ORIGINAL ARTICLE
Diclehan Orhan · Gülsev Kale · Melda Çağlar · Safiye Göğüş · Ergun Karaağaoğlu
Histone mRNA in situ hybridization and Ki 67 immunohistochemistry in pediatric adrenocortical tumors
Received: 26 October 2005 / Revised: 9 December 2005 / Accepted: 22 December 2005 / Published online: 18 February 2006 C Springer-Verlag 2006 ☐
Abstract Adrenocortical tumors in the pediatric popula- tion are rare. Classification of these tumors as adenomas or carcinomas using histological criteria is often difficult. Immunohistochemical expressions of proliferative markers are currently under investigation for utilization in the differential diagnosis and prediction of clinical outcomes. The value of histone proteins as prognostic markers in adrenocortical tumors has not yet been elucidated. We evaluated the histological features, immunohistochemical staining of Ki 67, and in situ hybridization for histone mRNA in 30 pediatric adrenocortical tumors. We investi- gated the relationship between these parameters and the prognosis. Using the classification proposed by Weiss, 19 tumors were classified as carcinomas and 11 as adenomas. Ki 67 and histone mRNA labeling indices (LIs, the percentage of Ki 67-positive and histone mRNA-positive tumor cells, respectively) were significantly higher in carcinomas than in adenomas (Ki 67 LI was 14.62±5.79 in adenomas and 20.35±6.23 in carcinomas, p=0.02. Histone mRNA LI was 1.73±1.71 in adenomas and 6.62±2.28 in carcinomas, p=0.00). The proliferative activity assessed by histone mRNA was lower than that assessed by Ki 67 in both diagnostic groups. The cut off point for the diagnosis of malignancy was found to be 14.55 for Ki 67 LI and 5.75 for histone mRNA LI. A correlation was found between a histone mRNA LI ≥5 and poor prognosis (recurrence, metastasis, or death). We concluded that the proliferative activity of the tumor assessed by Ki 67 and histone mRNA may assist in differentiating adrenocortical adenomas and
D. Orhan ☒ Division of Pediatric Pathology, Hacettepe University Faculty of Medicine, İhsan Doğramacı Children’s Hospital, 06100 Sıhhiye/Ankara, Turkey e-mail: diclehan@hacettepe.edu.tr Tel .: +90-312-3051304 Fax: +90-312-3118226
(X) · G. Kale · M. Çağlar · S. Göğüş
E. Karaağaoğlu Department of Biostatistics, Hacettepe University Faculty of Medicine, Ankara, Turkey
carcinomas. In addition, our results suggest that the most reliable parameter to predict prognosis in pediatric adre- nocortical tumors is the histone mRNA LI.
Keywords Adrenocortical tumors · Pediatric · Ki 67 . Histone mRNA · Prognosis · In situ hybridization
Introduction
Adrenocortical tumors are rare in children with an incidence of 0.3 cases per 1,000,000 children per year, and these tumors represent 0.2% of all childhood tumors [17]. The distinction between benign and malignant adrenocortical tumors is difficult [6, 17]. Despite histolog- ical criteria by which such a distinction may be possible, the behavior of adrenocortical tumors in childhood is usually unpredictable, and distinction based only on histological classification is difficult. Many reports con- clude that the adrenocortical tumors in children and adolescents differ from those occurring in adults, both clinically and histologically [3, 17, 19, 25]. Due to the low incidence of these tumors in childhood, there is paucity of data on the histological and molecular markers that could help in differential diagnosis or in predicting clinical outcomes. In recent studies, the proliferative activity of tumor cells is being investigated to assess their role in differentiating benign and malignant tumors. For this purpose, immunohistochemical expressions of Ki 67 and PCNA were evaluated in many human tumors. The diagnostic and prognostic value of the immunohistochem- ical Ki 67 expression in adrenocortical tumors remains unclear [11, 17]. A recent method for the identification of proliferating cells in tissue sections is the use of in situ hybridization for the localization of histone mRNA. Nonisotopic histone mRNA in situ hybridization allows the detection of cells in the S phase of the cell cycle in paraffin-embedded tissues [5, 15]. There have been studies evaluating the in situ hybridization of histone mRNA in cerebral astrocytomas, head and neck tumors, and benign tissues [2, 5, 16]. To our knowledge, in situ hybridization
of histone mRNA in adrenocortical neoplasms has not undergone investigation as a diagnostic tool and a prog- nostic marker.
The aim of this study was to evaluate the relevance of the immunohistochemical staining with Ki 67 and non isotopic in situ hybridization of histone mRNA in the differential diagnosis and clinical course of pediatric adrenocortical tumors.
Materials and methods
Pathological examination
Thirty pediatric adrenocortical tumors diagnosed in Hacet- tepe University, Faculty of Medicine, Children’s Hospital, Division of Pediatric Pathology, were included in the study group. All patients had undergone a minimum of 4 years of clinical follow-up. All cases were reviewed (DO) and classified as either adenoma or carcinoma using the histo- pathological criteria according to the methods of Weiss [26]. Hematoxylin and eosin-stained histological sections of each case were evaluated for nine histopathological features. The nuclear grade was assessed using the criteria of Fuhrman et al. [8]. Mitosis was evaluated by counting at random 50 high ☒ power fields (hpf; magnification at ×40 with a ×10 objective lens using an Olympus BX51 microscope) in the area of the greatest number of mitotic figures. Presence of atypical mitosis was also noted. The percentage of clear cells resembling the normal zona fasciculata was evaluated. Architecture was defined as diffuse if more than one-third of the tumor formed sheets of cells with no pattern. Necrosis was considered to be positive when it involved confluent nests of cells. Invasion of sinusoids (endothelium-lined vessels with no smooth muscle in the wall) and veins (vessels with smooth muscle in the wall) was recorded. Capsular invasion was scored as present when nests or cords of tumor extended into or through the capsule.
A tumor with two or fewer of the criteria below was classified as an adenoma; tumors with three or more criteria were considered carcinomas:
1. High nuclear grade (III or IV) assessed using the criteria of Fuhrman et al
2. Mitotic rate >5/50 hpf
3. Presence of a typical mitosis
4. Clear tumor cell cytoplasm (<25% of the tumor cells)
5. Diffuse architecture
6. Necrosis
7. Venous invasion
8. Sinusoidal invasion
9. Capsular invasion
Immunohistochemistry
Immunohistochemical staining was performed using the primary antibody against Ki 67 (Neomarkers, CA, USA; clone: MIB-1, dilution:1/100) with streptavidin-biotin
peroxidase method (Dako LSAB 2 System, Peroxidase, CA, USA). Pretreatment in citrate buffer was performed for antigen retrieval by heating in a microwave oven. For negative control, primary antibody was replaced with normal mouse serum. Sections of tonsils were used as positive control. All tumor cell nuclei with dark brown staining, even those with vague speckling patterns, were considered to be Ki 67-positive. Ki 67 counts were performed according to these staining criteria. The count of Ki 67-positively stained nuclei was performed in areas of greatest labeling density within each tumor. About 1,000 tumor cells were counted at a magnification of ×400. Ki 67 labeling index (LI) was defined as the number of Ki 67- positive tumor cell nuclei per 100 tumor cells.
Nonisotopic in situ hybridization assay
For the nonisotopic in situ hybridization, 10-um sections were obtained from paraffin-embedded tissue blocks and placed on 3-aminopropyltriethoxy-sialine slides (Novocastra Laboratories, New Castle upon Tyne, UK). A commercial fluorescein-labeled oligonucleotide cocktail was used for the detection of histone H2B, H3, and H4 mRNA sequences (Novocastra Laboratories). The nonisotopic in situ hybrid- ization protocol was the same as that recommended by the manufacturer. Tissue sections were deparaffinized in xylene, rehydrated through ethanol series and immersed in distilled water. The sections were then treated with proteinase K (Novocastra Laboratories) (20 µg/ml in 5 mM Tris-HCl buffer, pH 7.6) for 30 min at 37℃. After washing in distilled water and dehydrating through 95% and 99% ethanol, the sections were air dried and hybridization with fluorescein isothiocyanate (FITC)-labeled cocktail was performed in the supplied buffer overnight at 37℃. The Novocastra histone probe ISH kit (Novocastra Laboratories), which included alkaline phosphatase conjugated rabbit Fab anti-FITC, enzyme substrate (5-bromo-4-chloro-3-indolyl-phosphate and nitro blu tetrazolium), and an inhibitor (levamisole), was used for the detection of histone mRNA-positive cells. All solutions and equipment used for the stages up to the hybridization were treated with H2O2 and diethyl pyrocar- bonate to remove ribonuclease, and all procedures were carried out while wearing gloves.
Histone mRNA-expressing tumor cells were counted by light microscope at ×400 magnification. The total number of cells counted was 1,000. Histone mRNA LI was defined as the number of histone mRNA-positive tumor cells per 100 tumor cells.
Statistics
All statistical analyses were performed with the Statistical Package of Social Sciences for Windows 12.0. A p value <0.05 was considered statistically significant. An indepen- dent t test was used to determine whether Ki 67 LI and histone mRNA LI differed significantly between adenoma- carcinoma groups and prognostic groups.
| Nuclear grade III, IV | Mitosis (>5/50 hpf) | Abnormal mitosis | Clear cells (<25%) | Diffuse architecture | Necrosis | Capsular invasion | Venous invasion | Sinusoidal invasion | |
|---|---|---|---|---|---|---|---|---|---|
| Carcinoma (n=19) | 17 | 11 | 10 | 16 | 15 | 17 | 17 | 11 | 12 |
| Adenoma (n=11) | 9 | 4 | 0 | 2 | 1 | 2 | 0 | 0 | 0 |
Receiver operating characteristics (ROC) curve analyses were used to determine how well Ki 67 LI and histone mRNA LI discriminate between adenoma and carcinoma patients. The prognostic value of Ki 67 LI and histone mRNA LI was also analyzed by ROC curve analysis. Comparison of Ki 67 LI and histone mRNA LI was made using the area under the ROC curves. The best cut off points for both diagnostic and prognostic groups were found at the points where the sum of false positive and false negative rates was at the minimum value.
Results
Based on the histopathological classification proposed by Weiss, 19 tumors were classified as adrenocortical carcinomas and 11 as adenomas. Microscopic findings of the tumors are summarized in Table 1. All adenomas had two or fewer Weiss criteria. No single histological criterion could be used to distinguish a carcinoma from an adenoma. However, none of the adenomas showed abnormal mitosis, capsular, venous, or sinusoidal invasion.
Among the patients with adenoma, one had local recurrence during follow-up. Three of the patients that had carcinomas were alive with no evidence of disease, and the others showed poor prognoses (recurrence, metastasis, or death).
The staining of nonisotopic histone mRNA in situ hybridization was localized to the cytoplasm and nuclear membrane (Fig. 1). Immunohistochemical expression of Ki 67 was observed as nuclear immunostaining (Fig. 2). There
was a statistically significant difference of Ki 67 LI between adenoma (mean 14.62±5.79) and carcinoma (20.35±6.23) groups (p=0.02). Histone mRNA LI also differed significantly between adenoma (1.73±1.71) and carcinoma (6.62±2.28) groups (p=0.00) (Table 2). Individ- ual results of Ki 67 and histone mRNA labeling are shown in Table 3.
To evaluate the discriminatory power of histone mRNA and Ki 67, the area under the ROC curves was evaluated. The area under the ROC curve for Ki 67 LI was 0.78 (p=0.012), and for histone mRNA LI, the area was 0.957 (p=0.00), indicating that there is a statistically significant difference between the two parameters (p=0.049) (Table 4).
The best cut off point for the diagnosis of malignancy was found to be 14.55 for Ki 67 LI and 5.75 for histone mRNA LI. Considering these cut off points, the sensitivity and the specificity of Ki 67 LI were 94.7% and 63.6%, respectively. A histone mRNA LI ≥5.75 was significant for the diagnosis of malignancy with 84.2% sensitivity and 100% specificity (Fig. 3).
Both Ki 67 and histone mRNA LIs were significantly higher in tumors with poor prognosis. Mean Ki 67 LI and histone mRNA LI with standard deviations for good and poor prognostic groups are shown in Table 5. The cut off point for poor prognosis was also determined for both Ki 67 and histone mRNA LIs. Tumors with a Ki 67 LI ≥12.65 showed poor prognosis with 100% sensitivity and 38.5% specificity. As a prognostic marker, the best cut off point for histone mRNA LI was 5. Ki 67 turned out to be almost significant when the area under the ROC curve was considered. Histone mRNA appeared to be a gold standard
| Diagnosis | n | Mean | Standard deviation | ||
|---|---|---|---|---|---|
| Ki 67 LI | Adenoma | 11 | 14.62 | 5.79 | t=2.49, p=0.02 |
| Carcinoma | 19 | 20.35 | 6.23 | ||
| Histone mRNA LI | Adenoma | 11 | 1.73 | 1.71 | t=6.14, p=0.00 |
| Carcinoma | 19 | 6.62 | 2.28 |
test for prediction of prognosis (Fig. 4). The difference between the discriminatory power of the two parameters was also statistically significant (p=0.00) (Table 6).
Three patients with carcinoma who showed good prognoses had Ki 67 LIs (18.8, 19.6, and 16.4) greater than the cut off point for Ki 67 LI, but the histone mRNA LIs of these tumors were all lower than the cut off point for mRNA LI (1.4, 1.2, and 4.6). An adenoma with local recurrence had a Ki 67 LI of 13.1 and a histone mRNA LI of 5.4 (higher than the cut off point for poor prognosis for histone mRNA LI).
| Patient no | Diagnosis | Ki 67 LI | Histone LI |
|---|---|---|---|
| 1 | Carcinoma | 15.2 | 7.3 |
| 2 | Carcinoma | 23.6 | 8.4 |
| 3 | Carcinoma | 24.1 | 6.2 |
| 4 | Carcinoma | 15.8 | 9.3 |
| 5 | Carcinoma | 18.8 | 1.4 |
| 6 | Carcinoma | 19.6 | 1.2 |
| 7 | Carcinoma | 19.9 | 6.3 |
| 8 | Carcinoma | 19.3 | 7.6 |
| 9 | Carcinoma | 38.0 | 10.2 |
| 10 | Carcinoma | 14.9 | 6.1 |
| 11 | Carcinoma | 29.7 | 8.7 |
| 12 | Carcinoma | 15.3 | 6.1 |
| 13 | Carcinoma | 19.8 | 7.3 |
| 14 | Carcinoma | 17.1 | 6.4 |
| 15 | Carcinoma | 17.3 | 7.1 |
| 16 | Carcinoma | 29.2 | 8.1 |
| 17 | Carcinoma | 13.0 | 6.2 |
| 18 | Carcinoma | 19.8 | 7.4 |
| 19 | Carcinoma | 16.4 | 4.6 |
| 20 | Adenoma | 11.2 | 0.1 |
| 21 | Adenoma | 8.4 | 0.3 |
| 22 | Adenoma | 9.7 | 1.1 |
| 23 | Adenoma | 13.1 | 5.4 |
| 24 | Adenoma | 18.3 | 2.4 |
| 25 | Adenoma | 16.5 | 3.2 |
| 26 | Adenoma | 26.2 | 3.7 |
| 27 | Adenoma | 22.6 | 0.2 |
| 28 | Adenoma | 14.2 | 1.1 |
| 29 | Adenoma | 12.3 | 0.7 |
| 30 | Adenoma | 8.4 | 0.9 |
Discussion
Despite the presence of well-established histological criteria in the differential diagnosis of adrenocortical tumors in adults, it may be extremely difficult to distinguish between benign and malignant tumors [7, 14]. Weiss, Hough et al. and Van Slooten et al. proposed a variety of parameters to differentiate these tumors [11, 23, 26]. Aubert et al. confirmed the value of the Weiss system [1]. Because of the lack of histological criteria for the discrimination of adrenocortical adenomas and carcinomas in the pediatric population, the current classifications of adrenocortical tumors based on adult criteria are used in pediatric patient studies [17, 18, 24]. We too applied the Weiss criteria to classify the tumors in our study group as adenomas or carcinomas. Pediatric adrenocortical tumors are histologically different from those in adults [4, 12, 17, 19]. Although Weieneke et al. found only 31% of histologically malignant pediatric tumors behaving badly, Sawin and colleagues reported the long-term survival rate of children with adrenocortical carcinomas to be between 10 and 46%. Complete resection was associated with an improved survival rate (67%) despite a 38% local recur- rence rate. Mortality, recurrence, and metastasis rates of the current series are within the limits mentioned in literature.
Weiss et al. reported that in pediatric patients, adreno- cortical adenomas showed mitotic activity, broad fibrous bands, and nuclear pleomorphism more frequently than in adult adenomas [26, 27]. Supporting this knowledge, in our series, 4 of 11 adenomas had mitosis >5/50 hpf and 2 had necrosis. Similar to the results of the study by Weieneke and colleagues, there was no adenoma with capsular or vascular invasion in our series. As reported previously, we found no single histologic feature that reliably identified malignancy, either. All adenomas but one in our study diagnosed according to the Weiss criteria had a good clinical outcome. One patient with adenoma showed local recurrence during follow-up. Three patients with diagnosis of adrenocortical carcinoma were alive with no evidence of disease, and the others had poor prognoses. These results suggest that the Weiss criteria were not very valid in all pediatric adrenocortical tumors and that other parameters are also needed to predict the clinical outcomes of these tumors.
| Area | Standard error | p | |
|---|---|---|---|
| Ki 67 LI | 0.78 | 0.098 | 0.012 |
| Histone mRNA LI | 0.95 | 0.033 | 0.000 |
1,0
,8
,5
Sensitivity
,3
Histone mRNA LI ☐
0,0
☐ KI 67 LI
0,0
,3
,5
,8
1,0
1 - Specificity
Reports on the role of molecular markers, such as cellular DNA content and proliferative activity of tumor cells, in the differential diagnosis and prognoses of adrenocortical tumors are seen in literature. There are several studies suggesting flow cytometric DNA analysis to be a reliable prognostic factor in adrenocortical tumors [3, 21]. Some other studies, however, indicate that ploidy is not a significant indicator of histological diagnosis or prognosis in these tumors [6, 14].
Another approach to make differential diagnosis and to find reliable prognostic factors for adrenocortical tumors involves the use of immunohistochemical proliferation markers such as Ki 67. Ki 67 is a nuclear protein expressed by proliferating cells in G1, G2, S, and M phases of the cell cycle. Ki 67 expression can be observed immunohisto- chemically in paraffin-embedded material by using the primary antibody MIB 1 [7, 9, 13, 16, 19, 20, 24]. Goldblumet et al. and Vargas et al. found a strong correlation between Ki 67 expression and the malignancy of adrenocortical tumors in adults [9, 22]. Vargas et al. showed that the tumor’s proliferative activity, determined by the percentage of Ki 67-positive tumor cells, was 1.49% in adenomas, 20.8% in carcinomas, and 16.1% in recurrent and metastatic tumors [22]. Aubert and colleagues also observed a statistically significant difference in Ki 67 labeling between adenomas (2.4±1.3%) and carcinomas (21.2±18.44%) [1]. Weieneke et al. and Sbragia and colleagues have investigated Ki 67 expression in pediatric adrenocortical tumors, and they have reported that Ki 67 immunostaining did not have a value in predicting clinical outcomes [19, 25]. In our series, we found that Ki 67 expression was significantly higher in carcinomas than
| Prognosis | n | Mean | Standard deviation | ||
|---|---|---|---|---|---|
| Ki 67 LI | Good | 13 | 15.58 | 5.53 | t=2.04, p=0.05 |
| Poor | 17 | 20.30 | 6.76 | ||
| Histone | Good | 13 | 1.60 | 1.42 | t=11.38, p=0.00 |
| mRNA LI | Poor | 17 | 7.30 | 1.30 |
1,0
,8
,5
Sensitivity
,3
☐ Histone mRNA LI
0,0
☐
KI 67 LI
0,0
,3
,5
,8
1,0
1 - Specificity
adenomas, and a Ki 67 LI ≥14.55 was found to be an indicator of malignancy for pediatric adrenocortical neoplasms. The assessment of cells in the S phase of the cell cycle is an alternative method for evaluating proliferating cells [2, 5, 10]. In situ hybridization of histone H3 and/or H4 mRNA is an assay used to detect S-phase cells in archival paraffin-embedded material [15, 16]. Histones are a class of nuclear proteins that play an essential role in the packing and expression of the eukaryotic genome. Synthesis of histone protein is restricted to the S phase. During the S phase of the cell cycle, the level of histone mRNA increases over 50-fold and rapidly disappears at the start of G2. The rapid degradation of the histone mRNA after the S phase provides an opportunity to distinguish the proliferating cells in S phase by in situ hybridization for histone mRNA [2, 10, 15]. Rauttiainen et al. reported that nonisotopic histone mRNA in situ hybridization is a reproducible method for measuring the proliferative activity of tumor cells and an independent prognostic factor in astrocytomas [16]. In our study, we found that all carcinomas had a higher histone mRNA LI than adenomas, and that a histone mRNA LI ≥5.75 was significant for malignancy with 84.2% sensitivity and 100% specificity. Both Ki 67 and histone mRNA LIs exhibited a striking difference between adenomas and carcinomas. In our series, the number of Ki67-positive cells was greater than histone mRNA-bearing cells both in adenomas and carcinomas. This may be because Ki 67 is expressed not only in the S phase of the cell cycle, but also in G1, G2, and M phases. In contrast, histone mRNA in situ hybridization labels the cells only in the S phase [16]. Ki 67 LI did not demonstrate a sensitivity or specificity of 100% in the differential diagnosis. However, histone mRNA showed 100% specificity for both differential diagnosis and clinical outcome.
| Area | Standard error | p | |
|---|---|---|---|
| Ki 67 LI | 0.706 | 0.098 | 0.057 |
| Histone mRNA LI | 1.000 | 0.000 | 0.000 |
A single patient with an adenoma diagnosed according to the histological criteria showed local recurrence, and this patient’s tumor had a histone mRNA LI >5. All other adenomas had histone mRNA LIs lower than 5. Three patients having carcinomas with histone mRNA LIs of less than 5 had disease-free survivals of at least 4 years. All these carcinoma patients had Ki 67 LIs higher than the cut off point for Ki 67 LI for poor prognosis. These data could indicate that children with adrenocortical tumors with histone mRNA LIs of less than 5 may have good prognoses. Statistical analysis showed that nonisotopic in situ hybridization of histone mRNA was a better predictor of prognosis in pediatric adrenocortical tumors than immunohistochemical Ki 67 expression.
The results of this study indicate that the assessment of Ki 67 and histone mRNA LIs complements histological criteria in the diagnosis of malignancy in adrenocortical tumors. In addition, histone mRNA LI may also prove to be a valuable marker in predicting the clinical outcomes of pediatric patients with adrenocortical neoplasms.
Acknowledgements This study was supported by the Scientific research Project of Hacettepe University (02G013). The authors are grateful to Ayşe Yüzbaşıoğlu and Köksal Özgül for their technical assistance.
References
1. Aubert S, Wacreiner A, Leroy X, Devos P, Bruno C, Proye C, Wemeau JL, Lecomte-Houcke M, Leteurtre E (2002) Weiss system revisited. A clinicopathological and immunohistochem- ical study of 49 adrenocortical tumors. Am J Surg Pathol 26:1612-1619
2. Bosch FX, Udvarhelyi N, Venter E, Herold-Mende C, Schuhmann A, Maier H, Weidauer H, Born AI (1993) Expression of the histone H3 gene in benign semi-malignant lesions of the head and neck: a reliable proliferation marker. Eur J Cancer 29A:1454-1461
3. Bowlby LS, Bebault LE, Abraham SR (1986) Flow cytometric analysis of adrenal cortical tumor DNA: relationship between cellular DNA and histopathologic classification. Cancer 58:1499-1505
4. Cagle PT, Hough AJ, Pysher TJ, Johnson EH, Kirkland RT, Holcombe JH, Hawkins EP (1986) Comparison of adrenal cortical tumors in children and adults. Cancer 57:2235-2237
5. Chou MY, Chang ALC, McBride J, Donoff B, Gallagher GT, Wong DTW (1990) A rapid method to determine proliferation patterns of normal and malignant tissues by H3 mRNA in situ hybridization. Am J Pathol 136:729-733
6. Cibas ES, Medeiros LJ, Weinberg DS, Gelb AB, Weiss LM (1990) Cellular DNA profiles of benign and malignant adrenocortical tumors. Am J Surg Pathol 14(10):948-955
7. DeLellis RA, Mangray S (2004) The adrenal glands. In: Mills SE, Carter D, Greenson JK, Oberman HA, Reuter V, Stoler MH (ed) Sternberg’s diagnostic surgical pathology. Lippincott Williams & Wilkins, Philadelphia, pp 622-677
8. Fuhrman SA, Lasky LC, Limas C (1982) Prognostic sig- nificance of morphologic parameters in renal cell carcinoma. Am J Surg Pathol 6:655-663
9. Goldblum JR, Shannon R, Kaldjian EP, Thiny M, Davenport R, Thompson N, Lloyd LV (1993) Immunohistochemical assess- ment of proliferative activity in adrenocortical neoplasms. Mod Pathol 6:663-668
10. Hayama M, Ota H, Toki T, Ishii K, Honda T, Momose M, Nakata R (2002) Cell kinetic study of the endometrium by nonisotopic in situ hybridization for histone H3 messenger RNA and immunohistochemistry for Ki-67 and for estrogen and progesterone receptors. Anat Rec 266:234-240
11. Hough AJ, Hollifield JW, Page DL, Hartmann WH (1979) Prognostic factors in adrenal cortical tumors. A mathematical analysis of clinical and morphologic data. Am J Clin Pathol 72:390-399
12. Liou LS, Kay R (2000) Adrenocortical carcinoma in children: review and recent innovations. Urol Clin North Am 27: 403-421
13. Martins ACP, Cologna AJ, Tucci S, Suaid HJ, Falconi RAR (2005) Clinical features and immunoexpression of p53, MIB-1 and Proliferating Cell Nuclear Antigen in adrenal neoplasms. J Urol 173:2138-2142
14. Medeiros LJ, Weiss LM (1992) New developments in the pathologic diagnosis of adrenal cortical neoplasms: a review. 97:73-83
15. Millar MR, Sharpe RM, Maguire SM, Saunders PT (1993) Cellular localisation of messenger RNAs in rat testis: applica- tion of digoxigenin-labelled ribonucleotide probes to embedded tissue. Cell Tissue Res 273:269-277
16. Rautiainen E, Haapasssalo H, Sallinen I, Helen P, Helin H (1998) Histone mRNA in-situ hybridization in astrocytomas: a comparison with PCNA, MIB 1 and mitosis in paraffin- embedded material. Histopathology 32:43-50
17. Ribeiro RC, Figueiredo B (2004) Childhood adrenocortical tumors. Eur J Cancer 40:1117-1126
18. Sawin RS (1997) Functioning adrenal neoplasms. Semin Pediatr Surg 6:156-163
19. Sbragia L, Oliveira-Filho AG, Vassallo J, Pinto GA, Guerra- Junior G, Bustorff-Silva J (2005) Adrenocortical tumors in Brazilian children: immunohistochemical markers and prog- nostic factors. Arch Pathol Lab Med 129:1127-1131
20. Stojadinovic A, Brennan MF, Hoos A, Omeroglu A, Leung DH, Dudas ME, Nissan A, Cordon-Cardo C, Ghossein RA (2003) Adrenocortical adenoma and carcinoma: histopatholog- ical and molecular comparative analysis. Mod Pathol 16:742-751
21. Taylor SR, Roederer M, Murphy RF (1987) Flow cytometric DNA analysis of adrenocortical tumors in children. Cancer 59:2059-2063
22. Vargas MP, Vargas HI, Kleiner De, Merino MJ (1997) Adrenocortical neoplasms: role of prognostic factors MIB-1, p53 and RB. Am J Surg Pathol 21:556-562
23. Von Slooten H, Schaberg A, Smeenk D, Moolenaar AJ (1985) Morphological characteristics of benign and malignant adrenal cortical tumors. Cancer 55:766-773
24. Wachenfeld C, Beuschlein F, Zwermann O, Mora P, Fassnacht M, Allolio B, Reincke M (2001) Discerning malignancy in adrenocortical tumors: are molecular markers useful? Eur J Endocrinol 145:335-341
25. Weieneke JA, Thompson LD, Heffess CS (2003) Adrenal cortical neoplasms in the pediatric population: a clinicopatho- logic and immunophenotypic analysis of 83 patients. Am J Surg Pathol 27:867-881
26. Weiss LM (1984) Comparative histologic study of 43 metastasizing and nonmetastasizing adrenocortical tumors. Am J Surg Pathol 8:163-169
27. Weiss LM, Medeiros LJ, Vickery AL Jr (1989) Pathologic features of prognostic significance in adrenocortical carcinoma. Am J Surg Pathol 13:202-206