Original articles
DNA cytophotometry in adrenocortical tumours: a clinicomorphological study of 66 cases *
Barbara-Christina Padberg1, Ines Lauritzen1, Eike Achilles1, Konstanze Holl1, Max Bressel2, Günter Klöppel 3, Henning Dralle4, and Sören Schröder1
1 Institute of Pathology, University of Hamburg, Martinistrasse 52, W-2000 Hamburg 20, Federal Republic of Germany
2 Department of Urology, General Hospital Hamburg-Harburg, Federal Republic of Germany
3 Institute of Pathology, Free University of Brussels, Brussels, Belgium
4 Department of Surgery, Hannover School of Medicine, Hannover, Federal Republic of Germany
Received March 13, 1991 / Received after revision April 22, 1991 / Accepted April 23, 1991
Summary. Surgical specimens of 66 adrenocortical tu- mours were investigated by conventional microscopy and DNA cytophotometry. Histologically, 50 neoplasms were classified as adenomas and 16 as carcinomas. In only 8 of the latter cases were distant metastases and/or a lethal outcome recorded. On single cell scanning cyto- photometry either non-euploid or aneuploid DNA histo- grams were identified in 24 of 50 adenomas (48%) and in 14 of 16 carcinomas (88%). The two carcinomas ex- hibiting euploid DNA distributions fell into the group of 7 malignancies which are recurrence-free so far. From these findings it is concluded that DNA measurements have no diagnostic and only limited prognostic value in neoplasms of the adrenal cortex.
Key words: Adrenocortical tumours - DNA cytopho- tometry - Differential diagnosis - Prognosis
Introduction
In several neoplasms of endocrine and neuroendocrine lineage, there is the dilemma of assessing their biological potential correctly (benign versus malignant). This ap- plies especially for adrenocortical tumours (ACTs), in which malignancy cannot be demonstrated by a single morphological variable. Different authors have ad- dressed this problem by defining a histological index of malignancy, according to which a variety of micro- scopic features has to be evaluated, graded and added to make a score which results in the diagnosis of carcino- ma when a specifically defined threshold is crossed (Hough et al. 1979; Weiss 1984; van Slooten et al. 1985).
Some of these schemes are, however, hampered by the fact that the biological potential of individual ACTs re- mains indeterminate even after application of the respec- tive classification procedures (Hough et al. 1979; Weiss 1984). In addition, aggressive behaviour has been de- scribed for ACTs which had only been typed as border- line lesions histologically (Gandour and Grizzle 1986). The failure of conventional light microscopy to predict the clinical course of any given ACT precisely has stimu- lated several groups to perform DNA measurements on this type of neoplasm. Considerable discrepancies came to light when the DNA data of ACTs reported in the literature were compared, and all of the respective stu- dies so far utilized the method of flow cytometry. We performed DNA scanning cytophotometry on a selec- tion of 66 clinically well-documented ACTs and correlat- ed our results with the findings of conventional histology and the further evolution of disease.
Material and methods
Formalin-fixed, paraffin-embedded material obtained from surgi- cal specimens of 66 primary ACTs was analysed using conventional histology and single cell scanning DNA cytophotometry. Micro- scopically, the criteria listed in Table 1 (e.g. nuclear features, mitot- ic activity, structural findings and indicators of microinvasive growth) were evaluated and added to histological indices of malig- nancy as defined by Hough et al. (1979), Weiss (1984) and van Slooten et al. (1985).
Information relating to the preoperative history was noted from the patient’s medical reports. The values for tumour diameters and weights were taken from the original pathological descriptions. In each case, the development of disease following surgery was documented until spring 1990 (mean postoperative observation pe- riod 29 months, range 1-96 months).
Cytophotometric determinations of DNA content were per- formed as described previously (Padberg et al. 1990). Briefly, paraf- fin sections of the primary tumours were cut at a 6-um thickness and stained according to the Feulgen technique. Coverslips were mounted with “Eukitt” (refraction index 1.494). Single cell DNA measurement was performed in the scanning mode on a Leitz-
* Dedicated to Prof. G. Seifert, former director of the Institute of Pathology, University of Hamburg, on the occasion of his 70th birthday
Offprint requests to : B .- C. Padberg
| Criterion | Parameter | Score | ||
|---|---|---|---|---|
| Weiss 1984 | Hough et al. 1979 | van Slooten et al. 1985 | ||
| Nuclear grade | Moderate or | 1 | 0.39 | 2.1 |
| strong | ||||
| Nuclear | Moderate or | 2.6 | ||
| hyperchromasia | marked | |||
| Nucleoli | Abnormal | 4.1 | ||
| structure | ||||
| Mitotic | > 5/50 HPF | 1 | ||
| activity | ||||
| >10/100 HPF | 0.60 | |||
| >2/10 HPF | 9.0 | |||
| Atypical | Present | 1 | ||
| mitoses | ||||
| Clear cells | <25% of | 1 | ||
| the tumour | ||||
| Architecture | Diffuse | 1 | 0.92 | |
| pattern | ||||
| Structure | Mainly | 1.6 | ||
| abnormal | ||||
| Venous | Invasion | İ 1 | ||
| structures | present | 0.92b | ||
| Sinusoidal | Invasion | 1 | 3.3c | |
| structures | present | |||
| Tumour capsule | Invasion | 1 | 0.37 | |
| present | ||||
| Necroses | Present | 1 | 0.69 | |
| Regressive | Present | 5.7 | ||
| changes | ||||
| Broad fibrous | Present | 1.00 | ||
| bands |
ª Weiss (1984): 1-3, benign; 4, borderline; 5-9, malignant; Hough et al. (1979): 0.17±0.26, benign; 1±0.58, indeterminate; 2.91 ± 0.9, malignant; van Slooten et al. (1985): < 8, benign; ≥8, malig- nant
b Vascular invasion
” Capsular and/or vascular invasion
MPV-cytophotometer based on a Leitz-Orthoplan microscope. The measuring spot was 2.54 um2; the steps of the scanning process were 0.5 um wide. Absorption of the probes was determined at a wave length of 560.0±9.5 nm. The number of cells examined was 70-100 for each case. Data were processed on-line by a Euro- cos computer using specially adopted commercial software (Leitz, Wetzlar, FRG). Determination of diploid values was performed using 210 normal adrenocortical cells from 3 control specimens; the mean±2 SD of their cellular DNA content was defined as the 2 c region. Evaluation of the data was done using two different schemes: (i) classifying the DNA histograms according to Auer et al. (1980) into four different types I-IV (for details, see Results); (ii) determining the P90 value according to Forsslund et al. (1984).
Results
On reclassification in line with the three above-men- tioned criteria catalogues (Hough et al. 1979; Weiss 1984; van Slooten et al. 1985: Table 1), the results of which were identical each time, 50 tumours were typed as adenomas and 16 as carcinomas. The mean tumour
weights were 67 (8-1080) g for adenomas and 674 (20- 3000) g for carcinomas; 9 adenomas weighed more, 2 carcinomas less than 50 g. At the resection of the prima- ry tumour, the mean patient age was 53 (range 20-79) years in the former and 45 (range 2-76) years in the latter group. For both types of neoplasm, a clear pre- dominance of female patients was found (M:F ad- enomas 1:3.5; carcinomas 1:2.2). Preoperatively, in the adenoma patients the following hormonal symptoms had been documented: hyperaldosteronism (Conn’s syn- drome) in 20 of 50; hypercortisolism (Cushing’s syn- drome) in 8 of 50; virilization in 1 of 50; feminization in 1 of 50; 20 of 50 patients had not shown endocrine abnormalities (non-functioning ACTs). Clinicomorpho- logical findings in the total of 16 malignant ACTs are given in Table 2. At the end of the observation period, 48 adenoma patients showed symptom-free survival; 2 patients had previously died from unrelated causes. Of the carcinoma patients, 6 had died from the tumour, 1 from other causes; 2 patients were alive with persistent tumour manifestations, while 7 continued to live symp- tom-free 4-61 months after surgery.
On cytophotometry, 5 of 50 adenomas (10%) but none of the 16 carcinomas were characterized by single distinct modal values in the diploid region (diploid type I histograms, Fig. 1a) as recorded for each of the three control specimens. Twenty-one of the 50 adenomas (42%) and 2 of 16 carcinomas (12.5%) exhibited DNA distributions defined as having either a distinct modal value in the tetraploid or near-tetraploid region or show- ing two well-defined peaks around the 2c and 4c regions (euploid type II histograms, Fig. 1b). In 19 of 50 ad- enomas (38%) and 6 of 16 carcinomas (37.5%), non- euploid type III populations (Fig. 1c) were found, hav- ing two peaks but differing from the type II populations in that the histograms exhibited a sizeable number of cells with DNA amounts similar to those of the control cells in DNA synthesis. The position of the two peaks, as a rule, deviated somewhat from the 2c and 4c values of normal populations. In 5 of 50 adenomas (10%) and 8 of 16 carcinomas (50%), aneuploid type IV specimens (Fig. 1d) were seen with a pronounced irregularity and DNA amounts per cell ranged from levels near 2c up to values beyond 8c or even 12c.
To summarize, diploidy was observed exclusively in a small number of benign adrenocortical neoplasms, while aneuploidy occurred five times more frequently among malignant when compared with benign ACTs. Systematic differences could, however, not be detected between these two groups, either upon DNA histogram typing or when determining the respective P90 values (data for the group of the malignant ACTs included in Table 2). Neither was any association demonstrable between the DNA content of tumours and the presence or absence of hormonal function. When comparing the cytophotometric findings to the follow-up data, no clear- cut correlation was found between the DNA value of tumours and the further course of disease. Yet both of the two carcinoma cases exhibiting euploid DNA distri- butions fell into the group of 7 malignancies which are recurrence-free so far.
| No. | Age, sex | Histological indices of malignancy DNA | Follow-up (months) | |||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Tumour weight | Clinical symptoms | Hough et al. (1979) | Weiss (1984) | van Slooten et al. (1985) | Auer | p90 | ||||
| 1 | 47 M | 130 g | Cush. | 3.60 | 6 | 16.8 | III | 41.4 | 7 | DFC |
| 2 | 69 F | 58 g | Cush., Vir. | 2.97 | 7 | 24.3 | III | 41.4 | 2 | DFC |
| 3 | 44 F | 1500 g | Cush., Vir. | 2.60 | 6 | 18.4 | IV | 43.0 | 90 | AWT |
| 4 | 02 M | 34g | Cush., | 2.31 | 3 | 9.4 | III | 47.1 | 28 | NED |
| Pub. pr. | ||||||||||
| 5 | 62 M | 190 g | Cush. | 2.97 | 7 | 21.7 | III | 48.6 | 25 | NED |
| 6 | 51 F | 113 g | Cush. | 3.50 | 7 | 19.6 | III | 72.9 | 26 | DFC |
| 7 | 57 F | 1300 g | Cush., Vir. | 4.52 | 6 | 21.7 | IV | 75.7 | 1 | DFC |
| 8 | 02 F | 20 g | Cush., Vir. | 2.83 | 6 | 22.7 | IV | 77.1 | 7 | NED |
| 9 | 76 F | 250 g | Cush. | 3.97 | 6 | 15.3 | IV | 98.6 | 61 | NED |
| 10 | 31 F | 1440 g | Vir. | 3.92 | 5 | 11.1 | IV | 90.0 | 28 | AWT |
| 11 | 40 F | 750 g | Non-funct. | 3.60 | 5 | 9.4 | II | 22.9 | 55 | NED |
| 12 | 40 F | 54 g | Non-funct. | 2.60 | 6 | 22.5 | II | 24.3 | 35 | NED |
| 13 | 44 F | 995 g | Non-funct. | 3.97 | 6 | 24.3 | III | 50.0 | 96 | DFC |
| 14 | 61 F | 3000 g | Non-funct. | 3.97 | 7 | 21.7 | IV | 77.1 | 57 | DOC |
| 15 | 45 M | 280 g | Non-funct. | 3.89 | 7 | 28.4 | IV | 85.7 | 1 | DFC |
| 16 | 24 M | 1890 g | Non-funct. | 4.52 | 7 | 28.4 | IV | 92.5 | 31 | NED |
Cush., Cushing’s syndrome; Vir., virilization; Non-funct., non-functioning; Pub. pr., pubertas praecox; DFC, death from carcinoma; AWT, alive with tumour; NED, no evidence of disease; DOC, death from other causes
30
a
30
b
20
20
10
10
0
10c
0
1
30
2c
4c
6c
8c
30
2c
4c
6c
8c
10c
c
d
20
20
10
10
0
0
L
1.
2c
10c
4c
6c
A
4c
6c
8c
2c
8c
10c
Discussion
Our series underlines the relevance of the three common- ly used classification schemes for the histological dis- crimination of benign and malignant ACTs (Hough et al. 1979; Weiss 1984; van Slooten et al. 1985). All neoplasms typed in this way as adenomas followed an uneventful postoperative course, although 9 of these weighed more than 50 g and some considerably more - a point to remember in the differential diagnosis with adrenocortical carcinoma (Schteingart et al. 1968), as re- cently re-emphasized by Rosai (1989). Surprisingly, how- ever, 7 of our 16 ACTs typed as carcinomas have re- mained recurrence-free so far. Hence, the malignancies in our material exhibited an altogether more favourable behaviour than such neoplasms in several other reports communicated in the literature.
As far as we have been able to ascertain, nine different groups have communicated DNA findings in ACTs, all of them utilizing the method of flow cytometry on either paraffin-embedded or on fresh material. The flow cyto- metric data reported in the literature and the cytophoto- metric findings presented herein are summarized in Ta- ble 3. With regard to the detection of aneuploidy exclu- sively among carcinomas three of the groups (Klein et al. 1985; Bowlby et al. 1986; Kojima et al. 1988) assumed DNA measurements to be superior to conventional his- tology in the discrimination between benign and malig- nant tumours of the adrenal cortex. This assumption was, however, refuted by the findings of Joensuu and Klemi (1988), who demonstrated euploidy and aneup- loidy each to occur in approximately 50% of benign ACTs - a result in line with our observations and those of Cibas et al. (1990) - reflecting the fact that by no
| References | nª | Adenomas | Carcinomas | ||
|---|---|---|---|---|---|
| Euploid | Aneuploid | Euploid | Aneuploid | ||
| Klein et | 6 | 2/2 | 0/2 | 0/4 | 4/4 |
| al. (1985)b | |||||
| Bowlby et | 22 | 16/16 | 0/16 | 1/6 | 5/6 |
| al. (1986)e | |||||
| Amberson et | 48 | 1/5 | 4/5 | ||
| al. (1987)e | |||||
| Hosaka et | 52 | 22/52 | 30/52 | ||
| al. (1987)€ | |||||
| Taylor et | 10 | 5/10 | 5/10 | ||
| al. (1987)b+c | |||||
| Joensuu and | 17 | 8/17 | 9/17 | ||
| Klemi (1988)℃ | |||||
| Kojima et | 15 | 9/9 | 0/9 | 0/6 | 6/6 |
| al. (1988)℃ | |||||
| Rainwater et al. (1989)e | 26 | 19/20 | 1/20 | 3/6 | 3/6 |
| Cibas et | 43 | 24/30 | 6/30 | 4/13 | 9/13 |
| al. (1990)c | |||||
| Present | 66 | 26/50 | 24/50 | 2/16 | 14/16 |
| studyd | |||||
a Total number of adrenocortical tumours investigated
b Flow cytometry using fresh material
” Flow cytometry using paraffin-embedded material
d Cytophotometry
means all carcinomas exhibit aneuploid DNA distribu- tions. Several more recent cytometric studies concur with this statement by describing euploid DNA values for up to 50% of adrenocortical carcinomas (Amberson et al. 1987; Hosaka et al. 1987; Taylor et al. 1987; Rain- water et al. 1989). In each of these four series, however, an association between the DNA content of tumours and the aggressiveness of disease was suggested since aneuploidy was reported to be a phenomenon only en- countered among ACTs with virulent behaviour. It has yet to be stressed in this context that Cibas et al. (1990), in their comprehensive study of 43 ACTs, failed to dem- onstrate such differences in survival between patients with diploid versus aneuploid tumours. Our data appear to parallel the findings of the latter study, since not only our 2 euploid, but also 3 of our 8 aneuploid adrenocorti- cal carcinomas followed an uneventful course.
The discrepancies between our results and that of others might partly be explained by differences in the cytometric methodology; yet with regard to the small number of cases included in some of the cited studies selection bias could also have contributed to this discor- dance. Since none of the two evaluation schemes chosen to interpret our DNA data provided significant results, we conclude that cytophotometry is of no diagnostic and of only limited prognostic value in adrenocortical tumours. Rather, from our results, conventional histolo- gy remains the most effective means of assessing the malignancy of individuals ACTs, while DNA measure- ments - in order to determine euploidy - might only be regarded as a supplementary tool to recognize carci-
nomas of the adrenal cortex with low malignant poten- tial.
Acknowledgements. Supported by the Deutsche Forschungsge- meinschaft (Schr 274/4-1) and the Hamburger Krebsgesellschaft (No. 384).
References
Amberson JB, Vaughan ED, Gray GF, Naus GJ (1987) Flow cyto- metric analysis of nuclear DNA from adrenocortical neo- plasms. A retrospective study using paraffin-embedded tissue. Cancer 59:2091-2095
Auer G, Caspersson TO, Wallgren AS (1980) DNA content and survival in mammary carcinoma. Anal Quant Cytol Histol 2:161-165
Bowlby LS, Debault LE, Abraham SR (1986) Flow cytometric analysis of adrenal cortical tumour DNA. Relationship between cellular DNA and histopathologic classification. Cancer 58:1499-1505
Cibas ES, Medeiros LJ, Weinberg DS, Gelb AB, Weiss LM (1990) Cellular DNA profiles of benign and malignant adrenocortical tumors. Am J Surg Pathol 14:948-955
Forsslund G, Cedermark B, Öhmann U, Ehrhardt K, Zetterberg A, Auer G (1984) The significance of DNA distribution patterns in rectal carcinoma. Dis Colon Rectum 27:579-584
Gandour MJ, Grizzle WE (1986) A small adrenocortical carcinoma with aggressive behaviour. Arch Pathol Lab Med 110:1076- 1079
Hosaka Y, Rainwater LM, Grant CS, Young WF, Farrow GM, Heerden JA van, Lieber MM (1987) Adrenocortical carcinoma: nuclear deoxyribonucleic-acid ploidy studied by flow cyto- metry. Surgery 102:1027-1034
Hough AJ, Hollifield JW, Page DL, Hartmann WH (1979) Prog- nostic factors in adrenal cortical tumors. A mathematical analy- sis of clinical and morphological data. Am J Clin Pathol 72:390-399
Joensuu H, Klemi PJ (1988) DNA aneuploidy in adenomas of endocrine organs. Am J Pathol 132:145-151
Klein FA, Kay S, Ratliff JE, White FKH, Newsome HH (1985) Flow cytometric determinations of ploidy and proliferation pat- terns of adrenal neoplasms: an adjunct to histological classifica- tion. J Urol 134:862-866
Kojima M, Kamachi M, Ashihara T, Watanabe H (1988) DNA- cytofluorometric study on adrenal cortical tumors (abstract). Proceedings of the 8th International Congress of Endocrinolo- gy, Kyoto, Japan. A 13-18-001, p 343
Padberg BC, Garbe E, Achilles E, Dralle H, Bressel M, Schröder S (1990) Adrenomedullary hyperplasia and pheochromocyto- ma. DNA cytophotometric findings in 47 cases. Virchows Arch [A] 416:443-446
Rainwater LM, Young WF, Farrow GM, Grant CS, Heerden JA van (1989) Flow cytometric analysis of deoxyribonucleic-acid ploidy in benign and malignant aldosterone-producing neo- plasms of the adrenal gland. Surg Gynecol Obstet 168:491-496 Rosai J (1989) Ackerman’s surgical pathology, 7th edn., vol 1. Mosby, St. Louis, p 791
Schteingart DE, Oberman HA, Friedman BA, Conn JW (1968) Adrenal cortical neoplasms producing Cushing’s syndrome. A clinicopathologic study. Cancer 22:1005-1013
Slooten H van, Schaberg A, Smeenk D, Moolenaar J (1985) Mor- phologic characteristics of benign and malignant adrenocortical tumors. Cancer 55:766-773
Taylor SR, Roederer M, Murphy RF (1987) Flow cytometric anal- ysis of adrenocortical tumors in children. Cancer 55:766-773 Weiss LM (1984) Comparative histologic study of 43 metastasizing and non-metastasizing adrenocortical tumors. Am J Surg Pa- thol 8:163-169