J Clin Pathol 1991;44:317-320 317 Congenital mesoblastic nephroma: Possible prognostic and management value of assessing J Clin Pathol: first published as 10.1136/jcp.44.4.317 on 1 April 1991. Downloaded from DNA content

J C Barrantes, C Toyn, K R Muir, S E Parkes, F Raafat, A H Cameron, H B Marsden, J R Mann

Abstract appears benign with a leiomyomatous pattern The case records and pathology of all composed of interlacing bundles of spindle children with kidney tumours treated in cells and few mitotic figures; congenital the West Midlands Health Authority mesoblastic nephroma is densely cellular, with Region (WMHAR) from 1957 to 1986 many mitotic figures. Both patterns can be were reviewed. The histology was re- found in the same tumour, referred to as a viewed by a panel of three paediatric mixed tumour.9 pathologists. Thirteen (6%) out of 211 In trying to explain the histogenesis of these cases were considered to have congenital tumours, Snyder et al"' proposed a theory mesoblastic nephroma (CMN). Nine using the "two hit" model, described by were of the conventional type, three of Knudson." In this model congenital meso- the atypical cellular type, and one blastic nephroma would occur after a neoplas- mixed. DNA ploidy was investigated and tic mutation in the early stages of embryogen- showed two of the tumours to be aneu- esis, whereas atypical cellular mesoblastic ploid and nine diploid (tissue was not nephroma would develop in the later stages, available in the two other cases). The two but before the blastema has undergone meta- aneuploid tumours were of atypical nephric differentiation. In both cases a second cellular and mixed histology, respec- mutation would be necessary to produce the tively; the diploid tumours were of the malignant transformation. Its origin from pre- conventional type in eight cases and metanephric blastema is supported by an atypical cellular in one. immunohistochemical study by Kumar et al.'2 http://jcp.bmj.com/ The atypical cellular type has been Furthermore, Haas et al"' suggested that, on reported to behave more aggressively, the basis of the highly cellular histological but the benefit of additional treatment appearances of atypical cellular mesoblastic after surgery to prevent recurrence nephroma, clear cell sarcoma of the kidney remains unclear. Measurement of DNA (CCSK) may be the malignant counterpart of content by flow cytometry, together with congenital mesoblastic nephroma and atypical histological subclassification, may be cellular mesoblastic nephroma. More recent on September 30, 2021 by guest. Protected copyright. useful in selecting patients who will studies, however, indicate that this is un- Department of benefit from further treatment after likely.9 Oncology, The surgery. With a view to investigating the biological Children's Hospital, characteristics and the possible factors that Ladywood, Birmingham B16 SET could influence the decision as to which J C Barrantes Congenital mesoblastic nephroma is one of patients would need further treatment in J R Mann the commonest kidney tumours in infants. It addition to surgical resection, we undertook Department of has been associated with a very good prog- review of patients diagnosed as having kidney Pathology, The nosis and, in most cases, surgery alone may tumours in the West Midlands Health Auth- Children's Hospital, Birmingham effect a cure.' ority Region between 1957-1986. As DNA C Toyn It has also been recognised, however, that ploidy has shown a correlation with prognosis F Raafat not all of the cases follow the same course.24 A in Wilms' tumour,'4 15 which explains its The West Midlands more aggressive subtype, atypical cellular aggressive behaviour in some patients, we also Regional Children's mesoblastic with evidence of local Tumour Research nephroma performed flow cytometry wherever possible. Group, The Children's recurrence and even distant pulmonary metas- The objectives of the present study were: Hospital, Birmingham tases, has been reported in some patients,56 (1) To review the incidence, clinical presen- K R Muir but the reasons for the more aggressive be- and outcome of with con- S E Parkes tation, patients A H Cameron haviour are not known precisely. The exis- genital mesoblastic nephroma in the West The Royal Manchester tence of a spectrum of tumours covering the Midlands over a 30 year period. Children's Hospital, whole span of biological behaviour has been (2) To determine if there is any difference in Manchester proposed,78 identifying congenital mesoblastic DNA ploidy between typical congenital H B Marsden nephroma as a low grade tumour and placing mesoblastic nephroma and atypical Correspondence to: atypical cellular mesoblastic nephroma in the cellular mesoblastic nephroma and any Dr J R Mann on histo- Accepted for publication intermediate category. This is based possible implications for the biological 21 November 1990 logical appearances: mesoblastic nephroma nature and behaviour of the subgroups. 318 Barrantes, Toyn, Muir, Parkes, Raafat, Cameron, Marsden, Mann

Methods Results We reviewed the case records of children aged From a total of 211 kidney tumours diagnosed 0-14 years, resident in the West Midlands in the study period, 13 (6-2%) cases of con-

Health Authority Region (average childhood genital mesoblastic nephroma were found. J Clin Pathol: first published as 10.1136/jcp.44.4.317 on 1 April 1991. Downloaded from population 1 17 million) who were diagnosed Table 1 shows the principal characteristics of as having kidney tumours between 1957 and the whole group. Although most children 1986. (eight of 13) presented during infancy, two The cases were identified from the West were over 1 year of age. There were eight males Midlands Regional Children's Tumour and five females, ratio 1 -6: 1. Research Group. Data are held on all cases of The most common presenting features were childhood cancer (0-14 years) in the abdominal distension and mass. Only one WMHAR from 1957, based on the records of patient had a congenital abnormality, present- the Birmingham and West Midlands Regional ing with a congenital mesoblastic nephroma in Cancer Registry, the Children's Cancer the left side of a horse-shoe kidney. Research Group, Oxford, and the files of the All patients had surgery as the principal Oncology and Histopathology Departments of treatment. Eleven had a nephrectomy, one had The Children's Hospital, Birmingham. the tumour excised from a horse-shoe kidney, Clinical and histological staging of these and one with a well localised tumour had a tumours were carried out following the same partial nephrectomy. In seven cases surgery staging system as for Wilms' tumour, pro- was the only treatment (including the child posed after the National Wilms' Tumour with the horse-shoe kidney). Three of the Study (NWTS-2).'6 The available pathology patients received chemotherapy after surgery, material was reviewed by a panel of three two had radiotherapy to the tumour bed, and paediatric histopathologists. one had both. Sections 30 mm thick from paraffin wax The tumours were stage I in four patients embedded pathological material were and stage II in eight, because of local spread obtained and prepared for flow cytometry, through the capsule into the perirenal tissue. according to the method described by One patient (case 6) had the primary tumour Hedley,'7 with minor modifications. confined to the kidney, but malignant cells were The nuclear DNA content of 20 000 cells suspected in the bone marrow, so he was was measured using a Becton Dickinson treated with chemotherapy and radiotherapy, FACS 440 flow cytometer with an excitation following the Medical Research Council wavelength of 448 nm, emission filter of Wilms' tumour protocol. Histological review 590 nm LP, and laser power of 0-2 watts. showed that nine had congenital mesoblastic Normal kidney tissue was used as an external nephroma of the conventional type, three had control, and stromal cells from each sample atypical cellular mesoblastic nephroma, and were taken as the normal internal standard. one had a mixed tumour. http://jcp.bmj.com/ The DNA content was considered to be Analysis of DNA content was possible in 11 diploid when only one gO/gl peak was present cases, nine ofwhich were diploid (fig 1) and two in the histogram and aneuploid when two or aneuploid (fig 2). The two aneuploid tumours more gO/gl peaks were obtained. S phase were of cellular and mixed types, respectively. (synthetic phase) was identified as the per- The results of the DNA studies are shown in centage of cells lying between the gO/gl and table 2. the g2M peaks. The DNA index (DI) repre- The PI was very high in one patient (case 9) s-ented the ratio of the modal channel number (table 2), mildly increased in another (case 1) on September 30, 2021 by guest. Protected copyright. of the aneuploid gO/gl peak to the modal (table 2), and the remainder of the patients had channel number of the diploid gO/gl peak. a PI of < 10%. The DI of the aneuploid The proliferation index (PI) was obtained tumours was 1-4 and 13, respectively. The by the sum of the cells in S phase and g2M. median CV of the whole group was 8-3 with a The coefficient of variation (CV) was obtained range of 4 7-10. The histograms of some of our using the resident Becton Dickinson software. patients exhibited considerable variation in the Evaluable information for DNA ploidy was position of the gO/gl and g2M peaks in other- obtained in 11 of the 13 patients. wise normal histograms. This effect has been Table 1 Congenital mesoblastic nephroma: general characteristics

Case Age Histological Follow up No (m) Sex Side Stage type Treatment* Outcome (years) 1 NB M L 2 CMN RT D 0 75 3 NB M L 1 CMN None A 14 5 4 NB F R 2 CMN None D 5 days 5 NB M L 2 CMN None A 11.6 8 NB M R 2 CMN None A 5 8 9 13 F L 2 CMN CT A 5 10 NB M L 2 CMN None A 0 4 11 NB F L 2 CMN None A 1-4 13 NB F L 1 CMN None A 2-8 6 19 M R ?4t ACMN CT; RT A 8-3 7 1 M L 1 ACMN CT A 5-8 12 7 F L 1 ACMN CT A 3-3 2 3 M R 2 Mixed RT A 25 *In addition to nephrectomy. tBone marrow: suspected malignant cells. NB = newborn; CT = chemotherapy; RT = radiotherapy; A = alive; D = dead. Assessment ofDNA content in congenital mesoblastic nephroma 319

Figure I Diploid DNA DY:H020.DIS NF 20-1-89..020 Counts= 20480 Table 2 DNA studies of congenital mesoblastic content in a conventional HS=256 VS=1024 nephroma (CMN) mesoblastic nephroma. 3 Case Histological No type Ploidy DI PI J Clin Pathol: first published as 10.1136/jcp.44.4.317 on 1 April 1991. Downloaded from 1 CMN D 16-0 3 CMN D < 1 0 4 CMN 5 CMN D 3 4 8 CMN D 6-9 E 9 CMN D 514 10 CMN D 7-5 z 11 CMN D 8-5 13 CMN D 5-0 6 ACMN D 3-6 7 ACMN 12 ACMN A 14 2 Mixed A 1-3 DI = DNA index; PI = proliferative index; D = diploid; A = aneuploid.

described as a consequence ofthe fixation ofthe highest PI value in the DNA studies (case 9) tissue in the paraffin wax blocks.'8 (table 2) and was one ofthe two with congenital Only two patients out of 13 died, one ofthem mesoblastic nephroma that received additional as a consequence of a postoperative complica- treatment after surgery. He remained well with tion four days after surgery. The other child no evidence ofrecurrence at five year follow up. died nine months after the nephrectomy with Congenital mesoblastic nephroma enjoys an uncontrollable ascites and clinical evidence of excellent prognosis and, as several studies have progressive disease, although neither histology shown, patients do very well with surgery nor cytology was obtained at the time of this alone.' 20 1 Beckwith first postulated,' later sup- probable recurrence and a post mortem exam- ported by. others5820 certain characteristics ination was not performed. which might indicate the need for additional Eleven patients were still alive and well at treatment after surgery. It was suggested that follow up of between one year and five months patients over three months of age, who have to 25 years (median five years nine months). residual disease and whose tumours show atypical features indicating more aggressive behaviour, may benefit from more extensive Discussion treatment. In our 30 year series congenital mesoblastic What makes this tumour vary in its biological nephroma represented 6-5% of all renal behaviour is unknown. There are differences in tumours. This incidence seems to be higher the macroscopic and histological appearances than has previously been reported by the which have been well documented,5 showing http://jcp.bmj.com/ National Wilms' Tumor Study'9 and in more aggressive features such as fleshy areas, England (H B Marsden, personal communica- haemorrhage and necrosis, hypercellularity tion). and high mitotic index. It is not as yet possible There were no familial cases, nor any sugges- on the basis of these characteristics, however, tion of predisposing factors in our group. Only to predict which tumours will behave more one patient was found to have a congenital agggressively and which will recur or meta- abnormality (horse-shoe kidney) which corre- stasise. Furthermore, despite recommenda- on September 30, 2021 by guest. Protected copyright. lates with the findings in other series.20 tions for the use of chemotherapy or radio- We found an almost equal male:female dis- therapy, or both, in some cases and the fact that tribution with 5:4 for the conventional congen- some tumours have recurred after inadequate ital mesoblastic nephroma group and 2:1 for or no chemotherapy2 while others have the atypical cellular mesoblastic group. It is achieved complete remission with chemo- interesting to note that two of the patients were therapy,20 the role of adjuvant treatment is not over 1 year old at presentation, one of them clear. Whether congenital mesoblastic with a congenital mesoblastic nephroma of the nephroma and its variant respond to chemo- conventional type. This patient also had the therapy or radiotherapy has not been convinc- ingly proved. Figure 2 Aneuploid DY:H051.DIS NF 24-2-89..003 Counts= 20480 As DNA ploidy has shown a good correlation tumour in atypical cellular HS=256 VS=1024 with prognosis in Wilms' tumour'4 we mesoblastic nephroma '5 showing the second peak decided to study the DNA content in congeni- with the hyperdiploid tal mesoblastic nephroma. Two of the tumours DNA content. were clearly aneuploid, showing two gO/gl peaks (fig 2). The histology for these was cellular and mixed, respectively. Of the other 0 two cases of atypical cellular mesoblastic

E nephroma, one had inadequate material for z study and the other was diploid. Unfortu- nately, the number ofpatients was too small to reach any definite conclusion. It is possible to speculate, however, that an increase in the DNA content may produce changes in the biology of the cellular variant. 320 Barrantes, Toyn, Muir, Parkes, Raafat, Cameron, Marsden, Mann

When we examined the outcome of the numbers of patients with congenital meso- patients with the atypical cellular mesoblastic blastic nephroma are required to clarify this. nephroma variant, we found that all four were alive with no evidence of but it is We thank the West Midlands Regional Health Authority, the recurrence, J Clin Pathol: first published as 10.1136/jcp.44.4.317 on 1 April 1991. Downloaded from Special Trustees of the Former United Birmingham Hospitals, notable that all received further treatment in and the British Council for financial support; Ms C Roginsky for addition to surgery (table 1). help with access to data in the Birmingham and West Midlands Regional Cancer Registry; and Dr G J Draper and Mr C A If the changes in DNA reflect a highly Stiller for access to data held by the Children's Cancer Research proliferative tumour with more aggressive Group. behaviour, but at the same time better response Thanks are also due to Mrs K Smith for technical support and to treatment, as has been proposed for neuro- to numerous medical records officers, consultant colleagues, and general practitioners for access to patients' medical records blastoma2223 and acute lymphoblastic leu- and histopathological material. kaemia,24 then aneuploidy would not neces- represent a poor prognostic feature for sarily 1 Howell CA, Othersen HB, Kiviat NE, et al. Therapy and congenital mesoblastic nephroma, as long as outcome in 51 children with congenital mesoblastic treatment in addition to surgery nephroma. A report of the NWTS. J Pediatr Surg 1982; appropriate 17:826-31. were given. If this were so, flow cytometry 2 Joshi VV, Kay S, Milsten R, et al. Congenital mesoblastic could become an important tool in defining nephroma in infancy. Report of a case with unusual clinical behavior. Am J Clin Pathol 1973;60:811-6. which patients would benefit from other meth- 3 Fu YS, Kay S. Congenital mesoblastic nephroma and its ods of treatment, and even to predict which recurrence. Arch Pathol 1973;96:66-70. 4 Walker D, Richard GA. Fetal hamartoma of the kidney. could sustain recurrence or develop metastases. Recurrence and death of the patient. J Urol 1973;110: Pettinato et al performed flow cytometric 352-3. 5 Joshi VV, Kasznica J, Walters T. Atypical mesoblastic studies in three patients with congenital nephroma: pathological characteristics of a potentially mesoblastic nephroma, one with a conventional aggressive variant of congenital mesoblastic nephroma. Arch Pathol Lab Med 1986;110:100-6. type, one with atypical cellular mesoblastic 6 Steinfield AD, Crowley CA, O'Shea PA, Tefft M. Recurrent nephroma, and on both components of a mixed and metastatic mesoblastic nephroma in infancy. J Clin Oncol 1984;2:956-60. tumour.25 The conventional and atypical 7 Beckwith JB. Mesenchymal renal neoplasm of infancy cellular mesoblastic nephroma types showed revisited. J Pediatr Surg 1974;9:803-5. 8 Gonzalez-Crussi F, Sotelo-Avila C, Kidd JH. Mesenchymal diploid DNA content, but the cellular area of renal tumours in infancy: a reappraisal. Hum Pathol 1981; the mixed tumour was aneuploid. He con- 12:78-85. 9 Marsden HB, Newton WA. New look at mesoblastic cluded that cytogenetics and ploidy analysis nephroma. J Clin Pathol 1986;39:508-13. did not seem to be beneficial in the selection of 10 Snyder HM, Lack EE, Chetty-Bektaviziam A, et al. Con- genital mesoblastic nephroma: relationship to other renal patients who will need additional treatment, tumors in infancy. J Urol 1981;126:513-6. although the number of cases included in his 11 Knudson AG Jr, Strong LC. Mutation and cancer: a model for Wilms' tumor of the kidney. JNCI 1972;48:313-24. analysis was too small to define their role 12 Kumar S, Carr T, Marsden HB, Calabuig-Crespo MC. accurately. It is interesting that his findings Study of childhood renal tumours, using antisera to fibronectin, laminin and epithelial membrane antigen. were similar to ours. J Clin Pathol 1986;39:51-7. One patient in our series with congenital 13 Haas JE, Bonadio JF, Beckwith JB. Clear cell sarcoma ofthe

kidney with emphasis on ultrastructural studies. Cancer http://jcp.bmj.com/ mesoblastic nephroma ofthe conventional type 1984;54:2978-87. and no adverse features died apparently with 14 Douglas EC, Look T, Webber B, et al. Hyperdiploidy and chromosomal rearrangements define the anaplastic variant progressive disease. He was initially diagnosed of Wilms' tumor. J Clin Oncol 1986;4:975-81. as having Wilms' tumour and received radio- 15 Kumar S, Marsden HB, Cowan RA, Barnes JM. Prognostic relevance ofDNA content in childhood renal tumours. Br therapy to the tumour bed. He subsequently J Cancer 1989;59:291-5. developed ascites and finally died nine months 16 Forewell VT, D'Angio GT, Breslow N, Norkool P. Retro- spective validation of a new staging system for Wilms' after diagnosis. Unfortunately there was no tumor. Cancer Clin Trials 1981;4:167-71. histological confirmation of the recurrence and 17 Hedley DW, Friedlander ML, Taylor IW, et al. Method for on September 30, 2021 by guest. Protected copyright. analysis of cellular DNA content of paraffin-embedded a post mortem examination was not carried out. pathological material using flow cytometry. J Histochem This is a most interesting case and could Cytochem 1983;31:1333-5. 18 Frierson HF. Flow cytometric analysis of ploidy in solid represent a patient with an aggressive congen- neoplasms: Comparison of fresh tissues with formalin- ital mesoblastic of the conventional fixed paraffin embedded specimens. Hum Pathol 1988;19: nephroma 290-4. type with fatal outcome. Without pathological 19 D'Angio GD, Evans A, Breslow N, et al. The treatment of evidence, however, this must remain specula- Wilms' tumor: Results of the second National Wilms' Tumor Study. Cancer 1981;47:2302-1 1. tion. 20 Chan HS, Cheng MY, Maucer K, et al. Congenital meso- Our series illustrates some ofthe characteris- blastic nephroma: a clinicoradiologic study of 17 cases representing the pathological spectrum of the disease. tics of congenital mesoblastic nephroma and J Pediatr 1987;111:64-70. its variant, atypical cellular mesoblastic 21 Bolande RP. Congenital mesoblastic nephroma in infancy. Perspect Pediatr Pathol 1973;1:227-50. nephroma. It is clear that most children with 22 Look T, Hayes A, Nitscht R, et al. Cellular DNA content as a this tumour will do well with surgery alone. For predictor of response to chemotherapy in infants with unresectable neuroblastoma. N Engl J Med 1984;311: the small proportion ofpatients whose tumours 231-5. show more aggressive behaviour it seems 23 Oppedal BR, Storm-Mathisen I, Lie SO, Brandtzaeg P. Prognostic factors in neuroblastoma: clinical, histopatho- logical to use additional chemotherapy or logical and immunohistochemical features and DNA radiotherapy, or both. Whether these two ploidy in relation to prognosis. Cancer 1988;62:772-80. 24 Tsurusawa M, Katono NM, Kawai S, Fugimoto T, Moede modalities of treatment are effective in con- M. Prognostic implications of cellular DNA content in mesoblastic nephroma is not completely acute lymphoblastic leukemia. Am J Ped Hematol Oncol genital 1988;10:75-80. clear. Perhaps it is in this area that flow 25 Pettinato G, Manivel JC, Wick MR, Dehner LP. Classical cytometry might be useful in the selection of and cellular (atypical) congenital mesoblastic nephroma: A clinicopathological, ultrastructural, immunohisto- patients who will need further treatment to chemical and flow cytometric study. Hum Pathol 1989; achieve a better response. Studies of larger 20:682-90.