Original Paper
Tumor Biol 1992;13:352-357
Institute of Surgery (2nd), University ‘La Sapienza’, Rome; Endocrine Surgery, Department of Experimental Medicine. University of L’Aquila; Surgical Pathology, University of Ancona, Italy
K ey W ords
Gastrointestinal neuroendocrine tumors Neuron-specific enolase Tumor marker Radioimmunoassay
Serum Neuron-Specific Enolase in Diagnosis and Follow-Up of Gastrointestinal Neuroendocrine Tumors
A b stra ct
Neuron-specific enolase (NSE), the glycolytic isoenzyme of the enolase gamma-gamma dimer, is a specific marker for the diffuse neuroendocrine system and derivative tumors (NET). Serum levels of NSE were measured in 39 patients with NET of the gastrointestinal tract (including 3 gastric and 13 intesti nal carcinoid tumors, 6 gastrinomas, 3 insulinomas, 1 glucagonoma, 2 mixed islet cell tumors, 11 neuroendocrine pan creatic carcinomas), in 15 healthy subjects and in 15 nonendocrine gastric, pancreatic, and intestinal tumors. Thirty-six of the 39 patients had elevated circulating levels of NSE, 2 insuli nomas and 1 gastrinoma had values below 12 ng/ml like healthy subjects and nonendocrine tumors. No significant dif ference of serum NSE was found between 23 ‘functioning’ and 16 ‘nonfunctioning’ NET. Fourteen of the NET were malig nant, and NSE circulating values were significantly higher than those of nonmalignant forms. After curative surgery serum NSE decreased significantly. NSE can be considered a reliable marker in the differential diagnosis between endo crine and nonendocrine neoplasms, in the clinical detection of silent endocrine tumors and in the follow-up of NET.
Introduction
The glycolytic enzyme enolase is widely distributed in the mammalian tissue as three immunologically distinct subunits: alpha,
Received: December 30. 1991 Accepted: August 24, 1992
beta and gamma, and in mammalian brain they are present as alpha-alpha, alpha-gamma and gamma-gamma forms [1,2]. The dimer form of enolase containing the gamma-subunit (gamma-gamma and alpha-
Mariadomenica D’Alcssandro. PhD II Clínica Chirurgica. Policlinico Umbcrto I Universitá La Sapienza víale del Policlinico I-0016I Rome(Italy)
© 1992 S. Karger AG. Basel 1010-4283/92/ 0136-035252.75/0
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/5/2018 7:35:06 AM
Mariadomenica D ’A lessandroa Pao!a M ariania Davide Lom anloa Francesco Carleé Emanuele Lezochec Vincenzo Speranzaa
Tumor type
Number of cases
Clinical features
Pancreas gastrinoma Glucagonoma Ileal carcinoid Ileal carcinoid Ileal carcinoid
4 1 1 2 3
Mixed islet cell
1
hepatic metastases hepatic metastases bone metastases large tumor and serosal spread multiorgan metastatic involvement and syndrome hepatic metastases and VIP, PP. insulin hypersecretion
Neuroendocrine pancreatic carcinoma
1
gamma dimer) is defined as neuron-specific enolase (NSE) and is selectively expressed by neurons and central or peripheral neuroendo crine cells [3-5]. Immunohistochemical studies and extrac tion techniques have demonstrated that NSE is present in all types of tumors arising from neuroendocrine cells (oat cell carcinoma, pheochromocytoma, islet cell tumor, medul lary thyroid carcinoma, neuroblastoma, gut and lung carcinoids) while it is absent or present in a very small amount in nonendocrine tumors [6]. Various authors have suggested that NSE can be evaluated in the serum of patients affected by neuroendocrine tumors, since tu mor cell turnover or secretory activity might result in increased serum NSE levels [7], Therefore, the gamma-gamma dimer of eno lase can be a useful marker for the diagnosis and monitoring of patients with neuroendo crine tumors [8, 9], The use of NSE as a tumor marker is well established for the diagnosis and monitoring of patients with small-cell carcinoma of the lung (SCCL) [10] and neuroblastoma in chil dren [5, 11], In the present study we have measured serum NSE levels in patients affected by var
diffuse metastases
ious neuroendocrine tumors of the gastroin testinal tract in order to assess a possible role of NSE as a tumor marker of these neo plasms.
M aterials and M eth o d s Thirty-nine patients (16 females, 23 males: aged from 18 to 74 years) with a previous diagnosis of neu roendocrine tumors were studied. Thirteen patients had intestinal neoplasms (3 rectal carcinoids, 10 ileal carcinoids), 23 patients had pan creatic neoplasms (6 Zollinger-Ellison syndromes, 3 insulinomas, 1 glucagonoma, 2 mixed islet cell tumors, 11 neuroendocrine carcinomas), and 3 patients had a gastric carcinoid. Of these patients, 23 were diagnosed as so called ‘nonfunctioning’ tumors since there was no clinical evidence of hormone hypersecretion (normal insulin, glucagon, gastrin, VIP, PP, somatostatin, serotonin, urinary 5-HIAA levels). In 13 patients a highly malignant neuroendocrine tumor was diagnosed on the basis of survival rate and presence of metastases (tabic 1). Fifteen healthy subjects and 15 non-endocrine tu mor patients: 1 carcinoma of the pancreas head, 6 ade nocarcinomas of the large bowel, 5 carcinomas of the stomach (3 intestinal, 2 diffuse), 2 villous adenomas of the sigmoid colon, and 1 adenocarcinoma of the lower esophagus were included as controls. Blood samples were collected at fasting, imme diately centrifuged (10 min at 1,000 g) and aliquots of
353
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/5/2018 7:35:06 AM
Table 1. Clinical data on 13 patients with malignant neuro endocrine tumors
250
-,
100
E a 50-
* *
NSE serum concentra tion according to localization of gastrointestinal neuroendocrine tu mors vs. controls. Fig. 1.
*
if \ Pancreas
Stomach
Neuroendocr ne tumors
sera were stored at -7 0 °C until assay. Since erythro cytes contain appreciable amounts of NSE. hemolysed samples were not used. Scrum from most patients was collected before and after tumor resection and chemo therapy. Serum levels of NSE were determined using a radioimmunoassay (RIA) kit from Eiken (Japan). Highly purified NSE from bovine brain and a poly clonal rabbit antiserum specific for the gamma subunit were used for a double antibody RIA technique. A 41.5% cross-reaction with the alpha-gamma isozyme and no cross-reaction between the alpha-alpha and beta-beta isozymes were observed. The detection limit is 2 ng/ml and the coefficient of interassay variation is less than 10%. Levels lower than 12.0 ng/ml are con sidered normal with this assay. Results were expressed as the mean ± standard error (M ± SEM): statistical analysis was performed according to Mann-Whitney U test and Wilcoxon’s signed-ranks matched-pairs test.
Results
The mean NSE level was 5.3 ± 0.5 ng/ml in healthy subjects and 6.5 ± 1 ng/ml in non endocrine tumor patients. The concentrations of NSE in the various cases of endocrine tumors are illustrated in figure 1.
354
V W Vv Intestine
Healthy
Nonendocrine tumors
Controls
Two insulinomas had values below 12 ng/ ml, 1 gastrinoma 10 ng/ml. The mean value of all cases considered was 50.2 ± 8.2 ng/ml (p < 0.01). The mean serum NSE level of the ‘func tioning’ neuroendocrine tumors was 52.6 ± 12.6 ng/ml and the mean of the ‘nonfunction ing' tumors was 48.6 ± 11.1 ng/ml. The dif ference was not significant (fig. 2). Serum NSE levels were significantly higher in the 13 patients with malignant tumors pre senting with diffuse metastases or local inva sion: the mean serum NSE in the malignant tumors was 76.6 ± 13.6 ng/ml and 37.1 ± 9.54 ng/ml in the nonmalignant forms (p < 0.01) (fig. 3), although the highest value was found in a nonmalignant intestinal carcinoid (253.8 ng/ml). The serum NSE values in 10 patients be fore and after surgery are shown in table 2. Of these patients, 7 had nonmalignant and 3 ma lignant tumors. Forty-eight hours after surgery a decrease in NSE levels was observed in all cases. Two months later, in the 7 patients in whom the surgery was curative, serum NSE decreased significantly (p < 0 .0 1 ) while the 3 cases with
D'Alcssandro/Mariani/Lomanto/ Carlei/Lezoche/Speranza
Radioimmunological Evaluation of Neuron-Specific Enolasc
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/5/2018 7:35:06 AM
to z
250-1
150-
Fig. 2. Scrum NSE levels in 23 patients with a diagnosis of ‘non functioning’ tumor and in 16 pa tients with ‘functioning’ tumor. The difference was not significant.
• • Functioning
Nonfunctioning
Fig. 3. Serum NSE values in be nign (26 patients) and malignant (13 patients) neuroendocrine tu mors. The difference was signifi cant (p < 0.01).
Case Tumor type No.
1 2 3 4 5 6 7 8 9 10
NSE, ng/ml pre48 h post after operative operative 2 months
19.7 insulinoma 25.3.8 ileal carcinoid ileal carcinoid 16.8 28.9 ileal carcinoid 37.2 rectal carcinoid 16.9 mixed islet cell endocrine pancreatic carcinoma 30.0 57.9 mixed islet cell endocrine pancreatic carcinoma 24.5 45.0 ileal carcinoid
15.7 9.7 13.3 16.2 5.2 7.5 14.6 39.8 11.0 17.9
7.9 8.2 7.4 8.6 5.3 7.8 9.0 83.2 15.3 27.3
355
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/5/2018 7:35:06 AM
Table 2. NSE values on 10 neuroendocrine tumors before and after surgery
5-FU + streptozotocin
Fig. 4. Serum NSE levels in 2 patients during response to chemo therapy. ▲= Nonfunctioning endo crine tumor of pancreas; • = func tioning mixed islet cell tumor.
D iscussion
Previous studies have demonstrated that gamma-gamma enolase is specifically present in all cells and nerves of the diffuse neuroen docrine system of the gut and pancreas and that NSE can be a specific marker of these structures [ 12]. The demonstration of NSE in different neuroendocrine tumors and its serum deter mination by a specific RIA [13] has suggested the use of this isoenzyme as a specific marker for the detection and therapeutic monitoring of neuroendocrine tumors. Subsequent studies have shown that serum NSE is a useful parameter in the diagnosis of different neoplasms such as neuroblastoma and SCCL, and in staging of the disease and monitoringoftheresponse to therapy [5,10,11],
356
In the present study we measured NSE lev els in various neuroendocrine tumors with interesting results. All considered cases have shown increased NSE levels except for 2 be nign insulinomas and 1 gastrinoma without metastases. No significant difference was observed be tween ‘functioning’ and ‘nonfunctioning’ tu mors. However, at present, serum NSE is a more reliable marker for ‘nonfunctioning’ tu mors, since this type of tumor cannot be detected by an increase of a specific hormone that could be used as a marker. These tumors are clinically inactive and are sometimes de tected incidentally. It is possible that these tumors secrete a still unknown hormone or an inactive prohormone. The use of serum NSE as a marker is suitable in these neoplasms for diagnosis and then for prognosis and response to chemotherapy [8, 14, 16], Significant dif ferences found between serum NSE levels in malignant and nonmalignant neoplasms are consistently reported in the literature [6, 15, 17], In our patients serum NSE levels seem to correlate well with tumor stage, local inva siveness and presence of metastasis, except for the benign ileal carcinoid with an enolase
D'Alcssandro/Mariani/Lomanto/ Carlei/Lezoche/Speranza
Radioimmunological Evaluation of Neuron-Specific Enolase
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/5/2018 7:35:06 AM
malignant tumors showed increased NSE val ues. Among these patients 1 died 3 months after operation and 2 received chemotherapy which was effective in 1. The serum NSE val ues of the latter patients are in agreement with this behavior as shown in figure 4.
level of 253.8 ng/ml. This patient disclosed a postoperative NSE value of 8.2 ng/ml; 2 years after operation the patient is in good general condition with a normal NSE value. A few of our patients with nonmalignant disease have shown only a slight increase in serum NSE values. This observation, while limiting the significance of the NSE assay (the gastrinoma with 10 ng/ml of NSE and the 2 insulinomas are false negatives), suggests that special attention should be paid also to a minor increase of this marker and that partic ular care should be taken of the assay tech nique. Normal serum values of enolase in patients with proven neuroendocrine tumors might be due to nondamaged tumor cells.
The results obtained in patients monitored after surgery, in spite of the limit due to the small series, suggest the usefulness of enolase as a marker of the extent of surgery, the response to therapy and the prediction of recurrence. NSE is present in all normal and neoplastic endocrine cells; therefore, this gly colytic enzyme can be considered a reliable marker also for those tumors that are not actively secreting any neuroendocrine peptide or are secreting abnormal molecular forms biologically inactive, and unrecognized by the available assays.
R eferences 7 Prinz RA. Marangos PJ: Use of neu ron specific enolase as serum marker for neuroendocrine neo plasm. Surgery 1982:92:887-889. 8 Iwase K, Nagasaka A, Kato K, et al: Enolase subunits in patients with neuroendocrine tumours. J Clin En docrinol Metab 1986;63:94-101. 9 Prinz RA. Bermes EW. Kimmel JR. Marangos PJ: Serum marker for pancreatic islet cell and intestinal carcinoid tumors: A comparison of neuron specific enolase, beta-hu man chorionic gonadotropin and pancreatic polypeptide. Surgery 1983;94:1019-1023. 10 Carney DN, Marangos PJ. Ihda DC. et al: Serum neuronspecific enolase: A marker for disease extent and re sponse to therapy of small-cell lung cancer. Lancet 1982;i:583—585. 11 Zelzer PM. Marangos PJ, Evans AE. Schneider SL: Serum neuron spe cific enolase in children with neuro blastoma: Relationship to stage and disease course. Cancer 1986:57: 1230-1234. 12 Bishop AE, PolakJM. Facer P, et al: Neuron specific enolase: A common marker for the endocrine cells and innervation of the gut and pancreas. Gastroenterology 1982;83:902—915.
13 Parma AM, Marangos PJ, Goodwin FK: A more sensitive radioimmu noassay for neuron specific enolase suitable for cerebrospinal fluid de termination. J Neurochem 1981:36: 1093-1097. 14 Prinz RA, Marangos PJ: Serum neu ron specific enolase: A serum marker for 'non-functioning' pan creatic islet cell carcinoma. Am J Surg 1983;145:77-81. 15 Jacques G. Bepler G. Holle R. et al: Prognostic value of pretreatment carcinoembryonic antigen, neuron specific enolase and creatine kinase -BB levels in sera of patients with small cell lung cancer. Cancer 1988; 62:125-134. 16 Plowman PN; Neuron specific cno-
lase as marker for medulloblastoma. Lancet 1990:336:1388. 17 Grouzmann F. Gicquel C, et al: Neuropeptide Y and neuron specific enolase levels in benign and malig nant pheochromocytomas. Cancer 1990;66:1833-1835.
357
Downloaded by: Univ. of California Santa Barbara 128.111.121.42 - 3/5/2018 7:35:06 AM
1 Rider CC, Taylor CB: Evidence for a new form of enolase in rat brain. Biochem Biophys Res Commun 1975;66:814-820". 2 Flecher L. Rider CC, Taylor CB: Enolase isoenzyme: 111. Chromatog raphy and immunological character istic of rat brain enolase. Biochcm Biophys Acta 1976;452:245-252. 3 Schmechel D, Marangos PJ. Brightman M. et al: Brain enolase as spe cific marker of neuronal and glial cell. Science 1978;199:313-315. 4 Schmechel D. Marangos PJ. Brightman M: Neuron specific enolase is a molecular marker for peripheral and central neuroendocrine cells. Nature 1978:276:834-836. 5 IshiguroY. KatoK. Ito I. et al: High levels of immunoreactive nervous system specific enolase in sera of patients with neuroblastoma. Clin Chim Acta 1982; 121:173-180. 6 Tapia PJ. Polak JM, Barbosa AJM, et al: Neuron specific enolase is pro duced by neuroendocrine tumours. Lancet 1981 ;i:808-811.