Q RT-PCR Detection of Chromogranin A

Abstract

Objective:

Message and protein expression of CgA was examined to evaluate the sensitivity of a PCR-based approach in the detection of covert neuroendocrine (NE) tissue.

Summary Background Data:

Immunohistochemical (IHC) measurement of chromogranin A (CgA) discriminates gastrointestinal (GI) carcinoids from epithelial tumors. IHC is, however, an insensitive technique to identify micrometastases or delineate subpopulations of NE cells.

Methods:

CgA gene expression was examined by Q-RT PCR in GI carcinoids (small intestinal and metastases, n = 17, gastric, n = 5), appendiceal tumors (n = 10), and adenocarcinomas (gastric, n = 5, colorectal, n = 6). CgA protein expression levels were quantitatively analyzed following IHC by automated quantitative analysis (AQUA) in 2 tissue microarrays (GI carcinoid and GI adenocarcinoma).

Results:

CgA gene was overexpressed (P < 0.001) in GI carcinoids compared with GI adenocarcinomas and normal mucosa. Elevated levels (P < 0.00001) were also identified in carcinoid liver and lymph node (LN) metastases. CgA levels were higher (∼2–4-fold) in NE appendiceal carcinoids than in adenocarcinoids, but in GI adenocarcinomas were identical to normal mucosa. Histologically normal lymph nodes expressed detectable CgA message in 30% of cases. CgA protein levels were highest in primary GI carcinoids and in liver metastases and significantly elevated (P < 0.005) compared with nonmetastatic lesions. Expression in liver and LN metastases was significantly elevated (P < 0.000001) compared with normal. Analysis of mRNA by Q-RT PCR was >200-fold more sensitive than by IHC.

Conclusions:

Overexpression of CgA mRNA and protein in GI carcinoids can identify metastatic cells; thus, PCR for CgA can be used to identify micrometastases not evident by light microscopy or IHC as well as define tumors of ambivalent morphologic phenotype. The use of this sensitive strategy to assess NETs and apparently normal LNs and liver may be of future utility in defining therapeutic strategy.

Q RT-PCR is > 200-fold more sensitive than immunohistochemistry for the identification of chromogranin A in gastrointestinal neuroendocrine tumors and their metastases. The technique allows detection of micrometastases in normal tissues and the delineation of tumors of indeterminate phenotype. The application of this molecular strategy will redefine future surgical strategy.

Chromogranin A (CgA), present in secretory dense core granules,¹ is widely used as an immunohistochemical (IHC) marker of neuroendocrine tumors (NETs). In addition, since CgA is cosecreted with the amines and peptides that are present in the neurosecretory granules, it is also useful as a serum marker.² Indeed, since the majority of peptide-producing endocrine neoplasms including gastrointestinal (GI) carcinoids secrete CgA, it is considered a definitive NE tumor marker for diagnosis and therapeutic evaluation.³ Observations that CgA serum or plasma levels reflect tumor load and may be an independent marker of prognosis in patients with midgut carcinoids have led to its widespread clinical utility in NET disease.⁴

A substantial body of data, however, has accumulated indicating that NE cells are also present in many tumors of nonendocrine origin.^5–9 Although the reasons for this are unknown, it has been proposed that this observation may reflect neuronal dedifferentiation of such lesions.¹⁰ Although there are limited data available concerning serum levels of CgA in subjects with nonendocrine tumors, elevated levels have been reported in association with prostatic,^11,12 non-small cell lung,¹³ ovarian, pancreatic, and colonic neoplasia.¹⁴

The sensitivity of detecting NE cells in non-NE tumors is dependent on the tumor type, the NE marker used, and the detection technique (IHC, serum analysis or PCR of NE markers). At this time, analysis of gene levels by PCR is the most sensitive method for the detection of cell-specific markers.¹⁵

Our hypothesis is that since CgA is a marker that specifically identifies gastrointestinal NETs, a more sensitive method for detection of this marker using a PCR approach would: 1) be more effective than conventional immunohistochemistry, 2) alter tumor staging, and 3) differentiate GI NETs from phenotypically indeterminate tumors and GI adenocarcinomas. We used a real-time quantitative reverse transcription-PCR (Q RT-PCR) approach to identify the presence of NET cells (by CgA) in primary tumors and metastases, and confirmed the expression, utility, and specificity of this marker. We then examined protein expression of CgA immunohistochemically to assess the utility of protein assessment and compared this with the real-time PCR technique.

MATERIALS AND METHODS

These studies were approved by the Human Investigations Committee at the Yale University School of Medicine.

Patients and Samples

Tissue specimens: Tumor tissue was collected from 43 patients (26 male, 17 female; median age, 44 years; range, 11–73 years) with histologically proven small bowel carcinoid tumors or metastases (n = 17), gastric carcinoid tumors (n = 5), NE appendiceal carcinoid tumors (n = 5), mixed cell (goblet) appendiceal adenocarcinoid tumors (n = 5), gastric adenocarcinomas (n = 5), and colorectal carcinomas (n = 6) who had either undergone resection of the primary tumor between 1997 and 2003 in the Yale University Department of Surgery or from the Cooperative Human Tissue Network, which is funded by the National Cancer Institute. Primary tumors (n = 35), liver metastases (n = 5), and 13 histologically positive lymph nodes were studied in total. Paired normal tissue samples were also obtained from adjacent, macroscopically normal, nontumor mucosa, liver, or lymph nodes in 26 patients. In addition, 10 other histologically negative lymph nodes were collected for Q RT-PCR and IHC analysis.

Construction and Processing of GI Carcinoid TMA (YTMA60)

Formalin-fixed, paraffin-embedded tissue blocks containing GI carcinoids (stomach, n = 6; duodenum, n = 4; small bowel, n = 55; appendix, n = 25 [these included samples from 5 NE appendiceal carcinoid tumors and 5 adenocarcinoids, goblet cell tumors, used in Q RT-PCR]; colon, n = 12) were retrieved, along with the corresponding hematoxylin and eosin-stained slides, from the archives of the Yale University School of Medicine Department of Pathology. Blocks were stored under ambient conditions within a temperature range of 18°C–37°C. To ensure uniformity of sectioning, older paraffin blocks were melted and re-embedded using modern-day plastic cassettes. Areas of carcinoid, distinct from normal tissue elements, were identified by a researcher (R.L.C.) and marked for subsequent retrieval and analysis. Core biopsies of 0.6 mm in diameter were taken from each donor block and arrayed into a recipient paraffin block (45 mm × 20 mm) using a tissue puncher/arrayer (Beecher Instruments, Silver Spring, MD) as described by Xie et al.¹⁶ The array contained 102 cases of primary GI carcinoids and matched normal tissue, 15 lymph node metastases. and 7 liver metastases diagnosed between 1965 and 2003. Follow-up information was available (median follow-up. 39 months; range, 1–456 months) for all patients (Table 1). The control comprised a multitissue array YTMA25, which included GI adenocarcinomas from the stomach, colon, and liver (n = 10 samples each) as well as normal tissue samples (n = 7/organ) from each site.

TABLE 1. Clinical Characteristics of Carcinoid Tissue Microarray (YTMA60) (n = 102 Patients)

Tissue Techniques

RNA Isolation

Total RNA was isolated from GI carcinoid tumor tissue (n = 32), GI carcinoid metastases (n = 20), GI adenocarcinomas (n = 11), and normal mucosa (n = 26) using TRIzol reagent (Invitrogen, Carlsbad, CA) as described.¹⁷ RNA was dissolved in DEPC water, measured spectrophotometrically, and an aliquot analyzed using Bioanalyzer to assess the quality of RNA isolated.

Q RT-PCR

Primary small intestinal carcinoid tumors (n = 7), metastases to the liver (n = 7) or lymph nodes (n = 13), gastric carcinoids (n = 5), gastric adenocarcinomas (n = 5), colorectal carcinomas (n = 6), and control tissues (small intestine [n = 5], stomach [n = 5], liver [n = 3], lymph node [n = 13]) were examined by Q RT-PCR. CgA and GAPDH message were quantitatively measured as described.¹⁷ Q RT-PCR was performed using the ABI 7900 Sequence Detection System. Total RNA from each sample was subjected to reverse transcription using the High Capacity cDNA Archive Kit (Applied Biosystems, Foster City, CA) following the manufacturer's suggestions. Briefly, 2 μg of total RNA in 50 μL of water was mixed with 50 μL of 2× RT mix containing Reverse Transcription Buffer, dNTPs, random primers and Multiscribe Reverse Transcriptase. RT reaction was carried out in a thermal cycler for 10 minutes at 25°C followed by 120 minutes at 37°C. Real-time PCR analysis was then performed in triplicate. Briefly, cDNA in 7.2 μL of water was mixed with 0.8 μL of 20× Assays-on-Demand primer (CgA = Hs00174938, GAPDH = Hs99999905) and probe mix, 8 μL of 2× TaqMan Universal Master mix in a 384-well optical reaction plate. The following PCR conditions were used: 50°C for 2 minutes, then 95°C for 10 minutes, followed by 40 cycles at 95°C/0.15 minutes and 60°C/1 minute. A standard curve was generated for each gene using cDNA obtained by pooling equal amounts from each sample. The expression level of target genes was normalized to internal GAPDH. Data were analyzed using Microsoft Excel and calculated using the relative standard curve method (ABI, User Bulletin #2).

Tissue Microarray Immunostaining, Image Acquisition, and Data Analysis

Slides from the carcinoid tissue microarray (TMA) and the adenocarcinoma array were stained as described.^17,18 For antigen retrieval purpose, sections were immersed in citrate buffer (10 mm sodium citrate, pH 6.0), and subjected to 1 × 10 minutes high temperature-high pressure treatment followed by the treatment with 0.3% H₂O₂ in methanol for 30 minutes at 37°C to inactivate endogenous peroxidase. Slides were incubated for 24 hours at 4°C with a 1:1000 dilution of the anti-CgA mouse monoclonal antibody (DAKO Corp, Carpinteria, CA). Goat anti-mouse antibodies conjugated to a horseradish peroxidase-decorated dextran polymer backbone (Envision; DAKO Corp) was used as a secondary reagent. For automated analysis, NET cells or normal mucosal epithelia were identified by the use of a fluorescently tagged anticytokeratin antibody cocktail (AE1/AE3; DAKO Corp), nuclei were visualized by 4′,6-diamidino-2-phenylindole (DAPI), and targets visualized with a fluorescent chromogen (Cy-5-tyramide; NEN Life Science Products, Boston, MA).^17,19 Briefly, monochromatic, high-resolution (1024 × 1024 pixel; 0.5-μm) images were obtained of each histospot. Areas of tumor or normal epithelia from stromal elements were distinguished by creating a mask from the cytokeratin signal. Coalescence of cytokeratin at the cell surface localized the cell membranes, and DAPI was used to identify nuclei. The CgA signal from the tumor cells or epithelial cells was scored and expressed as signal intensity divided by the cytokeratin mask area. Histospots containing <10% tumor, as assessed by mask area (automated), were excluded from further analysis. Previous studies have demonstrated that the staining from a single histospot provides a sufficiently representative sample for analysis.²⁰ Disc scores from the same tumor were averaged to produce a single score. It has been determined previously that analysis of 2 cores concurs with analysis of an entire tissue section with >95% accuracy.²⁰

Statistical Analysis

Results were expressed as mean ± SEM; n indicates the numbers of patients in each study group. Statistical significance was calculated by the 2-tailed Student t test for paired and unpaired values as appropriate, with a probability of <0.05 representing significance. The χ² test was performed to evaluate the statistical significance of 2 groups.

RESULTS

Real-time Q RT-PCR

Small Intestinal Carcinoids

Messenger RNA levels of CgA were >3 Log (>1000×) higher in small intestinal carcinoid tumors (P < 0.000001) compared with normal mucosa (Figure 1). Message levels were similarly elevated (P < 0.02) in liver and lymph node metastatic tissue compared with normal liver and normal lymph node material respectively.

FIGURE 1. Message levels of Chromogranin A (CgA) determined by Q RT-PCR. Levels were corrected against expression of the housekeeping gene, GAPDH. Levels of CgA were significantly overexpressed (3-Log) in small intestinal carcinoid tumor (ST) samples compared with normal mucosa (SN) (*P < 0.000001) and in liver metastases (LVM) compared with normal liver (LVN) (^#P < 0.02). CgA was elevated in lymph node metastases (LNM) compared with normal lymph nodes (LNN) (^#P < 0.02). Data are mean ± SEM. S, small bowel; LN, lymph node; LV, liver; N, normal; T, tumor; M, metastases.

Gastric Carcinoids

The expression of CgA was examined in 2 different types of gastric carcinoids: nonmetastatic type I and II tumors and malignant (capable of metastases) type III/IV tumors. Q RT-PCR analysis demonstrated that mRNA levels of CgA were significantly elevated in both the nonmalignant type I/II carcinoids (P < 0.01) and malignant type III/IV gastric carcinoid tumors (P < 0.005) compared with normal mucosa (Figure 2). A significant difference was noted in expression levels between the nonmetastatic and malignant tumor types; higher levels were recorded in the latter (P < 0.05).

FIGURE 2. Message levels of CgA determined by Q RT-PCR in gastric carcinoid specimens. Levels of CgA were significantly overexpressed (10×) in nonmetastatic (GNM) gastric carcinoid tumors (type I/II tumors) compared with normal mucosa (GN) and in malignant (GM) gastric carcinoids (>200×) (type III/IV). Levels in the latter were significantly elevated compared with the nonmalignant tumors (^#P < 0.05, *P < 0.01, **P < 0.005). Data are mean ± SEM. G, gastric; N, normal; NM, nonmetastatic; M, metastatic.

Specificity of CgA as a Marker Gene for GI Carcinoid Tumors

Having identified that the gene expression levels of the CgA were elevated in GI carcinoid tumors, we next examined the expression of this NE marker gene in gastric adenocarcinomas and colorectal carcinomas. In contrast to the gastric and small intestinal carcinoids that demonstrated significantly elevated CgA levels (P < 0.005), message for CgA was not elevated in adenocarcinomas of either the stomach or colon (Figure 3). Levels of CgA in adenocarcinomas were not significantly different from normal mucosa.

FIGURE 3. Message levels of CgA determined by Q RT-PCR in gastric adenocarcinomas and colorectal adenocarcinomas compared with gastric carcinoids and small intestinal carcinoids. Levels of CgA were elevated in carcinoids but were not different between normal mucosa and tumor samples (*P < 0.005). Data are mean ± SEM.

The log-difference in CgA expression between gastric carcinoids and gastric adenocarcinomas was ∼2-Log (>100×) while the difference between small intestinal carcinoids and colorectal carcinomas was ∼3-Log (>1000×). This highlights that PCR levels of CgA are at least >100 times higher in GI carcinoids than in GI carcinomas.

Appendiceal Carcinoids and Mixed Cell Appendiceal Tumors

The expression of CgA was examined in 2 different types of appendiceal tumors: “incidental” NE appendiceal carcinoid tumors (<1cm in size with no evidence of metastasis) and mixed (goblet) cell appendiceal adenocarcinoids (>1.5 cm, with evidence of metastatic invasion). Q RT-PCR analysis demonstrated that mRNA levels of CgA were significantly elevated in both the NE appendiceal carcinoids (P < 0.01) and adenocarcinoid tumors (P < 0.05) compared with normal mucosa (Figure 4A). Higher CgA levels were noted in the nonmetastatic compared with metastatic tumors.

FIGURE 4. Message levels of CgA determined by Q RT-PCR or IHC in appendiceal carcinoids and appendiceal adenocarcinoids compared with normal mucosa. A, Levels of CgA message were elevated in neuro endocrine appendiceal carcinoids (ANM; ∼200×) and adenocarcinoids (AM, ∼50×) compared with normal appendiceal mucosa (AN) (*P < 0.05; **P < 0.01). Data are mean ± SEM. B, On the TMA, CgA levels were higher in neuro endocrine appendiceal carcinoids (ANM; ∼4×) and adenocarcinoids (AM; 2×) compared with normal appendiceal mucosa (AN). Neuroendocrine carcinoids had significantly higher CgA levels than adenocarcinoids (*P < 0.05; **P < 0.001; ^#P < 0.0002). Data are mean ± SEM. AN, normal appendix; ANM, nonmalignant appendix; AM, malignant appendix.

CgA in Histologically Positive and Negative Lymph Nodes

Levels of CgA were highly expressed in histologically lymph node-positive samples compared with negative samples (P < 0.002) (Figure 5). Levels were low in all histologically documented lymph node-negative samples, except for 3 cases that exhibited significantly (P > 0.05) elevated CgA expression. Three of the 10 specimens (30%) could therefore be upgraded to exhibiting molecular positivity for CgA although they are characterized as histologically negative for metastases.

FIGURE 5. Message levels of CgA in histologically positive (LNM) and histologically negative (LNN) lymph nodes. CgA was significantly overexpressed in positive lymph nodes compared with negative lymph nodes. Three histologically lymph node-negative samples had elevated CgA message. ⋄, lymph node negative sample with abnormally elevated gene expression. *P < 0.002. Data are mean ± SEM. LNM, lymph node metastases; LNN, lymph node negative.

Protein Immunohistochemistry

CgA in GI Carcinoids and Adenocarcinomas

The immunohistochemical expression of CgA in GI carcinoids was examined by immunohistochemistry and AQUA (automated quantitative analysis) quantitation of staining intensity using YTMA60 (Figure 6A). Expression values of the CgA protein were correlated with clinical evidence of disease (liver metastases or lymph node metastases). Expression levels of CgA were elevated in primary carcinoid tumors that developed liver or lymph node metastases compared with both nonmetastatic primary tumors (P < 0.005) and to normal mucosa (P < 0.00001) (Figure 6B). Similarly, levels of CgA were elevated in liver metastatic tissue (P < 0.00001) and lymph node metastatic tissue (P < 0.000001) compared with normal tissues (Figure 6B). Levels of CgA in primary tumors that were nonmetastatic (no clinical or pathologic evidence of liver or lymph node metastases) was increased compared with normal mucosa (P < 0.00005) but was lower than CgA levels in metastatic tissue. This was significant for liver metastases (P < 0.005) but not lymph node metastases (P = 0.068).

FIGURE 6. Expression levels of CgA determined by immunohistochemistry and AQUA quantitation on YTMA60. A, Three-color image of a small bowel carcinoid demonstrating significant overlap between cytokeratin and cytoplasmic CgA staining in the carcinoid tumor (inset). Immunostaining of CgA was invariably cytoplasmic localized. Blue, nuclei (DAPI); green, tumor mask (cytokeratin–Alexa488); red, CgA (Cy5). Dual membrane staining (red and green) results in yellow. (100× magnification). B, AQUA levels of CgA were significantly overexpressed in primary malignant GI carcinoid tumors (TM), and primary nonmetastatic carcinoid tumors (TNM) compared with normal small intestinal mucosa (SN). Expression levels of CgA were elevated in lymph node metastatic tissue (LNM) and in liver metastases (LVM) compared with normal lymph nodes (LNN) and normal liver (LVN), respectively. In addition, CgA levels were elevated in liver metastases compared with primary nonmalignant tumors (*P < 0.00001; ^#P < 0.005; ^@P < 0.00005). Data are mean ± SEM. LV, liver; LN, lymph node; S, small bowel; N, normal; T, tumor; M, metastases; NM, no metastases.

Analysis of the gastric and small intestinal tumor groups individually demonstrated that CgA levels were significantly elevated (P < 0.0003 and P < 0.0000001, respectively) in each of these tumor types compared with normal mucosa.

Analysis of the appendiceal samples demonstrated that both NE appendiceal carcinoids (626 ± 70; P < 0.0001) and the more malignant adenocarcinoids (316 ± 20; P < 0.05) had elevated CgA levels compared with normal appendiceal mucosa (145 ± 24) (Figure 4B). The NE carcinoids had significantly elevated CgA expression compared with the appendiceal adenocarcinoids (P < 0.002).

In comparison, the adenocarcinoma array (YTMA25), demonstrated expression levels of CgA to be indistinguishable in adenocarcinomas of the stomach, liver, or colon when compared with normal mucosa from each of these sites (Figure 7).

FIGURE 7. Expression levels of CgA determined by immunohistochemistry and AQUA quantitation on YTMA25. A, Three-color image of a colorectal adenocarcinoma. No overlap was shown between cytokeratin and cytoplasmic CgA staining in the adenocarcinoma tumor (insets). Blue, nuclei (DAPI); green, tumor mask (cytokeratin–Alexa488); red, CgA (Cy5) (100× magnification). B, AQUA levels of CgA were not significantly elevated in gastric (GT) or colorectal adenocarcinomas (CT) or in hepatocellular carcinomas (LVT) compared with normal mucosa from each of these sites. Data are mean ± SEM. C, colorectal; G, gastric; LV, liver; N, normal; T, tumor.

The log-difference in CgA protein expression between carcinoid tumors and normal mucosa on YTMA60 using the DAKO antibody immunostaining was in the order of 1-Log (∼5×; tumor = 1000 versus normal small intestinal mucosa = 200) (Figure 7A). PCR analysis of CgA mRNA demonstrated a >3 Log (>1000×) difference between carcinoid tumors and normal samples (Figure 1). This further emphasizes the different sensitivities of these 2 methodologies and demonstrates that PCR is ∼200 times more sensitive than protein immunostaining in the detection of CgA in tumors.

DISCUSSION

These data demonstrate that the molecular biologic and peptide detection of CgA in GI tissue specimens is sensitive and specific for the identification of GI carcinoid tumors and that these techniques preferentially identify tumors and metastases that are derived from NE cells (carcinoid tumors) as opposed to epithelial cell-derived tumors (adenocarcinomas).

Message of CgA are expressed at ∼1000 × the level in malignant GI carcinoids compared with normal mucosa. Detectable levels of CgA in normal mucosa (small intestine/gastric) reflects the presence of CgA-expressing endocrine cells in these tissues and further emphasizes the sensitivity of the technique since it detects NE cells, which represent approximately 1 per 2000 epithelial cells.²¹ Since endocrine cells constitute ∼1% by volume of the GI mucosa,²² the detection in normal mucosa further confirms the sensitivity of PCR as an identification tool and emphasizes its ability to detect disease at a cellular level. Within non-NE tumors, NE cells are on occasion identifiable as diffusely scattered or multifocal aggregations located in small nests.²³ Message level of CgA in non-NE GI tumors by PCR was similar or below levels in normal mucosa. The identification of CgA-positive tumor cells (∼1%) on YTMA25 demonstrates that these tumors do contain NE cells. The sensitivity of the technique is, however, substantial enough that the low CgA PCR signal is able to identify these few cells. Irrespective of the origin of these cells (tumor dedifferentiation, sequestration of neighboring NE cell populations), they are identifiable by the PCR approach. An increasing CgA signal will reflect increasing numbers of NE cells. The PCR approach is therefore of utility in discriminating mixed tumors with both epithelial and NE cell components. This is of importance since the biologic behavior of such mixed lesions is often different from that of the pure phenotype and such information will enable reconfiguration of therapeutic strategy and may well have prognostic implications.

Message levels of CgA were elevated in all 4 types of gastric carcinoid tumors but were higher in the malignant (type III/IV) group as compared with the less aggressive type I/II group. This might also be a reflection of tumor size since type III/IV lesions tend to be large and solitary while the latter are small and multiple.²² This differential CgA expression also occurs in serum measurements where type III tumors have elevated levels compared with type I carcinoids.²⁴ This relationship, noted at both an mRNA and secreted protein level, indicates that transcription of the CgA gene and its secreted product is closely regulated and that message levels identified using PCR reflect intracellular and serum marker levels.

Message levels of CgA were elevated in both types of appendiceal carcinoid tumors but were higher in the NE appendiceal carcinoids compared with the more aggressive mixed cell goblet adenocarcinoids. This is a reflection of NE cell numbers since the latter lesions have a mixed phenotype with a partial NE differentiation and intestinal type goblet cell morphology in contrast to the carcinoids that are completely NE cell in origin.²⁵ These differences demonstrate the utility of a PCR approach in discriminating these 2 tumor types and have considerable clinical relevance in determining therapeutic strategy.

Three of 10 histologically negative lymph nodes were identified to express significant levels of CgA. Light microscopic examination of tissue sections, typically a 4-μm thickness, cannot accurately identify malignancy when micrometastases (a few cells) may be present within a lymph node. Our results suggest that 30% of lymph nodes could potentially be upgraded to histologically positive samples that contain carcinoid micrometastases. This percentage alteration in tumor (malignancy) stage is similar to other studies in different GI (gastric and colorectal) organ systems.^26,27

Although protein expression can be measured using Western blot, a tissue microarray-based immunohistochemical approach allows direct quantitation of expression of a protein within a tumor cell and reduces nonspecific signal from nontumor material. We therefore used this technique to investigate whether the differences in gene expression in CgA translated into differences in protein expression levels. Our studies demonstrated, similar to the PCR studies, that CgA was overexpressed in primary malignant small intestinal and gastric carcinoids compared with normal tissue. The sensitivity of this was ∼5-fold compared with the ∼100- to 1000-fold difference noted for PCR. This difference in detection level is predictable since CgA message (∼50-fold) and protein (∼2-fold) were also decreased in appendiceal adenocarcinoids, a tumor that contains fewer NE cells. While the multilog difference in CgA expression levels between PCR and an AQUA approach may represent different sensitivities, they may also represent true differences in protein versus RNA expression due to post-translational events (eg, protein degradation or sequestering in granules). Immunohistochemistry, which is often more readily available than frozen tissue and Q RT-PCR facilities, is a potentially useful approach for discriminating between tumor types. Its discriminant ability is, however, limited since it is less sensitive than a PCR-based approach.

CONCLUSION

Our data confirm overexpression of CgA mRNA and protein in tumor and metastatic GI carcinoid tissue and demonstrate the utility of CgA (mRNA or protein) to identify metastatic gastrointestinal carcinoid cells (tissue) in both liver and lymph nodes. In addition, this strategy can discriminate mixed phenotypic tumors and confirms that CgA is a cell-specific NE marker. The ability to identify metastatic tissue by CgA expression with such amplified sensitivity indicates that this technique has application in the identification of GI carcinoid tumor metastases at levels that cannot be identified by conventional (immunohistochemical) techniques. The implications for altering staging and therapeutic strategy are of considerable clinical relevance.