Prevalence of APC and PTEN Alterations in Urachal Cancer

Urachal carcinoma (UrC) is a rare tumor with remarkable histological and molecular similarities to colorectal cancer (CRC). Adenomatous polyposis coli (APC) is the most frequently affected gene in CRC, but the prevalence and significance of its alterations in UrC is poorly understood. In addition, loss of phosphatase and tensin homologue (PTEN) was shown to be associated with therapy resistance in CRC. Our primary aim was to assess specific genetic alterations including APC and PTEN in a large series of UrC samples in order to identify clinically significant genomic alterations. We analyzed a total of 40 UrC cases. Targeted 5-gene (APC, PTEN, DICER1, PRKAR1A, TSHR, WRN) panel sequencing was performed on the Illumina MiSeq platform (n = 34). In addition, ß-catenin (n = 38) and PTEN (n = 30) expressions were assessed by immunohistochemistry. APC and PTEN genes were affected in 15% (5/34) and 6% (2/34) of cases. Two of five APC alterations (p.Y1075*, p.K1199*) were truncating pathogenic mutations. One of the two PTEN variants was a pathogenic frameshift insertion (p.C211fs). In 29% (11/38) of samples, at least some weak nuclear ß-catenin immunostaining was detected and PTEN loss was observed in 20% (6/30) of samples. The low prevalence of APC mutations in UrC represents a characteristic difference to CRC. Based on APC and ß-catenin results, the Wnt pathway seems to be rarely affected in UrC. Considering the formerly described involvement of PTEN protein loss in anti-EGFR therapy-resistance its immunohistochemical testing may have therapeutic relevance.


Introduction
The urachus is an extraperitoneal fibromuscular band that connects the bladder dome and the umbilicus. During fetal development, the urachus ensures the communication between the forming bladder and the allantois. After the fourth month of embryonic life, the urachus usually transforms into a fibromuscular strand (i.e. median umbilical ligament). In up to one third of adults, this obliteration can be incomplete resulting in microscopic urachal residues. Incomplete regression of the urachal structure may lead to various diseases including urachal cancer (UrC) [1,2].
UrC is an extremely rare disease with an estimated annual incidence of one in one million adults. The majority of UrC present as adenocarcinomas (~90%) with mucinous, intestinal, signet ring, not otherwise specified (NOS) or mixed histology. Because of its hidden anatomical location, UrC is mostly detected in progressed stages when patients present with hematuria resulting from its invasion into the urinary bladder. At this advanced stage, the overall 5-year survival is only about 50% [3][4][5].
In CRC, the adenomatous polyposis coli (APC) tumor suppressor gene has a critical role in the initiation of tumorigenesis. APC mutations lead to abnormal ß-catenin accumulation in the nucleus. This nuclear ß-catenin gets in contact with a member of the TCF/LEF family and acts as a transcriptional regulator on specific proliferation-associated target genes. APC mutations occur in more than 80% of CRC [17]. As a consequence, nuclear ß-catenin expression can be observed in the generality of CRC making ß-catenin a well established diagnostic and prognostic biomarker in CRC [19,20]. The occurrence and role of APC-alterations and ß-catenin-expression are less well established in UrC.
Phosphatase and tensin homolog (PTEN) is a leading negative regulator of the PI3K signaling pathway and therefore known as a tumorsupressor gene. Its inactivating genetic mutations occur in up to 10% of CRCs [17]. In addition to genetic loss, PTEN is frequently downregulated by epigenetic silencing, leading to loss of PTEN expression [21]. Loss of PTEN protein expression was shown to be associated with resistance to anti-epidermal growth factor receptor (EGFR) therapy [22], while this therapy demonstrated effectiveness in UrC [14].
Therefore, we aimed to analyze relevant genetic alterations including the APC and PTEN genes as well as to assess the tissue protein expressions of β-catenin and PTEN in UrC.

Clinical Samples
Formalin-fixed paraffin embedded (FFPE) tumor tissues from 40 urachal adenocarcinoma patients were retrospectively collected from nine academic centers. Clinicopathological and follow-up data were retrieved from cooperating institutions using a uniform datasheet. Histopathological evaluation was performed according to criteria also adopted by the World Health Organization [23]. The study was performed in compliance with the Declaration of Helsinki and the institutional ethics committee approved the study protocol (16-6902-BO).

Sample Preparation and Targeted Next-Generation Sequencing
FFPE tumor tissue blocks were cut and 4 μm-thick slides were stained with hematoxylin and eosin. Tumor containing areas were marked by a pathologist (H.R.) and macrodissection was performed. Sample preparation and targeted next-generation sequencing were performed as previously described [7]. For targeted sequencing, a customized panel was used containing regions of interest. The customized panel contains exonic and flanking intronic regions from 6 genes (APC: exons 1-16, DICER1: exons 1-28, PRKAR1A: exons 1-11, PTEN: exons 1-9, TSHR: exons 1-10, WRN: exons 2-35). The regions were covered by a total of 108 amplicons. In all runs, an average coverage of approximately 5000x was obtained. Analyses were performed using CLC Biomedical Genomics Workbench.

Statistical Analysis
Statistical analyses were conducted with SPSS software package (v24 Chicago, IL). Kaplan-Meier survival analyses with the log-rank test were used to evaluate the impact of selected variables on overall survival. Furthermore, univariable Cox regression analyses were done. P values <0.05 were accounted statistically significant.

Cohort Characteristics and Primary Treatment
The median age of patients at diagnosis was 51.5 years (range 24-78). Twenty-three patients (58%) were male and 17 (42%) were female. The most frequent symptom of UrC was haematuria reported in 83% of patients. Other symptoms included abdominal pain (10%) and palpable tumor mass (3%) ( Table 1).

Genomic Alterations in Urachal Cancer
Sample purification in 34 of 40 cases yielded appropriate amount and quality of DNA sample for sequencing analysis. We analyzed the alterations of APC, DICER1, PRKAR1A, PTEN, TSHR and WRN genes. The APC and PTEN genes were affected in 15% (5/34) and 6% (2/34) of cases. Two of the APC alterations (p.Y1075*; c.3225 T > G, p.K1199*; c.3595A > T) were truncating mutations resulting in a truncated and non-functional APC protein and were therefore considered as pathogenic, while other

Survival Analysis
The median overall survival (OS) was 34 months (range: 2-212 months) and the median progression-free survival (PFS) was 22 months (range 0-117 months). Three-year and five-year OS was 71% and 64%, respectively. Postoperative progression was detected in 50% of cases (16/31). At the time of data evaluation, 65% of patient (22/34) were alive. Neither APC nor PTEN mutations were associated with OS, however these results must be interpreted with caution as only two cases demonstrated APC and PTEN mutations ( Table 3) (Fig. 3a, c). Presence of nuclear ß-catenin and loss of PTEN expressions were not associated with shorter OS (p = 0.606 and p = 0.869) (Fig. 3b, d).

Discussion
It is well known, that mutations of the APC tumor suppressor gene have a critical role already in early stages of CRC development. The most important function of APC is to establish an interaction with the ß-catenin protein, thus accelerating its degradation and regulation of the cadherin-mediated cellcell adhesion system. Functional loss of APC leads to abnormal ß-catenin accumulation and translocation from the plasma membrane to the nucleus [20].
In our study, analyzing 34 UrC cases, deleterious alterations of APC tumor suppressor gene were present in 2 UrC samples. Neither mutation has been previously reported in UrC: p.Y1075* (c.3225 T > G), p.K1199* (c.3595A > T). Both of these nonsense mutations are predicted to cause a truncated, dysfunctional APC protein.
Three additional alterations were considered benign. Presence of potentially pathogenic APC mutations tended to be associated with shorter OS, however, because of the low number of cases with APC mutation, this correlation has to be interpreted cautiously.
Summarizing our results with all currently available data on APC status in UrC, an overall number of 14 of 141 (10%) UrC samples exhibited APC alterations [6-16, 18, 24-26] which is in clear contrast to the high APC mutational rate (80%) found in CRC [17]. This finding represents a further characteristic difference in the molecular taxonomy between CRC and UrC. Therefore, our results demonstrate that the Wnt pathway is probably less frequently involved in the pathogenesis of UrC compared to CRC. ß-catenin is a well-established diagnostic and prognostic biomarker in CRC. Physiologically, ß-catenin staining is restricted to the membrane/cytoplasm and is involved in cadherin-mediated cell-cell adhesion and gene transcription regulation. Nuclear ß-catenin expression can be observed in CRC [19]. Wong et al. demonstrated positive nuclear staining for ß-catenin expression in the vast majority of colorectal adenocarcinomas which reflects the high APC mutation rate found in CRC [27]. In a recent review of the UrC literature including own data, we found a low incidence of nuclear ß-catenin staining in UrC (14%, 9/63) [4]. In the present study, we detected positive nuclear ß-catenin immunostaining in 29% (11/38) of UrC samples. In addition, ß-catenin nuclear expression was not associated with adverse OS (p = 0.606). Interestingly, we observed a nuclear ß-catenin localization only in one of the two samples with truncating APC mutation. As a potential genetic mechanism for ß-catenin nuclear accumulation, Alomar et al. identified an activating mutation in exon 3 of CTNNB1 (ß-catenin) gene which resulted in an amino acid change at phosphorylation sites of glycogen synthase kinase-3 (GSK-3β). Failing of phosphorylation was found to decrease sequestration of β-catenin by APC [28]. This effect might explain our finding with low APC mutational frequency but at the same time surprising high rate (29%) of nuclear ß-catenin positivity. It has also to be kept in mind that we used a 1% threshold for calling a case positive in case of nuclear β-catenin staining. In some cases, it is difficult to discriminate between a real nuclear staining event in the background of strong membranous/ Sex The tumor suppressor PTEN is a negative regulator of the PI3K signaling pathway. PTEN mutations occur in 4-10% of CRC and were suggested as potential markers of response to EGFR and mitogen-activated protein kinase (MAPK) inhibitor-based targeted therapies [17]. Perrone et al. showed that inactivation of PTEN protein by mutation (P103S, E99*) or deletion (hemizygous) was responsible of anti-EGFR resistance [29]. PTEN protein loss was detected in approximately 40% of all CRC patients, which is clearly higher compared to the rate of genomic loss. This suggests that PTEN is more frequently downregulated by epigenetic silencing. In accordance, PTEN methylation was found to be significantly correlated with PTEN expression [21]. Similar to the findings at the genomic level, Frattini et al. showed that loss of PTEN protein expression was associated with non-responsiveness to cetuximab [22].
In the literature, we found three whole exome sequencing studies with an overall number of 19 UrC samples. None of them reported any PTEN mutations [10,[14][15][16]. In addition, none of the targeted sequencing studies reported any PTEN mutations in UrC. Here, we identified a pathogenic frameshift PTEN mutation (p.C211fs) in one of the 34 UrC cancer samples (3%), suggesting a low incidence for PTEN mutation in UrC. Therefore, the PTEN mutational frequency in UrC seems to be similar to that of reported in CRC (4-10%) [17].
To the best of our knowledge, no published data is available on PTEN protein expression in UrC. In the present study, performing PTEN IHC analysis, we observed PTEN protein loss in 20% of UrC cases (6/30), which is somewhat lower than that of 30-40% described in CRC. The low rate of PTEN loss at the DNA level, but relative high rate at the protein level suggests that PTEN is  predominantly silenced by epigenetic downregulation in UrC [21]. The high rate of PTEN protein loss in UrC suggests that its immunohistochemical analysis may be important in order to predict potential inresponsiveness to anti-EGFR therapy. In contrast, the rare occurrence of PTEN inactivating mutations suggests that these alterations are less important to analyze when considering the administration of an anti-EGFR drug, while activating KRAS mutations are more common in UrC (28%) and are clinically important negative predictors of anti-EGFR therapy [6-16, 18, 24-26].

Conclusions
In summary, we analyzed a large cohort of UrC with a targeted next-generation sequencing approach supplemented with ß-catenin and PTEN immunohistochemical analyses.
Our results show that APC mutations are much less frequent in UrC (10%) compared to CRC (80%) suggesting that the Wnt pathway is involved in the pathogenesis of only a relatively small portion of UrC. In addition, both UrC and CRC have low rates genomic loss (3% and 4-10%) but relative high rates of PTEN protein loss (20% and 40%), suggesting an epigenetic regulation for this gene in both tumor types. Therefore, as the loss of PTEN protein expression was found to be associated with resistance against anti-EGFR therapy, its immunohistochemical testing should be considered when planning anti-EGFR therapy. Our data provides further clues for both the similarity (regarding PTEN) and difference (regarding APC) between UrC and CRC and suggest that these tumor types are similar yet distinct on the molecular level.
Acknowledgments Tibor Szarvas was supported by János Bolyai Research Scholarship of the Hungarian Academy of Sciences. This work was supported by the National Research, Development and Innovation Office -NKFIH / PD 115616 and NVKP_16-1-2016-004.
Funding Information Open access funding provided by Semmelweis University.

Compliance with Ethical Standards
Declarations of Interest None. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.