Skip to main content
NCBI home page
Lippincott Open Access logoLink to Lippincott Open Access
. 2024 Jan 25;48(4):377–386. doi: 10.1097/PAS.0000000000002186

GAD2 Is a Highly Specific Marker for Neuroendocrine Neoplasms of the Pancreas

Maximilian Lennartz *, Nick Benjamin Dünnebier *, Doris Höflmayer *, Sebastian Dwertmann Rico *, Simon Kind *, Viktor Reiswich *, Florian Viehweger *, Florian Lutz *, Christoph Fraune *, Natalia Gorbokon *, Andreas M Luebke *, Claudia Hube-Magg *, Franziska Büscheck *, Anne Menz *, Ria Uhlig *, Till Krech *,†, Andrea Hinsch *, Eike Burandt *, Guido Sauter *, Ronald Simon *,✉, Martina Kluth *, Stefan Steurer *, Andreas H Marx ‡, Patrick Lebok *,†, David Dum *, Sarah Minner *, Frank Jacobsen *, Till S Clauditz *, Christian Bernreuther *
PMCID: PMC10930383  PMID: 38271200

Abstract

Glutamate decarboxylase 2 (GAD2) is the most important inhibitory neurotransmitter and plays a role in insulin-producing β cells of pancreatic islets. The limitation of GAD2 expression to a few normal cell types makes GAD2 a potential immunohistochemical diagnostic marker. To evaluate the diagnostic utility of GAD2 immunohistochemistry, a tissue microarray containing 19,202 samples from 152 different tumor entities and 608 samples of 76 different normal tissue types was analyzed. In normal tissues, GAD2 staining was restricted to brain and pancreatic islet cells. GAD2 staining was seen in 20 (13.2%) of 152 tumor categories, including 5 (3.3%) tumor categories containing at least 1 strongly positive case. GAD2 immunostaining was most commonly seen in neuroendocrine carcinomas (58.3%) and neuroendocrine tumors (63.2%) of the pancreas, followed by granular cell tumors (37.0%) and neuroendocrine tumors of the lung (11.1%). GAD2 was only occasionally (<10% of cases) seen in 16 other tumor entities including paraganglioma, medullary thyroid carcinoma, and small cell neuroendocrine carcinoma of the urinary bladder. Data on GAD2 and progesterone receptor (PR) expression (from a previous study) were available for 95 pancreatic and 380 extrapancreatic neuroendocrine neoplasms. For determining a pancreatic origin of a neuroendocrine neoplasm, the sensitivity of GAD2 was 64.2% and specificity 96.3%, while the sensitivity of PR was 56.8% and specificity 92.6%. The combination of PR and GAD2 increased both sensitivity and specificity. GAD2 immunohistochemistry is a highly useful diagnostic tool for the identification of pancreatic origin in case of neuroendocrine neoplasms with unknown site of origin.

Key Words: GAD2, tissue microarray, immunohistochemistry, diagnostic marker, neuroendocrine neoplasms of the pancreas

Glutamate decarboxylase 2 (GAD2), also termed GAD65, is 1 out of 2 glutamate decarboxylases that are required for the decarboxylation of glutamate to gamma-aminobutyric acid (GABA).1 GABA is the most important inhibitory neurotransmitter in the central nervous system where it reduces neuronal excitability at nerve terminals and synapses (summarized in the study by Lee et al2). In the pancreas, GAD2 plays a role in insulin-producing β cells of pancreatic islets.3 GAD2 alterations play a critical role in several disease types. Downregulation of cerebral GAD2 expression has, for example, been described in autism4 and Alzheimer disease.5 GAD2 autoantibodies are found in up to 80% of patients with type 1 diabetes and have therefore been considered a possible cause of this disease.6 GAD2 autoantibodies are also associated with several rare neurological disorders, such as stiff-person syndrome, cerebellar ataxia, epilepsy, and limbic encephalitis (summarized in the study by Tsiortou et al7).

According to RNA expression data, GAD2 expression is strictly limited to the brain and the pancreas (https://www.proteinatlas.org/ENSG00000136750-GAD2/summary/rna 8–11). GAD2 immunohistochemistry (IHC) might therefore be of use for identifying tumors originating from the pancreas. Only a few authors have examined GAD2 expression in cancer and proposed a role for tumor progression in gastric and gallbladder cancer. Song et al12 reported a significantly higher GAD65 expression in 313 gastric cancers than in 60 adjacent nontumor tissues and found a link between high expression and the depth of tumor invasion, high TNM stage, and poor prognosis. For gallbladder adenocarcinoma, Deng and Pei13 found increased levels of GAD65 immunostaining as compared with peritumoral tissues and a significant link between high GAD65 expression and reduced patient survival. Other cancer entities have so far not been systematically analyzed for GAD2 protein expression, but Huang et al14 described GAD2 protein expression in 1 out of 3 cancers in patients with GAD2 autoantibodies and associated neurological disorders.

To comprehensively determine the prevalence of GAD2 protein expression in cancer and to assess the potential diagnostic utility of GAD2 IHC, we analyzed a pre-existing set of tissue microarrays (TMA) containing more than 19,000 tumor tissue samples from 152 different tumor types and subtypes as well as 76 non-neoplastic tissue categories for GAD2 expression by IHC in this study.

MATERIALS AND METHODS

Patient Samples

Two types of TMAs were included in our study. Our normal TMA was composed of 8 samples from 8 different donors from each of 76 different normal tissue types (608 samples on 1 slide). The tumor TMAs included a total of 19,202 primary tumors from 152 different tumor types and subtypes. The composition of normal and tumor TMAs is described in the Results section. All samples were from the archives of the Institutes of Pathology, University Hospital of Hamburg, Germany, the Institute of Pathology, Clinical Center Osnabrueck, Germany, and the Department of Pathology, Academic Hospital Fuerth, Germany. Tissues were fixed in 4% buffered formalin and then embedded in paraffin. The TMA manufacturing process was described earlier in detail.15,16 In brief, tissue spots (diameter: 0.6 mm) were transmitted from tumor-containing donor blocks to empty recipient paraffin blocks. Immunohistochemical data on progesterone receptor (PR) expression were available from a previous study.17 Conventional whole sections were taken from liver metastases of 15 cases of pancreatic neuroendocrine tumors (NETs) for the purpose of data validation. The use of archived remnants of diagnostic tissues for TMA manufacturing, their analysis for research purposes, and the use of patient data were according to local laws (HmbKHG, §12), and the analysis had been approved by the local ethics committee (Ethics Commission Hamburg, WF-049/09). All work has been carried out in compliance with the Helsinki Declaration.

Immunohistochemistry

Freshly prepared TMA sections were immunostained on one day in one experiment. Slides were deparaffinized with xylol, rehydrated through a graded alcohol series and exposed to heat-induced antigen retrieval for 5 minutes in an autoclave at 121°C in pH 9.0 DakoTarget Retrieval Solution (Agilent; #S2367). Endogenous peroxidase activity was blocked with Dako Peroxidase Blocking Solution (Agilent; #52023) for 10 minutes. Primary antibody specific for GAD2 (mouse monoclonal, MSVA-602M, MS Validated Antibodies, GmbH; #4975-602M) was applied at 37°C for 60 minutes at a dilution of 1:150. Only the TMA slides containing pancreatic neuroendocrine neoplasms were also stained by antibodies against insulin (recombinant rabbit monoclonal, HMV-308, MS Validated Antibodies, GmbH; cat#6515-308, dilution 1:150), glucagon (polyclonal rabbit, BioGenex, San Ramon; PU039-UP ) dilution 1:150, c-peptide (recombinant rabbit monoclonal, HMV-363, MS Validated Antibodies, GmbH; cat#6402-363, dilution 1:150), and pancreatic polypeptide (Abcam #ab272732, clone EPR-23320-10, dilution 1:450). For the purpose of antibody validation, the normal tissue TMA and the subset of our tumors were also analyzed by the rabbit monoclonal GAD2 antibody EPR22952-70 (cat. # ab239372, Abcam) at a dilution of 1:50,000 and an otherwise identical protocol. For the purpose of data validation, EPR22952-70 was also applied to 4 TMA sections containing 19 of our 29 tumors that had shown an unexpected GAD2 staining. Bound antibody was visualized using the EnVision Kit (Agilent; #K5007) according to the manufacturer’s directions. The sections were counterstained with hemalaun. For tumor tissues, the percentage of positive neoplastic cells was estimated, and the staining intensity was semiquantitatively recorded (0, 1+, 2+, 3+). For statistical analyses, the staining results were categorized into 4 groups. Tumors without any staining were considered negative. Tumors with 1+ staining intensity in ≤70% of tumor cells and 2+ intensity in ≤30% of tumor cells were considered weakly positive. Tumors with 1+ staining intensity in >70% of tumor cells, 2+ intensity in 31% to 70%, or 3+ intensity in ≤30% of tumor cells were regarded moderately positive. Tumors with 2+ intensity in >70% or 3+ intensity in >30% of tumor cells were considered strongly positive.

Statistics

Sensitivity and specificity for the detection of pancreatic neuroendocrine neoplasia and lung neuroendocrine neoplasia were calculated according to the following formulas: sensitivity = number of true positives divided by the number of true positives plus number of false negatives; specificity = number of true negatives divided by the number of true negatives plus number of false positives. The positive predictive value (PPV) was calculated using the formula PPV = (sensitivity × prevalence)/(sensitivity × prevalence + [1-specificity] × [1-prevalence]).

RESULTS

Technical Issues

A total of 17,507 (91.2%) of 19,202 tumor samples were interpretable in our tumor TMA analysis. Noninterpretable samples demonstrated a lack of unequivocal tumor cells or lack of the entire tissue spot. Sufficient numbers of samples (≥4) of each normal tissue type were evaluable.

GAD2 in Normal Tissue

A strong cytoplasmic GAD2 staining was seen in a large subset of cells of islets of Langerhans in the pancreas. A strong GAD2 staining also occurred in nerve fibers of the cerebrum and the cerebellum. All these findings were obtained by using the monoclonal mouse antibody MSVA-602M and the rabbit monoclonal antibody EPR22952-70 and therefore considered to be specific. Using MSVA-602M, GAD2 immunostaining was not seen in any other normal tissues including squamous epithelium, sebaceous glands, gastrointestinal epithelium, Brunner glands, gallbladder, exocrine pancreas, salivary glands, breast, endocervical glands, endometrium, fallopian tube, ovary, placenta, chorion cells, amnion cells, respiratory epithelium, lung, kidney, urothelium, prostate, seminal vesicles, testis, epididymis, thyroid (including C-cells), parathyroid, hypophysis, and the brain. There was, however, a staining of pigments (probably lipofuscin) in several organs including heart, adrenal gland, and the liver. Representative images of normal tissues are shown in Figure 1. EPR22952-70 did not show any pigment staining but resulted in significant nuclear staining in a broad range of different tissues. Pigment staining by MSVA-602M and nuclear staining by EPR22952-70 were considered antibody-specific cross-reactivities (Supplementary Fig. 1, Supplemental Digital Content 1, http://links.lww.com/PAS/B757).

FIGURE 1.

FIGURE 1

Open in a new tab

Glutamate decarboxylase 2 (GAD2) immunostaining of normal tissues. The panels show a strong fibrillar GAD2 staining of the gray matter of the cerebrum (while cell bodies of neurons are GAD2 negative; A) and a strong cytoplasmic GAD2 staining of pancreatic islet cells (B). GAD2 also labels pigment (probably lipofuscin) in samples of normal liver (C), adrenal gland (D), and the heart (E). GAD2 immunostaining is absent in colorectal epithelium (F).

GAD2 in Cancer

GAD2 immunostaining in tumors was always cytoplasmic. It was detectable in 114 (0.7%) of the 17,507 analyzable tumors, including 54 (0.3%) with weak, 29 (0.2%) with moderate, and 31 (0.2%) with strong immunostaining. Overall, 20 (13%) of 152 tumor categories showed detectable GAD2 expression, with 5 (3.3%) tumor categories including at least 1 case with strong positivity (Table 1). Representative images of GAD2-positive tumors are shown in Figure 2. GAD2 immunostaining was most commonly seen in neuroendocrine carcinomas (NECs) (58.3% of 12) and NETs (63.2% of 87) of the pancreas, followed by granular cell tumors (37.0%) and NETs of the lung (11.1%). GAD2 was only occasionally (<10% of cases) seen in 16 other tumor entities including paraganglioma, medullary thyroid carcinoma, ganglioneuroma, small cell NEC of the bladder, Ewing sarcoma, pancreatic/ampullary adenocarcinoma, ductal adenocarcinoma of the pancreas, gallbladder adenocarcinoma, gastric adenocarcinoma, adenocarcinoma of the prostate, serous carcinoma of the ovary, follicular thyroid carcinoma, anaplastic thyroid carcinoma, adrenal cortical adenoma, squamous cell carcinoma of the vulva, and Warthin tumor of the parotid gland. A graphical representation of a ranking order of GAD2-positive and strongly positive tumors is given in Fig. 3. For all 19 cases from tumor types with only occasional GAD2 positivity, the findings were confirmed by the rabbit monoclonal antibody EPR22952-70 (Supplementary Fig. 2, Supplemental Digital Content 2, http://links.lww.com/PAS/B758). These tumors included follicular thyroid carcinoma (n=1), medullary thyroid carcinoma (n=8), anaplastic thyroid carcinoma (n=1), pancreatic ampullary adenocarcinoma (n=2), ductal adenocarcinoma of the pancreas (n=2), paraganglioma (n=4), and ganglioneuroma (n=1). Whole-section analysis of liver metastases from pancreatic NETs revealed positive GAD2 staining in 10 of 15 cases (66.7%). Examples are shown in Supplementary Fig. 3, Supplemental Digital Content 3, http://links.lww.com/PAS/B759.

TABLE 1.

GAD2 Immunostaining in Human Tumors

GAD2 immunostaining
Tumor entity On TMA (n) Analyzable (n) Negative (%) Weak (%) Moderate (%) Strong (%)
Tumors of the skin Pilomatricoma 35 28 100.0 0.0 0.0 0.0
Basal cell carcinoma of the skin 89 81 100.0 0.0 0.0 0.0
Benign nevus 29 27 100.0 0.0 0.0 0.0
Squamous cell carcinoma of the skin 145 129 100.0 0.0 0.0 0.0
Malignant melanoma 65 61 100.0 0.0 0.0 0.0
Malignant melanoma lymph node metastasis 86 73 100.0 0.0 0.0 0.0
Merkel cell carcinoma 48 38 100.0 0.0 0.0 0.0
Tumors of the head and neck Squamous cell carcinoma of the larynx 109 97 100.0 0.0 0.0 0.0
Squamous cell carcinoma of the pharynx 60 50 100.0 0.0 0.0 0.0
Oral squamous cell carcinoma (floor of the mouth) 130 115 100.0 0.0 0.0 0.0
Pleomorphic adenoma of the parotid gland 50 44 100.0 0.0 0.0 0.0
Warthin tumor of the parotid gland 104 100 99.0 1.0 0.0 0.0
Adenocarcinoma, NOS (papillary cystadenocarcinoma) 14 12 100.0 0.0 0.0 0.0
Salivary duct carcinoma 15 12 100.0 0.0 0.0 0.0
Acinic cell carcinoma of the salivary gland 181 149 100.0 0.0 0.0 0.0
Adenocarcinoma NOS of the salivary gland 109 95 100.0 0.0 0.0 0.0
Adenoid cystic carcinoma of the salivary gland 180 140 100.0 0.0 0.0 0.0
Basal cell adenocarcinoma of the salivary gland 25 23 100.0 0.0 0.0 0.0
Basal cell adenoma of the salivary gland 101 88 100.0 0.0 0.0 0.0
Epithelial-myoepithelial carcinoma of the salivary gland 53 52 100.0 0.0 0.0 0.0
Mucoepidermoid carcinoma of the salivary gland 343 299 100.0 0.0 0.0 0.0
Myoepithelial carcinoma of the salivary gland 21 20 100.0 0.0 0.0 0.0
Myoepithelioma of the salivary gland 11 9 100.0 0.0 0.0 0.0
Oncocytic carcinoma of the salivary gland 12 12 100.0 0.0 0.0 0.0
Polymorphous adenocarcinoma, low grade of the salivary gland 41 36 100.0 0.0 0.0 0.0
Pleomorphic adenoma of the salivary gland 53 38 100.0 0.0 0.0 0.0
Tumors of the lung, pleura, and thymus Adenocarcinoma of the lung 196 190 100.0 0.0 0.0 0.0
Squamous cell carcinoma of the lung 80 73 100.0 0.0 0.0 0.0
Small cell carcinoma of the lung 16 15 100.0 0.0 0.0 0.0
Mesothelioma, epithelioid 40 36 100.0 0.0 0.0 0.0
Mesothelioma, biphasic 77 67 100.0 0.0 0.0 0.0
Thymoma 29 28 100.0 0.0 0.0 0.0
Lung, neuroendocrine tumor (NET) 29 27 88.9 11.1 0.0 0.0
Tumors of the female genital tract Squamous cell carcinoma of the vagina 78 64 100.0 0.0 0.0 0.0
Squamous cell carcinoma of the vulva 157 144 99.3 0.7 0.0 0.0
Squamous cell carcinoma of the cervix 136 127 100.0 0.0 0.0 0.0
Adenocarcinoma of the cervix 23 22 100.0 0.0 0.0 0.0
Endometrioid endometrial carcinoma 338 297 100.0 0.0 0.0 0.0
Endometrial serous carcinoma 86 64 100.0 0.0 0.0 0.0
Carcinosarcoma of the uterus 57 52 100.0 0.0 0.0 0.0
Endometrial carcinoma, high grade, G3 13 10 100.0 0.0 0.0 0.0
Endometrial clear cell carcinoma 9 8 100.0 0.0 0.0 0.0
Endometrioid carcinoma of the ovary 130 120 100.0 0.0 0.0 0.0
Serous carcinoma of the ovary 580 551 99.8 0.0 0.2 0.0
Mucinous carcinoma of the ovary 101 89 100.0 0.0 0.0 0.0
Clear cell carcinoma of the ovary 51 48 100.0 0.0 0.0 0.0
Carcinosarcoma of the ovary 47 47 100.0 0.0 0.0 0.0
Granulosa cell tumor of the ovary 44 38 100.0 0.0 0.0 0.0
Leydig cell tumor of the ovary 4 4 100.0 0.0 0.0 0.0
Sertoli cell tumor of the ovary 1 1 100.0 0.0 0.0 0.0
Sertoli Leydig cell tumor of the ovary 3 3 100.0 0.0 0.0 0.0
Steroid cell tumor of the ovary 3 3 100.0 0.0 0.0 0.0
Brenner tumor 41 41 100.0 0.0 0.0 0.0
Tumors of the breast Invasive breast carcinoma of no special type 1764 1663 100.0 0.0 0.0 0.0
Lobular carcinoma of the breast 363 336 100.0 0.0 0.0 0.0
Medullary carcinoma of the breast 34 32 100.0 0.0 0.0 0.0
Tubular carcinoma of the breast 29 25 100.0 0.0 0.0 0.0
Mucinous carcinoma of the breast 65 56 100.0 0.0 0.0 0.0
Phyllodes tumor of the breast 50 44 100.0 0.0 0.0 0.0
Tumors of the digestive system Adenomatous polyp, low-grade dysplasia 50 50 100.0 0.0 0.0 0.0
Adenomatous polyp, high-grade dysplasia 50 50 100.0 0.0 0.0 0.0
Adenocarcinoma of the colon 2483 2306 100.0 0.0 0.0 0.0
Gastric adenocarcinoma, diffuse type 215 202 100.0 0.0 0.0 0.0
Gastric adenocarcinoma, intestinal type 215 204 100.0 0.0 0.0 0.0
Gastric adenocarcinoma, mixed type 62 61 98.4 1.6 0.0 0.0
Adenocarcinoma of the esophagus 83 71 100.0 0.0 0.0 0.0
Squamous cell carcinoma of the esophagus 76 61 100.0 0.0 0.0 0.0
Squamous cell carcinoma of the anal canal 91 83 100.0 0.0 0.0 0.0
Cholangiocarcinoma 58 54 100.0 0.0 0.0 0.0
Gallbladder adenocarcinoma 51 49 98.0 2.0 0.0 0.0
Klatskin tumor 42 35 100.0 0.0 0.0 0.0
Hepatocellular carcinoma 312 274 100.0 0.0 0.0 0.0
Ductal adenocarcinoma of the pancreas 659 616 99.7 0.3 0.0 0.0
Pancreatic/ampullary adenocarcinoma 98 97 96.9 3.1 0.0 0.0
Acinar cell carcinoma of the pancreas 18 17 100.0 0.0 0.0 0.0
Gastrointestinal stromal tumor (GIST) 62 61 100.0 0.0 0.0 0.0
Appendix, neuroendocrine tumor (NET) 25 22 100.0 0.0 0.0 0.0
Colorectal, neuroendocrine tumor (NET) 12 11 100.0 0.0 0.0 0.0
Ileum, neuroendocrine tumor (NET) 53 52 100.0 0.0 0.0 0.0
Pancreas, neuroendocrine tumor (NET) 101 87 36.8 16.1 17.2 29.9
Colorectal, neuroendocrine carcinoma (NEC) 14 12 100.0 0.0 0.0 0.0
Ileum, neuroendocrine carcinoma (NEC) 8 8 100.0 0.0 0.0 0.0
Gallbladder, neuroendocrine carcinoma (NEC) 4 4 100.0 0.0 0.0 0.0
Pancreas, neuroendocrine carcinoma (NEC) 14 12 41.7 33.3 16.7 8.3
Tumors of the urinary system Noninvasive papillary urothelial carcinoma, pTa G2 low grade 177 173 100.0 0.0 0.0 0.0
Noninvasive papillary urothelial carcinoma, pTa G2 high grade 141 140 100.0 0.0 0.0 0.0
Noninvasive papillary urothelial carcinoma, pTa G3 219 175 100.0 0.0 0.0 0.0
Urothelial carcinoma, pT2-4 G3 735 652 100.0 0.0 0.0 0.0
Squamous cell carcinoma of the bladder 22 20 100.0 0.0 0.0 0.0
Small cell neuroendocrine carcinoma of the bladder 23 16 93.8 6.3 0.0 0.0
Sarcomatoid urothelial carcinoma 25 19 100.0 0.0 0.0 0.0
Urothelial carcinoma of the renal pelvis 62 55 100.0 0.0 0.0 0.0
Clear cell renal cell carcinoma 1287 1146 100.0 0.0 0.0 0.0
Papillary renal cell carcinoma 368 322 100.0 0.0 0.0 0.0
Clear cell (tubulo) papillary renal cell carcinoma 26 24 100.0 0.0 0.0 0.0
Chromophobe renal cell carcinoma 170 153 100.0 0.0 0.0 0.0
Oncocytoma of the kidney 257 223 100.0 0.0 0.0 0.0
Tumors of the male genital organs Adenocarcinoma of the prostate, Gleason 3+3 83 78 100.0 0.0 0.0 0.0
Adenocarcinoma of the prostate, Gleason 4+4 80 70 100.0 0.0 0.0 0.0
Adenocarcinoma of the prostate, Gleason 5+5 85 81 98.8 1.2 0.0 0.0
Adenocarcinoma of the prostate (recurrence) 258 237 100.0 0.0 0.0 0.0
Small cell neuroendocrine carcinoma of the prostate 19 12 100.0 0.0 0.0 0.0
Seminoma 682 674 100.0 0.0 0.0 0.0
Embryonal carcinoma of the testis 54 50 100.0 0.0 0.0 0.0
Leydig cell tumor of the testis 31 23 100.0 0.0 0.0 0.0
Sertoli cell tumor of the testis 2 2 100.0 0.0 0.0 0.0
Sex cord stromal tumor of the testis 1 1 100.0 0.0 0.0 0.0
Spermatocytic tumor of the testis 1 1 100.0 0.0 0.0 0.0
Yolk sac tumor 53 47 100.0 0.0 0.0 0.0
Teratoma 53 47 100.0 0.0 0.0 0.0
Squamous cell carcinoma of the penis 92 72 100.0 0.0 0.0 0.0
Tumors of endocrine organs Adenoma of the thyroid gland 113 108 100.0 0.0 0.0 0.0
Papillary thyroid carcinoma 391 349 100.0 0.0 0.0 0.0
Follicular thyroid carcinoma 154 145 99.3 0.7 0.0 0.0
Medullary thyroid carcinoma 111 106 92.5 1.9 4.7 0.9
Parathyroid gland adenoma 43 35 100.0 0.0 0.0 0.0
Anaplastic thyroid carcinoma 45 42 97.6 2.4 0.0 0.0
Adrenal cortical adenoma 50 48 97.9 2.1 0.0 0.0
Adrenal cortical carcinoma 28 28 100.0 0.0 0.0 0.0
Pheochromocytoma 50 50 100.0 0.0 0.0 0.0
Tumors of hemotopoetic and lymphoid tissues Hodgkin’s lymphoma 103 94 100.0 0.0 0.0 0.0
Small lymphocytic lymphoma, B-cell type (B-SLL/B-CLL) 50 39 100.0 0.0 0.0 0.0
Diffuse large B-cell lymphoma (DLBCL) 113 92 100.0 0.0 0.0 0.0
Follicular lymphoma 88 67 100.0 0.0 0.0 0.0
T-cell non-Hodgkin’s lymphoma 25 21 100.0 0.0 0.0 0.0
Mantle cell lymphoma 18 18 100.0 0.0 0.0 0.0
Marginal zone lymphoma 16 14 100.0 0.0 0.0 0.0
Diffuse large B-cell lymphoma (DLBCL) in the testis 16 15 100.0 0.0 0.0 0.0
Burkitt lymphoma 5 5 100.0 0.0 0.0 0.0
Tumors of soft tissue and bone Tendosynovial giant cell tumor 45 41 100.0 0.0 0.0 0.0
Granular cell tumor 53 46 63.0 23.9 10.9 2.2
Leiomyoma 50 50 100.0 0.0 0.0 0.0
Leiomyosarcoma 94 90 100.0 0.0 0.0 0.0
Liposarcoma 145 129 100.0 0.0 0.0 0.0
Malignant peripheral nerve sheath tumor (MPNST) 15 15 100.0 0.0 0.0 0.0
Myofibrosarcoma 26 26 100.0 0.0 0.0 0.0
Angiosarcoma 74 64 100.0 0.0 0.0 0.0
Angiomyolipoma 91 89 100.0 0.0 0.0 0.0
Dermatofibrosarcoma protuberans 21 19 100.0 0.0 0.0 0.0
Ganglioneuroma 14 14 92.9 0.0 0.0 7.1
Kaposi sarcoma 8 4 100.0 0.0 0.0 0.0
Neurofibroma 117 117 100.0 0.0 0.0 0.0
Sarcoma, not otherwise specified (NOS) 74 70 100.0 0.0 0.0 0.0
Paraganglioma 41 41 90.2 9.8 0.0 0.0
Ewing sarcoma 23 20 95.0 5.0 0.0 0.0
Rhabdomyosarcoma 7 6 100.0 0.0 0.0 0.0
Schwannoma 122 121 100.0 0.0 0.0 0.0
Synovial sarcoma 12 11 100.0 0.0 0.0 0.0
Osteosarcoma 44 35 100.0 0.0 0.0 0.0
Chondrosarcoma 40 37 100.0 0.0 0.0 0.0
Rhabdoid tumor 5 5 100.0 0.0 0.0 0.0
Solitary fibrous tumor 17 16 100.0 0.0 0.0 0.0
Open in a new tab

FIGURE 2.

FIGURE 2

Open in a new tab

GAD2 immunostaining in tumors. The panels show a cytoplasmic glutamate decarboxylase 2 (GAD2) staining of variable intensity in pancreatic neuroendocrine tumors (A–C) and carcinomas (D) including cases with a weak to moderate staining of most tumor cells while only few scattered tumor cells show strong GAD2 staining (C, D). A diffuse cytoplasmic GAD2 positivity also occurs in cases of medullary carcinoma of the thyroid (E), granular cell tumor (F), and ganglioneuroma (G). GAD2 staining is absent in a pancreatic acinar cell carcinoma (H).

FIGURE 3.

FIGURE 3

Open in a new tab

Ranking order of GAD2 immunostaining in tumors. Both the percentage of positive cases (blue dots) and the percentage of strongly positive cases (orange dots) are shown. GAD2 indicates glutamate decarboxylase 2.

Comparison With PR Staining (Neuroendocrine Tumors)

Data on the expression of PR were available for 475 of our neuroendocrine neoplasms for which GAD2 data were collected in our project. In this cohort, a GAD2 staining was seen in 61 (64.2%) of 95 NETs and NECs from the pancreas and in 14 (3.7%) of 380 neuroendocrine neoplasms from other sites. PR positivity had been recorded in 54 (56.8%) of 95 pancreatic NET/NECs and in 28 (7.4%) of 380 extrapancreatic neuroendocrine neoplasms.17 With respect to the capacity to identify a pancreatic origin in a cohort of neuroendocrine neoplasms, this resulted in a sensitivity of 64.2% and a specificity of 96.3% for GAD2, as well as a sensitivity of 56.8% and a specificity of 92.6% for PR. Both sensitivity and specificity could be improved by the combination of GAD2 and PR data (Table 2). For tumors having either PR or GAD2 positivity, the sensitivity rose to 78.9%. For tumors having PR and GAD2 positivity, the specificity increased to 99.2%. If medullary carcinomas of the thyroid were excluded because of their overrepresentation in our cohort, there was a sensitivity of 64.2% and a specificity of 97.8% for GAD2 as well as a sensitivity of 56.8% and a specificity of 97.8% for PR. A detailed comparison of GAD2 and PR data is given for NET entities in Figure 4.

TABLE 2.

Sensitivity and Specificity of GAD2 and Progesterone Receptor (PR) Immunostaining

Including medullary thyroid carcinomas Excluding medullary thyroid carcinomas
Sensitivity Specificity Sensitivity Specificity
GAD2 only 0.64 0.96 0.64 0.98
PR only 0.57 0.93 0.57 0.98
GAD2 and/or PR 0.79 0.90 0.79 0.96
GAD2 and PR 0.42 0.99 0.42 1.00
Open in a new tab

FIGURE 4.

FIGURE 4

Open in a new tab

GAD2 and PR immunostaining in neuroendocrine neoplasms. GAD2 indicates glutamate decarboxylase 2; PR, progesterone receptor.

Comparison With Other Pancreatic Markers (Pancreatic NETs)

A total of 53 NET/NECs of the pancreas, including 30 GAD2-positive and 23 GAD2-negative tumors, were also successfully stained for insulin, glucagon, pancreatic polypeptide, and c-peptide. Eleven (36.7%) of 30 GAD2-positive tumors and 14 (60.8%) of the 23 GAD2-negative tumors were negative for all the other markers. The staining results are summarized in Supplementary Figure 4, Supplemental Digital Content 4, http://links.lww.com/PAS/B760.

PPV of GAD2

To estimate the diagnostic utility of GAD2 immunostaining for the differential diagnosis of NETs originating from the lung versus pancreatic NETs, the PPV of GAD2 immunostaining was estimated. On the basis of our data for GAD2 to detect lung NETs (sensitivity: 6%, specificity: 83%, estimated prevalence 40% of all NETs18) and pancreatic NETs (sensitivity: 64%, specificity: 96%, estimated prevalence 10% of all NETs18), the calculated PPVs of GAD2 for lung and pancreatic NETs were 10.8% and 87.7%, respectively.

DISCUSSION

Our comprehensive analysis of 17,507 tumors from 152 different entities revealed significant rates of GAD2 positivity in only few tumor entities. That these primarily included NETs and NECs of the pancreas is not surprising, given the strong GAD2 immunostaining of normal pancreatic islet cells. The 37% positivity rate of granular cell tumors is consistent with the neural origin of this tumor entity.19 Medullary thyroid cancer, 7.5% GAD2 positive in our cohort, is the only tumor category with a significant frequency of GAD2-positive cases lacking a precursor cell with known GAD2 expression. That only few other tumor entities contained (weakly) GAD2-positive cases in <10% of cases fit well with the summarized data from the The Cancer Genome Atlas database where only very few nonpancreatic neoplasms displayed detectable but very low-level GAD2 RNA expression (https://www.proteinatlas.org/ENSG00000136750-GAD2/summary/rna). It is of note that other authors have described immunohistochemical GAD2 positivity in 71% of 313 gastric cancers,12 55% of 108 adenocarcinomas of the gallbladder,13 and in 30% of 46 parathyroidal adenomas.13 However, GAD2 immunostaining was only observed in 1 (0.2%) of 467 gastric and 1 (2%) of 49 gallbladder adenocarcinomas. The use of different reagents and experimental conditions is the most likely reason for these discrepant results.

On the basis of our data, GAD2 IHC may represent a highly useful additional tool for the distinction of neuroendocrine neoplasms from different sites of origin, resulting in a PPV of 87.7% for a pancreatic derivation. Up to 20% of NETs/NECs present as metastasis of unknown origin (summarized in the study by Bellizzi.20). Even in cases with disseminated metastasis, the determination of the organ of tumor origin is important for these patients because both treatment and prognosis depend on the site of tumor origin. The median survival of rectal, small intestinal, colon, gastric, bronchopulmonary, and pancreatic NETs was described to average 70, 51, 41, 41, 25, and 22 months, respectively.21 Everolimus and capecitabine/temozolomide are frequently given to patients with pancreatic NETs but almost never in case of NETs of the midgut.22 If the primary tumor is in the small intestine, tumor resection may be needed to prevent intestinal obstruction and bleeding. Although the histologic features can provide some clues on the origin of NETs/NECs, antibody panels are regularly used for the identification of the site of tumor origin (summarized in the study by Bellizzi23). The comparable rate of GAD2 positivity in metastatic lesions (67%), as in primary pancreatic NET/NECs (58% to 63%), suggests that the GAD2 expression status is largely retained during metastatic spread.

The most commonly used markers to suggest a pancreatic NET/NEC origin include islet 1, PR, PAX6, and polyclonal PAX8.23,24 Islet 1 has a reported sensitivity of up to 75% but was also found to be positive in duodenal and rectal NETs as well as in medullary carcinomas of the thyroid.23,25 PAX6, which may also be recognized by cross-reacting polyclonal PAX8 antibodies, has been described to have a sensitivity of up to 70% for pancreatic NETs, but it was also found positive in NETs of the duodenum, rectum, and the lung.24,26 PR positivity has been found in 58% to 67% of pancreatic NETs but was also seen in NETs of the tubal gut, small intestine, and lung.27,28 As we had previously analyzed PR immunostaining in the same cohort of patients,17 we were able to compare the performance of PR and GAD2 for the distinction of a pancreatic origin of neuroendocrine neoplasms. Although this comparison revealed a slightly superior sensitivity and specificity of GAD2 as compared with PR for detecting a pancreatic NET/NEC origin, the added value of GAD2 IHC appears to come from the combination with PR analysis. That 21 (22.1%) of our pancreatic NET/NECs were PR negative but GAD2 positive demonstrates that GAD2 IHC could potentially improve the existing panels for the distinction of neuroendocrine neoplasms. This is also supported by GAD2 positivity in 11 out of 25 pancreatic NET/NECs that were negative for typical pancreatic islet cell proteins (ie, insulin, c-peptide, glucagon, and pancreatic polypeptide). It is also of note that our 63% GAD2 positivity rate in pancreatic NETs would probably increase to some extent if >0.6 mm tissue was analyzed per tumor. IHC studies on TMAs often result in a slight underestimation of the positivity rates due to the occasional sampling of poorly immunoreactive tumor areas.29 Studies are now needed to evaluate GAD2 in combination with currently used antibody panels in larger cohorts of patients.

Given the large scale of our study, emphasis was placed on a thorough validation of our assay. The International Working Group for Antibody Validation has proposed that antibody validation for IHC on formalin-fixed tissues should include either a comparison of the findings obtained by 2 different independent antibodies or a comparison with expression data obtained by another independent method.30 RNA data obtained in 3 independent RNA screening studies, including the Human Protein Atlas RNA-sequencing tissue data set,11 the Functional ANnoTation Of the Mammalian genome project,9,10 and the Genotype-Tissue Expression project8 had identified GAD2 RNA only in the brain and the pancreas. That the immunohistochemical analysis of 76 different normal tissue categories by using MSVA-602M resulted in an almost complete restriction of the staining to brain and pancreas strongly validates our assay. The same cell types as detected by MSVA-602M were also positive by EPR22952-70 both in normal and in neoplastic tissues, which is an additional verification of our data. The occasional staining of pigments by MSVA-602M and the nuclear staining seen by EPR22952-70 in all organs were considered antibody-specific cross-reactivities. It is of note that the use of a very broad range of different tissues for antibody validation increases the likelihood of detecting undesired cross-reactivities because virtually all proteins occurring in normal cells of adult humans are subjected to the validation experiment. Our panel of 76 different normal tissues assures that virtually all proteins occurring in cells of adult humans are subjected to cross-reactivity screening. That unexpected GAD2 staining observed with our assay in individual tumors was always confirmed by EPR22952-70 further emphasizes the strength of our resource-demanding validation procedure.

In summary, our data provide a comprehensive overview on GAD2 expression in normal and neoplastic human tissues. The strong predilection of GAD2 immunostaining to normal and neoplastic neuroendocrine pancreatic cells strongly suggests a diagnostic utility of GAD2 IHC for determining a pancreatic origin of neuroendocrine neoplasms of unknown derivation.

Supplementary Material

pas-48-377-s001.pdf (1.2MB, pdf)
Open in a new tab

graphic file with name pas-48-377-s002.jpg

Open in a new tab

graphic file with name pas-48-377-s003.jpg

Open in a new tab

graphic file with name pas-48-377-s004.jpg

Open in a new tab

ACKNOWLEDGMENTS

The authors thank Laura Behm, Inge Brandt, Maren Eisenberg, and Sünje Seekamp for excellent technical assistance.

Footnotes

C.B., M.L., N.B.D., R.S., M.K., and G.S.: contributed to the conception, design, data collection, data analysis, and manuscript writing. M.L., C.B., N.B.D., D.H., S.D.R., S.K., V.R., F.V., F.L., C.F., N.G., A.M.L., F.B., A.M., R.U., T.K., A.H., E.B., S.S., A.H.M., P.L., D.D., S.M., F.J., and T.S.C.: participated in pathology data analysis, data interpretation, and collection of samples M.L., R.S., M.K., and C.H.M.: data analysis C.B., R.S., and G.S.: study supervision.

Conflicts of Interest and Source of Funding: The antibodies against GAD2 (MSVA-602M), insulin (HMV-308), and c-peptide (HMV-363) were provided by MS Validated Antibodies GmbH (owned by a family member of G.S.). For the remaining authors none were declared.

Supplemental Digital Content is available for this article. Direct URL citations are provided in the HTML and PDF versions of this article on the journal's website, www.ajsp.com.

Contributor Information

Maximilian Lennartz, Email: m.lennartz@uke.de.

Nick Benjamin Dünnebier, Email: Nick.Duennebier@gmx.de.

Doris Höflmayer, Email: d.hoeflmayer@uke.de.

Sebastian Dwertmann Rico, Email: s.dwertmann-rico@uke.de.

Simon Kind, Email: s.kind@uke.de.

Viktor Reiswich, Email: v.reiswich@uke.de.

Florian Viehweger, Email: f.viehweger@uke.de.

Florian Lutz, Email: f.lutz@uke.de.

Christoph Fraune, Email: c.fraune@uke.de.

Natalia Gorbokon, Email: n.gorbokon@uke.de.

Claudia Hube-Magg, Email: c.hube@uke.de.

Franziska Büscheck, Email: f.buescheck@uke.de.

Anne Menz, Email: a.menz@uke.de.

Ria Uhlig, Email: r.uhlig@uke.de.

Till Krech, Email: t.krech@uke.de.

Andrea Hinsch, Email: a.hinsch@uke.de.

Eike Burandt, Email: e.burandt@uke.de.

Guido Sauter, Email: g.sauter@uke.de.

Ronald Simon, Email: r.simon@uke.de.

Martina Kluth, Email: m.kluth@uke.de.

Stefan Steurer, Email: s.steurer@uke.de.

Andreas H. Marx, Email: Andreas.Marx@klinikum-fuerth.de;a.marx@uke.de.

Patrick Lebok, Email: p.lebok@uke.de.

David Dum, Email: d.dum@uke.de.

Sarah Minner, Email: s.minner@uke.de.

Frank Jacobsen, Email: f.jacobsen@uke.de.

Till S. Clauditz, Email: t.clauditz@uke.de.

Christian Bernreuther, Email: c.bernreuther@uke.de.

REFERENCES

  • 1. Erlander MG, Tillakaratne NJ, Feldblum S, et al. Two genes encode distinct glutamate decarboxylases. Neuron. 1991;7:91–100. [DOI] [PubMed] [Google Scholar]
  • 2. Lee SE, Lee Y, Lee GH. The regulation of glutamic acid decarboxylases in GABA neurotransmission in the brain. Arch Pharm Res. 2019;42:1031–1039. [DOI] [PubMed] [Google Scholar]
  • 3. Baekkeskov S, Aanstoot HJ, Christgau S, et al. Identification of the 64K autoantigen in insulin-dependent diabetes as the GABA-synthesizing enzyme glutamic acid decarboxylase. Nature. 1990;347:151–156. [DOI] [PubMed] [Google Scholar]
  • 4. Wiebe S, Nagpal A, Truong VT, et al. Inhibitory interneurons mediate autism-associated behaviors via 4E-BP2. Proc Natl Acad Sci USA. 2019;116:18060–18067. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5. Schwab C, Yu S, Wong W, et al. GAD65, GAD67, and GABAT immunostaining in human brain and apparent GAD65 loss in Alzheimer’s disease. J Alzheimers Dis. 2013;33:1073–1088. [DOI] [PubMed] [Google Scholar]
  • 6. Atkinson MA, Maclaren NK, Scharp DW, et al. 64,000 Mr autoantibodies as predictors of insulin-dependent diabetes. Lancet. 1990;335:1357–1360. [DOI] [PubMed] [Google Scholar]
  • 7. Tsiortou P, Alexopoulos H, Dalakas MC. GAD antibody-spectrum disorders: progress in clinical phenotypes, immunopathogenesis and therapeutic interventions. Ther Adv Neurol Disord. 2021;14:17562864211003486. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8. Consortium GT. The genotype-tissue expression (GTEx) project. Nat Genet. 2013;45:580–585. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9. Lizio M, Abugessaisa I, Noguchi S, et al. Update of the FANTOM web resource: expansion to provide additional transcriptome atlases. Nucleic Acids Res. 2019;47(D1):D752–D758. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10. Lizio M, Harshbarger J, Shimoji H, et al. Gateways to the FANTOM5 promoter level mammalian expression atlas. Genome Biol. 2015;16:22. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11. Thul PJ, Akesson L, Wiking M, et al. A subcellular map of the human proteome. Science. 2017;356:820–eaal3321. [DOI] [PubMed] [Google Scholar]
  • 12. Song Y, Li S, Wang Z, et al. The role of glutamate decarboxylase 65 in gastric cancer development, progression, and prognosis. Histopathology. 2013;63:334–342. [DOI] [PubMed] [Google Scholar]
  • 13. Deng X, Pei D. Ornithine decarboxylase and glutamate decarboxylase 65 as prognostic markers of gallbladder malignancy: a clinicopathological study in benign and malignant lesions of the gallbladder. Mol Med Rep. 2013;7:413–418. [DOI] [PubMed] [Google Scholar]
  • 14. Huang J, Li H, Zhou R, et al. Clinical heterogeneity in patients with glutamate decarboxylase antibody. Neuroimmunomodulation. 2019;26:234–238. [DOI] [PubMed] [Google Scholar]
  • 15. Bubendorf L, Nocito A, Moch H, et al. Tissue microarray (TMA) technology: miniaturized pathology archives for high-throughput in situ studies. J Pathol. 2001;195:72–79. [DOI] [PubMed] [Google Scholar]
  • 16. Kononen J, Bubendorf L, Kallioniemi A, et al. Tissue microarrays for high-throughput molecular profiling of tumor specimens. Nat Med. 1998;4:844–847. [DOI] [PubMed] [Google Scholar]
  • 17. Viehweger F, Tinger LM, Dum D, et al. Diagnostic and prognostic impact of progesterone receptor immunohistochemistry: a study evaluating more than 16,000 tumors. Anal Cell Pathol (Amst). 2022;2022:6412148. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18. Caldarella A, Crocetti E, Paci E. Distribution, incidence, and prognosis in neuroendocrine tumors: a population based study from a cancer registry. Pathol Oncol Res. 2011;17:759–763. [DOI] [PubMed] [Google Scholar]
  • 19. Singh VA, Gunasagaran J, Pailoor J. Granular cell tumour: malignant or benign? Singapore Med J. 2015;56:513–517. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20. Bellizzi AM. Assigning site of origin in metastatic neuroendocrine neoplasms: a clinically significant application of diagnostic immunohistochemistry. Adv Anat Pathol. 2013;20:285–314. [DOI] [PubMed] [Google Scholar]
  • 21. Man D, Wu J, Shen Z, et al. Prognosis of patients with neuroendocrine tumor: a SEER database analysis. Cancer Manag Res. 2018;10:5629–5638. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22. Kunz PL, Reidy-Lagunes D, Anthony LB, et al. Consensus guidelines for the management and treatment of neuroendocrine tumors. Pancreas. 2013;42:557–577. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23. Bellizzi AM. Immunohistochemistry in the diagnosis and classification of neuroendocrine neoplasms: what can brown do for you? Hum Pathol. 2020;96:8–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24. Lai JP, Mertens RB, Mirocha J, et al. Comparison of PAX6 and PAX8 as immunohistochemical markers for pancreatic neuroendocrine tumors. Endocr Pathol. 2015;26:54–62. [DOI] [PubMed] [Google Scholar]
  • 25. Agaimy A, Erlenbach-Wunsch K, Konukiewitz B, et al. ISL1 expression is not restricted to pancreatic well-differentiated neuroendocrine neoplasms, but is also commonly found in well and poorly differentiated neuroendocrine neoplasms of extrapancreatic origin. Mod Pathol. 2013;26:995–1003. [DOI] [PubMed] [Google Scholar]
  • 26. Lorenzo PI, Jimenez Moreno CM, Delgado I, et al. Immunohistochemical assessment of Pax8 expression during pancreatic islet development and in human neuroendocrine tumors. Histochem Cell Biol. 2011;136:595–607. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27. Viale G, Doglioni C, Gambacorta M, et al. Progesterone receptor immunoreactivity in pancreatic endocrine tumors. An immunocytochemical study of 156 neuroendocrine tumors of the pancreas, gastrointestinal and respiratory tracts, and skin. Cancer. 1992;70:2268–2277. [DOI] [PubMed] [Google Scholar]
  • 28. Stashek KM, Czeczok TW, Bellizzi AM. Extensive evaluation of immunohistochemistry to assign site of origin in well-differentiated neuroendocrine tumors: a study of 10 markers in 265 tumors. Modern Pathol. 2014;27:160a–160aa. [Google Scholar]
  • 29. Fraune C, Burandt E, Simon R, et al. MMR deficiency is homogeneous in pancreatic carcinoma and associated with high density of Cd8-positive lymphocytes. Ann Surg Oncol. 2020;27:3997–4006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30. Uhlen M, Bandrowski A, Carr S, et al. A proposal for validation of antibodies. Nat Methods. 2016;13:823–827. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from The American Journal of Surgical Pathology are provided here courtesy of Wolters Kluwer Health

RESOURCES