Article Text

Download PDFPDF

Mature renal teratoma and a synchronous malignant neuroepithelial tumour of the ipsilateral adrenal gland
  1. D Govender,
  2. L M Nteene,
  3. R Chetty,
  4. G P Hadley
  1. Division of Anatomical Pathology, School of Pathology and Laboratory Medicine, University of Natal Medical School, Private Bag 7, Congella, Durban 4013, South Africa govendh{at}
  2. Department of Paediatric Surgery, University of Natal Medical School

    Statistics from

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    The main primary childhood renal neoplasms are nephroblastoma, mesoblastic nephroma, clear cell sarcoma, and rhabdoid tumour. Other primary renal neoplasms include primitive neuroectodermal tumour (PNET), renal cell carcinoma, and angiomyolipoma. Nephroblastoma is the most common renal tumour in children. It is a complex embryonal tumour of metanephric blastemal derivation, which often contains diverse epithelial and stromal tissues. Diagnostic problems are often encountered when tumours contain a variety of heterologous elements. The term teratoid nephroblastoma has been used to describe a variant of nephroblastoma with a predominance of heterologous tissues.1 It is this variant that can be confused with a teratoma. Renal teratomas are rare and most have been dismissed as cases of teratoid nephroblastomas or retroperitoneal teratomas secondarily invading the kidney.2 The differentiation between these two neoplasms in the kidney is often problematic.

    Neurogenic tissues in the kidney can be found in primary tumours or as part of metastatic tumours. The primary tumours are nephroblastoma, which may contain ganglion cells, neuroblast, and neuroglial tissue,2 and PNET.3, 4 Adrenal neuroblastomas can directly invade the adjacent kidney.5 We describe the pathology of a right renal mass in a 3 year old child and discuss the differential diagnosis.

    A 3 year old girl presented with abdominal pain and diarrhoea. On examination she was found to have signs of pulmonary tuberculosis and was started on antituberculous treatment. Subsequently, a large, firm, tender, right flank mass clearly separate from the liver was detected and she was referred to the Regional Paediatric Surgical Unit for further investigation and management.

    On admission, the child was apyrexial, emaciated, and weighed 13 kg. She had bilateral coarse crackles and a wheeze. The abdomen was distended and a non-tender 3 cm hepatomegaly was palpated. Furthermore, a 10 × 12 cm non-tender, firm, non-pulsatile right flank mass was detected.

    Results of routine laboratory tests were as follows: haemoglobin, 90 g/litre (normal, 112–143); white blood cell count, 8.2 × 109/litre (normal, 5.5–15.5); and platelet count, 224 × 109/litre. Urinary catecholamine values were as follows: noradrenaline, 0.279 μM/mM creatinine (CRT) (normal, 0–0.08); adrenaline, 0.023 μM/mM CRT (normal, 0–0.035); dopamine, 0.67 μM/mM CRT (normal, 0–1.13); vanillylmandelic acid 9 μM/mM CRT (normal, 0–15); and homovanillic acid, 11 μM/mM CRT (normal, 0–15). Renal and liver function tests were normal.

    Computed tomography (CT) scan of the abdomen revealed a large tumour involving the right side of the abdomen. There were also multiple hepatic lesions consistent with metastases. A fine needle aspiration biopsy of the mass was performed. After a cytological diagnosis of neuroblastoma the child was started on the appropriate chemotherapy protocol consisting of vincristine, actinomycin, cyclophosphamide, and adriamycin. She suffered seizures while on chemotherapy, which was subsequently decreased to 75% dosage. A CT scan of the brain was normal.

    The child died two weeks after commencement of chemotherapy. A necropsy was performed.

    Necropsy revealed an enlarged right kidney measuring 16 × 10 × 8 cm and weighing 1200 g. There was a well demarcated tumour mass in the upper pole of the kidney, which measured 7 x 6.5 x 5 cm (fig 1). The upper pole mass was encapsulated (fig 1; arrows), predominantly firm in consistency, and had yellow and white areas. There was a small area of soft, friable tumour present within the mass, close to the junction of the adjacent kidney. On the capsular surface the upper pole mass was clearly demarcated from the remaining kidney. It was not possible to recognise normal renal tissue because the remaining kidney was diffusely swollen, pale, soft, and contained focal areas of necrosis (fig 1). The right adrenal gland was not identified despite serial sectioning. There was extension of the tumour into the right renal vein and inferior vena cava. Tumour spread into the ureter was also present. There were multiple greatly enlarged para-aortic lymph nodes containing metastatic tumour. Metastatic tumour deposits were present in the liver, vertebrae, and both lungs. In addition, caseous nodules were identified in the lungs and hilar lymph nodes. The left kidney and adrenal gland were normal. The rest of the postmortem examination was normal.

    Histological examination of the well demarcated upper pole mass showed a tumour composed of multiple heterologous tissue elements. Tissues derived from all three germ layers—ectoderm, mesoderm, and endoderm—were present. The yellow areas corresponded to mature adipose tissue microscopically. In addition, striated muscle, smooth muscle, and fibrous tissue were present, as were small islands of neuroglial tissue. The epithelial component was variable and consisted of small tubular structures lined by cuboidal epithelial cells with clear cytoplasm. Also present were larger cystic structures lined by respiratory epithelium. Adjacent to these cysts were small closely packed acini composed of mucin containing epithelium. A prominent feature was the presence of several cysts lined by keratinising stratified squamous epithelium with hair follicles and hair shafts (fig 2). Small round tumour cells diffusely infiltrated the remaining kidney. These were arranged predominantly in solid sheets, but focal neuroectodermal canals and rosettes were also identified (fig 3). In addition, there was microscopic evidence of metastases in the lungs, liver, ovaries, and vertebrae. The metastatic tumour in all sites consisted of the primitive small, round cell component. Sections taken from the region of the right adrenal gland showed a diffuse infiltrate of small round blue cells, but no residual adrenal gland tissue was identified. There were no nephrogenic rests in the kidney.

    A panel of immunohistochemical stains was performed on the small cell malignant tumour to detect cytokeratins (AE1/AE3, CAM5.2), synaptophysin, neurone specific enolase (NSE), chromogranin, WT1, desmin, muscle specific actin, S100 protein, glial fibrillary acidic protein, O13, leucocyte common antigen, and epithelial membrane antigen. The small round cells showed immunoreactivity for NSE, synaptophysin, and chromogranin. The cells were non-reactive for the remaining markers.

    Histology confirmed tuberculosis of the lungs and lymph nodes. Acid fast bacilli were identified.

    To diagnose a renal teratoma, the primary tumour should be unequivocally of renal origin and the tumour should exhibit unequivocal heterotopic organogenesis clearly recognised as evidence of attempts to form organs other than the kidney.2 It is the second criterion that often presents a problem. The question is: what constitutes unequivocal organogenesis? The presence of bone, cartilage, muscle, fat, neuroglial tissue, and mature epithelium cannot on their own be regarded as evidence of organogenesis. Indeed, all of these tissues can be present in both teratomas and teratoid nephroblastomas. In mature teratomas, skin with the dermal appendages, bronchial structures with bronchial glands and cartilage, brain (neuroglial tissue), and teeth are commonly present, and regarded as evidence of organogenesis.

    Those neuroblastomas that are characterised by the presence of neuropil and ganglion cells can be readily differentiated from PNET. It is the undifferentiated neuroblastoma, in which neuropil and ganglion cells are sparse or absent, that is sometimes difficult to differentiate from PNET. The immunohistochemical profile and cytogenetics are helpful in this instance. PNET characteristically demonstrates diffuse membrane positivity with O13 (mic2 gene product) and is also immunoreactive for vimentin, cytokeratin, and NSE.6 Another highly characteristic feature of PNET/Ewing's sarcoma is the presence of a specific reciprocal translocation: t(11;22)(q24;q12).7

    Our case presents a unique constellation of pathological features, which pose a diagnostic problem. The well circumscribed encapsulated upper pole mass could be interpreted either as a renal teratoma or a teratoid nephroblastoma. After much deliberation and consultation, we prefer the diagnosis of a primary intrarenal teratoma. Most of the heterologous elements present can, as alluded to earlier, occur in a nephroblastoma. The presence of hair shafts is evidence of terminal differentiation, which is seen in teratomas. Although structures resembling hair follicles have been described in nephroblastoma, to the best of our knowledge hair shafts have not been described in nephroblastoma.

    The small round cell tumour presents a more difficult problem. We have demonstrated neural differentiation in these cells with immunoreactivity to NSE, synaptophysin, and chromogranin. It is also quite clear on morphology and immunohistochemistry that this component represents a malignant neuroepithelial neoplasm. The problem however is the derivation or origin of this component. First, it might be a primitive neuroepithelial component arising from the teratoma, a phenomenon that is known to occur. However, we consider this unlikely because the mature teratoma was well circumscribed and not diffusely infiltrated by the neuroepithelial component. Furthermore, there is no evidence of merging of the mature and immature components. Second, it might be a PNET of the kidney. The negative O13 immunoreactivity and the diffuse infiltrative nature of the tumour makes PNET less likely. Cytogenetic studies for the characteristic t(11;22) translocation were not done in this case. Finally, the inability to identify the ipsilateral adrenal gland and the extensive extrarenal tumour in the region of the adrenal gland must be deemed important. We feel that the most plausible explanation is an undifferentiated neuroblastoma of the adrenal gland with permeation of the kidney. The incidence of renal parenchymal invasion by neuroblastoma has been reported to be as high as 20.4% in one series.5 This same study found that invasion of the kidney occurred more frequently with undifferentiated neuroblastomas and those with extensive abdominal involvement.

    In trying to clarify the definition of “unequivocal organogenesis”, the concept of mimicry in embryonal tumours needs considering. Clearly, mimicry refers to structures or tissue resembling the normal developmental stages in organogenesis. Therefore, unequivocal organogenesis could be defined as the presence of immature or mature tissues arranged in a manner that is comparable to the “normal” development of the organ or the mature appearance of that organ. If—for example, structures resembling hair follicles and sweat glands are present, but do not demonstrate their normal arrangement in skin, then this should be referred to as mimicry rather than evidence of unequivocal organogenesis. In our case, hair shafts are regarded as evidence of the end stage of differentiation. The cysts lined by squamous epithelium (epidermis) with eccrine sweat glands, hair follicles, and hair shafts warrant classification as a renal teratoma.

    The pathological features were finally interpreted as representing two distinct neoplasms: a mature renal teratoma and a malignant neuroepithelial tumour. The latter is probably an undifferentiated neuroblastoma arising in the adrenal gland and invading the kidney, mainly along the hilum.

    Our case posed a problem often faced by pathologists who regularly examine renal tumours in children. Furthermore, it highlights the need for a refinement of the diagnostic criteria for renal teratoma. We believe that the differentiation between these two neoplasms must lie in their respective genetics. For example, deletion of the short arm of chromosome 11 should favour a nephroblastoma, as would the presence of nephrogenic rests.

    Figure 1

    An enlarged right kidney with an encapsulated tumour mass (arrows) in the upper pole. The rest of the kidney is also diffusely infiltrated.

    Figure 2

    Keratinous cysts, hair follicle containing hair shaft (arrow), and adjacent adipose tissue. Inset: hair follicle with hair shaft.

    Figure 3

    Malignant neuroepithelial tumour with focal neuroectodermal canals and rosettes.


    We thank Professor LP Dehner, St Louis, USA for reviewing the histology.