Accepted Manuscript
Title: Analysis of histological and immunohistochemical patterns of benign and malignant adrenocortical tumors by computerized morphometry
Authors: Paolo Dalino Ciaramella, Maurizio Vertemati, Duccio Petrella, Edgardo Bonacina, Erika Grossrubatscher, Eleonora Duregon, Marco Volante, Mauro Papotti, Paola Loli
PATHOLOGY RESEARCH PRACTICE ND
Ca-sponsored by the Canadian Auncacion of ihchologie Co-sponsored by the Jiganne Sudety of Pathology
211/7
Dujiang/På China
| PII: | S0344-0338(16)30242-4 |
| DOI: | http://dx.doi.org/doi:10.1016/j.prp.2017.03.004 |
| Reference: | PRP 51768 |
| To appear in: | |
| Received date: | 11-7-2016 |
| Revised date: | 2-3-2017 |
| Accepted date: | 4-3-2017 |
Please cite this article as: Paolo Dalino Ciaramella, Maurizio Vertemati, Duccio Petrella, Edgardo Bonacina, Erika Grossrubatscher, Eleonora Duregon, Marco Volante, Mauro Papotti, Paola Loli, Analysis of histological and immunohistochemical patterns of benign and malignant adrenocortical tumors by computerized morphometry, Pathology - Research and Practice http://dx.doi.org/10.1016/j.prp.2017.03.004
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Analysis of histological and immunohistochemical patterns of benign and malignant adrenocortical tumors by computerized morphometry
Paolo Dalino Ciaramella, MD, PHD, ** Maurizio Vertemati MD, PHD,*§ Duccio Petrella, MD,1 Edgardo Bonacina, MD, Erika Grossrubatscher, MD,¡ Eleonora Duregon, MD,11 Marco Volante, MD, PhD, 11 Mauro Papotti, MD, 1] Paola Loli, MD;
* Department of Internal Medicine, Endocrinology Unit, Azienda Ospedaliera Niguarda Ca’ Granda, Milano
§ Department of Biomedical and Clinical Sciences “L. Sacco”, Milano
1 Department of Laboratory Medicine, Pathology Unit, Azienda Ospedaliera Niguarda Ca’ Granda, Milano
11 Department of Oncology, Pathology Unit, University of Torino at Azienda Ospedaliero- Universitaria San Luigi Gonzaga, Torino
For correspondence:
Dr. Paolo Dalino Ciaramella Address: Piazza Ospedale Maggiore 3, 20162, Milano - Italy Fax Number: +390264442082 or +390331592068
Telephone Number: +393284492029 (mobile) or +390264444578 (landline) Email: paolo.dalinociaramella@ospedaleniguarda.it
Work partially supported by grants from AIRC Milan to MP, no IG 14820/2013
ACCEPTED MANUSCRIPT
ABSTRACT
Diagnosis of benign and purely localized malignant adrenocortical lesions is still a complex issue. Moreover, histology-based diagnosis may suffer of a moment of subjectivity due to inter- and intra-individual variations. The aim of the present study was to assess, by computerized morphometry, the morphological features in benign and malignant adrenocortical neoplasms.
Eleven adrenocortical adenomas (ACA) were compared with 18 adrenocortical cancers (ACC). All specimens were stained with H&E, cellular proliferation marker Ki-67 and reticulin. We generated a morphometric model based on the analysis of volume fractions occupied by Ki-67 positive and negative cells (nuclei and cytoplasm), vascular and inflammatory compartment; we also analyzed the surface fraction occupied by reticulin. We compared the quantitative data of Ki- 67 obtained by morphometry with the quantification resulting from pathologist’s visual reading.
The volume fraction of Ki-67 positive cells in ACCs was higher than in ACAs. The volume fraction of nuclei in unit volume and the nuclear/cytoplasmic ratio in both Ki-67 negative cells and Ki-67 positive cells were prominent in ACCs. The surface fraction of reticulin was considerably lower in ACCs.
Our computerized morphometric model is simple, reproducible and can be used by the pathologist in the histological workup of adrenocortical tumors to achieve precise and reader- independent quantification of several morphological characteristics of adrenocortical tumors.
ACCEPTED MANUSCRIPT
INTRODUCTION
Tumors of the adrenal cortex are mostly adenomas (ACA), often found incidentally (incidentalomas) during workup for unrelated indication [1, 2]. ACA is usually a well- circumscribed nodular lesion with cells organized in alveoli, cords or nests similar to those of the normal adrenal cortex, with minimal pleomorphism and generally unremarkable mitotic activity [3].
On the other side, adrenocortical carcinoma (ACC) is a highly aggressive malignancy with an estimated worldwide prevalence of 4-12 cases per million adults and a five-year-survival ranging between 16 and 38% [4]. Microscopically, ACCs are composed of cell cords organized in large bands in a fine sinusoid network and have capsular and vascular infiltration. Necrosis and fibrosis are frequent. Nuclear pleomorphism is often prominent and the mitotic rate is quite variable [5].
Although several different scoring systems have been proposed to assess malignancy in adrenocortical tumors, Weiss score remains the most used in separating benign from malignant adrenocortical neoplasms [6]. This score counts 9 histopathologic criteria: eosinophilic (“dark”) cytoplasm in more than 75% of tumor cells, a “patternless” diffuse architecture, necrosis, nuclear atypia, mitotic index above 5 per 50 high-power fields, atypical mitoses, sinusoidal, venous, and capsular invasion [7]. An adrenocortical neoplasm is classified as malignant when it meets 3 or more criteria [8]. However, low Weiss score (2 to 3) generates ambiguous results, especially in small sized and purely localized lesions and in large tumors without invasive features or cellular atypia in which well-differentiated cells resemble those seen in ACA [9]. This “grey zone” needs further investigation to obtain a more precise characterization and to eventually reveal, if possible, an intermediate class of adrenocortical tumors. In this context, several studies already reported that ACCs with Weiss score 3 often do not recur; on the other side, adrenocortical tumors with Weiss score 2 with lung metastases have been described [10, 11].
Several immunohistochemical markers have been proposed to improve the histological recognition of malignancy. Among molecular and phenotypic markers, the expression of Ki-67, a non-histonic protein involved in DNA replication, has been indicated as a useful marker of
ACCEPTED MANUSCRIPT
malignancy for adrenocortical tumors [12, 13]. However, some overlaps between benign and malignant lesion have been described [14, 15].
Furthermore, the disruption of the reticular network has been demonstrated to be present in the majority of ACCs [16] and an algorithm including reticulin disruption and additional parameters (mitotic index, necrosis and vascular invasion) has been proposed to simplify the diagnostic workup of adrenocortical tumors [17].
Nonetheless, histology-based diagnosis may suffer of a moment of subjectivity due to inter- and intra-individual variability. This is particularly evident when certain histological features (i.e. cell density, nuclear atypia and nuclear/cytoplasmic ratio, extent of architectural disruption) need to be quantified and subsequently integrated for interpretation in a semi-quantitative way. In this setting, relevant interobserver and intraobserver variability in histopathological evaluation of Ki-67 in several tumors has been documented [18, 19, 20, 21].
To address this problem, morphometric analysis could minimize the subjectivity of the individual morphological observations.
In our study, we generated a computerized morphometric model to evaluate the morphological features of adrenocortical lesions stained with reticulin method and Ki-67. The quantitative data of Ki-67 expression obtained by computerized morphometry have also been compared with the quantification obtained by the pathologist. Furthermore, to assess the reproducibility of the morphometric evaluation, 10 randomly selected specimens were recounted by another blinded operator and compared with the initial counting.
ACCEPTED MANUSCRIPT
MATERIALS AND METHODS
In this study, we retrospectively examined 11 ACAs (median size= 6.2 cm, range 5-16 cm) and 18 ACCs (median size= 11 cm, range 4-19 cm). Specimens were obtained from the archival files of 29 patients (9M/20F, 18 ACCs and 5 ACAs from Niguarda Ca’ Granda Hospital and 6 ACAs from Azienda Ospedaliera Universitaria San Luigi Gonzaga Orbassano) submitted to adrenal surgery (Table I). According to Weiss score, it was found that all ACC but one possessed 3 or more of these criteria of malignancy, and 10 (55,5%) had six or more. The most frequent criteria seen in all the carcinomas were nuclear grade III or IV based on Fuhrman criteria (83,3%), diffuse architecture (77,7%) and clear or vacuolated cells comprising 25% or less of tumor (77,7%). One ACC case had Weiss score 2 (nuclear grade III and capsular invasion) with invasion of vena cava. ENSAT stage in ACC was I (61,1%), II (27,8%) and III (11,1%). Five patients (Weiss score 7, ENSAT III-IV) had metastases, 6 patients (Weiss score 4 to 7, ENSAT II-IV) had recurrence after treatment.
The ACA showed 0-2 criteria of malignancy (high nuclear grade was the most frequent feature).
All patients were regularly followed up after surgery.
ACAs and ACCs were defined grossly and microscopically following the criteria and the nomenclature system of pathological features proposed by Weiss et al.(3) All primary malignant adrenal tumors reviewed as part of this study demonstrated three or more of the histopathologic criteria needed for the diagnosis of ACC as defined by Weiss.
All specimens have been reviewed by two experienced pathologists blinded to clinical history or outcome.
In all cases investigated, from a representative formalin-fixed, paraffin-embedded block three consecutive 4 um thick sections were obtained.
Each section series was stained with different methods:
ACCEPTED MANUSCRIPT
o Hematoxylin-Eosin (HE) to confirm the diagnosis of adrenal nodular lesions. ☐
o Monoclonal antibodies against Ki-67 (Rabbit anti-human Ki-67 monoclonal Ab Clone ☐ SP6, 1:400 - Thermo Fisher Scientific, CA - USA) to assess volume fractions of Ki-67 positive cells [22].
o Silver impregnation for reticulin fibers using the Gordon and Sweets method to assess ☐ the tumor structural network [23].
The variables assessed by morphometry are listed in Table II and include cellular compartment, fibrous stroma, and vascular supply in both groups of lesions. The morphometric analysis was performed at two magnification levels using an interactive approach with a high- resolution computerized image analyzer (Kontron-Zeiss KS 400) that included a color video camera (JVC TK-C1381EG) attached to a light microscope equipped with a motorized stage with 10X and 40X objectives and auto-focusing software. The software system, tailored on the research needs of our team, consisted of different programs to control interactively the scanning stage and autofocus functions [24, 25, 26].
The analysis works as follows. Images acquired by video camera are displayed on the monitor. The analyzer automatically superimposes to each microscopic field different grids of points and lines, included in a test area 504 x 504 pixels, allowing an evaluation of the stereological variables [27, 28].
The observer can interactively apply techniques of enhancement for a better definition of the different structures. It is also possible to exclude fields in which the tissue section may not be suitable for analysis due to technical artifact. An algorithm automatically controls the scanning stage operation in order to avoid duplicate measurements of the same structures.
Two different grids have been used: a 144-points square grid to evaluate the volume fractions of the components investigated (Fig. 1) and a 4-lines grid to evaluate reticulin surface (Fig. 2).
ACCEPTED MANUSCRIPT
To examine the microscopic fields, 10X and 40X objectives have been used. On immunostained sections at 10X magnification, surface area of reticulin was evaluated by superimposing on each microscopic field displayed on the monitor a grid of lines: surface densities were calculated by differential intersection counting [29]. The grid was then rotated by 45°, 90°, and 135° and counts repeated each time.
On stained slices at 40X, volumetric analysis of Ki-67 negative and Ki-67 positive cells was performed by differential point counting and more than 250 microscopic fields systematically selected were examined [30]. During point counting procedure, a single experienced operator blinded to pathologist’s diagnosis or clinical history identified the different structures.
Ki-67 assessment by morphometry was compared with the histological semi-quantitative scoring by regression analysis.
Lastly, to assess the reproducibility of our morphometric model, 10 specimens were randomly selected from the entire pool of specimens, reassessed as above by another blinded operator and compared to the initial counting.
Statistical analysis
SPSS 22.0 software was used for statistical analysis.
For each parameter, a comparison between the two groups of lesions was performed by variance analysis. Regression lines were obtained by least squares method.
Statistical significance was established at the p<0.05 level. Furthermore, regression analysis was used to test reproducibility of the morphometric analysis, according to the results obtained by two blinded operators.
ACCEPTED MANUSCRIPT
RESULTS
Results are shown in Tables III-IV and in Figures 2-8.
On Ki-67 stained sections (Fig. 4 Panel A and Panel B), the volume fraction of nuclei in unit volume showed significantly higher values in ACCs, both in Ki-67 negative cells and in Ki-67 positive cells. On the other hand, the volume fraction of cytoplasm in Ki-67 negative cells was significantly higher in ACAs, whereas the volume fraction of cytoplasm in Ki-67 positive cells resulted higher in ACCs (Table III, p<0.0001).
The volume fraction of the inflammatory infiltrate and vessels (Vvother) was significantly higher in ACCs (Table III, p<0.05).
When considering the nuclear/cytoplasmic ratio (N/C), ACCs showed the highest values in both Ki-67 negative cells and Ki-67 positive cells (Table III, p<0.0001; Fig. 5).
The surface fraction of reticulin was significantly lower in ACCs when compared with ACAs (Table III, p<0.0001; Fig. 6).
Moreover we found a correlation between pathologist’s visual assessment of the percentage of Ki-67 positive cells and our computerized morphometric evaluation: the linear regression shows that, when comparing morphometric analysis to pathologist’s scores, the data of the point grid analysis revealed significantly lower values with respect to conventional histopathology evaluation (Fig. 7, p<0.001).
We found a correlation between volume fraction of Ki-67 positive cells (Vycellpos) referred to parenchymal compartment and surface fraction occupied by reticulin (SyRet) in ACA and ACC groups: only the regression line referred to ACCs was significant (Fig. 8, p< 0.05).
Summarizing our results, we could individuate some morphometric parameters that more discriminated among benign and malignant nodular lesions (Table IV).
Lastly, to evaluate the reproducibility of our morphometric model, 10 randomly selected specimens were analyzed in a blinded fashion by another operator and compared with the results of
ACCEPTED MANUSCRIPT
the initial point counting: a high inter-observer agreement in assessing the morphometric characteristics of ACAs and ACCs was found (Table V).
ACCEPTED MANUSCRIPT
DISCUSSION
In this study, morphometry and computer image analysis showed robust and powerful capabilities to evaluate and measure morphologic features of ACC that may be overlooked by routine histological examination, enabling to describe structures in quantitative terms and achieve a precise and reader-independent quantification.
Scarpelli et al. found that computer-assisted analysis of nuclear characteristics proved to be useful in identifying and describing differences between groups of adrenocortical lesions [31, 32]. Nonetheless, an overlap between the different lesions, with 20-25% of ACCs nuclei resulting in the range of ACAs, has been reported. Shirata et al. found that none of the nuclear parameters evaluated by computer-assisted image analysis discriminated between malignant and benign adrenocortical lesions in adults, whereas there was a marginal significant correlation of DNA-ploidy with clinical behaviour in children [33].
The proliferation marker Ki-67 is expressed in the nuclei of cells in the G1, S, and G2 phases of the cell division cycle and in mitosis, but is absent in quiescent or resting cells [34]. Its quantification becomes crucial when the tumor prognosis and the clinical choice depend on the Ki- 67 index [35]. The examination of at least 500 cells (optimally 1,000) is generally considered the gold standard for a correct estimation of the Ki-67 labeling index, either for histological samples or core biopsy samples [36, 37, 38, 39]. The visual microscopic reading of Ki-67 stained histological samples is a time consuming procedure, markedly operator-dependent and still lacks an adequate standardization. The use of digital images improved the reproducibility of Ki-67 expression, allowing a faster and more convenient procedure, mainly when performed with automated analyzers [40,41, 42, 43].
Ki-67 has been proposed as diagnostic sensitive marker also to discriminate between benign and malignant adrenocortical tumors, the expression resulting significantly higher in carcinomas. Additionally, a significant inverse correlation between Ki-67 expression and overall survival has been reported in patients affected by adrenal carcinoma. A Ki-67 index above 5-7% is considered a
ACCEPTED MANUSCRIPT
sensitive and specific indicator of ACC and may be useful in the differentiation of adenomas from carcinomas [12, 13, 44, 45].
Recently, a multicenter study performed in patients with localized ACCs (ENSAT stage I to III) following R0 resection, identified Ki-67 as the single most important factor predicting recurrence [46]. Despite the high number of patients, the Authors concluded that one of the limitations of the study was the variability of Ki-67 index, due to possible related bias (i.e., preanalytic variations, different antibodies and staining reagents, quantification underestimating tumor heterogeneity).
We generated a computerized morphometric model including staining with reticulin to assess tumor architecture framework and Ki-67 expression to identify and quantify the proliferating tumor cell compartment in ACCs and ACAs. To the best of our knowledge, our study is the first performed with a morphometric semi-automated computerized method on both Ki-67 and reticulin stained sections of adrenocortical tumors. In our analysis, we have examined a total of ~ 250-350 microscopic fields to make a complete scanning of the sample and to avoid potential suboptimal evaluation related to intralesional heterogeneity of Ki-67 expression.
On Ki-67 sections ACCs showed a significantly higher volume fraction of nuclei (which appears as “nuclear clouding” in histological specimens) both in Ki-67 negative cells (Vynucneg) and in Ki-67 positive cells (Vvnucpos).
In ACCs the volume fraction of cytoplasm was higher in Ki-67 positive cells (Vvcytpos), while in ACAs the volume fraction of cytoplasm was higher in Ki-67 negative cells (Vvcytneg). Overall, when compared to the ACAs cellular populations, ACCs showed the highest values of nuclei/cytoplasm ratio in both Ki-67 negative and Ki-67 positive cells. The volume fraction of Ki- 67 negative cells (Vycelneg) was higher in ACAs than in ACCs; conversely, the volume fraction of Ki-67 positive cells (Vycelpos) was higher in ACCs. The volume fractions of the other compartments (vascular structures, inflammatory infiltrate) were higher in ACCs.
ACCEPTED MANUSCRIPT
Sampling, subjectivity and interobserver variability (affected also by varying tumor cellularity, positive immunoreactivity in inflammatory cells) can be sources of pitfalls in the quantification of Ki-67 in histopathologic samples and can limit semi-quantitative grading (40); some improvements has been obtained by web-based virtual microscopy and a pathologist training program [47].
Data obtained by our morphometric analysis were compared with the pathologist’s “eyeballed” quantification of Ki-67 resulting significantly lower with respect to conventional histopathology (Figure 7). It must be underlined that, in assessing of Ki-67 labeling index in histological samples, pathologists usually try to visually estimate the proportion of Ki-67 positive cells, so measuring a cell count fraction. This estimation is not the same as measuring by morphometry the total area occupied by Ki-67 positive cells, where the results are often given as the percentage of histologic sample area. Moreover, pathologists usually count Ki-67 positive cells in the areas of greatest concentrations of positively stained cells (also known as “hotspot areas”), being these areas manually selected by pathologists using visual examination of whole mounted Ki- 67 stained section at a low magnification. This process is considered a “gold standard” in histopathology but it might lack reproducibility being operator-dependent and consequently affecting Ki-67 quantification: in this setting, an automated selection of hotspot areas based on automated segmentation has been recently proposed [48].
To control this potential source of discrepancy when the two approaches are compared, volume fraction of Ki-67 positive cells has been referred only to the cellular compartment, i.e. the reference volume, so estimating the “real” amount of Ki-67 positive cells across tumor specimen (Vvcellpos, Table III).
According to our results, all ACAs showed Ki-67 values < 1%.
One ACC case (Weiss score 2, ENSAT III stage with invasion of vena cava) showed low Ki-67 value on both computerized morphometric analysis and pathologist’s visual evaluation, fitting in the range of ACAs group. The outcome of this patient was favorable despite the staging of
ACCEPTED MANUSCRIPT
the disease. Although this is not a clinical study, this data could suggest that a low Ki-67 index could be related to an “intermediate-low aggressive” category of ACCs. To better investigate this assessment further larger studies, including multiple centres, are needed
On reticulin stains, the surface fraction of reticulin (Syret) resulted markedly reduced in ACCs when compared to ACAs (Fig. 6), moreover in ACCs there was a significant negative relationship between Syret and volume fraction of Ki-67 positive cells (Fig. 8). Our results confirm previous observations underlying that the disruption of reticulin framework could be considered a highly sensitive pathologic feature of ACC [16, 17]. As a matter of fact, figure 6 shows that one ACC had Syret well away, i.e. more than 3 times the 75-25 interquartile distance from the rest of the tumors. In our opinion this proves that morphometry may also identify outliers within a certain cell population.
Certainly, adding a morphometric parameter to semi-quantitative features, introduces numerical values that in future studies may actually help in identifying subpopulations of adrenocortical tumors (Table IV).
Lastly, another operator analyzed 10 randomly selected specimens (in different times) comparing the results to the initial counting: as from Table V, we could demonstrate a high interobserver agreement between these two morphometric evaluations, confirming the reproducibility of the method.
In conclusion, our computerized morphometric model is an efficient and simple method to quantify the morphologic characteristics of adrenocortical lesions. It is reproducible and lacks of observer or subjective bias. Therefore, it could represent an additional tool to complement conventional histological analysis and to conceivably enhance the reliability and uniformity of the diagnostic workup of adrenocortical nodular lesions. In future, the data obtained by the morphometry of ACCs could be integrated with clinical-surgical-radiological and conventional histological data (e.g. Weiss score) trying to explain the inconsistencies occasionally arising from the clinical course of the disease with respect to the histological grading and the staging; this
ACCEPTED MANUSCRIPT
“histomorphological fingerprint” might contribute to give an accurate prognostic stratification of these heterogeneous tumors.
The computerized morphometric model proposed here is intended to achieve a precise and reader-independent quantification of morphological features by using histochemical markers (i.e., Ki-67 and reticulin) in adrenocortical tumors. Our paper provides a morphologic assay that is efficient, precise and statistically rigorous and it gives a detailed and reader-independent quantification of the morphological characteristics of adrenocortical lesions particularly concerning Ki-67 and reticulin (which need semi-quantitative assessment), minimising the variability due to the subjectivity of morphological observations. Lastly, data obtained by our computerized morphometry could be integrated with clinical-surgical-radiological data and supplement the conventional histological evaluation trying to explain the inconsistencies occasionally arising from the clinical course of the disease with respect to the histological grading and the staging.
ACCEPTED MANUSCRIPT
1) Grumbach MM, Biller BM, Braunstein GD, Campbell KK, Carney JA, Godley PA, Harris EL, Lee JK, Oertel YC, Posner MC, Schlechte JA, Wieand HS, Management of the clinically inapparent adrenal mass (incidentaloma”), Ann Intern Med. 138(5) (2003) 424-429.
2) Barzon L, Sonino N, Fallo F, Palù G, Boscaro M, Prevalence and natural history of adrenal incidentaloma. Eur J Endocrinol, 149(4) (2003) 273-285.
3) Young WF Jr. Clinical practice, The incidentally discovered adrenal mass. N Engl J Med, 356(6) (2007) 601-610.
4) Allolio B, Fassnacht M, Adrenocortical carcinoma: clinical update. J Clin Endocrinol Metab, 91(6) (2006) 2027-2037.
5) Maitra A. The endocrine system. In: Kumar V, Abbas AK, eds. Robbins and Cotran Pathologic Basis of Disease. 8 th ed., Ch. 24. New Delhi, India: Elsevier; 2010, pp 1148-1161.
6) Weiss LM. Comparative histologic study of 43 metastasizing and nonmetastasizing adrenocortical tumors, Am J Surg Pathol, 8(3) (1984) 163-169.
7) Lau SK, Weiss LM, The Weiss system for evaluating adrenocortical neoplasms: 25 years later. Hum Pathol, 40(6) (.2009) 757-768.
8) Aubert S, Wacrenier A, Leroy X, Devos P, Carnaille B, Proye C, Wemeau JL, Lecomte- Houcke M, Leteurtre E, Weiss system revisited. A clinicopathologic and immunohistochemical study of 49 adrenocortical tumors. Am J Surg Pathol, 26(12) (2002) 1612-1619.
9) Grondal S, Cedermark B, Eriksson B, Grimelius L, Harach R, Kristoffersno A, Rastad J, Udèn P, Akerström G, Adrenocortical carcinoma: a retrospective study of a rare tumor with a poor prognosis. Eur J Surg Oncol. 16 (1990) 500-506.
10) Tissier F, Classification of adrenal cortical tumors: what limits for the pathological approach? Best Pract Res Clin Endocrinol Metab, 24(6) (2010) 877-885.
ACCEPTED MANUSCRIPT
11) Pohlink C, Tannapfe A, Eichfeld U, Führer D, Paschke R, Koch CA, Does tumor heterogeneity limit the use of the Weiss criteria in the evaluation of adrenocortical tumors? J Endocrinol Invest, 27 (2004) 565-569.
12) Stojadinovic A, Brennan MF, Hoos A, Hoos A, Omeroglu A, Leung DHY, Dudas ME, Nissan A, Cordon-Cardo C, Ghossein RA, Adrenocortical adenoma and carcinoma: histopathological and molecular comparative analysis. Mod Pathol, 16(8) (2003) 742-751.
13) Terzolo M, Boccuzzi A, Bovio S, Cappia S, De Giuli P, Ali A, Paccotti P, Porpiglia F, Fontana D, Angeli A, Immunohistochemical assessment of Ki-67 in the differential diagnosis of adrenocortical tumors. Urology, 57(1) (2001) 176-182.
14) McNicol AM, Assessment of malignancy in adrenal cortical tumors. Endocr Pathol, 17(2) (2006) 131-136.
15) Morimoto R, Satoh F, Murukami O, Suzuki T, Abe T, Tanemoto M, Abe M, Uruno A, Ishidova S, Arai Y, Takahashi K, Sasano H, Ito S, Immunohistochemistry of a proliferation marker Ki67/MIB1 in adrenocortical carcinomas: Ki67/MIB1 labeling index is a predictor for recurrence of adrenocortical carcinomas. Endocr J, 55(1) (2008) 49-55.
16) Volante M, Bollito E, Sperone P, Tavaglione V, Daffara F, Porpiglia F, Terzolo M, Berruti A, Papotti M, Clinicopathological study of a series of 92 adrenocortical carcinomas: from a proposal of simplified diagnostic algorithm to prognostic stratification. Histopathology, 55(5) (2009) 535-543.
17) Duregon E, Fassina A, Volante M, Nesi G, Santi R, Gatti G, Cappellesso R, Dalino Ciaramella P, Ventura L, Gambacorta M, Dei Tos AP, Loli P, Mannelli M, Mantero F, Berruti A, Terzolo M, Papotti M, The reticulin algorithm for adrenocortical tumor diagnosis: a multicentric validation study on 245 unpublished cases. Am J Surg Pathol, 37(9) (2013) 1433-1440.
18) Dhall D, Frishberg DP, Galliano G, Chiles L, Chung F, Ines D, Wang H, Interobserver variability in assessing Ki67 proliferative index in gastrointestinal well-differentiated neuroendocrine neoplasms. Mod Pathol, 22:116A (2009) 515.
ACCEPTED MANUSCRIPT
19) De Azambuja E, Cardoso F, de Castro G Jr, Colozza M, Mano MS, Durbecq V, Sotiriou C, Larsimont D, Piccart-Gebhart M J, Paesmans M, Ki-67 as prognostic marker in early breast cancer: a meta-analysis of published studies involving 12,155 patients. Br J Cancer, 96(10) (2006) 1504- 1513.
20) Stojadinovic A, Ghossein RA, Hoos A, Nissan A, Marshall D, Dudas M, Cordon-Cardo C, Jacques DP, Brennan MF, Adrenocortical Carcinoma: Clinical, Morphologic, and Molecular Characterization. J Clin Oncol, 20(4) (2002) 941-950.
21) Duregon E, Molinaro L, Volante M, Ventura L, Righi L, Bolla S, Terzolo M, Sapino A, Papotti MG, Comparative diagnostic and prognostic performances of the hematoxylin-eosin and phospho-histone H3 mitotic count and Ki-67 index in adrenocortical carcinoma. Mod Pathol. 27(9) (2014) 1246-1254.
22) Key G, Kubbutat MH, Gerdes J, Assessment of cell proliferation by means of an enzyme- linked immunosorbent assay based on the detection of the Ki-67 protein. J Immunol Methods, 28;177 (1994) 113-117.
23) Culling CFA, Allison RT, Barr WT, Cellular pathology technique. Ed. 4. London, England. Butterworths, 1985.
24) Vizzotto L, Vertemati M, Gambacorta M, Sabatella G, Spina V, Minola E, Analysis of histological and immunohistochemical patterns of the liver in posthepatitic and alcoholic cirrhosis by computerized morphometry. Mod Pathol, 15 (2002) 798-806.
25) Vertemati M, Minola E, Goffredi M, Sabatella G, Gambacorta M, Vizzotto L, Computerized morphometry of the cirrhotic liver: comparative analysis in primary biliary cirrhosis, alcoholic cirrhosis, and posthepatitic cirrhosis. Microsc Res Tech, 65 (2004) 113-121.
26) Vertemati M, Goffredi M, Moscheni C, Callegari S, Vizzotto L, Human visceral fat in different anthropometric patterns and in diabetes: a morphometric study. Anal Quant Cytol Histol, 30(1) (2008) 39-46.
ACCEPTED MANUSCRIPT
27) Meek GA, Elder HY. Analytical and quantitative methods in microscopy. Cambridge, England: Cambridge University Press, 1977.
28) Mouton PR. Principles and practices of unbiased stereology: An introduction for bioscentists. Baltimore, Maryland: Johns Hopkins University Press, 2002.
29) Elias H, Henning A, Schwartz DE, Stereology: applications to biomedical research. Physiol Rev, 51 (1971) 158-200.
30) Williams MA, Quantitative methods in biology. Amsterdam New York Oxford, North- Holland Publishing Co, 1977.
31) Scarpelli M, Baccarini MG, Colanzi P, Arnaldi G, Montironi R, Thompson D, Bartels PH, Chromatin texture analysis of cortical adrenal gland adenomas, including incidentalomas, and adjacent normal-appearing cortical tissue. Anal Quant Cytol Histol, 22(3) (2000) 235-243.
32) Scarpelli M, Montironi R, Mazzucchelli R, Thompson D, Bartels PH, Distinguishing cortical adrenal gland adenomas from carcinomas by their quantitative nuclear features. Anal Quant Cytol Histol, 21(2) (1999) 131-138.
33) Shirata NK, Sredni ST, Castelo A, Santinelli A, Mendonca B, Montironi R, Longatto Filho A, Zerbini MCN, Texture Image Analysis in Differentiating Malignant from Benign Adrenal Cortical Tumors in Children. Anticancer Res, 29(8) (2009) 3365-3368.
34) Gerdes J, Lemke, H, Baisch, H, Wacker HH, Schwab U, Stein H, Cell cycle analysis of a cell proliferation-associated human nuclear antigen defined by the monoclonal antibody Ki-67. J Immunol, 133 (1984) 1710-1715.
35) Falconi M, Bartsch DK, Eriksson B, Günter Klöppel G, Lopes JM, O’Connor JM, Salazar R, Taal BG, Vullierme MP, O’Toole D, Consensus Guidelines for the management of patients with digestive neuroendocrine neoplasms of the digestive system: well-differentiated pancreatic non- functioning tumors. Neuroendocrinology, 95 (2012) 120-134.
36) Dowsett M, Nielsen TO, A’Hern R, Bartlett J, Coombes RC, Cuzick J, Ellis M, Lynn Henry N, Hugh JC, lively T, Mc Shane L, Paik S, Penault-lorca F, prudkin L, Regan M, Salter J, Sotirou
ACCEPTED MANUSCRIPT
C, Smith IE, Viale G, Zujewski JA, Hayes DF, Assessment of Ki67 in breast cancer: recommendations from the International Ki67 in Breast Cancer Working Group. J Natl Cancer Inst, 103 (2011) 1-9.
37) Gil J, Wu H, Wang BY. Image analysis and morphometry in the diagnosis of breast cancer, Microsc Res Tech, 15;59(2) (2002) 109-118.
38) Słodkowska J, MarKiewicz T, Grala B, Kozłowski W, Papierz W, Pleskacz K, Murawski P, Accuracy of a remote quantitative image analysis in the whole slide images. Diagn Pathol, 30;6 (2011) Suppl 1:S20.
39) Menter DG, Hoque A, Motiwala N, Sahin AA, Sneige N, Lieberman R, Lippman SM, Computerized image analysis of Ki-67 in ductal breast carcinoma in situ. Anal Quant Cytol Histol, 23(3) (2011) 218-228.
40) Tang LH, Gonen M, Hedvat C, Modlin IM, Klimstra DS, Objective quantification of the Ki67 proliferative index in neuroendocrine tumors of the gastroenteropancreatic system: a comparison of digital image analysis with manual methods. B Am J Surg Pathol, 2012;36(12) (2012) 1761-1770.
41) Vörös A, Csörgő E, Kővári B, Lázár P, Kelemen G, Cserni G, The use of digital images improves reproducibility of the Ki-67 labeling index as a proliferation marker in breast cancer. Pathol Oncol Res, 20(2) (2014) 391-397
42) Mizuno Y, Fuchikami H, Natori T, Takeda N, Inoue Y, Yamada J, Abe H, Seto H, Sato K, Standardized Assessment of Ki-67 in Breast Cancer Patients Using Virtual Slides and an Automated Analyzer in Comparison to Central/Local Pathological Assessments. J Cancer Ther, 5 (2014) 141-146.
43) Seidal T, Balaton AJ, Battifora H. Interpretation and quantification of immunostains. Am J Surg Pathol, 25 (2001) 1204-1207.
ACCEPTED MANUSCRIPT
44) Wachenfeld C, Beuschlein F, Zwermann O, Mora P, Fassnacht M, Allolio B, Reincke M, Discerning malignancy in adrenocortical tumors: are molecular markers useful? Eur J Endocrinol, 145(3) (2001) 335-341.
45) Goldblum JR, Shannon R, Kaldjian EP, Thiny M, Davenport R, Thompson N, Lloyd RV, Immunohistochemical assessment of proliferative activity in adrenocortical neoplasms. Mod Pathol, 6(6) (1993) 663-668.
46) Beuschlein F, Weigel J, Saeger W, Kroiss M, Wild V, Daffara V, Libé R, Ardito A, Al Ghuzlan A, Quinkler M, Oßwald A, Ronchi CL, de Krijger R, Feelders RA, Waldmann J, Willenberg HS, Deutschbein T, Stell A, Reincke M, Papotti M, Baudin E, Tissier F, Haak HR, Loli P, Terzolo M, Allolio B, Müller HH, Fassnacht M, Major prognostic role of Ki67 in localized adrenocortical carcinoma after complete resection. J Clin Endocrinol Metab, 100(3) (2015) 841- 849.
47) Tissier F, Aubert S, Leteurtre E Al Ghuzlan A, Patey M, Decaussin M, Doucet L, Gobet F, Hoang C, Mazerolles C, Monges G, Renaudin K, Sturm N, Trouette H, Vacher-Lavenu MC, Viallon V, Baudin E, Bertagna X, Coste J, Libe R, Adrenocortical tumors: improving the practice of the Weiss system through virtual microscopy: a National Program of the French Network INCa- COMETE. Am J Surg Pathol, 36(8) (2012) 1194-1201.
48) Lu H, Papathomas TG, van Zessen D, Palli I, de Krijger RR, van der Spek PJ, Dinjens WNM, Stubbs AP, Automated Selection of Hotspots (ASH): enhanced automated segmentation and adaptive step finding for Ki67 hotspot detection in adrenal cortical cancer. Diagn Pathol, 9 (2014) 216-224.
ACCEPTED MANUSCRIPT
| Lesion | Dimensions (cm) | Age | M/F |
|---|---|---|---|
| ACA | 4-11 | 28-59 | 4/7 |
| ACC | 5-19 | 25-68 | 5/13 |
ACCEPTED MANUSCRIPT
Table II Morphometric model.
· Ki-67 (40 X magnification):
Volume fractions referred to the entire tissue occupied by: ☐
☒ nuclei of Ki-67 negative cells (Vynucneg)
☒ cytoplasm of Ki-67 negative cells (Vycytneg) nuclei of Ki-67 positive cells (Vvnucpos)
☒ cytoplasm of Ki-67 positive cells (Vycytpos)
☒ ratio nuclei/cytoplasm of Ki-67 negative cells (N/Cneg)
ratio nuclei/cytoplasm of Ki-67 positive cells (N/Cpos) other structure (vessels and inflammatory infiltrate) (Vyother) ☒
· Reticulin (10 X magnification):
☐
Surface fraction (surface in unit volume) of reticulin (Svret)
ACCEPTED MANUSCRIPT
| Variables | Dimension | ACA | ACC | p | F |
|---|---|---|---|---|---|
| Vynucneg | mm3/mm3 | .06637±.0153 | .12046±.0226 | <. 0001 | 45.718 |
| Vycytneg | mm3/mm3 | .71900±.0295 | .58985±.0571 | <. 0001 | 52.973 |
| Vynucpos | mm3/mm3 | .00104±.0004 | .01227±.0088 | <. 0001 | 20.656 |
| Vvcytpos | mm3/mm3 | .00399±.0011 | .01771±.0094 | <. 0001 | 26.409 |
| Vycelneg | mm3/mm3 | .78537±.0304 | .71031±.0656 | <. 001 | 15.209 |
| Vycelpos | mm3/mm3 | .00503±.0013 | .02988±.0186 | <. 0001 | 24.304 |
| Vyother | mm3/mm3 | .20963±.0299 | .25969±.0155 | <. 05 | 6.366 |
| Vycellpos* | mm3/mm3 | .00639±.0017 | .04039±.0245 | <. 0001 | 20.82 |
| N/C neg | .09260±.0228 | .20535±.0396 | <. 0001 | 69.151 | |
| N/C pos | .27281±.1449 | .68022±.1693 | <. 0001 | 44.493 | |
| Syret | mm2/mm3 | 37.24±9.29 | 10.43±5.89 | <. 0001 | 60.019 |
Volume fractions referred to the entire lesion in the test area = 504x504 pixels occupied by nuclei and cytoplasm respectively in Ki-67 cells negative (Vvnucneg and Vycytneg) and Ki- 67 cells positive (Vynucpos and Vycytpos), vascular and inflammatory infiltrate (Vyother), Ki-67 cells negative and positive (Vycelneg and Vycelpos); N/C represents the nuclei/cytoplasm ratio in both type of cells; surface fraction occupied by reticulin (SyRet).
*Vycellpos: Volume fraction of Ki-67 positive cells referred to cellular compartment.
ACCEPTED MANUSCRIPT
| Marker | ACA | ACC |
|---|---|---|
| Vynucpos | <. 00108 | >.01239 |
| Vycelpos | <. 00516 | >.01128 |
| Vycellpos | <. 00468 | >.01589 |
| N/Cpos | <. 41771 | >.51092 |
| Ret | >26.95 | <16.32 |
The column referred to ACA shows the values related to the average and upper deviation standard (Vynucpos, Vycelpos, Vycellpos, N/Cpos, Ret) and to the average and lower deviation standard for reticulin evaluated by morphometry for the variables that, according to our data, discriminate among the different lesions investigated;
in the column of ACC are reported the values of the average and lower deviation standard.
ACCEPTED MANUSCRIPT
| Vynuc neg | Vvcyt neg | Vynuc pos | Vycyt pos | Vyother | Svret | |
|---|---|---|---|---|---|---|
| F | 0.654 | 0.587 | 0.102 | 0.447 | 0.091 | 0.754 |
| p | 0.321 | 0.654 | 0.729 | 0.573 | 0.528 | 0.301 |
ACCEPTED MANUSCRIPT
Figure Legends
Figure 1: An example of a microscopic field (10X), with superimposed a four-lines grid included in a 504 × 504 pixel test area, used during determination of surface fractions (reticulin stain).
Figure 2: An example of a counting frame (40X) with a 144-point square lattice, included in a 504 × 504 pixel test area, used during determination of volume fractions (Ki-67 stain).
Figure 3: Percentage of the different subcomponents relative to Ki-67 negative and positive cells, inflammatory infiltrate and vessels compartments of nodular lesions evaluated on Ki-67 stain.
Figure 4: Volume fractions of nuclei and cytoplasm respectively in Ki-67 negative cells (Panel A) and Ki-67 positive cells (Panel B), in both lesions investigated.
Figure 5: Nuclear/cytoplasmic ratio (N/C) in both lesions investigated. The symbol º32 shows that, in Ki-67 negative cells, one ACA (out value) has a N/C ratio well away, i.e. more than 1.5 times the 75-25 interquartile distance from the rest of the tumors.
Figures 6: Reticulin surface fraction in the nodular lesions (ACA and ACC). The symbol *3923 shows that one ACC (extreme value; Weiss score 7 and ENSAT III) has a surface fraction of reticulin well away, i.e. more than 3 times the 75-25 interquartile distance from the rest of the tumors.
Figure 7: Linear regression of pathologist assessment of the percentage of Ki-67 positive cells on the morphometric evaluation. Pathologist’s evaluation shows a tendency to overemphasize the amount of Ki-67 positive cells in the lesions (regression equation is morphometry = 4.150AP; p<0.001).
Figure 8: Linear regression between SyRet and volume fraction of Ki-67 positive cells (Vycellpos) in ACC (regression equation is SyRet =- 88.87V/cellpos + 12.88 p<0.05) and ACA (regression equation is SyRet =- 1213.27Vycellpos + 39,61 p=0.063) groups.
e
9
1,0
0,8
Volume Fractions
0,6
0,4
0,2
0,0
ACC
Vvnucneg
Vvcytneg
Vvnucpos Vvcytpos Vvother
ACA
Vvnuc Ki-67 positive Vvcyt Ki-67 positive
0,03
Volume Fractions
0,02
0,01-
0,00
ACC
Panel B
ACA
Vvnuc Ki-67 negative Vvcyt Ki-67 negative
0,8
0,6-
Volume Fractions
0,4
0,2
0,0
ACC
Panel A
ACA
Ki-67 positive cells Ki-67 negative cells
1,0
0,8
0,6
N/C Ratio 0,4
0,2
32
0,0
ACC
ACA
60,00
50,00
Surface Fraction of Reticulin (SvRet)
40,00
30,00
3923 ®
20,00
10,00
,00
ACC
ACA
0,40
Ki67 cells positive
☒ ACC
ACA
0,30
☒
Pathologist
0,20
☒
0,10
☒
☒
☒
0,00
0,00
0,02
0,04
0,06
0,08
0,10
Morphometry