To the Editor:
Ossifying renal tumor of infancy (ORTI) is a rare pediatric benign renal neoplasm first described in 1980.1 In the latest review of 28 cases, whose diagnostic age was described,2 the median age of diagnosis was 4.3 months old, and 23 cases (82.1%) were diagnosed within the first year of life. These observations suggest in utero development of ORTI but no direct confirmation has been made. Here, we reported a newborn girl with ORTI whose cystic lesion was first noticed at 25 weeks of gestation by ultrasonography. Furthermore, we identified the CTNNB1 mutation in her tumor sample by whole exome sequencing (WES). To our knowledge, this is the first identification of CTNNB1 mutation as a genetic alteration involved in ORTI tumorigenesis.
The patient was transferred to our hospital 10 days after birth due to intermittent gross hematuria that appeared on the second day of life. In ultrasonography at 25 weeks of gestation, a cystic lesion was first noticed in her right kidney, and it was continuously enlarged in proportion to fetal growth and multiloculated in the follow-up evaluations (Figure 1A). She was born by scheduled cesarean section at 38 weeks of gestation with a birth weight of 2,672 g. There was no significant family history. Her general condition was fair, but a soft mass was palpable on her right abdomen. On admission, ultrasonography showed a mass lesion with multiple cystic components at the lower pole of the right kidney (Figure S1). An abdominal computed tomography image performed 5 days after birth revealed a 4.5 x 4 cm multi-cystic mass with a small linear calcification in the middle of the solid part (Figure S2) without any sign of metastasis. Considering the gradually increased tumor size and continuous hematuria, right nephrectomy was performed at 23 days old. In the resected kidney, the tumor was consisted of cystic and solid lesions and was well-demarcated from the adjacent kidney. Histopathologically, the tumor consisted of three main components, an osteoid core, osteoblast-like cells, and spindle cells (Figure 1B), without typical nephrogenic rest. Immunohistochemically, osteoblast-like cells were positive for cytoplasmic WT1 and epithelial membrane antigen (EMA) while spindle cells were positive for nuclear WT1 but not EMA (Figure S3). Based on these findings, a final diagnosis of ORTI was made. She had a good postoperative course and no local recurrence or metastasis was observed at 9 months of age.
Since genetic alterations involved in the ORTI tumorigenesis have not been fully elucidated, we conducted WES of the patient’s tumor sample and her peripheral blood mononuclear cells (PBMC) after written informed consent was obtained from her parents. In copy number variation (CNV) analysis, several tumor-specific CNVs were identified (Table S1) including a gain of 7q34 locus containing the EPHB6 gene that was reportedly overexpressed in thyroid and breast cancers,3 and a loss of 19p13.3 locus containingTLE2 and TLE6 , which was previously reported in one sporadic Wilms tumor case.4 However, trisomy 4, which was previously observed in some ORTI cases with a relatively low prevalence,5,6 was undetectable with +1.20 of threshold. In mutational analysis, c.121A>G (p.T41A) mutation of the CTNNB1 gene was confirmed in the tumor tissue with 0.38 of variant allele frequency (Figures 2A, S4 and Table S1) but not in her PBSC, suggesting the heterozygous somatic mutation in the tumor. Threonine 41 is one of the targets for GSK-3β-dependent phosphorylation and is one of the hotspot mutation sites in diverse tumors.7 Consistently, nuclear accumulations of β-catenin protein, encoded by CTNNB1 , were confirmed in both osteoblast-like and spindle cells of the tumor by immunohistochemistry (Figure 2B). No further mutations were detected in the tumor sample including the WT1 and WTX genes, which are frequently mutated in Wilms tumors.8 Loss of heterozygosity ofWT1 was also unlikely since heterozygous genotypes were observed in the same single nucleotide polymorphisms in the tumor sample and her PBSC (Table S2).
Hyperplastic intralobar nephrogenic rest (ILNR), which is known as a precursor lesion of Wilms tumors, is also considered as possible origin of ORTI.9 In the ILNR-associated Wilms tumors, co-clustering of WT1 and CTNNB1 mutations has been identified.10 In the Wilms tumors with these two mutations, ILNR lesions had WT1 mutations but no CTNNB1mutations, indicating that CTNNB1 mutation is a later event in tumorigenesis.11 These observations suggest that preceding CTNNB1 mutation might induce ORTI instead of Wilms tumor. Of note, the CTNNB1 mutations in the GSK-3β-phosphorylation sites have been previously identified in at least two types of childhood benign tumors characterized by nested proliferation of spindled and epithelioid cells with calcification or ossification (Figure 2C). In almost all cases of pilomatricoma, a skin adnexal tumor in children,12 the CTNNB1mutations including threonine 41 were identified.13Moreover, in all sequenced cases of calcifying nested stromal-epithelial tumor (CNSET), a rare hepatic tumor in children and young adults, theCTNNB1 mutations including threonine 41 were identified in combination with the TERT promoter mutations,14which were undetectable in our case (Figure S5).
In summary, our case indicates in utero development of ORTI and involvement of CTNNB1 mutation in its tumorigenesis. Our observations warrant further investigation of pathobiological similarities and differences among pediatric CTNNB1 -mutated tumors.