|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 25, No 15 (May 20), 2007: pp. 1967-1973 © 2007 American Society of Clinical Oncology. DOI: 10.1200/JCO.2006.10.1535 Chromogranin A: Is It a Useful Marker of Neuroendocrine Tumors?
From the Department of Internal Medicine and Gastroenterology, S. Orsola-Malpighi Hospital, University of Bologna, Bologna, Italy Address reprint requests to Paola Tomassetti, MD, Department of Internal Medicine and Gastroenterology, University of Bologna, Policlinico S. Orsola-Malpighi, Via Massarenti, 9, 40138, Bologna, Italy; e-mail: paola.tomassetti{at}unibo.it
Purpose We evaluated the pattern of chromogranin A (CgA) plasma levels in a large number of patients with neuroendocrine tumors (NETs), in a series of patients with chronic atrophic gastritis (CAG) with and without enterochromaffin-like (ECL) cell hyperplasia, and in healthy participants (HPs). Patients and Methods Two hundred thirty-eight patients with NETs, 42 patients with CAG with or without ECL cell hyperplasia, and 48 HPs were studied. All patients underwent a baseline visit, biochemical routine check-up, imaging techniques, endoscopy, and histologic determination. Results CgA plasma levels were higher in patients with NETs compared with CAG patients or HPs (P < .001). In the NET group, we observed higher CgA levels in patients with diffuse disease compared with patients with local or hepatic disease (P < .001). CgA plasma levels were significantly higher in patients with Zollinger-Ellison syndrome compared with other types of endocrine tumors (P < .001). We found the best cutoff range between HPs and NET patients to be 18 to 19 U/L (sensitivity, 85.3%; specificity, 95.8%). Comparing all participants without neoplasia (HPs, CAG patients, and disease-free patients) and patients with endocrine tumors, the best cutoff range was 31 to 32 U/L (sensitivity, 75.3%; specificity, 84.2%). Setting the specificity at 95%, the cutoff range was 84 to 87 U/L (sensitivity, 55%). Conclusion Our study confirms the high specificity and sensitivity of CgA in diagnosing an endocrine tumor. It is necessary to use a cutoff range of 84 to 87 U/L to obtain a high specificity in diagnosing NETs, with the aim of excluding patients in whom the CgA was elevated as a result of other non-neoplastic diseases.
Chromogranin A (CgA) is an acidic glycoprotein with a molecular mass of 49 kd that is widely expressed by neuroendocrine cells and constitutes one of the most abundant components of secretory granules.1 CgA is physiologically released by exocytosis and may be detected in the blood. In particular, when a tumor develops in an endocrine tissue, it becomes the main source of circulating CgA.2,3 High CgA levels have been demonstrated in the serum or plasma of patients with different types of endocrine tumors such as pheochromocytoma, medullary thyroid carcinoma, and enterochromaffin and pancreatic islet cell tumors.1,4-9 In particular, circulating CgA levels have been claimed to be useful markers for neuroendocrine tumors (NETs), with a high specificity and a sensitivity ranging from 27% to 81%.7,10 Recent studies indicate that circulating CgA levels correlate positively with an enterochromaffin-like (ECL) cell mass in patients with autoimmune chronic atrophic gastritis,11,12 gastrinoma, and multiple endocrine neoplasia syndrome type 17,13 who may develop potentially malignant gastric carcinoids.14 However, these studies included only a small number of patients, except for two recent prospective studies; one of these two studies was performed in patients with gastrinomas,15 and the other study was performed in patients with various types of gastroenteropancreatic (GEP) endocrine tumors.16 Furthermore, no studies have compared CgA plasma levels in patients with endocrine tumors with CgA levels in patients with chronic atrophic gastritis (CAG). The aim of this study was to evaluate the pattern of CgA plasma levels in a large number of patients with endocrine tumors localized in different tissues and at different stages of the disease, in a series of patients with CAG with and without ECL cell hyperplasia, and in healthy participants (HPs).
We evaluated 280 consecutive patients and 48 HPs at the Center for the Study and Treatment of Gastro-Entero-Pancreatic Tumors in the Department of Internal Medicine and Gastroenterology, University of Bologna, from January 2004 to June 2006. All 280 patients enrolled underwent a baseline visit including clinical history and a biochemical routine check-up. Imaging techniques (ultrasound, computed tomography, somatostatin receptor scintigraphy, and gastrointestinal endoscopy) were performed to evaluate the presence of neoplasia and its localization and stage. Histologic determinations with a Ki-67% index were performed. In all patients with CAG, we carried out an endoscopy with multiple biopsies in the antrum, body, and fundus to confirm the presence of atrophy and to detect ECL cell hyperplasia. On the basis of these examinations, the participants were divided into three different groups. Group A included 238 patients with NETs (130 men and 108 women; mean age, 59.1 years; range, 26 to 85 years) localized in the stomach in 14 patients (5.9%), in the lung in 20 patients (8.4%), in the GI tract in 85 patients (35.7%), and in the pancreas in 94 patients (39.5%). In 25 (10.5%) of 238 patients, Zollinger-Ellison syndrome (ZES) was present. These patients were considered as a separate subgroup because, in our previous study, we had already stated that they showed extremely higher CgA levels compared with patients with endocrine tumors without ZES.7 The patients were considered disease free when the tumor was totally removed by surgery and when there was no evidence of the disease at imaging techniques. Patients with local disease had a primary lesion with or without positive local lymph nodes, whereas hepatic disease was defined as hepatic metastases with or without local disease. Patients with diffuse disease had distant metastases (other than the liver) with or without local disease. Group B included 42 patients with CAG (15 men and 27 women; median age, 56.1 years; range, 24 to 79 years); 29 patients (69%) had ECL cell hyperplasia, and 13 patients (31%) did not. Group C included 48 participants without apparent pathologic alterations and considered to be in good health (24 men and 24 women; median age, 55.5 years; range, 19 to 86 years). None of the 280 patients or 48 HPs had renal insufficiency (evaluated based on the creatinine concentrations), and except the patients with ZES, none of the participants were treated with proton pump inhibitors at the time of the study. All three groups were comparable for sex (P = .076) and age (P = .170).
Ethics
CgA The CgA plasma level was measured with a technique previously validated in our laboratory7 and commercially available (DAKO CgA ELISA kit; Dako A/S, Copenhagen, Denmark). In brief, the CgA determination uses an immunoenzymatic sandwich methodology. Furthermore, three plasma samples obtained from a pool of our samples with low, medium, and high CgA concentrations were used for the calculation of the intra- and interassay coefficients. CgA was measured within a run and on consecutive days on 12 replicates. The mean coefficient of variation ranged from 2.2% (intra-assay) to 9.5% (interassay), and it was independent of the CgA concentration in the range of the selected samples (20 to 300 U/L). The recovery and the dynamic range of linearity are similar to those reported in the assay kit performance.
Statistics The analyses were performed using SPSS version 13.0 for Windows (SPSS Inc, Chicago, IL). Two-tailed P < .05 was considered statistically significant.
Differences Between the Three Groups The CgA data in the three groups of participants are listed in Table 1. CgA plasma levels were significantly different among the three different groups (P < .001); furthermore, we observed lower levels in HPs than in the other two groups (P < .001), whereas there was no statistically significant difference between the NET group and the CAG group (P = .114). Considering the overall population studied, there was no significant relationship between CgA and sex (P = .409) or age (P = .804).
Differences Between Subgroups The CgA data in the subgroups of NET and CAG patients are listed in Table 2 and shown in Figure 1. In the NET group, we observed higher CgA levels in patients with diffuse disease compared with patients with local (P < .001) or hepatic disease (P < .001). Otherwise, we did not observe a statistically significant difference in CgA levels between patients with local or hepatic disease (P = .563). In the disease-free patients, CgA levels were lower than the levels found in patients at any other stage of endocrine neoplastic disease (P < .001). In the CAG group, patients with hyperplasia had significantly higher CgA levels compared with patients who did not have hyperplasia (P = .032).
Other Subgroup Evaluations In the NET group, we excluded patients who were disease free at the time of sampling with the aim of better differentiating CgA levels in patients with endocrine tumors compared with patients with CAG to evaluate CgA as a diagnostic parameter of neoplasia. In 170 patients with endocrine tumors, the mean CgA level ± standard deviation was 594.4 ± 1,850.0. CgA levels in patients with neoplasia were significantly higher than the levels in patients with CAG (P < .001; effect = 308.5; 95% CI, 180.1 to 528.4); this was true even when we compared patients with local disease with patients with CAG (P < .006; effect = 186.0; 95% CI, 120.1 to 288.1).
Pharmacologic Treatment
Other Factors in the NET Group
Receiver Operating Characteristic In Table 4, we reported the area under the curve, the best cutoff range, and the sensitivity and specificity of CgA in the diagnosis of endocrine tumors, comparing different groups. Figure 2 shows the receiver operating characteristic curves. We excluded the 68 patients who were disease free at the time of sampling in the NET group to calculate the receiver operating characteristic curve.
The best cutoff range between HPs and NET patients was 18 to 19 U/L, with a sensitivity of 85.3% and a specificity of 95.8%, and the best cutoff range between CAG and NET patients was 53 to 54 U/L, with a sensitivity of 66.5% and a specificity of 71.4%. Considering the various factors that could have an impact on CgA levels, we also compared all of the participants without neoplasias (HPs, CAG patients, and disease-free patients) and patients with endocrine tumors. The best cutoff range was 31 to 32 U/L, with a sensitivity of 75.3% and a specificity of 84.2%. Setting the specificity at a 95% value, the cutoff was 84 to 87 U/L with a sensitivity of 55% (93 of 170 participants). To distinguish between patients with limited disease (local and/or liver) and patients with diffuse disease, another ROC curve was constructed, and the best cutoff point was calculated. The area under the curve was 0.805 ± 0.047. Using 281 to 282 U/L as the best cutoff range, the sensitivity and specificity of CgA were 71.1% and 78.8%, respectively. Setting the specificity at a 95% value, the cutoff was 564 to 603 U/L with a sensitivity of 55% (21 of 38 patients).
Circulating CgA levels have been said to be a useful marker for NETs, with a high specificity and sensitivity, but most studies included only a small number of patients.7,10 In the present study, we investigated the diagnostic value of plasma CgA in a large number of patients with endocrine tumors localized in different tissues and at different stages of the disease compared with a series of patients with CAG, with and without ECL cell hyperplasia, and HPs. In agreement with previous studies, our findings demonstrate that abnormally high plasma CgA levels are a characteristic feature of patients with endocrine tumors7,13,16 and patients with CAG.12 We have found that patients with endocrine tumors showed higher levels of CgA than HPs.7,13,16 Using a best cutoff range of 18 to 19 U/L, we obtained a specificity of 95.8% and a sensitivity of 85.3%. These findings are in agreement with others reported in the literature, where a variable cutoff value from 17 to 34 U/L was identified, with a variable specificity (83% to 91%)13,17-19 and sensitivity (79% to 92%).7,13,17-19 In our series, the higher specificity obtained was a result of the careful selection of HPs in whom the various causes of CgA elevation were not present. Similarly, we found that patients with tumors showed higher levels of CgA than patients with CAG, even if we only considered patients with localized disease. The best cutoff range between CAG and NET patients was 53 to 54 U/L, with a sensitivity of 66.5% and a specificity of 71.4%. The fact that there were higher levels of CgA in patients with tumors and also in patients with localized tumors compared with patients with CAG has never been mentioned before in the literature. Furthermore, according to Peracchi et al,12 patients with CAG showed levels of CgA significantly higher than HPs. In the present study, we have also observed that patients with endocrine tumors showed progressively higher CgA levels as the disease progressed.6,7,16 This increase in levels was not statistically significant between patients with local and hepatic disease, but it became significant between patients with localized and diffuse disease (distant lymph node, lung, bone, and spleen metastasis). The best cutoff range identified to distinguish between localized (local and/or liver) and diffuse disease was 281 to 282 U/L, with a sensitivity of 71.1% and a specificity of 78.8%. Setting the specificity at the 95% value, the cutoff was 564 to 603 U/L with a sensitivity of 55% (21 of 38 patients). This finding is important to speculate as to whether a disease is localized or diffuse based on the CgA levels because diffuse disease is related to a poorer prognosis and needs an aggressive therapeutic approach.20,21 In the CAG group, we observed higher CgA levels in patients with ECL cell hyperplasia, and this finding was in agreement with other data already published by Peracchi et al.12 In this article,12 plasma CgA levels in patients with hyperplasia were higher compared with levels in patients with CAG without hyperplasia, although this was not statistically significant. In the present study, we also compared patients with endocrine tumors with those without tumors, such as HPs, patients with CAG, and disease-free patients at the time of sampling, with the aim of obtaining the best cutoff value useful for clinical practice. We have determined a best cutoff range of 31 to 32 U/L, with a sensitivity of 75.3% and a specificity of 84.2%. To eliminate the various causes capable of increasing CgA levels, we set the specificity at 95%, obtaining a cutoff value of 84 to 87 U/L with a sensitivity of 55% (93 of 170 participants). This value is extremely high compared with previous data in the literature,7,13,16,18,22 but it allows us to distinguish between patients who need specific and more accurate diagnostic means to detect endocrine tumors and patients in whom it is first necessary to exclude other putative causes of increasing CgA levels. Regarding the primary localization of endocrine tumors, we would stress that, in patients with lung tumors, the CgA levels were significantly lower compared with the levels in patients who had GEP tumors. On the basis of this finding, we can speculate that lung tumors have a lower secretory activity compared with GEP tumors, but further studies based on a higher number of patients are required. Patients with ZES showed significantly higher levels of CgA compared with patients without gastrin-secreting tumors.7,16 This is a result of the trophic action of gastrin on the endocrine cells of the gastric mucosa, which leads to an increase in CgA levels.5,23 Therefore, it is important to combine the determination of CgA levels and gastrin when ZES is suspected. In fact, ZES patients do not only have high levels of gastrin, but their CgA levels are also much more elevated compared with patients with other conditions such as CAG.7,12,23 In our series, we demonstrated that, in CAG patients, the levels of CgA were higher compared with HPs but lower than the levels found in patients with endocrine tumors and, in particular, in patients with ZES. In our study, we did not observe statistically significant differences in CgA levels in relation to the primary localization of the endocrine tumor, such as the stomach, gut, or pancreas.7,13,17 We found lower CgA levels in patients treated with SST analogs compared with untreated patients, although these differences were not statistically significant.22,24 However, we did not consider the change in CgA levels before and after the first dose, but we compared treated versus untreated patients at the same stage of the disease. In conclusion, our study confirms the high specificity and sensitivity of CgA in the diagnosis of endocrine tumors, but this finding is limited by the choice of carefully selected HPs as a control group. We think that it is necessary to use a cutoff range of 84 to 87 U/L to obtain a high specificity in the diagnosis of endocrine tumors, with the aim of excluding patients in whom the CgA level was elevated as a result of other non-neoplastic conditions. To our knowledge, patients with CgA levels greater than 84 to 87 U/L need to be studied using specific diagnostic means to detect endocrine tumors. On the contrary, in patients with levels less than 84 to 87 U/L but greater than 18 to 19 U/L without evident tumors, it is necessary to exclude all conditions that could result in an increase of CgA, such as chronic renal failure,7 proton pump inhibitor treatment,25 and CAG,12 and only later on do these patients need to undergo specific imaging techniques. This finding is particularly important because it differs from previously reported cutoff values between 17 and 34 U/L. Using our value, we can exclude all patients with high non-neoplastic CgA levels who often undergo unnecessary examinations that are specific for endocrine tumors. In conclusion, in our study, we evaluated CgA levels using the DAKO ELISA kit. These data should not be compared with data obtained with other types of commercial kits because, as already stated in numerous studies, the sensitivity and specificity differ between the kits.17,19,26
The authors indicated no potential conflicts of interest.
Conception and design: Davide Campana, Roberto Corinaldesi, Paola Tomassetti Provision of study materials or patients: Davide Campana, Paola Tomassetti Collection and assembly of data: Davide Campana, Lidya Piscitelli, Paola Tomassetti Data analysis and interpretation: Davide Campana, Antonio Maria Morselli-Labate, Paola Tomassetti Manuscript writing: Davide Campana, Francesca Nori, Raffaele Pezzilli, Paola Tomassetti Final approval of manuscript: Davide Campana, Paola Tomassetti
Authors' disclosures of potential conflicts of interest and author contributions are found at the end of this article.
1. Deftos LJ: Chromogranin A: Its role in endocrine function and as an endocrine and neuroendocrine tumor marker. Endocr Rev 12: 181-187, 1991 2. O'Connor DT, Deftos LJ: Secretion of chromogranin A by peptide-producing endocrine neoplasms. N Engl J Med 314: 1145-1151, 1986[Abstract] 3. Eriksson B, Oberg K: Peptide hormones as tumor markers in neuroendocrine gastrointestinal tumors. Acta Oncol 30: 477-483, 1991[Medline] 4. Schurmann G, Raeth U, Wiedenmann B, et al: Serum chromogranin A in the diagnosis and follow-up of neuroendocrine tumors of the gastroenteropancreatic tract. World J Surg 16: 697-701, 1992[CrossRef][Medline] 5. Syversen U, Mignon M, Bonfils S, et al: Chromogranin A and pancreastatin-like immunoreactivity in serum of gastrinoma patients. Acta Oncol 32: 161-165, 1993[Medline] 6. Nobels FR, Kwekkeboom DJ, Coopmans W, et al: Chromogranin A as serum marker for neuroendocrine neoplasia: Comparison with neuron-specific enolase and the alpha-subunit of glycoprotein hormones. J Clin Endocrinol Metab 82: 2622-2628, 1997 7. Tomassetti P, Migliori M, Simoni P, et al: Diagnostic value of plasma chromogranin A in neuroendocrine tumours. Eur J Gastroenterol Hepatol 13: 55-58, 2001[CrossRef][Medline] 8. Stridsberg M, Oberg K, Li Q, et al: Measurements of chromogranin A, chromogranin B (secretogranin I), chromogranin C (secretogranin II) and pancreastatin in plasma and urine from patients with carcinoid tumours and endocrine pancreatic tumours. J Endocrinol 144: 49-59, 1995 9. Guignat L, Bidart JM, Nocera M, et al: Chromogranin A and the alpha-subunit of glycoprotein hormones in medullary thyroid carcinoma and phaeochromocytoma. Br J Cancer 84: 808-812, 2001[CrossRef][Medline] 10. Oberg K, Janson ET, Eriksson B: Tumour markers in neuroendocrine tumours. Ital J Gastroenterol Hepatol 31: S160-S162, 1999 (suppl 2) 11. Borch K, Stridsberg M, Burman P, et al: Basal chromogranin A and gastrin concentrations in circulation correlate to endocrine cell proliferation in type-A gastritis. Scand J Gastroenterol 32: 198-202, 1997[Medline] 12. Peracchi M, Gebbia C, Basilisco G, et al: Plasma chromogranin A in patients with autoimmune chronic atrophic gastritis, enterochromaffin-like cell lesions and gastric carcinoids. Eur J Endocrinol 152: 443-448, 2005 13. Peracchi M, Conte D, Gebbia C, et al: Plasma chromogranin A in patients with sporadic gastro-entero-pancreatic neuroendocrine tumors or multiple endocrine neoplasia type 1. Eur J Endocrinol 148: 39-43, 2003[Abstract] 14. Bashir S, Gibril F, Ojeaburu JV, et al: Prospective study of the ability of histamine, serotonin or serum chromogranin A levels to identify gastric carcinoids in patients with gastrinomas. Aliment Pharmacol Ther 16: 1367-1382, 2002[CrossRef][Medline] 15. Abou-Saif A, Gibril F, Ojeaburu JV, et al: Prospective study of the ability of serial measurements of serum chromogranin A and gastrin to detect changes in tumor burden in patients with gastrinomas. Cancer 98: 249-261, 2003[CrossRef][Medline] 16. Nehar D, Lombard-Bohas C, Olivieri S, et al: Interest of chromogranin A for diagnosis and follow-up of endocrine tumours. Clin Endocrinol 60: 644-652, 2004[CrossRef][Medline] 17. Stridsberg M, Eriksson B, Oberg K, et al: A comparison between three commercial kits for chromogranin A measurements. J Endocrinol 177: 337-341, 2003[Abstract] 18. Panzuto F, Severi C, Cannizzaro R, et al: Utility of combined use of plasma levels of chromogranin A and pancreatic polypeptide in the diagnosis of gastrointestinal and pancreatic endocrine tumors. J Endocrinol Invest 27: 6-11, 2004[Medline] 19. Ferrari L, Seregni E, Lucignani E, et al: Accuracy and clinical correlates of two different methods for chromogranin A assay in neuroendocrine tumors. Int J Biol Markers 19: 295-304, 2004[Medline] 20. Tomassetti P, Campana D, Piscitelli L, et al: Endocrine pancreatic tumors: Factors correlated with survival. Ann Oncol 16: 1806-1810, 2005 21. Tomassetti P, Campana D, Piscitelli L, et al: Endocrine tumors of the ileum: Factors correlated with survival. Neuroendocrinology 83: 380-386, 2006[CrossRef][Medline] 22. Seregni E, Ferrari L, Bajetta E, et al: Clinical significance of blood chromogranin A measurement in neuroendocrine tumours. Ann Oncol 12: S69-S72, 2001 (suppl 2) 23. Kleveland O, Syversen U, Slordahl K, et al: Hypergastrinemia as a cause of chromogranin a increase in blood in patients suspected to have neuroendocrine tumor. Digestion 64: 71-74, 2001[CrossRef][Medline] 24. Moattari AR, Deftos LJ, Vinik AI: Effects of sandostatin on plasma chromogranin-A levels in neuroendocrine tumors. J Clin Endocrinol Metab 69: 902-905, 1989 25. Giusti M, Sidoti M, Augeri C, et al: Effect of short-term treatment with low dosages of the proton-pump inhibitor omeprazole on serum chromogranin A levels in man. Eur J Endocrinol 150: 299-303, 2004[Abstract] 26. Leon A, Torta M, Dittadi R, et al: Comparison between two methods in the determination of circulating chromogranin A in neuroendocrine tumors (NETs): Results of a prospective multicenter observational study. Int J Biol Markers 20: 156-168, 2005[Medline] Submitted November 29, 2006; accepted February 26, 2007.
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2007 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
