Radiol Oncol 2004; 38(1): 21-6.
Neuron specific enolase - selective marker for small-cell lung cancer
Biljana Ilievska Poposka1, Mirko Spirovski2, Dean Trajkov2, Tome Stefanovski3, Sonja Atanasova1, Marija Metodieva1
1Institute for Lung Diseases and Tuberculosis, Clinical Center, Skopje
2Institute for Immunology, Clinical Center, Skopje 3Pulmology and Allergology Clinic, Clinical Center Skopje, Macedonia
Background. Neuron specific enolase (NSE) is an isomer of the glycolytic enzyme enolase, which was first found in extracts of brain tissue, and was later shown to be present in neuroendocrine cells and neuroen-docrine tumours. The aim of the study was to confirm the importance of serum NSE as a tumour marker in patients with small-cell lung cancer.
Patients and methods. Serum levels of NSE were measured by the radioimmunoassay in 71 patients with lung cancer (LC), in 24 patients with non malignant lung diseases and in 28 healthy adults. Results. According to the serum values in the group of healthy adults, 16.6 ng/ml was determined as a cut of level of NSE. By the specificity of 88.13 % in the group of non malignant lung diseases, the sensitivity of 47.82 % was obtained in patients with LC, which increased to 72.72 % in the patients with SCLC. In pa-tients with non-small-cell lung cancer (NSCLC) the sensitivity of NSE test was 38.89 %. The patients with SCLC-extensive disease had a significantly higher mean NSE level (290.48 ng/ml) than patients with the limited stage disease (46.94 ng/ml). Serial measurements in 16 patients receiving combined chemotherapy and/or radiotherapy showed an excellent correlation between serum NSE level and clinical response. Conclusions. These results indicate that serum NSE may be a useful marker for diagnosis, staging and for monitoring response to the therapy in patients with SCLC.
Key words: lung neoplasms; carcinoma small cell; tumor markers; biological; neuron specific enolase
Introduction
Received 25 February 2004 Accepted 10 March 2004
Correspondence to. Biljana Ilievska Poposka, MD, Institute for Lung Diseases and Tuberculosis; Vodnjanska 17, 1000 Skopje, Macedonia; Phone: +389 02 3147 616; Fax: +389 02 3229166; E-mail : biljana_ili@hotmail.com
Enolase is a glycolytic cytoplasmic enzyme, present in all human cells, catalysing the conversion of 2-phosphoglycerate to 2-phospho-enolpyruvate.1 The enzyme consists of thrree dimeric isoenzymes, called as a, ß and y. Neuron specific enolase (NSE) is y-y dimer and presents the neuronal form of the enolase.2 Originally extracted from the bovine
22
Ilievska Poposka B et al. / Neuron specific enolase
brain tissue, it was first considered that the gene coding for NSE was restricted to neurons, and that it was only present in the central nervous system. In 1978 Schmechel et al. have shown that NSE is also present in all pe-ripheral and central neuroendocrine cells, named APUD (amine precursor uptake and decarboxylation) cells.3 Tapia et al. have ex-tended this work with immunohistochemical and extraction techniques and showed that NSE is present in a wide variety of APUD neoplasms or APUDomas including: islet tu-mours of the pancreas, gastrinomas, VIPomas, medullary carcinoma of the thy-roid, pheochromocytoma, and small-cell car-cinoma of the lung (SCLC) among others.4 In contrast, they could not find NSE presence in any non-neuroendocrine tumours.4,5 High pre-treatment levels of NSE have been detec-ted by the radioimmunoassay in the sera of the patients with neuroendocrine tumours, including SCLC.6-8
Therefore, this study was designed to re-evaluate the role of serum NSE in the diag-nosis and differential diagnosis of the pa-tients with lung carcinoma; to re-evaluate whether serum NSE levels are in the correla-tion with the extent of tumour dissemination or stages of the disease, and to re-evaluate the role of NSE as a marker for monitoring a therapeutic response in the patients with lung carcinoma.
Patients and methods
Patients
In this study 123 persons divided in three groups were included: first group - 71 newly diagnosed untreated patients with a different type of lung carcinoma; second group - 24 pa-tients with non malignant lung diseases and third group - 28 healthy adults.
According to the histological types of lung carcinoma, the patients from the first group were further divided in two groups: 33 pa-Radiol Oncol 2004; 38(1): 61-6.
tients with SCLC and 38 patients with no-small-cell lung cancer (NSCLC).
Methods
Clinical assessment
The routine pre-treatment staging procedures consisted of physical examination; biochem-istry; chest X ray; lung functional tests; fiberoptic bronchoscopy (with bronchial biopsy and cytological examination of brush-ings and washings); ultrasound procedures; radionuclide scan of bone. Biopsy or fine nee-dle aspiration specimens of enlarged lymph-nodes, subcutaneus nodules and pleural effu-sions were taken when clinically indicated. According to these findings, the patients with SCLC were staged as having limited disease (tumour confined to one hemithorax, includ-ing the ipsilateral lymph nodes) or extensive disease (outside these limits). The patients with NSCLC were divided in five stages of TNM classification. Only 16 patients were followed up and a response to chemotherapy and/or radiotherapy was evaluated. The re-sponse was judged to be: complete (CR) when both clinical and pathological evidence of tumour totally disappeared; partial (PR) when there was a reduction of 50% or more in the sum of all measurable and evaluable tu-mour masses. Lesser degrees of tumour re-duction were judged as no response.
The diagnosis of all patients was con-firmed at the Institute for Lung Diseases and Tuberculosis; the patients were treated at the Institute for Radiology and Oncology, Medical Faculty in Skopje.
Immunoassay
Blood specimens were collected from each of the 71 patients with lung carcinoma at diag-nosis, as well as from the patients with non malignant lung diseases and healthy adults. Serial samples were obtained from 16 of 71 patients with lung carcinoma, usually at 6-week    intervals,    after    each    course    of
Ilievska Poposka B et al. / Neuron specific enolase
23
chemotherapy or after the end of radiothera-py. The serum was separated immediately after the collection and was stored at - 20 °C be-fore the assay. NSE levels were determined by a double-antibody solid phase radioim-munoassay technique (Pharmacia NSE-RIA test) at the Institute for Immunology, Medical Faculty, Skopje.
Student's t test, Newman- Keuls test and x2 test were used to determine the statistical significance between the mean values and be-tween raised frequencies separately.
Results
Twenty eight healthy adults had NSE serum level ranging from 2.58 to 17.41 ng/ml (mean level 8.01 +/- 4.40 ng /ml). The upper limit of the normal interval for serum NSE 16.6 ng/ml is defined as the mean value for healthy controls plus 1.96 standard deviations. Only two patients from this group (7.14%) had a raised serum level above the normal value.
In the group of patients with non-malig-nant lung diseases NSE serum level was ranged from 2.61ng/ml to 41.87ng/ml, with the mean level of 11.79 +/- 9.53 ng/ml. Among them, five were serum NSE positive (20.83%). On the basis of these findings 88.13% was determined as a specificity of NSE test.
The mean level of NSE in the group of 71
patients with LC was 127.96 +/- 442.53 ng/ml. Thirty-four of them were found to have raised serum NSE concentrations that determined the sensitivity of NSE in LC of 47.82%. A statistical analysis between the positive NSE findings in the three groups showed a significant difference (x2 = 18.19 p< 0.001). With Newman-Keuls test we obtained a statistical significant difference between the mean values in the three groups (F=1.84, p<0.05).
When the upper normal limit for serum NSE was taken to be 16.6 ng/ml, 73% of pa-tients with SCLC were found to have raised NSE concentrations compared with 38% of patients with NSCLC (x2 = 12.78, p<0.001).
Eighteen of 33 patients with SCLC had a limited stage disease and 56 had an extensive disease. NSE was raised in 10 of 18 (55.55%) limited-stage patients and in all 15 (100%) pa-tients with an extensive-stage disease (x2 = 8.8. p<0.005). The mean pre-treatment NSE in the limited-stage disease was 46.94 +/- 56.92 ng /ml, versus 290.48 +/- 325.24 ng/ml for the extensive-stage disease (Student's t test = 2.69, p<0.001; Table 1).
There was not a significant statistical dif-ference between the number of patients with NSCLC in different TNM stages who had the raised serum NSE level above 16.6 ng/ml. The results are shown in the Table 2 (x2 = 7.27, p non significant).
Serum NSE was measured again in 16 pa-
Table 1. Mean value and sensitivity of NSE in the patients with SCLC
Group	Mean value            Frequency of raised value >16,6ng/ml	Sensitivity
SCLC limited diseases          46.94 +/- 56.92                                        10/18 SCLC extensive diseases   290.48 +/- 325.24                                      15/15		55.5 % 100 %
Table 2. Sensitivity of NSE in patients with NSCLC in different TNM stages		
Group	Number               Freqensy of raised value> 16,6ng/ml	Sensitivity (%)
Stage I Stage II Stage IIIa Stage IIIb Stage IV	13                                                    4 / 13 12                                                     1 / 8 10                                          2 / 6 92 / 5 11                                          5 / 6	30.7 12.5 33.3 40.0 83.3
Radiol Oncol 2004; 38(1): 21-6.
24
Ilievska Poposka B et al. / Neuron specific enolase
tients after the end of each course of chemotherapy or at the end of radiotherapy when staging procedures were repeated. Eleven of these 16 patients were with SCLC, and 5 with NSCLC. In Figure 1a, there are the changes of serum NSE level in 13 patients who responded to therapy (responders); 9 were with SCLC and 2 with NSCLC. Eight pa-tients had the elevated serum NSE level above 16.6 ng/ml and had a predominantly extensive disease.
According to the clinical signs, when a complete or a partial response was obtained, the serum NSE level in 6 patients decreases to a normal range; in one patient the serum NSE level decreased, but between 16 and 20 ng / ml, and in one patient the level remained stable. In other five patients who had the serum NSE level in a normal range at diagno-sis, at the moment when CR or PR was clini-cally achieved, the serum NSE level de-creased in 2 patients, versus slightly elevation in three patients, even in the normal range. However, when they had a relapse, the serum NSE rose again: in 11 patients above a normal range and in 2 patients to a normal value.
Figure 1b shows the changes in the serum level of NSE in three patients who did not re-
sponse to chemotherapy. In one patient the pre-treatment level was in a normal range; in the other two patients NSE level was above 16.6 ng/ml. In all three patients there is a clear elevation in NSE during the chemother-apy that predicted the clinical recognition of relapse.
Discussion and conclusions
A raised serum NSE was observed in 47.8% of the patients with LC; the sensibility of NSE was 72.72% in patients with SCLC versus 38.89% in NSCLC. Our results correspond to the findings of other authors: Carney et al.6 -NSE sensibility of 69% in SCLC with the cut of level of 12 ng/ml, Cooper et al.8 - NSE sen-sibility of 77% in SCLC with the cut of level of 13 ng/ml, Esscher et al.1 - NSE sensibility of 85% in SCLC and 25% in NSCLC with the cut of level of 12 ng/ml, Lorenz et al.9 - NSE sen-sibility of 98% in SCLC and 4% in NSCLC with the cut of level of 15 ng/ml.
NSE is a marker specific for the neuroen-docrine system and for tumours that arise from it, the so-called ADUP neoplasms. The characteristics of these tumours are defined
ab
inn-
relapse
progress
ANSCLC   BSCLC Figure 1. Serum NSE level changes (ng/ml) in patients with SCLC and NSCLC during treatment.
Radiol Oncol 2004; 38(1): 21-6.
Ilievska Poposka B et al. / Neuron specific enolase
25
in vivo and in vitro studies: many neurosecre-tory granules, the ability for the production of different hormones and polypeptides, the high level of L-Dopa decarboxylasa and NSE.10 When the characteristic features of SCLC were first described, this tumour was considered to be an anaplastic malignancy with which a variety of paraneoplastic syn-dromes were associated.11 In the 1960, how-ever, the presence of neurosecretory granules within the cells was described, which led to the inclusion of SCLC within the APUD system. It was thus presumed that the anaplastic cells were derived, in the normal bronchial mucosa, from the Kulchitsky cells which posses APUD properties. It is interesting why serum NSE is elevated in some of patients with NSCLC. The mechanisms by which NSCLC cells are capable of producing APUD-derived enzymes and hormones are not known. According to the findings of Gazard, there is a possibility of in vitro "conversion" of small-cell lung cacinoma to large cell cancer morphology. Although "transformed" cells had lost most or all of their "amine precusor uptake and decarboxylation characteristics" certain neuroendocrine features such as NSE were retained. Thus, it is possible that the large cell lung cancers were originally small cell tumours that "changed" histology, but "re-tained" NSE activity or were mixed tumours with the small cell component. The appear-ance of heterogeneous cell populations in the carcinomas is the reason why different parts of the tumour tissue show a different im-munohistochemical expression for the same marker and why the tumour metastasises do not release some tumour marker which is re-leased by the primary tumour.12
The results regarding the extension of the disease in patients with SCLC are compara-ble with the results reported by other authors (mean value of limited disease versus mean value of extensive disease): Fischbah and Berthold13 - 8.4 +/- 0.8 / 47.7 +/- 8.8 ng/ml; Carney et al.6 - 13.8 / 59.0 ng/ml; Johnson et
al.7 - 33.4 +/- 4.7 / 94.5 +/- 13.8 ng/ml; Cooper et al.8 - 14 / 42 ng/ml; Splinter et al.14 - 25 / 51 ng/ml. According to Carney et al.6, Johnson et al7 and others, the serum level of NSE, is more in correlation with the tumour burden and the number of metastatic site than with the individual ability of tumour to produce NSE.
Our follow-up studies in 16 patients showed the correlation between the tumour burden, the clinical response and serum NSE concentrations: the elevated, pre-treatment NSE levels decline to normal or nearly normal when CR or PR was achieved, while 1 re-sponding patient maintained the essentially stable NSE level. When the relapse or progression in the disease occurred, the NSE level rose again; what is more important, raising was obtained before the clinical recognition of the relapse; a rising NSE level may predict relapse weeks to months in advance of other clinical evidence and signal the need to change a therapy sooner. A progressive rise in NSE levels during the treatment indicates the tumour resistance or relapse and also need to change the therapy. With serial NSE measurements we can follow up the disease course of the patients with SCLC, and give a different therapy at a time when it is likely to be most beneficial, i.e., when the tumour bur-den is low.7,8,15 NSE is a useful tumour marker for the diagnosis and the differential diag-nosis in patients with SCLC. This means that NSE measurements in patients with SCLC are at least a useful addition to standard in-vestigational methods. Serial NSE measure-ments are useful for monitoring the course of the disease and therapeutical response; they provide information relevant to patient management which could not be obtained by the physical examination or routine staging pro-cedures.
Radiol Oncol 2004; 38(1): 21-6.
26
Ilievska Poposka B et al. / Neuron specific enolase
References
1.   Essher T, Steinoltz L, Berg J, Nou E, Nilsson K, Pahlman S. Neuron specific enolase: a useful di-agnostic serum marker for small cell carcinoma of the lung. Thorax 1985; 40: 85-90.
2.   Marangos PJ, Zomzoly-Neurath C, Goodwin FK. Structural and immunological properties of neuron specific protein (NSP) from rat, cat and human brain: comparison to bovine 14-3-2. J Neurochem 1977; 28: 1097-107.
3.   Schmechell DE, Marangos PJ, Zis AP, Brightman M. Brain enolase and specific markers of neuronal and glial cells. Science 1978; 199: 313-5.
4.   Tapia FJ, Barbosa AJA, Marangos PJ, Polak JM, Bloom SR, Dermody C, et al. Neuron-specific eno-lase is produced by neuroendocrine tumors. Lancet 1981; 11: 808-11.
5.   Prinz RA, Marngos PJ. Use of neuron-specific eno-lase as a serum marker for neuroendocrine neoplasma. Surgery 1982; 92: 887-9.
6.   Carney DN, Marangos PJ, Ihde DC, Bunn PA Jr, Cohen MH, Minna JD, et al. Serum neuron- spe-cific enolase: a marker for disease extent and re-sponse to therapy of small cell lung cancer. Lancet 1982; 1: 583-5.
7.   Johnson DH, Marangos PJ, Forbes JT, Hainsworth JD, Van Welch R, Hande KR, et al. Potential utility of serum neuron specific enolase levels insmall cell carcinoma of the lung. Cancer Res 1984; 44: 5409-19.
8.   Cooper EH, Splinter TAW, Brown DA, Muers MF, Peake MD, Pearson SL. Evaluation of radioim-munoassay for neuron specific enolase in small cell lung cancer. Br J Cancer 1985; 51: 645-52.
9.   Lorenz J, Mai GA, Schulz V. Neuronenspezifische enolase - ein selektiver marker des kleinzelligen bronchialkarzinoma. Prax Klin Pneumol 1986; 40: 100-2.
10. Marangos PJ, Gazdar AF, Carney DN. Neuron spe-cific enolase in human small cell carcinoma cul-tures. Cancer Lett 1982; 15: 67-71.
11. Cole GA, Polak JM, Wharton J, Marangos PJ, Pearse AGE. Neuron specific enolase as a useful histochemical marker for the neuroendocrine sistem of the lung. J Pathol 1980; 132: 351-2.
12. Bombardieri E, Jotti GS, Cocciolo MG, Mori M, Rusconi A, Rusca M, et al. Tissue polypeptide anti-gen and carcinoembruonic antigen in colon tu-
mors: serum levels and immunohistochemical lo-calization. Cancer Detect Prev 1985; 8: 219-26.
13. Fischbach W, Berthold J. Neuron-specific enolase in the diagnosis and therapy monitoring of lung cancer: a comparison with CEA, TPA, ferritin and calcitonin. Int J Biol Markers 1986; 1: 129-36.
14. Splinter TAW, Cooper EH, Kho GS. Neuron spe-cific enolase as a guide to the treatment of small cell lung cancer. Cancer Clin Oncol 1986; 23: 171-6.
15. Akoun GM, Scarna HM, Milleron BJ, Benichou MP, Herman DP. Serum neuron specific enolase: marker for disease extent and response to therapy for small cell cancer. Chest 1985; 87: 39-43.
Radiol Oncol 2004; 38(1): 21-6.