Role of the ratio of NSE serum concentration in evaluating the therapeutic effect on metastatic neuroendocrine neoplasms of the liver

Abstract

Background:

Neuron-specific enolase (NSE) is one of the biomarkers of neuroendocrine neoplasms (NEN). Its level of evidence is significantly lower than some other biomarkers. However, the ratio of NSE serum concentration (NSE ratio) before and after the treatment cycle may be a good tool for evaluating the therapeutic effect of metastatic neuroendocrine neoplasms of the liver (MNENOL).

Methods:

We collected clinical cases of NEN with liver metastases, calculating the ratio of NSE in each case before and after the treatment cycle, using thin-slice computed tomography or magnetic resonance imaging as a reference to evaluate the therapeutic effect. We analyzed the correlation between NSE ratio and NSE serum concentration and curative effect, and then compared the evaluation performance of the two.

Results:

We found that increase in the NSE ratio is a risk factor for the progression of MNENOL. Compared with NSE, NSE ratio has a greater advantage in evaluating the effect of MNENOL. NSE ratio is related to the curative effect of NEN, and the correlation is better than that of NSE. When judging whether NEN has new metastasis, the NSE ratio shows a similar effect to NSE, and there is no significant difference between the two.

Conclusion:

NSE ratio is more effective than NSE in evaluating the therapeutic effect of MNENOL, but it is not significantly different from NSE in terms of predicting new metastases.

Keywords

Neuron-specific enolase ratio metastatic neuroendocrine neoplasms of the liver therapeutic effect

Introduction

Neuron-specific enolase (NSE) is a soluble brain protein that can provide information about nerve, neuroendocrine, and paraneuronal cells. Neuroendocrine neoplasm (NEN) is a rare tumor originating from peptidergic neurons and neuroendocrine cells. It was called carcinoid in the past. It is highly complex and heterogeneous.^1,2 In recent years, its incidence has gradually increased.^3–5

NSE is a clinical biomarker of NEN, which is directly related to the differentiation, aggressiveness, and size of NEN, but its level of evidence is not as high as that of another biomarker, chromogranin A (CgA).^6,7 The specificity of NSE is similar to that of CgA, but the sensitivity is lower.⁸ Although CgA is more sensitive,⁷ the popularity of CgA testing is very low in China, the testing of this protein can only be performed by a few large hospitals, and the test results are difficult to standardize. CgA is often inconsistent with morphologic examination results.

In tumors of neuroendocrine origin, the staining intensity and number of cells expressing NSE are the highest, and the expression of NSE is similar to other neuroendocrine markers expressed in each tumor. There is a positive correlation between the numbers of other neuroendocrine markers expressed in each tumor.⁹ NSE serum concentration test is a common and popular test in various hospitals in China, which may be related to its use as a biomarker for diseases such as heart and brain-related diseases and small cell lung cancer.^10–13 For NEN, NSE is a relatively cost-effective tumor indicator in China, although its level of evidence is not as high as CgA, mRNA,¹⁴ and imaging (computed tomography [CT], magnetic resonance imaging [MRI], and positron emission tomography [PET]/CT).¹⁵ However, the role of NSE in the diagnosis and treatment of NEN is irreplaceable.

About 46%–93% of patients with NEN have liver metastases when they are diagnosed. Because the liver is rich in blood flow, wealthy in nutrients, and tumors grow vigorously, it is also the place where NEN often metastasizes.¹⁶ Therefore, it is important to pay attention to metastatic neuroendocrine neoplasms of the liver (MNENOL).

There are large differences in the expression of NSE in the population, and the same is true of its expression in patients with NEN. As a tumor marker with low-level evidence, it is difficult to improve upon the value of clinical evaluation of efficacy. However, blood NSE level is relatively independent. NSE serum concentration after a period of treatment is divided by the serum concentration before the treatment to obtain the ratio inferred to reflect the growth and metastases of NEN (NSE ratio). For this reason, we used imaging as a reference to explore the relationship between NSE ratio and the therapeutic effect of MNENOL.

Methods

Patients

Patients with NEN with liver metastases who received treatment since the establishment of the Neuroendocrine Tumor Diagnosis and Treatment Center of the First Affiliated Hospital of Nanjing Medical University (December 2018–December 2020) were included in our study. Inclusion criteria were NEN, liver metastases, treatment period longer than 3 months, and MRI or enhanced CT examination and NSE serum concentration detection simultaneously before and after the treatment period (within 3 days). Exclusion criteria were unclear location of the primary lesion, no systematic antitumor treatment, cases where the size of tumor lesions could not be assessed, or death.

NSE and tumor biomarkers serum concentration detection, NSE ratio calculation, and NEN therapeutic effect evaluation

Before and after the treatment cycle, CT or MRI (CT NET-specific triphasic scan, MRI gadolinium selenate disodium injection enhanced) examinations were performed on the enrolled patients, and slices of less than 5 mm were obtained to assess the size, number, and burden of the tumors. All image diagnosis reports were reevaluated by independent double-blind reviewers. At the same time, as CT or MRI examination, within 3 days, the serum concentrations of NSE, CYFRA211 (the soluble fragment of cytokeratin 19), CA72-4 (carbohydrate antigen 72-4), CA19-9 (carbohydrate antigen 19-9), CEA (carcinoembryonic antigen), and AFP (α-fetoprotein) were detected. These six tumor biomarkers were determined by the Roche Cobas e 602 analyzer system (Roche Diagnostics GmbH, Mannheim, Germany) based on electrochemiluminescence immunoassay and detected using standardized methods at the high-specification testing center. The normal value of NSE is <16.3 ng/mL. The NSE serum concentration after the treatment cycle is divided by the serum concentration before the treatment cycle to obtain the NSE ratio. The NSE serum concentration after the treatment cycle and NSE ratio were used to evaluate the tumor treatment effect.

Analysis of the relationship among NSE ratio, NSE and NEN curative effect, and tumor metastases

Logistic regression was used to analyze the relationship among NSE serum concentration, NSE ratio, other common tumor biomarkers, and tumor progression. We compared the value of NSE ratio and NSE serum concentration in evaluating the therapeutic effect of MNENOL. We drew the receiver operating characteristic (ROC) curve and analyzed and compared the ability of NSE ratio and NSE serum concentration to judge tumor progression and metastases. The relationships between stratified NSE ratio, ranked NSE serum concentration, and tumor progression and metastases were studied separately, and the differences between the two were compared.

Statistical analysis

The data from this study do not conform to the normal distribution. The data were compared to the two groups and multiple groups using the Mann-Whitney rank sum test and the Kruskal-Wallis test. The Kruskal-Wallis test was used for comparison between ranked data. The Spearman rank correlation test was used to analyze the correlation between graded data. Unconditional binary logistic regression was used to analyze the risk factors of tumor progression. p < 0.05 was considered significant. IBM SPSS software (version 25.0, IBM SPSS Inc., Chicago, IL) was used for all statistical evaluations.

Results

This study collected 57 NEN-diagnosed patients with liver metastases who met the criteria. Among them, the number of patients with original tumors from the stomach, pancreas, small intestine, colon, and rectum was significantly different (p = 0.000). There were more cases of first lesions from the rectum and pancreas than from other parts. According to the WHO NEN classification based on the Ki67 index,¹⁷ the cases in this study were divided into grade 1 (G1), grade 2 (G2), grade 3 (G3), and NEC, among which grade G2 had the largest number of patients (35 in total) (Table 1). The evaluation of the treatment effect showed that no patient achieved complete response, and the lower the NEN grade, the higher the disease control rate and objective response rate, and the higher the ratio of partial response (PR) to stable disease (SD) (p = 0.017), while progressive disease (PD) was the opposite. NEN classification and therapeutic effect have a rank correlation (p = 0.004, r = 0.376).

Table 1.

Characteristics of neuroendocrine neoplasm (NEN) with liver metastases.

NEN grade	Primary lesion, number of patients							Evaluation of the therapeutic effect
NEN grade	Stomach	Pancreas	Small intestine	Colon	Rectum	Other	Total	CR	PR	SD	PD	DCR, %	ORR, %
G1	1	2	0	0	4	1	8	0	3	3	2	75.0	37.5
G2	2	13	2	0	16	2	35	0	8	14	13	62.9	22.9
G3	0	0	1	1	3	1	6	0	1	1	4	33.3	16.7
NEC	1	0	1	3	2	1^a	8	0	0	1	7^b,c	12.5	12.5
Total	4	15	4	4	25	5	57	0	12	19	26	54.4	22.8

CR: complete response; DCR: disease control rate; NEC: neuroendocrine cancer; ORR: objective response rate; PD: progressive disease; PR: partial response; SD: stable disease.

Comparison of the number of patients with primary lesion (p = 0.000).

Comparison of the therapeutic effect of different grades of NEN (p = 0.017).

Correlation between NEN grade and therapeutic effect (p = 0.004; r = 0.376).

NSE ratio and NSE were both risk factors for NEN progression and the contribution of NSE ratio was greater than that of NSE

Based on whether the tumor had progressed, we assigned NSE ratio and NSE serum concentration. The assignment method was as follows: NSE ratio <0.5 was 1, 0.5 ⩽ NSE ratio < 1 was 2, 1 ⩽ NSE ratio < 1.5 was 3, 1.5 ⩽ NSE ratio < 2 was 4, NSE ratio ⩾2 was 5, NSE <16.3 was 1, 16.3 ⩽ NSE < 32.6 was 2, 32.6 ⩽ NSE < 48.9 was 3, 48.9 ⩽ NSE < 65.2 was 4, NSE ⩾65.2 was 5. After the assignment, together with other commonly used tumor biomarkers (CYFRA211, CA72-4, CA19-9, CEA, and AFP), they were substituted into the univariate unconditional binary logistic regression equation. The results showed that the odds ratio (Exp[B] = 5.623) of NSE ratio (p = 0.001) was significantly higher than that of NSE (p = 0.017, Exp[B] = 2.137) (Tables 2 and 3). Stepwise regression analysis showed that except for NSE ratio (p = 0.001, Exp[B] = 4.948), other commonly used tumor biomarkers were all kicked out of the equation in turn (Table 4).

Table 2.

Univariate nonconditional binary logistic regression analysis of neuron-specific enolase (NSE) ratio and common tumor biomarkers.

	B	SE	Wald	Significance	Exp(B)
CYFRA211, ng/mL	−0.028	0.212	0.017	0.895	0.972
CA72-4, U/mL	0.043	0.036	1.432	0.231	1.044
CA19-9, U/mL	−0.027	0.048	0.324	0.569	0.973
CEA, ng/mL	0.001	0.108	0.000	0.990	1.001
AFP, ng/mL	0.02	0.044	0.213	0.645	1.021
NSE ratio	1.727	0.498	12.015	0.001	5.623
Intercept	−4.734	1.408	11.305	0.001	0.009

AFP: α-fetoprotein; CEA: carcinoembryonic antigen.

Table 3.

Univariate nonconditional binary logistic regression analysis of neuron-specific enolase (NSE) and common tumor biomarkers.

	B	SE	Wald	Significance	Exp(B)
AFP, ng/mL	0.095	0.188	0.253	0.615	1.099
CEA, ng/mL	0.027	0.036	0.567	0.452	1.028
CA19-9, U/mL	−0.029	0.038	0.573	0.449	0.972
CA72-4, U/mL	−0.038	0.098	0.147	0.701	0.963
CYFRA211, ng/mL	0.006	0.034	0.03	0.862	1.006
NSE, ng/mL	0.759	0.318	5.7	0.017	2.137
Intercept	−1.876	0.871	4.644	0.031	0.153

AFP: α-fetoprotein; CEA: carcinoembryonic antigen.

Table 4.

Multivariate unconditional binary logistic stepwise regression analysis results of neuron-specific enolase (NSE) ratio and commonly used tumor biomarkers.

	B	SE	Wald	Significance	Exp(B)
NSE ratio	1.606	0.449	12.818	0	4.984
Intercept	−4.415	1.171	14.202	0	0.012

Both NSE ratio and NSE were related to tumor progression or remission: NSE ratio has advantages over NSE in judging whether NEN was progressing or in remission

NSE ratio, NSE (using log₂NSE, nonparametric test results are the same as NSE), and PR, SD, and PD correlation analysis results showed that in each group of PR, SD, and PD, the levels of NSE ratio (p = 0.000) and NSE (p = 0.006) were different, and NSE ratio (p = 0.000) and NSE (p = 0.001) showed a rank correlation in each group; the rank correlation coefficients were 0.669 and 0.437, respectively (Figure 1). According to the value of NSE ratio and NSE, they were divided into 5 layers, and the relationship between them and tumor progression (with and without PD) was explored. The result is that both the NSE ratio (p = 0.001) and NSE (p = 0.010) are related to tumor progression. The coefficients were 0.598 and 0.384, respectively, and the NSE ratio was more closely related to tumor progression (Figure 2).

We drew the ROC curve of NSE ratio and NSE according to whether the tumor was progressing. The results show that the areas under the ROC curve of NSE ratio (p = 0.000) and NSE (p = 0.002) were 0.859 and 0.737, respectively. Sensitivity and specificity of the NSE ratio ROC curves were 0.654 and 0.936, respectively; the sensitivity and specificity of the NSE ROC curves were 0.692 and 0.733, respectively. Comparing the performance of the two ROC curves, the difference was significant (p = 0.037), and the ROC curve of the NSE ratio was significantly better than that of NSE (Figure 3). We calculated the cutoff value through the ROC curve, and the results showed that the medical reference value of the NSE ratio for MNENOL was 1.176.

Both NSE ratio and NSE are correlated with new metastases, and there is no significant difference between the two in judging whether there are new metastases

The levels of NSE ratio (p = 0.008) and NSE (p = 0.006) were different in the groups with new metastases and those without metastases. The ROC curves of NSE ratio and NSE were drawn according to whether fresh metastases appeared. The results show that the areas under the ROC curves of NSE ratio (p = 0.004) and NSE (p = 0.005) were 0.752 and 0.763, respectively. The sensitivity and specificity of the NSE ratio ROC curve were 0.500 and 0.956; the sensitivity and specificity of the NSE ROC curve were 0.583 and 0.911. Comparing the performance of the two ROC curves, there was no significant difference (p = 0.238). According to the value of NSE ratio and NSE, they were divided into 5 layers, and the relationship between them and new tumor metastases was explored. NSE ratio (0.001) and NSE (0.009) were both related to tumor metastases, and the correlation coefficients were 0.433 and 0.342 (Figure 4).

Discussion

NEN is a rare disease. Its serum marker is mainly CgA.^6,7,18,19 However, due to the small number of cases, CgA detection cannot be popularized in most hospitals in China. In the process of diagnosis and treatment, NSE is often used as a clinical serologic tumor marker in patients with NEN. In recent years, another intermediate level of evidence NETest¹⁴ has been developed, but its high cost is prohibitive. This study explored the ratio of serum NSE concentration before and after the treatment cycle and described a new way to evaluate the curative effect of NEN. Our study tried to find the value of this ratio.

Because NEN is a rare disease, only 57 cases were collected in this study. MNENOL are a special kind of NEN. There are no patients in complete remission among these cases, which may be related to the difficulty of complete remission of liver tumor. The results showed that the relationship between NSE ratio and PR, SD, and PD was better than NSE in evaluating the therapeutic effect of MNENOL, with a correlation coefficient of 0.669 vs 0.437. Even if only tumor progression was evaluated, NSE ratio was better than NSE, with a correlation coefficient of 0.589 vs 0.384. In logistic regression analysis, NSE ratio and NSE were assigned to 5 grades; the odds ratios of NSE ratio and NSE were 5.623 and 2.137, respectively. The NSE ratio was clearly more valuable. Drawing ROC curve to screen for tumor progression, the NSE ratio was significantly different from NSE (p = 0.032). The area under the curve was 0.859 vs 0.737, and the specificity was 0.936 vs 0.733. These data suggest that the evaluation value of the NSE ratio for the treatment effect is much higher than that of NSE itself. NSE ratio, which is a comparison of the same person before and after treatment, eliminates heterogeneity,^20,21 whereas NSE is based on medical reference values and has poor sensitivity.²²

We also explored the relationship between NSE ratio and new metastases and found that when fresh metastases appeared, both NSE ratio and NSE increased significantly. There is no difference in performance between the two in this regard. There is no significant difference in either the ROC curve or the correlation coefficient, but both the NSE ratio and the NSE have high performance in diagnosing the specificity of new tumor metastases, which reflects the unique evaluation value of NSE ratio and NSE in the treatment process of MNENOL.

In conclusion, the value of the NSE ratio exceeds that of serum concentration of NSE, can better evaluate the therapeutic effect of MNENOL, and has good clinical value. It can also be compared with other NEN biomarkers in the future. Due to the rarity of this disease, as reflected by the small sample size in this study, more cases need to be collected for research.

Footnotes

Author contributions

All the authors contributed to the conception of the study. Jingbao Kan performed the data collection and composed the draft of the manuscript. All the authors take responsibility for the accuracy of the results and agreed to publish the manuscript in its final form.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Jingbao Kan

References

Garcia-Carbonero

Rinke

Valle

, et al. ENETS consensus guidelines for the standards of care in neuroendocrine neoplasms: systemic therapy 2: chemotherapy. Neuroendocrinology 2017; 105: 281–294.

Pavel

Valle

Eriksson

, et al. ENETS consensus guidelines for the standards of care in neuroendocrine neoplasms: systemic therapy: biotherapy and novel targeted agents. Neuroendocrinology 2017; 105: 266–280.

Hallet

Law

Cukier

, et al. Exploring the rising incidence of neuroendocrine tumors: a population-based analysis of epidemiology, metastatic presentation, and outcomes. Cancer 2015; 121: 589–597.

Leoncini

Boffetta

Shafir

, et al. Increased incidence trend of low-grade and high-grade neuroendocrine neoplasms. Endocrine 2017; 58: 368–379.

Hauso

Gustafsson

Kidd

, et al. Neuroendocrine tumor epidemiology. Cancer 2008; 113: 2655–2664.

Georgantzi

Skoldenberg

Stridsberg

, et al. Chromogranin A and neuron-specific enolase in neuroblastoma: correlation to stage and prognostic factors. Pediatr Hematol Oncol 2018; 35: 156–165.

Giovanella

La Rosa

Ceriani

, et al. Chromogranin-A as a serum marker for neuroendocrine tumors: comparison with neuron-specific enolase and correlation with immunohistochemical findings. Int J Biol Markers 1999; 14: 160–166.

Herrera-Martinez

Hofland

Galvez Moreno

, et al. Neuroendocrine neoplasms: current and potential diagnostic, predictive and prognostic markers. Endocr Relat Cancer. 2019; 26: R157–R179.

Mjones

Sagatun

Nordrum

, et al. Neuron-specific enolase as an immunohistochemical marker is better than its reputation. J Histochem Cytochem 2017; 65: 687–703.

10.

Bagnato

Andriolo

Boccagni

, et al. Reduced neuron-specific enolase levels in chronic severe traumatic brain injury. J Neurotrauma 2020; 37: 423–427.

11.

Chung-Esaki

Mui

Mlynash

, et al. The neuron specific enolase (NSE) ratio offers benefits over absolute value thresholds in post-cardiac arrest coma prognosis. J Clin Neurosci 2018; 57: 99–104.

12.

Vondrakova

Kruger

Janotka

, et al. Association of neuron-specific enolase values with outcomes in cardiac arrest survivors is dependent on the time of sample collection. Crit Care 2017; 21: 172.

13.

Streitberger

Leithner

Wattenberg

, et al. Neuron-specific enolase predicts poor outcome after cardiac arrest and targeted temperature management: a multicenter study on 1,053 patients. Crit Care Med 2017; 45: 1145–1151.

14.

Öberg

Califano

Strosberg

, et al. A meta-analysis of the accuracy of a neuroendocrine tumor mRNA genomic biomarker (NETest) in blood. Ann Oncol 2020; 31: 202–212.

15.

de Mestier

Dromain

d'Assignies

, et al. Evaluating digestive neuroendocrine tumor progression and therapeutic responses in the era of targeted therapies: state of the art. Endocr Relat Cancer 2014; 21: R105–R120.

16.

Harring

Nguyen

Goss

, et al. Treatment of liver metastases in patients with neuroendocrine tumors: a comprehensive review. Int J Hepatol 2011; 2011: 154541.

17.

Nagtegaal

Odze

Klimstra

, et al. The 2019 WHO classification of tumours of the digestive system. Histopathology 2020; 76: 182–188.

18.

Modlin

Bodei

Kidd

Neuroendocrine tumor biomarkers: from monoanalytes to transcripts and algorithms. Best Pract Res Clin Endocrinol Metab 2016; 30: 59–77.

19.

Modlin

Oberg

Taylor

, et al. Neuroendocrine tumor biomarkers: current status and perspectives. Neuroendocrinology 2014; 100: 265–277.

20.

Pavel

Oberg

Falconi

, et al. Gastroenteropancreatic neuroendocrine neoplasms: ESMO clinical practice guidelines for diagnosis, treatment and follow-up. Ann Oncol 2020; 31: 844–860.

21.

Haque

Polcyn

Matzelle

, et al. New insights into the role of neuron-specific enolase in neuro-inflammation, neurodegeneration, and neuroprotection. Brain Sci 2018; 8.

22.

Baudin

Gigliotti

Ducreux

, et al. Neuron-specific enolase and chromogranin A as markers of neuroendocrine tumours. Br J Cancer 1998; 78: 1102–1107.