Sage Journals: Discover world-class research

Abstract

BACKGROUND:

The significance of serum dipeptidyl peptidase-IV (DPP-IV) in papillary thyroid carcinoma (PTC) has not been elucidated.

OBJECTIVE:

This study aimed to assess the role of serum DPP-IV in the carcinogenesis and prognosis of PTC.

METHODS:

The serum DPP-IV concentration was measured in 171 male patients with PTC, 81 male patients with a benign thyroid nodule (BTN), and 52 male healthy controls (HCs). Multivariate logistic regression and Cox regression analyses were performed to evaluate the correlations between variables. Receiver operating characteristic (ROC) curves were used to calculate the diagnosis accuracy.

RESULTS:

The ROC curve indicated a good performance of DPP-IV for discriminating PTC from BTN, with an area under the curve (AUC) of 0.881 (95% CI, 0.840–0.922). Serum DPP-IV demonstrated a modest performance in predicting nonstructurally persistent disease/recurrent disease (NSPRD) survival, with an AUC of 0.778 (95% CI, 0.635–0.922). A serum DPP-IV level $\geqslant$ 250 nkat/L (HR, 6.529; 95% CI, 2.090–20.398; $P=$ 0.001) and an advanced tumor, lymph node, metastasis (TNM) stage (HR, 4.677; 95% CI, 1.498–14.605; $P=$ 0.008) were found to be independent factors for predicting SPRD. PTC patients with a DPP-IV level $\geqslant$ 250 nkat/L had a worse outcome than those with a DPP-IV level $<$ 250 nkat/L ( $P<$ 0.001).

CONCLUSIONS:

Serum DPP-IV may be a predictive biomarker for PTC diagnosis and prognosis in Chinese male patients.

Keywords

Dipeptidyl peptidase-IV CD26 papillary thyroid carcinoma structurally persistent recurrent disease

1. Introduction

Thyroid carcinoma is the most common endocrine malignancy, and its incidence has increased rapidly in recent years [1, 2]. In particular, papillary thyroid carcinoma (PTC) accounts for 80–85% of all thyroid malignancies [1, 3]. Currently, effective and reliable biomarkers allowing the early prediction and prognostic intervention for PTC are not available, although several biomarkers, such as high-mobility group A1, long noncoding RNA cancer susceptibility candidate 2, and survivin DEx3, have been studied for thyroid carcinomas [4, 5, 6]. Therefore, it is very important to identify novel biomarkers that can accurately diagnose PTC and predict its prognosis.

Dipeptidyl peptidase IV (DPP-IV), also referred to as CD26 (EC 3.4.14.5) [7], is a glycosylated transmembrane type II glycoprotein [8, 9, 10]. It is able to cleave N-terminal dipeptides with either L-proline or L-alanine at the penultimate position [11]. A soluble form of DPP-IV exists in bodily fluids [12], and the tissue origin of circulating DPP-IV includes bone marrow cells [13], adipocyte cells [14], and human skeletal muscle cells [15]. DPP-IV is implicated in a variety of biological processes, including lipid metabolism [16], glucose regulation [17], autoimmunity and inflammation [18], carcinogenesis [9], and aggressiveness of cancer [19]. In addition, overexpression of DDP-IV has been found in thyroid cancers and other malignancies [1]. Moreover, DPP-IV is involved in various aspects of carcinogenesis, including tumor cell proliferation and invasion [20], metastasis [10, 21], and prognosis [22], and has been studied as a therapeutic target [7, 9]. The diverse biological activities of DPP-IV largely depend on the cell type and the microenvironment of the carcinoma [11]. Both in vitro and in vivo experiments have verified that DPP-IV accelerates endometrial carcinoma progression [9]. Furthermore, DPP-IV has been implicated in the pathogenesis of thyroid carcinomas as it exhibits significantly increased activity in PTC tissue than in follicular thyroid adenoma, thyroid nodular hyperplasia, and nontumorous tissue [23]. DPP-IV staining in tissue for the diagnosis of PTC has yielded a sensitivity of 97–100% [24] and a specificity of 95% [25]. Additionally, high DPP-IV expression in PTC tissue has been positively correlated with extrathyroidal extension and an advanced tumor stage, and negatively correlated with no evidence of disease during follow-up [1]. Given a certain role of tissue DPP-IV in the diagnosis and prognosis of various carcinomas, there is growing interest in the potential of serum DPP-IV for predicting carcinomas; however, various studies have yielded conflicting results. For example, circulating DPP-IV has been shown to be overexpressed in patients with pancreatic ductal adenocarcinoma [26]. In addition, serum DPP-IV, as an independent prognostic factor for colorectal cancer, has been found to be correlated with an unfavorable overall and disease-free survival [22]. In contrast, another study found that, compared with healthy controls, the serum DPP-IV level was lower in patients with esophageal squamous cell carcinoma, and a low DPP-IV level predicted a poor prognosis [27]. Furthermore, for patients with Ewing sarcoma, there is no significant difference in serum DPP-IV activity between healthy controls and patients with osteosarcoma [28]. Thus, for PTC, the diagnostic applicability of circulating DPP-IV has not been established.

In this study, we investigated the role of the serum DPP-IV level in patients with PTC versus patients with a benign thyroid nodule (BTN) and healthy individuals, and its predictive performance for the diagnosis and prognosis of PTC was also assessed.

2. Materials and methods

2.1 Study population

Between March 2012 and March 2013, male patients with PTC at The First Hospital of Jilin University were consecutively enrolled. In addition, consecutive subjects with a BTN from the Physical Examination Department were recruited. The diagnosis of PTC or BTN was based on radiological evidence from ultrasound, computed tomography (CT) scanning, and magnetic resonance imaging as well as histopathological results of fine-needle aspiration biopsy or thyroid tissue after operation. Volunteers without thyroid diseases or malignant diseases were enrolled as healthy controls (HCs) and donated their blood samples at the Physical Examination Department. All the patients with PTC had newly diagnosed primary PTC. The exclusion criteria included the following: 1) carcinomas other than PTC; 2) metastases from another carcinoma to the thyroid; 3) radiation therapy at any time before admission; 4) severe comorbidities, such as renal failure, heart failure, liver cirrhosis, and malignant hematologic disease; 5) insufficient data. Approval was obtained from the Ethics Committee of The First Hospital of Jilin University, and all subjects gave their informed consent to participate.

Tumor, lymph node, metastasis (TNM) stages of PTC were graded according to the 7 ${}^{\text{th}}$ edition of the American Joint Committee on Cancer staging system [1]. Early-stage was denoted as TNM stage I or II, and advanced-stage as stage III or IV. The bilateral central-compartment neck dissection was routinely performed during the total thyroidectomy. Lateral or modified neck dissection was conducted in patients with clinically definite or imaging-detected metastatic cervical lymph nodes. For suspected lateral lymph node metastasis (LLNM) found preoperatively and/or during operation, lateral cervical node dissection was performed. All operations were performed by the same surgical team.

Patients were routinely followed up every three months after surgery during a four-year follow-up until March 2017. All patient data were originated from a thyroid carcinoma database. During follow-up, neck ultrasound and lung CT were used to monitor for PTC recurrence. Both serum thyroglobulin (Tg) and thyroglobulin antibody (TgAb) levels were measured every three months. The disease status during the follow-up was defined based on the modified criteria as follows [1]: 1) no evidence of disease, Tg $<$ 1 ng/mL and no structurally detectable disease; 2) indeterminate response, persistently detectable TgAb ( $>$ 0 IU/mL) and Tg $<$ 1 ng/mL but absence of structural disease; 3) biochemically persistent disease, Tg $\geqslant$ 1 ng/mL but absence of structural disease; 4) structurally persistent disease, metastasis (locoregional or distant) irrespective of the Tg level; 5) recurrence events, structural evidence of disease identified after a period of no evidence of disease. The nonstructurally persistent disease/recurrent disease (NSPRD) survival time was defined as the interval between the initial treatment date and structurally persistent disease/recurrent disease (SPRD) or the last follow-up.

2.2 Blood samples and determination of serum DPP-IV concentration

Venous blood samples were taken from each patient with PTC at the time of diagnosis and from the controls (BTN and HCs). Blood samples were placed at room temperature for 30 min for coagulation, followed by centrifugation at 3000 $\times$ g for 10 min. The supernatant was stored frozen at $-$ 70 ${}^{\circ}$ C until analysis.

As described previously [29], the assay for the determination of DPP-IV enzymatic activity was conducted in triplicate in 50 mM Hepes buffer (200 mM NaCl, 10 mM ZnCl ${}_{2}$ , and 1% DMSO, pH 6.8). Enzymatic reactions were initiated by adding the substrate (H-Gly-Pro- $\beta$ -naphthylamide) of the DPP-IV enzymatic reaction in 96-well microtiter plates (Nunclon, Denmark). Then, Lys-Z-NO-thiazolidid, an inhibitor of DPP-IV, was added to terminate the DPP-IV activity. The optical density was measured at 405 nm with an excitation at 620 nm by an Infinite M200 microtiter plate reader (Tecan, Crailsheim, Germany).

Table 1
Clinicopathological characteristics of the male subjects, $n$ (%)

Characteristic	PTC	BTN	HC
	( $n=$ 171)	( $n=$ 81)	( $n=$ 52)
Age
$<$ 45 years old	103 (60.2)	11 (13.6)	27 (51.9)
$\geqslant$ 45 years old	68 (39.8)	70 (86.4)	25 (48.1)
Tumor size
$\leqslant$ 1 cm	114 (66.7)
$>$ 1 cm	57 (33.3)
Capsule invasion
No	53 (31.0)
Yes	118 (69.0)
Multifocality
Unifocal	79 (46.2)
Multifocal	92 (53.8)
Nodal status
N0	64 (37.4)
CLNM	105 (61.4)
LLNM	39 (22.8)
Extrathyroidal invasion
Negative	137 (80.1)
Microscopic	27 (15.8)
Macroscopic	7 (4.1)
Vascular invasion
No	157 (91.8)
Yes	14 (8.2)
Distant metastasis
No	162 (94.7)
Yes	9 (5.3)
TNM stage
I $+$ II	133 (77.8)
III $+$ IV	38 (22.2)

BTN, benign thyroid nodule; HC, healthy control; PTC, papillary thyroid carcinoma; CLNM, central lymph node metastasis; LLNM, lateral lymph node metastasis; TNM, tumor, lymph node, metastasis.

Table 2

The correlation between the serum DPP-IV level and clinicopathological characteristics by binary univariate logistic regression analysis

	OR	95% CI	$P$ value
Malignancy (PTC vs. HC $+$ BTN)	1.074	1.055–1.092	$<$ 0.001
Tumor size ( $\geqslant$ 1 cm vs. $<$ 1 cm )	1.002	0.995–1.010	0.507
Capsule invasion (yes vs. no)	1.009	1.001–1.017	0.028
Multifocality (yes vs. no)	1.003	0.996–1.010	0.360
CLNM (yes vs. no)	1.012	1.004–1.020	0.004
LLNM (yes vs. no)	0.998	0.990–1.006	0.629
Extrathyroidal invasion (yes vs. no)	1.003	0.995–1.012	0.471
Vascular invasion (yes vs. no)	1.014	1.002–1.026	0.021
Distant metastasis (yes vs. no)	1.010	0.996–1.025	0.150
TNM stage (III $+$ IV vs. I $+$ II)	1.024	1.014–1.034	$<$ 0.001

DPP-IV, dipeptidyl peptidase-IV; OR, odds ratio; CI, confidence interval; PTC, papillary thyroid carcinoma; BTN, benign thyroid nodule; CLNM, central lymph node metastasis; LLNM, lateral lymph node metastasis.

Table 3

Diagnostic performance of serum DPP-IV in differentiating PTC from BTN and/or HC

	Optimal cutoff value	AUC	Sensitivity	Specificity	PPV	NPV	LR $+$	LR $-$
	(nkat/L)	(95% CI)	(%)	(%)	(%)	(%)
PTC vs. BTN $+$ HC	174.70	0.891 (0.855–0.926)	73.7	93.2			10.838	0.282
PTC vs. BTN	174.70	0.881 (0.840–0.922)	73.7	95.1	96.9	63.1	15.040	0.277
PTC vs. HC	161.70	0.906 (0.862–0.949)	86.0	82.7			4.971	0.169

DPP-IV, dipeptidyl peptidase-IV; AUC, area under the curve; BTN, benign thyroid nodule; HC, healthy control; PPV, positive predictive value; NPV, negative predictive value; LR $+$ , positive likelihood ratio; LR $-$ , negative likelihood ratio; PTC, papillary thyroid carcinoma.

2.3 Statistical analysis

SPSS (version 18.0) and GraphPad Prism software (version 5.01) were used for statistical analyses. Only male subjects were enrolled to avoid the influence of gender. Continuous data are expressed as the mean $\pm$ standard deviation when normally distributed, or median and interquartile range when not normally distributed. Categorical data are expressed as the number of patients and percentages. Continuous data were compared using an independent t test or analysis of variance (ANOVA) when normally distributed, or a nonparametric test (Mann-Whitney U test or Kruskal-Wallis H(K) test) when not normally distributed. The proportions were compared using the chi-squared test or Fisher’s exact test. Receiver operating characteristic (ROC) curve analysis was conducted to evaluate the performance of DPP-IV in the prediction of PTC. The sensitivity against the false-positive rate (1-specificity) was plotted. The optimal cutoff value for diagnosis was identified according to the Youden index [30]. The cumulative NSPRD survival rate was evaluated by the Kaplan-Meier method, and comparisons among groups were analyzed using the log-rank test. Binary univariate logistic regression was employed to examine the correlations between the serum DPP-IV level and the clinicopathological characteristics, and the data are presented as odds ratios (ORs) and 95% confidence intervals (CIs). Univariate and multivariate Cox proportional hazard analyses were performed to identify independent prognostic factors for SPRD and are presented as hazard ratios (HRs) and 95% CIs. Variables with a $P$ value $<$ 0.15 according to the univariate analysis were introduced into the multivariate model. All statistical tests were two-sided, and statistical significance was determined as $P<$ 0.05.

3. Results

3.1 Patient characteristics

In this study, a total of 171 patients with PTC, 81 with a BTN, and 52 HCs were included. The clinicopathological characteristics are summarized in Table 1. Subjects with a BTN (53.30 $\pm$ 11.29 years old) were significantly older than those with PTC (42.70 $\pm$ 10.42 years old) and HCs (42.02 $\pm$ 13.36 years old, $P<$ 0.05). Among patients with PTC, 105 (61.4%) had central lymph node metastasis (CLNM) and 39 (22.8%) had LLNM. In addition, 139 (77.8%) patients were TNM stage I or II, and 38 (22.2%) were stage III or IV.

3.2 Comparison of serum DPP-IV levels among groups

The serum DPP-IV concentration was 207.9 $\pm$ 44.0 nkat/L in the PTC group, which was significantly greater than those of the BTN (155.3 $\pm$ 17.6 nkat/L), HC (153.9 $\pm$ 15.6 nkat/L), and BTN $+$ HC (154.7 $\pm$ 16.8 nkat/L) groups, all $P<$ 0.001, Fig. 1. There was no significant difference in the serum DDP-IV concentration between the BTN and HC groups ( $P\geqslant$ 0.05).

3.3 Correlation between the serum DPP-IV level and clinicopathological characteristics

Binary univariate logistic regression analysis showed that the serum DPP-IV level was positively correlated with malignancy, capsule invasion, CLNM, vascular invasion, and TNM stage (III $+$ IV), all $P<$ 0.05, Table 2.

Figure 1.

Serum dipeptidyl peptidase-IV (DPP-IV) levels in healthy controls (HCs), patients with a benign thyroid nodule (BTN), and patients with papillary thyroid carcinoma (PTC). The black horizontal lines indicate the mean values of measurements, and the error bars indicate standard deviations. Data were statistically analyzed by the ANOVA test.

Figure 2.

The capability of serum DPP-IV in the differential diagnosis between PTC vs. HCs and/or BTN by ROC curves. A. PTC vs. HC $+$ BTN (AUC $=$ 0.891); B. PTC vs. BTN (AUC $=$ 0.881); C. PTC vs. HCs (AUC $=$ 0.906). DPP-IV, dipeptidyl peptidase-IV; HCs, healthy controls; BTN, benign thyroid nodule; PTC, papillary thyroid carcinoma; ROC, receiver operating characteristic; AUC, area under the ROC curve.

3.4 Performance of serum DPP-IV in the diagnosis of PTC

ROC curves were used to assess the validity of the serum DPP-IV level in diagnosing PTC (Table 3, Fig. 2). ROC curves revealed that serum DPP-IV yielded an AUC of 0.881 (95% CI, 0.840–0.922) at the cutoff value of 174.70 nkat/L for differentiating PTC from BTN. Moreover, at the same cutoff value, DPP-IV yielded an AUC of 0.891 (95% CI, 0.855–0.926) for the differential diagnosis of PTC from BTN $+$ HC. In addition, an AUC of 0.906 (95% CI, 0.862–0.949) was yielded at the cutoff value of 161.70 nkat/L for diagnosing PTC from the HC group.

3.5 Clinicopathological features of patients with PTC by serum DPP-IV stratification

Considering the optimal cutoff value of DPP-IV as 174.70 nkat/L for differentiating PTC from BTN, patients with PTC were stratified into two subgroups: those with a low level of serum DPP-IV ( $<$ 174.70 nkat/L) and those with a high level of serum DPP-IV ( $\geqslant$ 174.70 nkat/L). Patients with a high DPP-IV level were more likely to have CLNM ( $P<$ 0.001) and an advanced TNM stage ( $P=$ 0.037, Table 4).

Table 4
Clinicopathological features of patients with PTC by serum DPP-IV stratification at 174.70 nkat/L, $n$ (%)

	Serum DPP-IV
	Low level	High level	$P$ value
	$<$ 174.70	$\geqslant$ 174.70
	( $n=$ 45)	( $n=$ 126)
Age			0.751
$<$ 45 years old	28 (62.2)	75 (59.5)
$\geqslant$ 45 years old	17 (37.8)	51 (40.5)
Tumor size			0.065
$\leqslant$ 1 cm	35 (77.8)	79 (62.7)
$>$ 1 cm	10 (22.2)	47 (37.3)
Capsule invasion			0.128
No	18 (40.0)	35 (27.8)
Yes	27 (60.0)	91 (72.2)
Multifocality			0.942
No	21 (46.7)	58 (46.0)
Yes	24 (53.3)	68 (54.0)
CLNM
No	29 (64.4)	37 (29.4)	$<$ 0.001
Yes	16 (35.6)	89 (70.6)
LLNM			0.601
No	36 (80.0)	96 (76.2)
Yes	9 (20.0)	30 (23.8)
Extrathyroidal invasion			0.428
No	39 (86.7)	98 (77.8)
Microscopic	5 (11.1)	22 (17.5)
Macroscopic	1 (2.2)	6 (4.7)
Vascular invasion			0.167
No	44 (97.8)	113 (89.7)
Yes	1 (2.2)	13 (10.3)
Distant metastasis			0.499
No	44 (97.8)	118 (93.7)
Yes	1 (2.2)	8 (6.3)
TNM stage			0.037
I $+$ II	40 (88.9)	93 (73.8)
III $+$ IV	5 (11.1)	33 (26.2)

DPP-IV, dipeptidyl peptidase-IV; CLNM, central lymph node metastasis; LLNM, lateral lymph node metastasis; TNM, tumor, lymph node, metastasis.

Figure 3.

Serum DPP-IV levels in the patients with CLNM or different TNM stages. A. Patients with CLNM vs. without CLNM; B. Patients with an early TNM stage (I $+$ II) vs. an advanced TNM stage (III $+$ IV). DPP-IV, dipeptidyl peptidase-IV; CLNM, central lymph node metastasis; TNM, tumor, lymph node, metastasis. The black horizontal lines indicate the mean values of measurements, and the error bars indicate standard deviations.

Figure 4.

Receiver operating characteristic curve analysis for structurally persistent disease and recurrent disease in patients with papillary thyroid carcinoma predicted by the level of serum dipeptidyl peptidase-IV (AUC $=$ 0.778).

Patients with CLNM had a higher DPP-IV level than patients without (215.80 $\pm$ 44.67 nkat/L vs.195.35 $\pm$ 40.09 nkat/L, $P=$ 0.003). Also, patients with an advanced TNM stage (stage III or IV) had a higher DPP-IV level than those with TNM stage I or II (242.29 $\pm$ 52.71 nkat/L vs. 198.08 $\pm$ 35.71 nkat/L, $P<$ 0.001; Fig. 3).

3.6 Serum DPP-IV for prediction of clinical outcomes

The patients were followed up for four years. By ROC curve analysis, serum DPP-IV demonstrated a moderate performance in predicting SPRD at the cutoff value of 250.0 nkat/L, with an AUC of 0.778 (95% CI, 0.635–0.922), a sensitivity of 66.7%, a specificity of 85.3%, a positive predictive value of 66.7%, a negative predictive value of 85.3%, a positive likelihood ratio of 4.537, and a negative likelihood ratio of 0.390 (Fig. 4).

Patients with PTC had a mean NSPRD survival time of 46.65 $\pm$ 5.01 months and a SPRD rate of 9.9%. Patients with DPP-IV $\geqslant$ 250.0 nkat/L had a shorter NSPRD survival time and a higher rate of SRPD than those with DPP-IV $<$ 250.0 nkat/L (42.45 $\pm$ 9.31 months vs. 47.65 $\pm$ 2.38 months, $P=$ 0.003; and 36.4% vs. 3.6%, $P<$ 0.001). By the Kaplan-Meier plot, patients with DPP-IV $\geqslant$ 250.0 nkat/L had a worse outcome than those with DPP-IV $<$ 250.0 nkat/L (log-rank $P<$ 0.001; Fig. 5).

Table 5
Clinicopathological features of patients with SPRD or NSPRD, $n$ (%)

Variable	NSPRD	SPRD	$P$ value
Age			0.006
$<$ 45 years old	98 (63.6)	5 (29.4)
$\geqslant$ 45 years old	56 (36.4)	12 (70.6)
Tumor size			0.071
$\leqslant$ 1 cm	106 (68.8)	8 (47.1)
$>$ 1 cm	48 (31.2)	9 (52.9)
Capsule invasion			0.018
No	52 (33.8)	1 (5.9)
Yes	102 (66.2)	16 (94.1)
Multifocality			0.143
No	74 (48.1)	5 (29.4)
Yes	80 (51.9)	12 (70.6)
CLNM			0.062
No	63 (40.9)	3 (17.6)
Yes	91 (59.1)	14 (82.4)
LLNM			0.818
No	118 (76.6)	14 (82.4)
Yes	36 (23.4)	3 (17.6)
Extrathyroidal invasion			0.087
Negative	127 (82.5)	10 (58.8)
Microscopic	22 (14.3)	5 (29.4)
Macroscopic	5 (3.2)	2 (11.8)
Vascular invasion			0.004
No	145 (94.2)	12 (70.6)
Yes	9 (5.8)	5 (29.4)
Distant metastasis			1.000
No	146 (94.8)	16 (94.1)
Yes	8 (5.2)	1 (5.9)
TNM stage			$<$ 0.001
I $+$ II	128 (83.1)	5 (29.4)
III $+$ IV	26 (16.9)	12 (70.6)

CLNM, central lymph node metastasis; LLNM, lateral lymph node metastasis; SPRD, structurally persistent disease/recurrent disease; NSPRD, nonstructurally persistent/recurrent disease.

Figure 5.

Nonstructurally persistent disease/recurrent disease (NSPRD) survival time curve in the male patients with papillary thyroid carcinoma based on the cut-off value of dipeptidyl peptidase-IV (DPP-IV) at 250.0 nkat/L. SPRD was observed in 17 out of 171 subjects, among which 5 of 138 patients had a DPP-IV level $<$ 250 nkat/L and 12 of 33 patients had a DPP-IV level $\geqslant$ 250 nkat/L.

Older patients ( $\geqslant$ 45 years old) were more likely to have SPRD events than younger patients ( $<$ 45 years old). In addition, patients with SPRD were more likely to develop capsule invasion, vascular invasion, and an advanced TNM stage (III $+$ IV), as compared to patients with NSPRD (all $P<$ 0.05, Table 5).

Table 6

Risk factors for predicting SPRD in patients with PTC by Cox proportional hazards regression analysis

	Univariate analysis		Multivariate analysis
	HR (95% CI)	$P$ value	HR (95% CI)	$P$ value
Age ( $\geqslant$ 45 years vs. $<$ 45 years)	3.912 (1.378–11.107)	0.010	0.000 (0.000–39.451)	0.920
BMI ( $\geqslant$ 25 kg/m ${}^{2}$ vs. $<$ 25 kg/m ${}^{2}$ )	1.777 (0.579–5.451)	0.315
Tumor size ( $>$ 1 cm vs. $\leqslant$ 1 cm)	2.276 (0.878–5.899)	0.089	1.227 (0.336–4.484)	0.757
Capsule invasion (yes vs. no)	7.699 (1.021–58.064)	0.048	3.690 (0.478–28.507)	0.211
Multifocality (yes vs. no)	2.074 (0.731–5.889)	0.170
CLNM (yes vs. no)	3.055 (0.878–10.630)	0.079	0.544 (0.113–2.616)	0.447
LLNM (yes vs. no)	0.692 (0.199–2.409)	0.563	0.604 (0.125–2.914)	0.530
Extrathyroidal invasion (yes vs. no)	2.806 (1.068–7.375)	0.036	1.677 (0.530–5.303)	0.379
Vascular invasion (yes vs. no)	4.739 (1.668–13.458)	0.003	1.536 (0.437–5.398)	0.503
Distant metastasis (yes vs. no)	1.013 (0.146–8.315)	0.924	0.247 (0.018–3.293)	0.290
TNM stage (III $+$ IV vs. I $+$ II)	9.860 (3.470–28.015)	0.000	4.677 (1.498–14.605)	0.008
DPP-IV ( $\geqslant$ 250 nkat/L vs. $<$ 250 nkat/L)	12.305 (4.327–34.987)	0.000	6.529 (2.090–20.398)	0.001

SPRD, structurally persistent disease/recurrent disease; PTC, papillary thyroid carcinoma; OR, Odds ratio; CI, confidence interval; BMI, body mass index; CLNM, central lymph node metastasis; LLNM, lateral lymph node metastasis.

3.7 Risk factors for predicting SPRD in patients with PTC by Cox proportional hazards regression analysis

Analyzed by the univariate Cox proportional hazard regression model, age ( $P=$ 0.010), capsule invasion ( $P=$ 0.047), extrathyroidal invasion ( $P=$ 0.036), vascular invasion ( $P=$ 0.004), advanced TNM stage ( $P<$ 0.001), and serum DPP-IV level ( $\geqslant$ 250.0 nkat/L) ( $P<$ 0.001) were all determined to be risk factors of SPRD. Variables with a $P$ value $<$ 0.15 in the univariate analysis were introduced into the multivariate model. Also, variables that were considered to be closely related to a dependent variable (e.g., SPRD) in clinical work, such as distant metastasis, were enrolled in the multivariate models, irrespective of the $P$ value. According to the multivariate Cox regression model, a serum DPP-IV level $\geqslant$ 250 nkat/L (HR, 6.529; 95% CI, 2.090–20.398; $P=$ 0.001) and an advanced TNM stage (HR 4.677; 95% CI 1.498–14.605; $P=$ 0.008) were found to be independent factors for predicting SPRD (Table 6).

4. Discussion

Although PTC typically has a good prognosis, there is a risk of it differentiating into aggressive and lethal thyroid carcinomas, e.g. poorly differentiated thyroid carcinoma or anaplastic thyroid carcinoma [31]. Moreover, PTC has shown a high recurrence rate (30%) and a high cancer-related mortality rate (8%) over a long-term follow-up period [32]. Therefore, early diagnosis and intervention is of pivotal significance. In this study, we found that the serum DDP-IV expression was significantly higher in patients with PTC than in patients with a BTN and HCs. Serum DPP-IV cutoff values of 174.70 nkat/L for PTC diagnosis and 250 nkat/L for SPRD were identified. Patients were more likely to develop SPRD if they had a larger tumor size ( $\geqslant$ 1 cm), capsule invasion, vascular invasion, and an advanced TNM stage (III $+$ IV). The serum DPP-IV level and advanced TNM stage were found to be independent factors for predicting PTC prognosis. Thus, serum DPP-IV may serve as a novel potential biomarker for PTC diagnosis and prognostic prediction in clinical applications.

Tissue DPP-IV can be utilized as a diagnostic biomarker for discriminating benign neoplasms from malignancies. An immunohistochemistry assay has demonstrated that DPP-IV has a sensitivity of 82.81% for the diagnosis of epithelial ovarian carcinoma, which was increased to 97.67% if detected by mRNA analysis [33]. In addition, tissue DPP-IV is associated with cancer progression and aggressiveness. In colorectal cancer, DPP-IV activity is positively associated with an advanced TNM stage, the degree of differentiation, and the development of metastasis [34]. Moreover, a higher tissue DPP-IV level predicts a worse overall survival as compared to a lower level in colorectal cancer or PTC patients [34]. Taken together, these findings verify the importance of DPP-IV expression in carcinogenesis, carcinoma aggressiveness, and prognosis of diverse cancers.

However, the use of molecular analysis of tumor tissues and cells is restricted in daily clinical work due to its invasive properties and unavailability. Herein, relative to the detection of tissue DPP-IV, determination of the serum DPP-IV level is more convenient, less invasive, and more cost-efficient. Owing to the heterogeneity of previous studies, no consensus has been achieved regarding the level of serum DPP-IV in the prediction of carcinogenesis and progression. Some studies have revealed unfavorable outcomes attributed to a high DPP-IV level, while others have supposed inauspicious consequences resulting from a low DPP-IV level. In colorectal carcinoma, serological DPP-IV activity has been inversely correlated with worse overall and disease-free survivals [22]. On the contrary, patients with metastatic prostate cancer had a reduced serum DPP-IV activity compared to those with localized disease or HCs [35]. In esophageal squamous cell carcinoma, a high serum DPP-IV level represented a survival advantage compared with a low serum DPP-IV level [27]. In this study, serum DPP-IV was positively associated with PTC occurrence, aggressiveness, and prognosis. Especially, serum DPP-IV was able to serve as a biomarker for PTC diagnosis at a cutoff value of 174.70 nkat/L and for prognostic prediction at a cutoff value of 250 nkat/L. Busek et al. found that newly diagnosed diabetic patients with pancreatic ductal adenocarcinoma have higher serum DPP-IV levels than those of patients with diabetes mellitus only [26]. This result was similar with ours, which showed that the serum DPP-IV level was higher in PTC patients than in controls. Agreement was achieved from another study showing that an elevated serum DPP-IV level was associated with poor outcomes in colorectal cancer prognosis [22]. However, different results from ours were found in a study performed by Boccardi et al. [36], which indicated that a decrease in the serum DPP-IV level could serve as an early diagnostic marker for gastric cancers. These discrepancies may be explained by the fact that the investigators studied different types of malignancies with various assessment tools among a heterogeneous population (e.g., gender) [37]. In addition, the presence of diabetes mellitus also affects the serum DPP-IV level [26]. Therefore, diverse proportions of diabetic subjects among studies may partially contribute to the discrepancies. The findings from our study helped to make a timely diagnosis of PTC and to predict the prognosis of such patients. However, a further study with a larger sample size is needed to verify our current data.

Although DPP-IV has been implicated in carcinogenesis, the molecular mechanism is still ambiguous. Several molecular mechanisms have been studied to determine the involvement of DPP-IV in carcinogenesis. Nuclear accumulation of DPP-IV is inhibited by the loss of tumor suppressor P53, which consequently incurs plasma-membrane-associated DPP-IV-dependent lipid peroxidation, resulting in ferroptosis [16]. A high DPP-IV level is associated with the BRAF mutation in PTC, and a positive BRAF mutation is found in approximately 50% of PTC patients [38]. DPP-IV promotes epidermal growth factor-induced epithelial cell transformation and mammary carcinogenesis through induction of PIN1 expression [39]. In addition, in in vitro experiments of malignant mesothelioma cells, DPP-IV has demonstrated its tumor cell invasion ability through its interaction with alpha5beta1 integrin [11]. Further investigation is needed to elucidate the exact mechanism underlying the role of DPP-IV in carcinogenesis in the future.

This study has several strengths. First, to the best of our knowledge, it is the first study to analyze the significance of serum DPP-IV in PTC diagnosis and prognosis. Second, only male subjects were enrolled to avoid the influence of gender. Third, the eligible patients were consecutively enrolled to minimize a selection bias. Our study also has some limitations, including a relatively small sample size and the lack of female participants. Additionally, due to the lack of a standard serum biomarker for PTC, we did not set a control with a known serum PTC biomarker. Finally, considering that this was a single-center study conducted in China, caution must be exercised when generalizing the results of this study in different populations.

In conclusion, serum DPP-IV may be a predictive biomarker for male patients with PTC for both diagnosis and prognosis. Confirmation of these findings will require further investigation in a prospective study with a larger sample size, including female participants. Also, the mechanisms by which DPP-IV is involved in the progression and aggressiveness of PTC need to be elucidated.

Footnotes

Acknowledgments

This study was supported by Special Projects for Health in Jilin Province (2018SCZWSZX-039; 2018 SCZWSZX-050), Projects of the Education Department of Jilin Province (20150125; 2016453; 2016482), and a Project of the Health Department of Jilin Province (20150125).

Conflict of interest

The authors have no conflicts of interest to declare in relation to this article.

References

Lee

J.J.

Wang

T.Y.

Liu

C.L.

Chien

M.N.

Chen

M.J.

Hsu

Y.C.

Leung

C.H.

and Cheng

S.P.

, Dipeptidyl peptidase IV as a prognostic marker and therapeutic target in papillary thyroid carcinoma, J Clin Endocrinol Metab (2017).

Wang

Chen

Zou

and Yang

, Associations between body mass index and lymph node metastases of patients with papillary thyroid cancer: A retrospective study, Medicine (Baltimore) 96 (2017), e6202.

Huang

L.L.

Wang

Cao

C.J.

Z.F.

Wang

Luo

C.Q.

and Wang

L.T.

, AEG-1 associates with metastasis in papillary thyroid cancer through upregulation of MMP2/9, Int J Oncol 51 (2017), 812–822.

Zhong

Liu

Zhang

Q.H.

Chen

Shen

Y.Y.

Chen

Y.J.

Zeng

X.Y.

and Cao

R.X.

, TGF-beta1 induces HMGA1 expression: The role of HMGA1 in thyroid cancer proliferation and invasion, Int J Oncol 50 (2017), 1567–1578.

Xiong

Zhu

and Chen

, Low expression of long noncoding RNA CASC2 indicates a poor prognosis and promotes tumorigenesis in thyroid carcinoma, Biomed Pharmacother 93 (2017), 391–397.

Waligorska-Stachura

Sawicka-Gutaj

Zabel

Andrusiewicz

Gut

Czarnywojtek

and Ruchala

, Survivin DEx3 as a biomarker of thyroid cancers: A study at the mRNA and protein level, Oncol Lett 13 (2017), 2437–2441.

Aliyari Serej

Ebrahimi Kalan

Mehdipour

and Nozad Charoudeh

, Regulation and roles of CD26/DPPIV in hematopoiesis and diseases, Biomed Pharmacother 91 (2017), 88–94.

Rohrborn

Eckel

and Sell

, Shedding of dipeptidyl peptidase 4 is mediated by metalloproteases and up-regulated by hypoxia in human adipocytes and smooth muscle cells, FEBS Lett 588 (2014), 3870–7.

Yang

Zhang

Huang

Jiang

Qin

Yao

Jin

and Zhang

, DPPIV promotes endometrial carcinoma cell proliferation, invasion and tumorigenesis, Oncotarget 8 (2017), 8679–8692.

10.

Tian

Yue

Yan

Guan

Wang

and Hao

, Suppression of CD26 inhibits growth and metastasis of pancreatic cancer, Tumour Biol (2016).

11.

Okamoto

Iwata

Yamazaki

Hatano

Komiya

Dang

N.H.

Ohnuma

and Morimoto

, CD9 negatively regulates CD26 expression and inhibits CD26-mediated enhancement of invasive potential of malignant mesothelioma cells, PLoS One 9 (2014), e86671.

12.

Vliegen

Raju

T.K.

Adriaensen

Lambeir

A.M.

and De Meester

, The expression of proline-specific enzymes in the human lung, Ann Transl Med 5 (2017), 130.

13.

Wang

Grigo

Steinbeck

von Horsten

Amann

and Daniel

, Soluble DPP4 originates in part from bone marrow cells and not from the kidney, Peptides 57 (2014), 109–17.

14.

Lamers

Famulla

Wronkowitz

Hartwig

Lehr

Ouwens

D.M.

Eckardt

Kaufman

J.M.

Ryden

Muller

Hanisch

F.G.

Ruige

Arner

Sell

and Eckel

, Dipeptidyl peptidase 4 is a novel adipokine potentially linking obesity to the metabolic syndrome, Diabetes 60 (2011), 1917–25.

15.

Raschke

Eckardt

Bjorklund Holven

Jensen

and Eckel

, Identification and validation of novel contraction-regulated myokines released from primary human skeletal muscle cells, PLoS One 8 (2013), e62008.

16.

Xie

Zhu

Song

Sun

Fan

Liu

Zhong

Yuan

Zhang

Billiar

T.R.

Lotze

M.T.

Zeh

H.J.

, 3rd Kang

Kroemer

and Tang

, The tumor suppressor p53 limits ferroptosis by blocking DPP4 activity, Cell Rep 20 (2017), 1692–1704.

17.

Zhang

Chen

Zhang

Jia

Shao

Xie

Yang

and Yu

, Incretin-based agents in type 2 diabetic patients at cardiovascular risk: Compare the effect of GLP-1 agonists and DPP-4 inhibitors on cardiovascular and pancreatic outcomes, Cardiovasc Diabetol 16 (2017), 31.

18.

Yazbeck

Jaenisch

S.E.

and Abbott

C.A.

, Potential disease biomarkers: Dipeptidyl peptidase 4 and fibroblast activation protein, Protoplasma (2017).

19.

Liang

P.I.

Yeh

B.W.

W.M.

Chan

T.C.

Chang

I.W.

Huang

C.N.

C.C.

H.L.

Yeh

H.C.

W.J.

and Li

C.F.

, DPP4/CD26 overexpression in urothelial carcinoma confers an independent prognostic impact and correlates with intrinsic biological aggressiveness, Int J Cancer 8 (2017), 2995–3008.

20.

Teichgraber

Monasterio

Chaitanya

Boger

Gordon

Dieterle

Jager

and Bauer

, Specific inhibition of fibroblast activation protein (FAP)-alpha prevents tumor progression in vitro , Adv Med Sci 60 (2015), 264–72.

21.

Beckenkamp

Willig

J.B.

Santana

D.B.

Nascimento

Paccez

J.D.

Zerbini

L.F.

Bruno

A.N.

Pilger

D.A.

Wink

M.R.

and Buffon

, Differential expression and enzymatic activity of DPPIV/CD26 affects migration ability of cervical carcinoma cells, PLoS One 10 (2015), e0134305.

22.

Larrinaga

Perez

Sanz

Beitia

Errarte

Fernandez

Blanco

Etxezarraga

M.C.

Gil

and Lopez

J.I.

, Dipeptidyl-peptidase IV activity is correlated with colorectal cancer prognosis, PLoS One 10 (2015), e0119436.

23.

Larrinaga

Blanco

Errarte

Beitia

Sanz

Perez

Irazusta

Sanchez

C.E.

Santaolalla

Andres

and Lopez

J.I.

, Altered peptidase activities in thyroid neoplasia and hyperplasia, Dis Markers 35 (2013), 825–32.

24.

Kotani

Aratake

Ogata

Umeki

Araki

Hirai

Kuma

and Ohtaki

, Expression of dipeptidyl aminopeptidase IV activity in thyroid carcinoma, Cancer Lett 57 (1991), 203–8.

25.

Maruta

Hashimoto

Yamashita

and Noguchi

, Diagnostic applicability of dipeptidyl aminopeptidase IV activity in cytological samples for differentiating follicular thyroid carcinoma from follicular adenoma, Arch Surg 139 (2004), 83–8.

26.

Busek

Vanickova

Hrabal

Brabec

Fric

Zavoral

Skrha

Kmochova

Laclav

Bunganic

Augustyns

Van Der Veken

and Sedo

, Increased tissue and circulating levels of dipeptidyl peptidase-IV enzymatic activity in patients with pancreatic ductal adenocarcinoma, Pancreatology 16 (2016), 829–38.

27.

Xinhua

Xiangting

Lingling

and Guohong

, Prognostic significance of serum CD26 concentration in patients with esophageal squamous cell carcinoma, Arch Med Res 47 (2016), 299–303.

28.

Tilan

J.U.

Krailo

Barkauskas

D.A.

Galli

Mtaweh

Long

Wang

Hawkins

Jeha

Izycka-Swieszewska

Lawlor

E.R.

Toretsky

J.A.

and Kitlinska

J.B.

, Systemic levels of neuropeptide Y and dipeptidyl peptidase activity in patients with Ewing sarcoma–associations with tumor phenotype and survival, Cancer 121 (2015), 697–707.

29.

Wolke

Teumer

Endlich

Rettig

Stracke

Fiene

Aymanns

Felix

S.B.

Hannemann

and Lendeckel

, Serum protease activity in chronic kidney disease patients: The GANI_MED renal cohort, Exp Biol Med (Maywood) 242 (2017), 554–563.

30.

Ruopp

M.D.

Perkins

N.J.

Whitcomb

B.W.

and Schisterman

E.F.

, Youden Index and optimal cut-point estimated from observations affected by a lower limit of detection, Biom J 50 (2008), 419–30.

31.

Yue

Bai

Wang

Zhao

and Hou

, N-cadherin promotes thyroid tumorigenesis through modulating major signaling pathways, Oncotarget 8 (2017), 8131–8142.

32.

Mazzaferri

E.L.

and Jhiang

S.M.

, Long-term impact of initial surgical and medical therapy on papillary and follicular thyroid cancer, Am J Med 97 (1994), 418–28.

33.

Zhang

Wang

Sun

and Ding

, Expression levels of seprase/FAPalpha and DPPIV/CD26 in epithelial ovarian carcinoma, Oncol Lett 10 (2015), 34–42.

34.

Lam

C.S.

Cheung

A.H.

Wong

S.K.

Wan

T.M.

Chow

A.K.

Cheng

N.S.

Pak

R.C.

H.S.

Man

J.H.

Yau

T.C.

O.S.

Poon

J.T.

Pang

R.W.

and Law

W.L.

, Prognostic significance of CD26 in patients with colorectal cancer, PLoS One 9 (2014), e98582.

35.

Nazarian

Lawlor

S.S.

Philip

Ghosh

Yaneva

Villanueva

Saghatelian

Assel

Vickers

A.J.

Eastham

J.A.

Scher

H.I.

Carver

B.S.

Lilja

and Tempst

, Inhibition of circulating dipeptidyl peptidase 4 activity in patients with metastatic prostate cancer, Mol Cell Proteomics 13 (2014), 3082–96.

36.

Boccardi

Marano

Rossetti

R.R.

Rizzo

M.R.

di Martino

and Paolisso

, Serum CD26 levels in patients with gastric cancer: A novel potential diagnostic marker, BMC Cancer 15 (2015), 703.

37.

Beckenkamp

Davies

Willig

J.B.

and Buffon

, DPPIV/CD26: A tumor suppressor or a marker of malignancy? Tumour Biol 37 (2016), 7059–73.

38.

Shi

R.L.

Liao

Wei

W.J.

Z.W.

Wang

Y.L.

and Ji

Q.H.

, Relationship of body mass index with BRAF (V600E) mutation in papillary thyroid cancer, Tumour Biol 37 (2016), 8383–90.

39.

Choi

H.J.

Kim

J.Y.

Lim

S.C.

Kim

Yun

H.J.

and Choi

H.S.

, Dipeptidyl peptidase 4 promotes epithelial cell transformation and breast tumourigenesis via induction of PIN1 gene expression, Br J Pharmacol 172 (2015), 5096–109.

Predictive significance of serum dipeptidyl peptidase-IV in papillary thyroid carcinoma

Abstract

BACKGROUND:

OBJECTIVE:

METHODS:

RESULTS:

CONCLUSIONS:

Keywords

1. Introduction

2. Materials and methods

2.1 Study population

2.2 Blood samples and determination of serum DPP-IV concentration

Table 1 Clinicopathological characteristics of the male subjects, n (%)

3. Results

3.1 Patient characteristics

3.2 Comparison of serum DPP-IV levels among groups

3.3 Correlation between the serum DPP-IV level and clinicopathological characteristics

3.5 Clinicopathological features of patients with PTC by serum DPP-IV stratification

Table 4 Clinicopathological features of patients with PTC by serum DPP-IV stratification at 174.70 nkat/L, n (%)

Table 5 Clinicopathological features of patients with SPRD or NSPRD, n (%)

4. Discussion

Footnotes

Acknowledgments

Conflict of interest

References

Table 1
Clinicopathological characteristics of the male subjects, $n$ (%)

Table 4
Clinicopathological features of patients with PTC by serum DPP-IV stratification at 174.70 nkat/L, $n$ (%)

Table 5
Clinicopathological features of patients with SPRD or NSPRD, $n$ (%)