The relationship between multiple clinicopathological features and nerve invasion in pancreatic cancer

Shanghai, China

The relationship between multiple clinicopathological features and nerve invasion in pancreatic cancer

Pao-Hsun Wang, Ning Song, Liu-Bin Shi, Qun-Hua Zhang and Zong-You Chen

Shanghai, China

BACKGROUND:Nerve invasion is a specif i c type of tumor expansion and characteristic manifestation of pancreatic cancer (PC), with an incidence rate ranging from 50% to 100%. It is an important prognostic factor for pancreatic cancer, and its early detection is helpful in the management of the disease. This study was undertaken to analyze retrospectively the relationship between neural invasion and multiple clinicopathological features and to provide evidences for clinicians in the management of neural invasion in patients with PC.

METHODS:Formalin-f i xed paraff i n-embeded specimens of PC taken from 215 patients were examined for the presence of neural invasion under a light microscope. Analyzed was the relationship between neural invasion and multiple clinicopathological feature including preoperative fasting blood glucose level, amylase level, serum CA19-9 level, abdominal pain, lumbar and back pain, and the expressions of p53 and Ki67 in tumor tissues.

RESULTS:Preoperative fasting blood glucose level, serum CA19-9 level and p53 positive cells in cancer tissue were increased with the rise of pathological grade (P<0.05). These indices were signif i cantly higher in patients with neural invasion than in those without (P<0.05). Further analysis revealed a positive correlation between p53 and Ki67 overexpression and lymphatic metastasis (P<0.05). Referred pain was positively correlated with neural invasion (P<0.05). Patients with PC perineural invasion were more likely to have a higher pathological grade (P<0.05).

CONCLUSIONS:Our data indicated that the preoperative fasting blood glucose level, serum CA19-9 level, and referred pain are novel predictive markers for neural invasion in patientswith PC. p53 and Ki67 play important roles in neural invasion of PC. Management of hyperglycemia may serve as an auxiliary treatment to curb neural invasion in PC.

(Hepatobiliary Pancreat Dis Int 2013;12:546-551)

pancreatic cancer;neural invasion; hyperglycemia; referred pain

Introduction

Pancreatic cancer (PC) is a malignant tumor with extremely poor prognosis.[1]It is the fourth most common cause of cancer-related death in both genders in the United States.[2]Neural invasion represents a way of tumor metastasis and is a characteristic manifestation of PC, with an incidence rate ranging from 50% to 100%. Moreover, it is an important prognostic factor for PC.[3]

The mechanism of neural invasion in adenocarcinoma is unclear. But PC is characterized by abdominal pain and neuropathy. Nerve inf i ltration inside or outside of the pancreas, which can be observed in 71%-98% of PC specimens, is a typical sign of neuropathy. To some extent, however, pain is due to the neuropathy in PC. Recent report[4]has shown that 34%-40% of PC patients are complicated with diabetes. Hyperglycemia has different inf l uence on nerve invasion in PC patients.

With a sensitivity of 77.4%, CA19-9 has been used as the most effective early diagnostic marker of PC.[5]Both p53 and Ki67 are reliable biomarkers of PC. The overexpression of mutant p53 can be observed in 50%-60% PC patients, indicating that the overexpression of mutant p53 is a common genetic abnormality in PC and is closely associated with the development of PC. p53 is useful in the diagnosis of PC. Ki67 is also highly expressed in PC tissues,indicating that Ki67 is related to carcinogenesis and the progression of PC. Fan and coworkers[6]reported that the level of Ki67 expression ref l ects the degree of tumor cell proliferation.

Neural invasion affects not only the prognosis of PC patients but also their life quality.[7]Therefore, early detection of neural invasion is helpful in the management of PC. The relationship is not clear between pathologic features and clinical manifestations in patients with PC such as pain, referred pain, serum CA19-9 level, serum amylase, p53 and Ki67 expression, etc. Little information is available about whether these clinical features are associated with neural invasion in patients with PC. Researchers proposed some mechanisms in neural invasion in PC patients, but there is no def i nite conclusion because of limited number of studies and sample sizes. In this retrospective study, data were used to assess the relationship between neural invasion and multiple clinicopathological features.

Methods

Tissue samples and data collection

A total of 215 patients who had been admitted to Huashan Hospital, Fudan University from January 2005 to December 2010 were histologically diagnosed as PC. The cancer involved the pancreatic head in 124 patients (Whipple or total pancreatectomy) and the body and tail in 91 patients (left pancreatectomy). Histologically, neural invasion was shown in 118 patients (extrapancreatic in 15 patients and 103 intrapancreatic). The other 97 patients had no evidence of neural invasion. In this series, 138 patients were male and 77 female; their age ranged from 48 to 74 years (mean 63). Fasting blood glucose level, amylase level, and serum CA19-9 level were collected before operation. Symptoms of abdominal pain, referred lumbar and back pain were reviewed. The patients with pain scaled less than 4 were def i ned as having negative results and those with pain greater than or equal to 4 were considered to have positive results.

Histological variables including cancer stage, grade, nerve invasion, lymph node metastasis, and adjacent tissue inf i ltration were determined by two independent pathologists via the examination of HE stained sections.

Immunohistochemistry

Formalin-f i xed paraff i n-embedded specimens were cut into 5 μm sections. The expressions of p53 and Ki67 protein were detected by mouse anti-human p53 monoclonal antibody and rabbit anti-human Ki67 monoclonal antibody (1:50 dilution), respectively (Shanghai Changdao Biotechnology, Shanghai, China). Colorectal cancer specimens were used as positive controls. Tris-buffered saline (TBS) was used in replacement of primary antibody as the negative control. Immunohistochemistry was performed according to the protocol recommended by the manufacturer. The sections were fi rst de-waxed and rehydrated by dimethylbenzene and alcohol gradient. After antigen retrieval, the sections were incubated with 0.3% hydrogen peroxide for 5 minutes. After a PBS wash (three times, two minutes each), the sections were incubated with primary antibodies for 60 minutes at room temperature. Then they were washed again with PBS (same wash as above). One drop of the ready-to-use supervision HRP-conjugated anti-mouse or anti-rabbit secondary antibody (Shanghai Changdao Biotechnology, Shanghai, China) was added to each sample, followed by 30-minute incubation at room temperature. After PBS washes, the sections were incubated with 3, 3'-diaminobenzidine tetrahydrochloride (DAB) for 5 to 10 minutes. Finally, the sections were dehydrated and sealed after counterstaining with haematoxylin.

Evaluation

Pathological stage was determined according to the TNM PC staging (UICC/AJCC, 2002).[1]Pathological grade was divided into well, moderate, or poor differentiation groups (grades 1, 2 or 3).[7]

p53 positive staining was brown in the nucleus. The degree of positive staining and the number of positive cells were taken into account. Five high power fi elds (HPF) of positive areas (×400) were chosen to count the number of positive cells. The proportion of immunopositive cells was categorized as follows: 0 (positive cells <10%), 1 (positive cells 10%-20%, pale yellow), 2 (positive cells 26%-50%, brown), and 3 (positive cells >50%, dark brown).

Ki67 positive staining was pale yellow or brown also in the nucleus. Proliferation index (PI) was calculated as the number of positive cells per 1000 cells in 15 random vision fi elds (positive cells/1000×100%). Five HPF positive areas (×400) were chosen and the number of positive cells was counted. PI scores were graded as follows: 0 (positive cells <10%), 1 (positive cells 10%-20%, pale yellow), 2 (positive cells 26%-50%, brown); and 3 (positive cells >50%, dark brown).

Statistical analysis

The data were analyzed using the SPSS 17.0 software package. Student'sttest was adopted for continuous variables, and the Kruskal-Wallis test was used to estimate the difference among the groups above three. Correlation coeff i cients were calculated using Fisher'sexact test and the Chi-square test. APvalue less than 0.05 was considered statistically signif i cant.

Fig. 1.PC (HE, original magnif i cation ×400).A: without neural invasion;B: with neural invasion.

Fig. 2.Expression of p53 (HE, original magnification ×400).A: negative;B: positive.

Fig. 3.Expression of Ki67 (HE, original magnif i cation ×400).A: negative;B: positive.

Results

PC without neural invasion and metastasis of tumor cells to the nerve-tract, even to the whole nerve fi ber are shown in Fig. 1.

p53 positive cells in the nucleus were stained brown by immunohistochemistry. Negative and positive expressions are shown in Fig. 2. The positive expression rate of p53 was 51.63% (111/215, Table 1).

Ki67 positive cells in the nucleus were stained pale yellow or brown by immunohistochemistry. Negative and positive expressions are shown in Fig. 3. The positive expression rate of Ki67 was 71.16% (153/215, Table 1).

The proportion of neural invasion in PC was notincreased with the elevation of pathological stage. However, there was a signif i cant relationship between neural invasion and pathological grade, but no relationship between neural invasion and lymphatic metastasis (Table 2). Preoperative fasting blood glucose level was signif i cantly higher in patients with neural invasion compared with those without. Moreover, preoperative fasting blood glucose level was correlated with the pathological grade of cancer, but not with pathological stage or lymphatic metastasis. Serum amylase level before operation was not increased with tumor stage, pathological grade, neural invasion or lymphatic metastasis. Preoperative serum CA19-9 level was signif i cantly correlated with the pathological grade of PC and neural invasion, but not with pathological stage or lymphatic metastasis. Interestingly, p53 expression was signif i cantly correlated with pathological stage, lymphatic metastasis and neural invasion, but not with pathological grade. Ki67 expression was signif i cantly higher in patients without neural invasion. There was no correlation between Ki67 and pathological stage, pathological grade, or lymph node metastasis. There was no signif i cant correlation between abdominal pain and neural invasion in PC patients. The proportion of referred pain in patients with neural invasion was signif i cantly higher than that in patients without neural invasion (Table 3).

Table 1.Expression of p53 and Ki67 (n=215)

Table 2.The relationship between neural invasion of PC (PNI), pathological stage, pathological grade, and lymphatic metastasis

Discussion

PC is characterized by a high incidence of neural invasion.[1]One of the most important biologicalcharacteristics of PC, especially in pancreatic ductal adenocarcinoma, is neurotropism.[8]Anatomically, there are multiple nerves in the pancreas, including sympathetic, parasympathetic and sensory nerves. These nerves all penetrate into the pancreas and participate in the normal activities of the pancreas. Our study showed that tumor cells can invade into the axis fi bers of the nerves.

Table 3.The relationship between multiple clinicopathological features and neural invasion in PC (mean±standard error, SEM)

It has been shown that neural invasion in PC is not correlated with tumor size, location or histology.[7]In our study, neural invasion in PC was found to be signif i cantly correlated with pathological grade, but not with tumor stage. The results suggest that neural invasion in PC is not limited by the pancreatic capsulae. Instead, the malignancy of PC cells plays an important role in neural invasion. Patients with lymphatic metastasis have a high incidence of peripancreatic neural invasion, whereas lymphatic metastasis does not initiate peripancreatic neural invasion. Therefore, neural invasion and lymphatic metastasis may represent the metastasis of two different types of tumor cells, with neural invasion more related to the capability of tumor cells to inf i ltrate nearby tissues.

Preoperative high blood glucose level after fasting is a common clinical manifestation of PC. However, little is known about the relationship between neural invasion and preoperative fasting glucose level in PC. A recent study showed that hyperglycemia may be related to neural invasion.[9]Our data showed that preoperative fasting blood glucose level is signif i cantly higher in patients with neural invasion than those without. The correlation among neural invasion, pathological grade (malignancy) and hyperglycemia indicated that hyperglycemia and PC neural invasion may be a vicious cycle. Moreover, hyperglycemia may be associated with the malignant nature of pancreatic tumor cells. Preoperative fasting glucose level is not correlated with tumor stage or lymphatic metastasis, indicating that hyperglycemia is not correlated with the tumor size or the inf i ltration of pancreas capsulae. The incidences of tumor recurrence, metastasis, diffusion, and critical conditions were higher in PC patients with hyperglycemia or diabetes than in PC patients without metabolic diseases.[10]Three factors affecting human PC microenvironment were neural invasion, hyperglycemia, and tumor cells. Two mechanisms have been proposed to address the relationship between neural invasion and hyperglycemia.[11]First, hyperglycemia enhances the proliferation of PC cells and the overexpression of certain cytokines, such as nerve growth factor, which sequentially stimulates the interaction between tumor cells and nerves. Second, hyperglycemia leads to axonal degeneration and nerve demyelination, which will facilitate the inf i ltration of pancreatic tumor cells into deep nerves. Hyperglycemia is one of the important factors affecting the recovery of patients with PC.[12]Statistical analysis showed that the preoperative fastingblood glucose level could be an indicator of neural invasion in PC, suggesting that the management of hyperglycemia may help to reduce neural invasion.

PC is characterized by abdominal pain and neuropathy.[13]Glial cells are likely to play an important role in neuropathic pain of PC.[14]Activation of glial cells is thought to be a pathogenic factor for neuropathic pain in PC.[15]Tumor cells which are not conf i ned to the peripheral nerves can penetrate into the perineurium to form close ties with Schwann cells of axon inside of the endoneurium. Ceyhan et al[16]found that PC has neurotropic features and can induce neural plasticity and morphological changes, resulting in a specif i c pancreatic neuropathy. These fi ndings suggest that PC-related pain may be due to the stimulation of the neural network by intrapancreatical neuropathy.[14]When the nerve is injured, Schwann cells and glial cells of the peripheral nervous system are activated and begin to proliferate and dedifferentiate. Sequentially, the increasing activity of Schwann cells will increase the expression of GAP-43. And overexpression of GAP-43 in PC is strongly associated with neuropathic changes and pain.[14]In the present study, the proportion of referred pain in patients with nerve invasion was signif i cantly higher than that in the control group, suggesting that referred pain may be another indicator for neural invasion of PC.

Preoperative serum CA19-9 level is the most effective diagnostic parameter of PC.[16,17]In our study, the preoperative serum level of CA19-9 in patients with neural invasion was higher than that in the control group. Moreover, the level of CA19-9 was increased with the increase of pathological grade, suggesting that preoperative serum CA19-9 level may serve as a biomarker for neural invasion in PC patients. In addition, CA19-9 level is also associated with the poor prognosis in PC patients with neural inf i ltration and the malignancy of tumor cells. Different results have been found in detecting the relationship between preoperative serum CA19-9 level and PC stage.[18]Our study showed that the preoperative serum CA19-9 level in patients with PC is not related to pathological stage or lymph node metastasis. Moreover, the preoperative serum CA19-9 level may be an indicator of neural invasion in PC patients. Thus, screening for preoperative serum CA19-9 level may contribute to the early detection of neural invasion in PC patients.

p53 is an important tumor suppressor. Conformation changes due to gene mutation lead to the loss of its regulatory functions in DNA repair, cell growth and apoptosis, and may transform p53 into an oncogene. p53 expression is signif i cantly different in PC tissues from that in non-pancreatic tumor tissue.[19]In our study, the positive expression rate of p53 was higher in patients with pathological stage III or IV than that in those with stage I or II. However, it was not related to pathological grade as reported previously.[20]We also found the association between p53 and neural invasion in PC. Ki67 as a biomarker for cell proliferation is closely associated with cell mitosis or a poor prognosis of PC.[21]In our study, the positive expression rate of Ki67 was signif i cantly higher in patients without neural invasion. The correlation between neural invasion and the expression of p53 and Ki67 in PC suggested that both p53 and Ki67 may be important factors for neural invasion in PC.

In conclusion, preoperative fasting blood glucose level, CA19-9 level and referred pain are novel biomarkers for neural invasion in PC. p53 and Ki67 may play important roles in neural invasion in PC. The management of hyperglycemia could be an auxiliary treatment to control neural invasion in patients with PC.

Contributors:SLB, ZQH and CZY proposed the study. WPH and SN performed research and prepared the manuscript under the supervision of SLB and ZQH. WPH and SN contributed equally to this article. All authors contributed to the study design and data analysis and revision of manuscript. SLB is the guarantor.

Funding:None.

Ethical approval:Not needed.

Competing interest:No benef i ts in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article.

1 Zhang QH, Ni QX; Coordination Group of The Committee on Pancreatic Cancer. Clinical analysis of 2340 cases of pancreatic cancer. Zhonghua Yi Xue Za Zhi 2004;84:214-218.

2 Siegel R, Naishadham D, Jemal A. Cancer statistics, 2012. CA Cancer J Clin 2012;62:10-29.

3 Yi SQ, Miwa K, Ohta T, Kayahara M, Kitagawa H, Tanaka A, et al. Innervation of the pancreas from the perspective of perineural invasion of pancreatic cancer. Pancreas 2003;27: 225-229.

4 Chari ST, Leibson CL, Rabe KG, Timmons LJ, Ransom J, de Andrade M, et al. Pancreatic cancer-associated diabetes mellitus: prevalence and temporal association with diagnosis of cancer. Gastroenterology 2008;134:95-101.

5 M?ssner J. What's new in therapy of pancreatic cancer? Dig Dis 2010;28:679-683.

6 Fan J, Tang F, Li Q, Zhou ZW, Xu ZD. The relationship between regional lymph node metastases and expression of p53 and ki67 in pancreatic carcinoma and its prognostic signif i cance. Chin J Clin Med 2007;14:825-828.

7 Ni QX, Zhang QH, Fu DL, Cao GH, Yao QY, Jin C, et al. Curative resection of pancreatic head carcinoma in recent 30 years: report of 377 cases. Hepatobiliary Pancreat Dis Int2002;1:126-128.

8 Takahashi H, Ohigashi H, Ishikawa O, Gotoh K, Yamada T, Nagata S, et al. Perineural invasion and lymph node involvement as indicators of surgical outcome and pattern of recurrence in the setting of preoperative gemcitabine-based chemoradiation therapy for resectable pancreatic cancer. Ann Surg 2012;255:95-102.

9 Li J, Ma Q. Hyperglycemia promotes the perineural invasion in pancreatic cancer. Med Hypotheses 2008;71:386-389.

10 Sahin IH, Shama MA, Tanaka M, Abbruzzese JL, Curley SA, Hassan M, et al. Association of diabetes and perineural invasion in pancreatic cancer. Cancer Med 2012;1:357-362.

11 Li J, Ma Q, Liu H, Guo K, Li F, Li W, et al. Relationship between neural alteration and perineural invasion in pancreatic cancer patients with hyperglycemia. PLoS One 2011;6:e17385.

12 Tezel E, Kaneko T, Sugimoto H, Takeda S, Inoue S, Nagasaka T, et al. Clinical signif i cance of intraportal endovascular ultrasonography for the diagnosis of extrapancreatic nerve plexus invasion by pancreatic carcinoma. Pancreatology 2004;4:76-81.

13 Bapat AA, Hostetter G, Von Hoff DD, Han H. Perineural invasion and associated pain in pancreatic cancer. Nat Rev Cancer 2011;11:695-707.

14 Ceyhan GO, Bergmann F, Kadihasanoglu M, Altintas B, Demir IE, Hinz U, et al. Pancreatic neuropathy and neuropathic pain--a comprehensive pathomorphological study of 546 cases. Gastroenterology 2009;136:177-186.

15 Kurtz KA, Hoffman HT, Zimmerman MB, Robinson RA. Perineural and vascular invasion in oral cavity squamous carcinoma: increased incidence on re-review of slides and by using immunohistochemical enhancement. Arch Pathol Lab Med 2005;129:354-359.

16 Ceyhan GO, Michalski CW, Demir IE, Müller MW, Friess H. Pancreatic pain. Best Pract Res Clin Gastroenterol 2008;22:31-44.

17 Kaneko T, Nakao A, Inoue S, Nomoto S, Nagasaka T, Nakashima N, et al. Extrapancreatic nerve plexus invasion by carcinoma of the head of the pancreas. Diagnosis with intraportal endovascular ultrasonography. Int J Pancreatol 1996;19:1-7.

18 Liebig C, Ayala G, Wilks JA, Berger DH, Albo D. Perineural invasion in cancer: a review of the literature. Cancer 2009; 115:3379-3391.

19 Hirai I, Kimura W, Ozawa K, Kudo S, Suto K, Kuzu H, et al. Perineural invasion in pancreatic cancer. Pancreas 2002;24: 15-25.

20 Ozawa F, Friess H, Künzli B, Shrikhande SV, Otani T, Makuuchi M, et al. Treatment of pancreatic cancer: the role of surgery. Dig Dis 2001;19:47-56.

21 Watson JJ, Allen SJ, Dawbarn D. Targeting nerve growth factor in pain: what is the therapeutic potential? BioDrugs 2008;22:349-359.

Received May 28, 2013

Accepted after revision August 3, 2013

Liu-Bin Shi, MD, PhD, Department of General Surgery, Huashan Hospital, Fudan University, 12 Mid-Urumchi Road, Shanghai 200040, China (Tel: 86-21-52888015; Email: shiliubin@medmail. com.cn)

10.1016/S1499-3872(13)60086-7

AuthorAff i liations:Department of General Surgery, Huashan Hospital, Fudan University, Shanghai 200040, China (Wang PH, Song N, Shi LB, Zhang QH and Chen ZY)

? 2013, Hepatobiliary Pancreat Dis Int. All rights reserved.