Comparison of Immunomarkers for the Identification of Adrenocortical Cells in Cytology Specimens
Mehri Jalali, M.D., and Savitri Krishnamurthy,* M.D.
We studied the immunoreactivity of three antibodies-A103, cal- retinin, and inhibin alpha in destained Papanicolaou (Pap) smears and cell-blocks of 40 fine-needle aspiration biopsy cases of adrenocortical lesions (35 cases of hyperplasialadenoma and 5 cases of carcinoma). Five cases of carcinoma (4) and mela- noma (1) metastases to the adrenal gland and five cases of renal-cell carcinoma were also included for comparison. In benign adrenocortical lesions, A103 staining was noted in 82% of the destained Pap smears and in 92% of cell-blocks. In malignant adrenocortical lesions, A103 staining was noted in 50% of the destained Pap smears and in 80% of cell-blocks. In comparison, calretinin staining was noted in 6% and 50% of destained smears and in 78% and 60% of the cell-blocks of benign and malignant adrenocortical lesions. Inhibin alpha was not positive in any of the smears and showed the lowest level of positivity in the cell-block sections, namely in 11% of the benign lesions and 25% of the malignant lesions. The sensitivity of A103 was 90% on cell-blocks and 74% on smears, that of calre- tinin 75% on cell-blocks and 11% on smears, and that of inhibin alpha, 13% on cell-blocks alone. The specificity of A103 was lower than the other two makers, 90% vs. 100% because of pos- itivity in metastatic melanoma in the adrenal gland.
Our data show A103 to be the immunomarker with the highest sensitivity for identifying cells of adrenocortical origin in destained Pap’s smears and cell-block sections with, however, a lower specificity when compared with calretinin and inhibin alpha. Calretinin is comparable in sensitivity with A103 on cell- block sections alone and not on smears. The results of this study suggest that if metastatic melanoma in adrenal gland is not a consideration then A103 is the marker of choice for identifying cells of adrenocortical origin in the limited material available for diagnostic purposes in cytology specimens. Diagn. Cytopa- thol. 2005;33:78-82. @ 2005 Wiley-Liss, Inc.
Key Words: fine-needle aspiration biopsy; adrenocortical tumors; inhibin alpha; calretinin; A103
*Correspondence to: Savitri Krishnamurthy, M.D., Department of Path- ology, Box 53, The University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Blvd., Houston, TX 77030, USA.
E-mail: skrishna@mdanderson.org
Published online in Wiley InterScience (www.interscience.wiley.com).
The three most widely used markers for the identification of adrenocortical cells are inhibin alpha, A103, and calreti- nin. Inhibin alpha acts as a modulator of steroidogenesis by inhibiting the release of follicle-stimulating hormone from the pituitary gland.1,2 Inhibin alpha is positive in ovarian granulosa cells and their tumors, testicular sertoli cells and their tumors, and normal and neoplastic adreno- cortical cells. A103 is a monoclonal antibody against MART-1, a melanoma antigen.3 A103 was originally studied as a melanoma marker, and was subsequently found to be expressed in all steroid-producing cells, includ- ing sex cord-stromal and adrenocortical cells, because of sharing of epitopes between these cells and melanoma cells. Calretinin, a calcium-binding protein, was first dis- covered in neural tissue and was subsequently found to be consistently expressed in mesothelial, sex cord-stromal, and adrenocortical cells.46 Calretinin has now proven to be a popular marker for these cells.
The relative sensitivity of inhibin alpha, A103, and cal- retinin as markers for adrenocortical cells have been tested mainly in tissue sections of different types of adre- nocortical lesions. Very few studies have compared the diagnostic utility of these markers in cytology specimens, and none have compared their sensitivity in these speci- mens. Such a comparison would be particularly valuable in cytology because of the limited availability of diagnos- tic material for immunostaining. Therefore, we performed the study described herein to compare the immunoreactiv- ity of inhibin alpha, A103, and calretinin with the use of destained Papanicolaou (Pap) smears and cell-blocks of benign and malignant adreno-cortical lesions.
Materials and Methods
The cytopathology files of the Department of Pathology at The University of Texas M. D. Anderson Cancer Cen- ter were searched to identify fine-needle aspiration biopsy (FNAB) proven cases of adrenal gland after gaining insti- tutional research board approval (IRB) of the study. FNAB was performed in all the cases by a radiologist
| Extent of staining“ | Positive cases | |||||
|---|---|---|---|---|---|---|
| 0 | 1+ | 2+ | 3+ | No. | Percentage | |
| A103 | ||||||
| Hyperplastic/adenoma (n = 26) | 2 | 2 | 7 | 15 | 24 | 92 |
| Adrenocortical carcinoma (n = 5) | 1 | — | 1 | 3 | 4 | 80 |
| Metastatic carcinoma (n = 4) | 4 | 0 | 0 | |||
| Metastatic melanoma (n = 1) | — | — | — | 1 | 1 | 100 |
| Calretinin | ||||||
| Hyperplastic/adenoma (n = 23) | 5 | 2 | 5 | 11 | 18 | 78 |
| Adrenocortical carcinoma (n = 5) | 2 | — | 2 | 1 | 3 | 60 |
| Metastatic carcinoma (n = 4) | 4 | 0 | 0 | |||
| Metastatic melanoma (n = 1) | 1 | 0 | 0 | |||
| Inhibin alpha | ||||||
| Hyperplastic/adenoma (n = 18) | 16 | 1 | 1 | — | 2 | 11 |
| Adrenocortical carcinoma (n = 5) | 4 | — | 1 | — | 1 | 25 |
| Metastatic carcinoma (n = 4) | 4 | 0 | 0 | |||
| Metastatic melanoma (n = 1) | 1 | 0 | 0 | |||
ªExtent of staining scored semiquantitatively as follows: 0, no staining; 1+, <25% with positive staining; 2+, 25%-50% with positive staining; and 3+, >50% with positive staining.
under computed tomography (CT) guidance with a 22- gauge needle. Direct smears were made and fixed in Car- noy’s solution for Pap staining and air-dried for Diff-Quik staining. The syringe used for the aspiration was subse- quently rinsed in RPMI solution. The RPMI solution was centrifuged; depending on the size of the pellet after cen- trifugation, cytospin smears (when the pellet was very small) or cell-blocks (when the cell pellet was relatively larger) were prepared. The cell-blocks were routinely processed, embedded in paraffin, and subsequently cut to produce 4-um-thin tissue sections. A representative Pap- stained smear and an unstained cell-block section were selected for immunostaining. The Pap smears were destained with 1% acid alcohol. Immunostaining was per- formed with the use of the avidin-biotin-complex method in a Dako Autostainer with DAB as the chromogen and appropriate positive and negative controls in each batch. Antigen retrieval was performed only for immunostaining of cell-block sections and not for the destained Pap smears because of the possibility of losing tissue from the smear during the retrieval. Primary antibodies against inhibin alpha (R1, 1:50; Oxford Bio-Innovation, Oxford- shire, UK), A103 (AB.4, 1:25; Neomarkers, Fremont, CA), and calretinin (1:20, Zymed, San Francisco, CA) were used. Positive staining was defined as finely granular cytoplasmic staining for inhibin alpha, coarsely granular
| Extent of stainingª | Positive cases | |||||
|---|---|---|---|---|---|---|
| 0 | 1+ | 2+ | 3+ | No. | Percentage | |
| A103 | ||||||
| Hyperplastic/adenoma (n = 17) | 3 | — | 6 | 8 | 13 | 82 |
| Adrenocortical carcinoma (n = 2) | 1 | — | — | 1 | 1 | 50 |
| Metastatic carcinoma (n = 3) | 3 | 0 | 0 | |||
| Renal-cell carcinoma (n = 0) | 0 | 0 | 0 | |||
| Calretinin | ||||||
| Hyperplastic/adenoma (n = 17) | 16 | 1 | — | — | 1 | 6 |
| Adrenocortical carcinoma (n = 2) | 1 | 1 | 0 | 0 | ||
| Metastatic carcinoma (n = 3) | 3 | 0 | 0 | |||
| Renal-cell carcinoma (n = 0) | 0 | 0 | ||||
| Inhibin alpha | ||||||
| Hyperplastic/adenoma (n = 21) | 21 | — | — | — | 0 | 0 |
| Adrenocortical carcinoma (n = 2) | 2 | 0 | 0 | |||
| Metastatic carcinoma (n = 3) | 3 | 0 | 0 | |||
| Renal cell carcinoma (n = 0) | 0 | 0 | ||||
ªExtent of staining scored semiquantitatively as follows: 0, no staining; 1+, <25% with positive staining; 2+, 25%-50% with positive staining;
cytoplasmic staining for A103, and finely granular cyto- plasmic and nuclear staining for calretinin. The proportion of positively stained cells was given a grade of 1+, 2+, or 3+ based on the presence of immunostaining in less than 25%, 25-50%, and more than 50% of positively stained cells, respectively. The immunostaining results were eventually used to calculate the sensitivity and spe- cificity of each of the markers for the identification of adrenocortical cells.
Results
Forty cases of FNAB proven adrenal gland lesions consis- ting of 35 cases diagnosed as benign adrenocortical tissue with a differential diagnosis of adrenocortical hyperplasia or adenoma and 5 cases of adrenocortical carcinoma were included. Five cases of carcinoma (4) or melanoma (1) metastases in the adrenal gland and five cases of renal-cell carcinoma (conventional/clear cell type) were also in- cluded in the study for comparison with the staining of the markers in the adrenocortical lesions.
The results of immunostaining of the cell-block sec- tions and destained Pap smears with A103, calretinin, and inhibin alpha are summarized in Tables I and II, respec- tively. Seventeen smears and 24 cell-blocks of benign adrenocortical lesions; 2 smears and 5 cell-blocks of adre- nocortical carcinoma; and 5 cell-blocks of renal-cell carci-
A
B
4
B
J
A 103
A 103
C
D
D
Calretinin
Inhibin
Calretinin
Inhibin
Fig. C-1
Fig. C-2
A
B
B
A 103
A 103
D
C
D
Calretinin
Inhibin
Calretinin
Inhibin
Fig. C-3
Fig. C-4
noma; 3 smears of adenocarcinoma metastases in the adrenal gland from the lung, bladder, and liver and cell- blocks of the same including 1 more case of esophageal carcinoma metastasis and 1 case of melanoma metastasis were available for immunostaining. Characteristic coarsely granular staining for A103 was noted in 82% and 50% of the destained Pap smears of benign adreno- cortical lesions and in adrenocortical carcinoma, respec- tively. None of the cases of metastatic carcinoma or renal-cell carcinoma were positive for A103. The only case of metastatic melanoma to adrenal gland was strongly positive for A103. The results of A103 in the cell-block sections was better than that in the smears and
was observed in 92% and 80% of the benign and malig- nant adrenocortical lesions, respectively. Similarly, calre- tinin stained more cell-block sections than destained Pap smears. Notably, 78% and 60% of the cell-blocks of the benign adrenocoritcal lesions and adrenocortical carci- noma showed positivity for calretinin in contrast with 6% and 50% of the destained Pap smears of benign adreno- cortical lesions and adrenocortical carcinoma. Inhibin alpha had the lowest level of positivity in the cell-block sections; 11% and 25% in the benign adrenocortical lesions and adrenocortical carcinomas, respectively. Inter- estingly, none of the destained Pap smears exhibited staining for inhibin alpha.
| Markers | Cell-block (%) | Smears (%) | ||
|---|---|---|---|---|
| Sensitivity | Specificity | Sensitivity | Specificity | |
| A103 | 90 | 90 | 79 | 100 |
| Calretinin | 75 | 100 | 11 | 100 |
| Inhibin alpha | 13 | 100 | 0 | — |
Overall, the intensity of staining for both A103 and calre- tinin was usually either strong or moderately strong, whereas that for inhibin alpha was weak. In addition, A103 and calretinin staining was often diffusely positive in both smears and cell-blocks in contrast with inhibin alpha, which was usually more focal and patchy. The illustration of the staining pattern and intensity of A103, calretinin, and inhibin alpha on destained Pap smears and cell-blocks of FNAB of a case of benign and malignant adrenocortical lesion is depicted in Figs. C-1-C-4. A103 proved to have the highest sensitivity in the destained Pap smears and cell- blocks: 74% and 90%, respectively. Because of positive staining in a case of metastatic melanoma the specificity of A103 was 90% in comparison with 100% for the other two markers. The relative sensitivity and specificity of these three immunomarkers in the identification of adrenocortical cells are listed in Table III.
Discussion
The usefulness of calretinin, A103, and inhibin alpha as markers of cells of adrenocortical origin has been well documented by several investigators mainly using tissue sections obtained from surgical specimens of the adrenal gland. The extent of the immunopositivity of each of the markers in adrenocortical lesions was very varied in these studies, even when the antibody vendors were similar. The immunopositivity ranged from 73% to 96% for calre- tinin, from 50% to 100% for A103, and from 73% to 100% for inhibin alpha.7,3,8-10 Very few studies have, however, compared the relative sensitivities of these anti- bodies as markers for adrenocortical cells. In one study, Renshaw and Granter9 compared the immunoreactivity of A103 and inhibin alpha in 22 benign and 4 malignant adrenocortical tumors. They found that 68% and 50% of the benign and malignant tumors, respectively, were posi- tive for A103 in comparison with 86% and 75% of the same tumors for inhibin alpha. They concluded that although both of these antibodies are useful in immuno- histochemical diagnosis of adrenocortical tumors, inhibin alpha is slightly more sensitive than A103. In addition, because rare cases of hepatocellular and renal carcinoma were also positive for inhibin alpha, A103 was reported to be marginally more specific.
In addition, Jorda et al.1 compared calretinin and inhibin alpha in the diagnosis of adrenocortical neoplasms
and found the sensitivity to be comparable: 73% and 70%, respectively. The only study comparing all three markers in tissue sections of adrenocortical lesions was performed by Zhang et al.12 who found calretinin to be the most sensitive. Notably, the extent of positivity for calretinin was 96%, whereas that for A103 and inhibin alpha was 89% and 92%, respectively.
Only two studies have used FNAB specimens, includ- ing tissue sections of cell-blocks alone and not smears, to study the use of inhibin alpha and A103 in cytology specimens. In the first study, Fetsch et al.13 studied the immunoreactivity of inhibin alpha in 45 archived for- malin-fixed, paraffin-embedded cell-block sections of FNAB specimens of known primary and metastatic adre- nocortical lesions. They found that all of the benign and malignant adrenocortical lesions studied showed a distinct cytoplasmic granular staining pattern. Based on the intense, specific immunostaining pattern in cells of adre- nocortical origin, these authors concluded that inhibin alpha is an excellent marker for adrenocortical cells in FNAB specimens. In the second study, Shin et al.14 exam- ined Melan-A as detected by the A103 antibody in 24 for- malin-fixed, paraffin-embedded cell-blocks of FNAB specimens of the adrenal gland. Seven of eight cases with normal, hyperplastic, and neoplastic adreno cortical cells were positive for A103, whereas none of the cases of metastatic carcinoma were positive for A103. Therefore, this small study found A103 monoclonal antibody to be a sensitive marker for adrenocortical cells.
Although our results with these markers in cell-block sections of adrenocortical FNAB specimens are similar to some of the prior studies, they are also strikingly different from those of many others. A103 staining in 90% of the cell-block sections of adrenocortical lesions in our study is comparable with the findings of Shin et al.14 and Zhang et al.12 who found positive staining in 87% and 89% of cases, respectively. Similarly, although the calretinin pos- titivity in 75% of the adrenocortical lesions in our study is similar to that in the study of Jorda et al., it is how- ever lower than that in the study by Zhang et al.12 The most striking difference between our study and previous studies concerns inhibin alpha. Unlike these prior studies, we found that only 13% of the cases had patchy, weak heterogeneous staining for this marker. It should be noted that the vendor for this antibody that we used was differ- ent from that used by other researchers (Oxford Bio-Inno- vation vs. Serotec), which may explain the difference to some extent. However, Zhang et al.12 used inhibin alpha obtained from the vendor we used, but found that 92% of their cases were positive for this marker. The heterogene- ous staining pattern of inhibin alpha may account for the discrepancy between our results in the limited material in the cell-block sections and those in the tissue sections used by Zhang and colleagues.
JALALI AND KRISHNAMURTHY
Our study is the first to test A103, calretinin, and inhibin alpha in destained Pap smears. We found A103 to be the best marker, as 74% of the smears were positive for it compared with 11% positive for calretinin. Interestingly, none of the smears were positive for inhibin alpha. It is possible that with antigen retrieval, the results for all of the markers could be improved. However, we did not use antigen retrieval because of the fear of losing the tissue during this process.
In summary, we found A103 to be the immunomarker with the highest sensitivity for identifying cells of adreno- cortical origin in destained Pap smears and cell-block sec- tions when compared with calretinin and inhibin alpha. Calretinin is comparable with A103 in sensitivity when used on cell-block sections alone and not on destained Pap smears. We have also found inhibin alpha to be the immunomarker with the lowest sensitivity in both cell- block sections and destained Pap smears. Both A103 and calretinin exhibited diffuse staining of good intensity in the majority of the andrenocortical lesions in comparison with inhibin alpha that exhibited weak and patchy stain- ing. Because A103 was positive in metastatic melanoma in the adrenal gland, the specificity of A103 was lower than the other two markers in this study. The results of this study suggest that A103 is the marker of choice for identifying cells of adrenocortical origin in limited diag- nostic material in cytology specimens if metastatic mela- noma is not a consideration.
References
1. Lappohn RE, Burger HG, Bouma J, Bangah M, de Bruijn HW. Inhibin as a marker for granulosa cell tumors. N Engl J Med 1989; 321:790-793.
2. Cooke I, O’Brien M, Charnock FM, Groome N, Ganesan TS. Inhibin as a marker for ovarian cancer. Br J Cancer 1995;71:1046-1050.
3. Busam KJ, Iverson K, Coplan KA, et al. Immunoreactivity for A 103, an antibody to melan-A (mart-1), in adrenocortical and other steroid tumors. Am J Surg Pathol 1998;22:57-63.
4. Papka RE, Collins J, Copelin T, Wilson K. Calretinin-innunoreactive cells in the uterus, pelvic autonomic ganglia, lumbrosacral spinal. Cell Tissue Res 1999;298:63-74.
5. Doglioni C, Tos AP, Laurino L, Iuzzolino P, Chiarelli C, Celio MR. Calretinin: a novel immunocytochemical marker for mesothelioma. Am J Surg Pathol 1996;20:1037-1046.
6. Rogers JH. Calretinin: a gene for a novel calcium-binding protein expressed principally in neurons. J Cell Biol 1987;105:1343-1353.
7. Cho EY, Ahn GH. Immunoexpression of inhibin alpha-subunit in adrenal neoplasm. Appl Immunohistochem Mol Morphol 2001;9: 222-228.
8. Pelkey TJ, Frierson HF, Mills SE, Stoler MH. The alpha sub- unit of inhibin in adrenal cortical neoplasia. Mod Pathol 1998;11: 516-24.
9. Renshaw AA, Granter SR. A comparison of A103 and inhibin reac- tivity in adrenal cortical tumors: distinction from hepatocellular car- cinoma and renal tumors. Mod Pathol 1998;11:1160-1164.
10. McCluggage WG, Burton J, Maxwell P, Sloan JM. Immunohisto- chemical staining of normal, hyperplastic, and neoplastioc adrenal cortex with a monoclonal antibody against alpha monoclonal anti- body against alpha monoclonal antibody against alpha inhibin. His- topathology 1997;30:518-522.
11. Jorda M, De Madeiros B, Nadji M. Calretinin and inhibin are useful in separating adrenocortical neoplasms from pheochromocytomas. Appl Immunohistochem Mol Morphol 2002;10:67-70.
12. Zhang PJ, Genega EM, Tomaszewski JE, Pasha TL, LiVolsi VA. The role of calretinin, inhibin, melan A, BCL-2, and C-kit in differ- entiating adrenal cortical and medullary tumors: an immunohisto- chemical study. Mod Pathol 2003;16:591-597.
13. Fetsch PA, Powers CN, Zakowski MF, Abati A. Anti-alpha-inhibin: marker of choice for the consistent distinction between adrenocorti- cal carcinoma and renal cell carcinoma in FNA. Cancer 1999;87: 168-172.
14. Shin SJ, Hoda RS, Ying L, Delellis RA. Diagnostic utility of the monoclonal antibody A103 in FNAB of the adrenal. Am J Clin Pathol 2000;113:295-302.