acid substitution; the remaining 5 patients lacked sequencing changes within the examined region of exon 15. When comparing EFHWTs with MA features with (7) versus without (5) BRAF exon 15 mutations, the average age (8.02 versus 8.75 years), gender ratios (5 male to 3 female versus 2 male to 3 female), average tumor size (6.9 versus 5.44 cm), and average mitotic rate (1.3 versus 2.0 mitoses per 10 hpf) did not reveal statistical differences, although these numbers are small. None of the 23 EFHWTs without MA features had BRAF exon 15 mutations.

BRAF exon 15 mutations have been described in numerous benign and malignant neoplasms, including benign nevi, melanoma, papillary thyroid carcinoma, several glioma subtypes, and disseminated Langerhans cell histiocytosis.⁸  Within the kidney, BRAF mutations have been thus far restricted to tumors showing features of MA or the related metanephric stromal tumors.^2–4,9  While WT with MA-like areas have been shown to harbor BRAF exon 15 mutations in both adults and children,¹⁶  prior studies of WT without MA-like features, including the present study of EFHWT without MA-like features, have not identified BRAF exon 15 mutations.^10,11 BRAF exon 15 mutation analysis has been performed in several studies of adults with MA, and BRAF V600E was present at incidences ranging from 15 of 24 (62%)¹²  to 26 of 29 (90%)⁴ ; until now only 3 MAs in children have been previously studied (2 of 3 had BRAF mutations)³  and the rate in the present study is 7 of 12 (58%). It is important to note that the current study does not assess the larger number of cases of unequivocal MAs identified and originally classified as such in children. In the current study of tumors previously classified as EFHWT with features of MA, there were no histologic features that distinguished those cases with from those without BRAF mutations.

BRAF mutations cause RAS/MAP kinase pathway signaling in the absence of growth factors, bestowing self-sufficiency in growth signals.^12,13  However, it has been suggested that in the absence of other mutations, uncontrolled growth caused by BRAF mutations induces senescence.¹⁴  This may help to explain the appearance of BRAF mutations in both benign and highly malignant tumors. In MA with or without BRAF mutation, the senescence marker p16^INK is expressed,²  suggesting that senescence is a core component of MA biology. The driver mutation in those MA lacking BRAF mutation has not been identified.

The current study addresses questions regarding certain criteria currently used for the diagnosis of MA, including the presence of a fibrous capsule and mitoses. The clinicopathologic and molecular findings we present suggest that epithelial nephroblastic lesions in children with features of MA should be classified as MA despite the focal presence of a fibrous capsule and modest mitotic activity. This represents a harmonization with diagnostic practice applied to adult MA.^2,15  The remaining histologic features of MA are the homogeneous presence of small, uniform acini and tubules with bland nuclei. Although this study suggests that the presence of a partial fibrous pseudocapsule and/or modest mitotic activity should not preclude the diagnosis of MA, it should be appreciated that these features are certainly significantly less prominent than what is seen in EFHWT lacking features of MA, in which the mitotic activity is generally greater than 20/10 hpfs and in which a thick, complete capsule is present. Because of the nonuniversal prevalence of BRAF exon 15 mutations in MA, the absence of histologic features separating those that have from those that lack BRAF mutations, and the absence of clear differences in outcome, we do not suggest requiring molecular testing as a criterion in this differential diagnosis. Similarly, although BRAF immunohistochemistry reliably identifies those MAs that have BRAF mutations, the currently available information is insufficient to recommend excluding the diagnosis based on its absence.

It should be noted that the clinical significance of this suggested change in criteria for MA versus EFHWT may be less than expected. Recent studies have shown excellent survival for all stage 1 epithelial predominant WTs.⁷  A clinical trial is in development that will assign such patients to observation only following nephrectomy, although this strategy is not yet proven.