Wilms tumor (WT) in adult patients is rare and has historically been a diagnostic and therapeutic conundrum, with limited data available in the literature.
To provide detailed diagnostic features, molecular profiling, and patient outcomes in a multi-institutional cohort of adult WT patients.
We identified and retrospectively examined 4 adult WT cases.
Two patients presented with metastatic disease, and diagnoses were made on fine-needle aspiration of their renal masses. The aspirates included malignant primitive-appearing epithelioid cells forming tubular rosettes and necrosis, and cell blocks demonstrated triphasic histology. In the remaining 2 cases, patients presented with localized disease and received a diagnosis on resection, with both patients demonstrating an epithelial-predominant morphology. Tumor cells in all cases were patchy variable positive for PAX8 and WT1 immunohistochemistry. Next-generation sequencing identified alterations previously reported in pediatric WT in 3 of 4 cases, including mutations in ASXL1 (2 of 4), WT1 (1 of 4), and the TERT promoter (1 of 4), as well as 1q gains (1 of 4); 1 case showed no alterations. Three patients were treated with pediatric chemotherapy protocols; during follow-up (range, 26–60 months), 1 patient died of disease.
WT is an unexpected and difficult entity to diagnose in adults and should be considered when faced with a primitive-appearing renal or metastatic tumor. Molecular testing may help exclude other possibilities but may not be sensitive or specific because of the relatively large number of driver mutations reported in WT.
Wilms tumor (WT; ie, nephroblastoma), a commonly encountered diagnosis in the evaluation of pediatric renal tumors, is an exceedingly rare and challenging diagnosis in adult patients.1 Patients with adult WT require appropriate and timely diagnosis because improved outcomes have been reported when it is managed with standardized pediatric protocols that include tumor stage–based chemotherapy.2–5 Histologically, adult WTs can demonstrate patterns characteristic of their pediatric counterparts, but multiple biologic differences have been noted, including lack of association with nephrogenic rests and with other developmental conditions and syndromes associated with pediatric WT.6 Here, we report detailed clinicopathologic and molecular findings of 4 adult WT cases identified in the pathology archives of the authors' institutions.
MATERIALS AND METHODS
Institutional review board approvals and cases were obtained at the authors' respective institutions. The requirement for obtaining individual patient consent was waived in accordance with institutional policies. Clinical and pathologic information was provided by each contributor, in addition to a representative formalin-fixed, paraffin-embedded (FFPE) tissue block for molecular testing. All cases underwent formal review at multiple institutions during original evaluation. Immunohistochemical stain results were based on those performed during original clinical evaluation. Molecular analysis was performed on FFPE tissue containing tumor with paired normal kidney using the UCSF500 Cancer Gene Test (UCSF Genomic Sequencing Services Laboratory, San Francisco, California), a clinically validated platform that uses capture-based next-generation sequencing as previously described (gene list: https://genomics.ucsf.edu/content/ucsf-500-cancer-gene-panel-test-ucsf500-uc500, version 1-3) on cases 1 to 3.7 In case 4, tumor-only UCSF500 Cancer Gene Test molecular testing had been performed during original clinical evaluation of the case.
RESULTS
Clinical, pathologic, molecular, and follow-up characteristics of the 4 adult WT patients are detailed in the Table. All patients were women, with a mean age of 33.5 years (range, 23–44 years). Two cases (cases 1 and 2) were presumed metastatic at presentation and diagnosed by fine-needle aspiration (FNA) of the patients' kidney masses. In case 3, the patient underwent initial core biopsy at another hospital and was definitively diagnosed following radical nephrectomy, and case 4 was diagnosed by radical nephrectomy. The 2 patients who received a diagnosis by FNA and 1 patient who received a diagnosis on resection were treated on pediatric chemotherapy protocols. During follow-up (range, 26–60 months), 1 of the patients who had undergone FNA died of disease, and the remaining patients showed no evidence of disease.
Diagnostic Histologic Features
Cases 1 and 2
FNA aspirates revealed discohesive clusters of malignant epithelioid cells with a distinct tubular and rosette pattern in a background of sparse necrosis and acute inflammation (Figure, A and B). The tumor cells had scant, delicate, ill-defined cytoplasm with hyperchromatic nuclei, irregular nuclear contours, and coarse granular chromatin. Mitotic figures were readily identified. Based on rapid on-site evaluation (ROSE) of the FNA sample at the time of the biopsy, a portion of the aspirate was immediately fixed in formalin for cell block preparation using the collodion bag technique.8 The FFPE cell block demonstrated tumor cells in a triphasic pattern, including tubular structures (ie, epithelial component) and solid aggregates (ie, blastemal component), as well as small foci of stromal component (Figure, C and D). Tumor cells were patchy and variably positive for WT1, PAX8, and pan-keratin by immunohistochemistry (Figure, E and F). The triphasic pattern and positive WT1 and PAX8 were supportive of WT. Notably, in both cases, given the relatively limited sampling on FNA, the original diagnostic considerations were broad, and an extensive panel of immunohistochemical stains (including CAIX, p63, SF1, CK7, CK20, CD117, synaptophysin, chromogranin, CD99, calretinin, inhibin, and β-catenin—all negative) had been performed to consider other possibilities, including adult renal cell carcinomas (eg, papillary renal cell carcinoma, clear cell renal cell carcinoma, collecting duct carcinoma), urothelial carcinoma, neuroendocrine neoplasms, neuroectodermal neoplasms, synovial sarcoma, and metastases (eg, Müllerian, thyroid), among others.
Diagnostic findings in adult Wilms tumor. A through F, Case 1: fine-needle aspiration cell aspirates (A and B); cell block hematoxylin-eosin (H&E) stain (C and D); PAX8 stain (E); WT1 stain (F). G and H, Case 3: representative H&E stains. I through L, Case 4: representative H&E (I and J); PAX8 (K); and WT1 (L) stains. In D, green asterisks highlight epithelial component; green arrowhead highlights stromal component (original magnifications ×20 [A, D through G, I, and J], ×4 [C], and ×40 [B, H, K, and L]).
Diagnostic findings in adult Wilms tumor. A through F, Case 1: fine-needle aspiration cell aspirates (A and B); cell block hematoxylin-eosin (H&E) stain (C and D); PAX8 stain (E); WT1 stain (F). G and H, Case 3: representative H&E stains. I through L, Case 4: representative H&E (I and J); PAX8 (K); and WT1 (L) stains. In D, green asterisks highlight epithelial component; green arrowhead highlights stromal component (original magnifications ×20 [A, D through G, I, and J], ×4 [C], and ×40 [B, H, K, and L]).
Case 3
Initial core biopsy of the kidney mass was diagnosed as an epithelioid neoplasm with nephrogenic features, suggestive of WT, but the differential diagnosis also included renal cell carcinoma. Definitive diagnosis was made at the time of resection, in which left radical nephrectomy revealed 2 masses (4.2 and 1.4 cm), each showing a triphasic but epithelial-predominant WT with histomorphology ranging from moderately differentiated tubular elements to poorly differentiated rosettelike structures and scattered admixed blastemal components (Figure, G and H). Focal anaplasia as defined for pediatric WT was seen. Vascular invasion was present in the large veins in intrarenal sinus tissue. Nephrogenic rests were not identified. Similar to cases 1 and 2, a broad immunohistochemical panel was performed, demonstrating the tumor cells to be patchy positive for WT1 and PAX8 and negative for synaptophysin, chromogranin, desmin, ER, PR, CK7, CK20, P504S, BRAF, CDX2, TTF1, melan-A, inhibin, calretinin, GATA3, TFE3, and cathepsin K. CD99 showed very focal equivocal staining. The overall morphologic features and immunoprofile supported a diagnosis of adult WT.
Case 4
No biopsy was performed. Based on the clinical presentation and imaging findings of a renal mass suspicious for renal cell carcinoma, the patient underwent right radical nephrectomy. Sections of the kidney mass revealed a cellular tumor composed of primitive-appearing epithelioid cells with solid, tubular, and trabecular architecture (Figure, I and J). The tumor cells had scant delicate cytoplasm and irregular nuclear contours with coarse chromatin. Mitotic figures were readily identified, and lymphovascular invasion was present. No nephrogenic rests were identified. This tumor was somewhat unusual for a WT: it did not show triphasic features, and the tumor cells were not blastemal or definitely epithelial. Based on the morphologic findings, the original differential diagnosis was broad and included WT, metanephric adenoma, synovial sarcoma, CIC-DUX4 sarcoma, clear cell sarcoma of the kidney, solitary fibrous tumor, and neuroendocrine tumor. The tumor showed positive immunohistochemical staining for PAX8 (diffuse variable weak to strong), WT1 (patchy), and keratin AE1/AE3 and was negative for GATA3, STAT6, BCOR, synaptophysin, chromogranin, and desmin; as in case 3, CD99 showed rare equivocal staining. Ultimately, the poorly differentiated epithelioid morphology, presence of nuclear staining for PAX8 and WT1, and exclusion of other tumors that would be considered in the differential diagnosis by ancillary testing (absence of BRAF mutation, see below), led to classification as WT, predominantly poorly differentiated epithelial type (Figure, K and L).
Molecular Findings
Pathogenic and likely pathogenic alterations and copy number analysis are shown in the Table and the supplemental digital content, containing 4 figures, at https://meridian.allenpress.com/aplm in the June 2024 table of contents. Notably, no BRAF p.V600E mutations or alterations suggestive of another adult renal tumor type were identified. No germ-line mutations were identified in the paired normal samples for cases 1 to 3.
DISCUSSION
WT in adult patients is a difficult and unexpected diagnosis, whereas renal cell carcinoma is the most common kidney tumor. In the literature, adult WT is described mostly in single case reports or small series9–12 ; the largest series focused on treatment presented by the Society of Pediatric Oncology suggests that central pathologic review is necessary to obviate the risk of misdiagnosis or misclassification.5 Our study of 4 cases highlights the challenges of rendering a diagnosis of WT in this age group, particularly at FNA when sampling may be limited, and provides further molecular characterization of these rare tumors.
In the vast majority of pediatric cases, the diagnosis is made clinically and by imaging, and the role of pathologic evaluation is to guide subsequent therapy.13 Pretreatment biopsy is generally discouraged because nephrogenic rests may be indistinguishable from WT based on cytologic features alone. In contrast, in our review of adult WT patients, 2 of 4 cases were diagnosed by FNA, and a third case underwent core biopsy prior to resection. To our knowledge, detailed FNA findings in adult WT have not been previously reported in the literature. In our FNA cases, the biopsy yielded diagnostic cytomorphology, including malignant epithelioid cells with tubular and rosettelike patterns, frequent mitotic figures, and necrosis. The FNA findings seen in our adult WT patients are similar to those reported in pediatric patients.14–16 However, unlike in pediatric patients, in adult patients there is a much broader differential diagnosis, thereby requiring significant additional ancillary testing to consider other possibilities. For example, the FNA findings of WT can sometimes mimic those of other tumors more commonly seen in the adult population, such as papillary adenoma/renal cell carcinoma and metanephric adenoma, both of which would have markedly different prognostic and therapeutic implications. In contrast to the cytomorphology of WT, papillary adenoma/renal cell carcinomas are more likely to show abundant papillary clusters with fibrovascular cores rather than single cells; the tumor cell nuclei are also typically small and uniform, with only mild to moderate hyperchromasia, with single and small nucleoli, and nuclear grooves.17 Metanephric adenoma, a typically benign tumor, usually shows bland uniform nuclei without necrosis or significant mitotic activity,18,19 yet it can show more cytomorphologic overlap with WT if a more differentiated epithelial area of WT is sampled. Although both metanephric adenoma and WT can have similar primitive-appearing epithelium that is WT1 positive, the presence of a blastemal component, mitotic figures, and necrosis indicate WT. In addition to other renal tumors, a very broad differential diagnosis should be considered, which includes urothelial carcinoma, neuroendocrine and neuroectodermal neoplasms, other soft tissue tumors, and metastases. Therefore, to make a diagnosis of adult WT on FNA, the most helpful pathologic features include malignant epithelioid cells with tubular and rosettelike architecture, frequent mitotic figures, and necrosis. Importantly, obtaining ample cell block material is critical for demonstrating a triphasic tumor and enabling immunohistochemical workup to further support WT and exclude other possibilities.
On a resection specimen, it may be easier to apply the criteria previously published in 1980 by Kilton et al20 for a diagnosis of adult WT: the tumor under consideration should (1) be a primary renal neoplasm; (2) demonstrate a primitive blastemic spindle or round cell component; (3) show abortive or embryonal tubules or glomerular structures; (4) not harbor an area diagnostic of renal cell carcinoma; (5) display pictorial confirmation of histology; and (6) arise in a patient older than 15 years. However, even in our 2 cases diagnosed on resection (cases 3 and 4), a broad differential diagnosis and extensive immunohistochemical workup was necessary to support a diagnosis of adult WT. Both of these cases were epithelial predominant.
The molecular findings in our 4 cases are also similar to those found in pediatric counterparts,21 including recurrent ASXL1 truncating mutations in cases 1 and 4, a WT1 mutation in case 1, and a 1q gain in case 2. A recent comprehensive genetic analysis of 14 adult WTs by Argani et al22 found a subset (5 of 14; 36%) harboring BRAF p.V600E mutations, with both histologic and molecular evidence for tumor progression from a metanephric adenoma to WT, which we had also reported previously.7 In our cases, no BRAF p.V600E mutations were identified; likewise, none contained well-differentiated metanephric adenoma-like areas, supporting that our cases were de novo WT (ie, not arising from a metanephric adenoma). Interestingly, case 3 was multifocal; however, no nephrogenic rests, germ-line mutations, syndromic association, or developmental conditions were identified in this or any of our cases.
Notably, case 1 demonstrated an activating TERT promoter (TERTp) mutation in addition to WT1 and ASXL1 mutations. TERT encodes the telomerase protein, and increased TERT expression and telomerase activity have been correlated with disease relapse, anaplastic histology, and increased percentage of blastemal elements in pediatric WT.23–25 In the Argani et al22 study of adult WT, TERTp mutations were only reported in the cases with BRAF p.V600E mutation, but they were not seen in pure adult WT. Our case with TERTp mutation presented with very advanced stage (Children's Oncology Group stage IV) and was posttreatment high-risk blastemal predominant, but it otherwise exhibited favorable histology (ie, no anaplasia) and has shown no evidence of disease after 60 months, responding well to pediatric treatment protocols.
In 1 case (case 4), the morphologic features were not classic for WT because of a lack of typical triphasic features. In addition, the tumor cells were fairly uniform and could not be definitively categorized as blastemal or epithelial. The tumor was therefore classified as “poorly differentiated,” a somewhat intermediate phenotype on the blastemal-epithelial spectrum. Positive staining for WT1 by immunohistochemistry supported a diagnosis of adult WT. In addition, a driver mutation in ASXL1 was identified, which has previously been reported to be recurrently involved in WT.21 This patient did not undergo additional chemotherapy and remains free of disease 26 months after resection.
Limitations to our study include only a small number of cases reported. However, at the primary author's institution, which sees approximately 150 kidney resections per year, only 1 in-house case was identified in a 30-year search (less than 0.05% of cases). In addition, 1 case showed a lack of any identified driver mutations, as well as no copy number changes. We speculate that case may harbor 1 of the additional driver mutations reported in the WT literature (eg, DGCR8, XPO5, SIX1, SIX2, MLLT1, or NONO) that is not covered by our gene panel.21 Although our study and others have shown that adult WTs do not have definitive characteristic genetic findings, molecular testing can still be considered to exclude other entities (eg, DICER1-associated sarcoma, clear cell sarcoma, etc) and may be used to obtain loss of heterozygosity information (1p/16q), associated with prognosis and used for risk assessment.26
In summary, WT is an unexpected and challenging entity to diagnose in adult patients and should be considered when faced with a primitive-appearing tumor, particularly on biopsy. The differential diagnosis is typically broad, and tissue may be collected for immunohistochemistry and other ancillary testing at FNA. Molecular testing can be helpful to exclude other possibilities, but sequencing results may not be specific because of the relatively large number of driver mutations reported in WT.
The authors would like to thank Shirley Kwok for technical assistance with tissue sectioning and members of the University of California, San Francisco (UCSF) Clinical Cancer Genomics Laboratory for sequencing work.
References
Author notes
Supplemental digital content is available for this article at https://meridian.allenpress.com/aplm in the June 2024 table of contents.
This research was supported and funded by the residents' teaching and research endowments from the UCSF Department of Pathology and Stanford Department of Pathology.
Competing Interests
The authors have no relevant financial interest in the products or companies described in this article.