Hepatic myelopathy Morphology of the thoracic and lumbar cord in liver cirrhosis and non-cirrhosis corpses and comparison of neuron functional protein1＊

Yu Lei, Zhen Liu, Huilong Huang, Suiliang Zhang, Yunheng Zhou, Shuai Wu, Xiaojun Hou,Jie Gong, Aiqun Wu

1Department of Clinical Medicine, the Second Military Medical University of Chinese PLA, Shanghai 200433, China

2Department of Anatomy, the Second Military Medical University of Chinese PLA, Shanghai 200433, China

3Department of Interventional Therapy, Eastern Hepatobiliary Surgery Hospital, the Second Military Medical University of Chinese PLA,Shanghai 200433, China

4Clinical Laboratory, Shanghai Armed Police Corps Hospital, Shanghai 201103, China

5Department of Neurology, Changhai Hospital, the Second Military Medical University of Chinese PLA, Shanghai 200433, China

lNTRODUCTlON

Hepatic myelopathy (HM) is a kind of neurological complication that occurs at advanced stages of liver disease due to portosystemic shunt and liver dysfunction.Its clinical manifestations include progressive symmetric spastic limb paralysis, young age at onset, and poor prognosis[1-2].Since HM was first reported by Leihet al[3]in 1949,many scholars have found that corticospinal tract degeneration occurs in the spinal cord lateral funiculus at the thoracic and lumbar nucleus in HM patients.However, a majority of these studies were case reports and explored the pathogenesis of disease.In addition, lack of a successful model resulted in inconclusive results[4].Therefore, the mechanism underlying pathogenesis of HM remains unclear.

The pathogenesis of HM is believed to be similar to hepatic encephalopathy (HE),which mainly occurs from high blood ammonia and malnutrition.The inability of the liver to detoxify and remove toxic substances, such as ammonia,viathe portosystemic shunt following liver cirrhosis allows ammonia to circulate in the blood and damage the brain and spinal cord[3,5].However, another study has shown that HM blood ammonia increases at a mild to moderate level, and that blood ammonia levels do not parallel with spinal cord injury.In addition, spinal cord injury has not improved despite the regulation of blood ammonia levels and better nutrition.These findings indicate that a portosystemic shunt may be the main cause of HM, rather than superfluous blood ammonia[6-10], which negates the need for surgical operation.This study aimed to compare the morphology of the thoracic and lumbar cord, the expression of functional proteins and changes in vessels between liver cirrhosis and non-cirrhosis corpses, in a broad attempt to elucidate the pathogenesis and clinical treatment of HM.

RESULTS

Structure of the abdominal cavity and spinal venous plexus, and liver tissue morphology in the cirrhosis group

The cirrhosis group was subdivided into severe and moderate subgroups according to the varicose severity of the spinal venous plexus.The abdominal gross anatomy showed that the hepatic portal veins expanded in both the cirrhosis subgroups, the average diameter was larger than that of the control group (2.52 cm ＞ 2.38 cm ＞ 1.70 cm; severe ＞ moderate ＞ control), and the gastrointestinal tract was full of coagulated blood (Figure 1).The lumbar veins thickened, the venous plexus was tortuous, and the ratio of vein/artery on the spinal cord surface increased in the cirrhosis group when compared with the control group (3: 1versus2: 1, respectively; Figures 2, 3).

Figure 1 Gross anatomy of the alimentary tract hemorrhage in patients with liver cirrhosis.Arrows represent the node-like changes on the liver surface,stomach and enteric cavity hematocele.

Figure 2 Histological changes of the liver in patients with liver cirrhosis (hematoxylin-eosin staining, × 20).HC:Hepatic cord; SC: sinusoidal congestion; CV: central vein;V: vasoformation.

Figure 3 Gross anatomy of the vertebral canal venous plexus, spinal dura mater and spinal cord surface veins after the vertebral canal and spinal dura mater were cut.a: Artery; v: vein.

Sinusoidal congestion was visible in both cirrhosis subgroups and was mainly located in the hepatic lobule of the severe subgroup.In addition, cells were slightly swollen, while false lobules formed in the moderate subgroup, liver cells were vacuole in the heavily congested areas, and intracellular structures were lost.

Morphological changes in spinal cord anterior horn neuronal cells and Nissl bodies in the liver cirrhosis and control group

Nissl staining showed that the average number of thoracic cord anterior horn motor neurons was 6-8, 6-7, 5-7 in the control group, liver cirrhosis moderate and severe subgroups, respectively, while average numbers in the lumbar cord were 12-19, 9-16, 8-14, respectively.The number of anterior horn neurons reduced in the lower thoracic and lumbar spinal cord segments of the liver cirrhosis groups compared with the control group, but the difference was not statistically significant (P＞ 0.05).The cell body area in the lower thoracic cord was 474.62 ±12.23, 456.24±18.17 and 375.71±16.24 in the moderate,server and control groups, respectively, while in the lumbar spinal cord, cell body area was 938.67±28.35, 667.69 ±23.15 and this 509.37±21.32, respectively.Cell bodies in the liver cirrhosis groups were significantly smaller than those in the control group (P＜0.05).Some neurons were degranulated and were of uneven size, and Nissl bodies were concentrated towards the center; the majority of nuclei were unclear or deviated, with obscure nucleoli;neurons appeared shrunken (Figure 4) and lumen widening or expansion was visible in white matter.

Figure 4 Anterior horn of thoracic and lumbar spinal cord segments in the liver cirrhosis group and control group(Nissl staining, × 40).The thoracic segment of moderate and severe cirrhosis groups (B-C) compared with the control group (A).The lumbar segment of the moderate and severe cirrhosis groups (E-F) compared with the control group (D).N: Normal neurons; P: shrunken neurons

Synapsin ll (Syn ll) and neurofilament (NF) levels in spinal cord anterior horn neurons in the liver cirrhosis group

Syn II and NF contribute to the function and structure of neurons and their constituent fibers.Immunofluorescence showed that the Syn II immunoreactive area in neurons was reduced, and that Syn II and NF immunoreactive intensity weakened in the liver cirrhosis group compared to the control group (Figures 5, 6).This result suggested that Syn II and NF levels decreased in the cirrhosis group.Differences were significant between the cirrhosis and the control group (P＜0.05 orP＜0.001).

Figure 5 Changes in synapsin and neurofilament protein content in anterior horn neurons of the spinal thoracic(A-C) and lumbar segments (D-F) (immunohistochemical staining, × 40).N: Normal neurons; P: shrunken neurons;1: green fluorescence of synapsin II; 2: red fluorescence of neurofilament; 3: yellow fluorescence of synapsin II and neurofilament.(A, D) Control group; (B, E) moderate cirrhosis group; (C, F) severe cirrhosis group

Figure 6 Synapsin II and neurofilament of the lateral funiculi in the lumbar lateral spinal cord in control (A),moderate cirrhosis (B), severe cirrhosis (C) groups(immunohistochemistry staining, × 40).N: Normal nerve fibers; P: degenerative nerve fibers.1: green fluorescence of synapsin II; 2: red fluorescence of neurofilament; 3:yellow fluorescence of synapsin II and neurofilament.

Compared with the control group, the cellular area, the Syn II-positive area and the intensity of staining all reduced following an increase in varicose veins in the liver cirrhosis group (P＜0.05).Staining intensity of the lumbar spinal cord significantly decreased compared with the thoracic cord, and the reduction of Syn II-positive staining intensity was greater than the Syn II-positive area.NF and Syn II immunoreactive intensity decreased along with the severity of intraspinal varicose veins in the moderate cirrhosis subgroup (P＜0.05).The staining intensity of the lumbar spinal cord decreased significantly compared with the thoracic cord, and the reduction in lumbar spinal NF-positive staining was more significant than that of Syn II (Tables 1, 2, Figure 6).

Table 1 Comparison of the total cellular area (cm2), synapsin II immunoreactive area (cm2) and absorbance value of synapsin II and neurofilament in thoracic cord cells between the liver cirrhosis group and control group

Table 2 Comparison of the total cellular area (cm2), synapsin II immunoreactive area (cm2) and absorbance value of synapsin II and neurofilament in lumbar cord cells between the liver cirrhosis group and control group

Compared with the control group, Syn II-positive staining intensity in lumbar spinal cell bodies declined similarly to Syn II staining in lumbar spinal lateral cord fibers(P＜0.05 orP＜0.001).Cell body Syn II-positive staining intensity decreased slowly, and NF-positive staining intensity in both the cell body and lateral cord fibers was weakened (Table 3).

Table 3 Comparison of synapsin II- and neurofilamentpositive immunofluorescence absorbance values in lateral lumbar spinal cord fibers

DlSCUSSlON

An autopsy report has found that pathological changes of HM greatly vary with that of HE, which affects the cerebral hemisphere and striatum resulting in neuronal loss.In contrast, HM gives rise to thoracolumbar corticospinal lateral tract demyelination and axonal degeneration,rarely affecting the upper cervical cord and brainstem.These differences indicate that HE and HM have different mechanisms of pathogenesis[9].Giangasperoet al[10]proposed that a portosystemic shunt evokes hemodynamic changes in the spinal cord and triggers congestion and hypoxia, which may be the pathological basis of HM.The results of the present study showed that the congestion elicited damages the function and structure of spinal cord anterior horn motor neurons and lateral funiculi fibers in patients with portal hypertension, and that the severity of the congestion determines the extent of spinal cord injury.

The experimental results also indicate that portal hypertension caused more severe damage to lumbar spinal anterior horn motor neurons than thoracic cord neurons.Moreover, damage to lumbar spinal neuron function occurred earlier than structural damage.

Therefore, we can clearly determine that the cirrhosis-portal hypertension-portosystemic shunt caused by hemodynamic changesviathe intraspinal venous plexus is the initiating factor for spinal cord injury.In contrast to a previous report that states HM only occurs in bilateral corticospinal tract injury, this study indicates that thoracic-lumbar spinal anterior horn motor neurons are also damaged, which is consistent with a recent animal experiment[8].The Syn II and NF content decreased in the lateral corticospinal tract and spinal cord anterior horn neurons, thus weakening the dominant function between upper and lower motor neurons, and between lower motor neurons and their dominant muscle.This is the main risk factor for early lower limb weakness and HM attack in liver sclerosis patients.

The diversity of clinical pathological reports on the spinal cord in HM patients may be associated to individual factors, nutrition and treatment procedures, such as poor blood supply in the thoracic and lumbar cord between individuals, and susceptibility differences in response to noxious stimuli in the spinal cord lateral corticospinal tract and anterior horn neurons in the same individual[11].Portal hemodynamic change is a factor for secondary liver injury.In this study, sinusoid congestion was found in both moderate and severe cirrhosis subgroups.The entire hepatic lobule and cord were not damaged by moderate cirrhosis, but central vein stenosis occurred and cells were slightly swollen; severe cirrhosis was widely visible in false lobules, and the hepatic cells lost intracellular structure in the heavy congestion areas.These findings indicated that liver inflammation occurred to varying degrees after liver cirrhosis and gastrointestinal bleeding.This may be due to the hypoxia caused by liver congestion (before bleeding) or ischemia (after bleeding).We speculate that the inflammatory response may accelerate the development of cirrhosis, promote opening of portal collateral circulation and the shunt, thus aggravating spinal cord hypoxia-induced injury.Sustained aggravation of portal hypertension and repeatable attacks of gastrointestinal congestion can not only induce ischemic hepatitis, but also damage the gastrointestinal mucosa, cause the absorption of toxins by the gastrointestinal tract and evoke toxic hepatitis.Therefore, viral hepatitis develops to hypoxic, toxic hepatitis repeatedly and progressively, further accelerating cirrhosis onset[12].Portal hemodynamic changes caused by liver cirrhosis aggravate the severity of cirrhosis, and promote a vicious cycle of secondary liver injury.

Relieving portal hypertension is a necessary measure to prevent HM.HM is evoked by chronic hepatitis-caused secondary portal hypertension and shunt following liver cirrhosis, and can be induced by gastrointestinal bleeding or is concurrent with HE.Because the pathogenesis of HM remains unclear, prevention focuses on protection of the liver and decreasing levels of ammonia, and surgical treatment of portal hypertension is conservative[13-15].However,the mortality of patients with esophageal-gastric variceal bleeding is as high as 30-50% under conservative treatment;the possibility of re-bleeding within 2 years is 70-80%, and survival rates after secondary surgery are poor[16].In this study, damage of spinal cord anterior horn neurons and lateral cord lesions were already present in patients with cirrhosis and bleeding, even if no HM attack was found prior to death.Based on the characteristics of digestive tract and spinal cord blood supply, we conclude that the presence of portal hypertension in the process of liver cirrhosis-portal hypertension-portal shunt is a hallmark indicator for the emergence of congestion in the digestive tract, spleen and liver.A portal shunt formation, such as esophageal varices, indicates that the spinal cord and other organs have blood circulation abnormalities and congestion.If portal hypertension was not resolved at this stage, it can affect the function of organs that may be compensated or decompensated.In this study, we found that spinal cord lesions existed and the HM seizure are not typically shown in HM cases, but that organ congestion and dysfunction symptoms such as abdominal distension, anorexia (gastrointestinal stasis), and heavy lower limb weakness (spinal cord dysfunction) are present[17].These findings indicate that hepatic spinal cord injury occurs as a result of portal hypertension and portal congestion-caused hypoxia following spinal cord injury,with atypical symptoms occurring first followed by the onset of typical symptoms.

HM prevention should focus on early resolution of portal hypertension and portal congestion/hypoxia.Clinical symptoms such as abdominal distension and signs of portal hypertension, should serve as indications for surgical treatment.We advocate a combined therapy for medically conservative treatment of chronic hepatitis and cirrhosis, a close follow-up visit and early surgical treatment of portal hypertension and congestion.Surgical treatment is rapidly developing and has achieved good results, but there are still issues of recurrent bleeding as a result of incomplete variceal ligation or sub-flow shortage[13-14], or excessive shunt induced HE and HM attacks[5,18].Although liver transplantation has resulted in impressive results[19-20], it is still limited due to the shortage of organs available for transplantation.Therefore, early shunt therapy of portal hypertension can relieve secondary injury and severe complications in the liver, spleen, and digestive tract caused by congestion and hypoxia, in a broader attempt to prevent HM.

MATERlALS AND METHODS

Design

An anatomical observation and non-randomized controlled trial.

Time and setting

The experiment was performed from May 2010 to May 2011 in the Laboratory, Department of Anatomy, the Second Military Medical University of Chinese PLA, China.

Materials

Six corpses were obtained as a voluntary donation, from the Chinese Red Cross, Department of Anatomy, the Second Military Medical University of Chinese PLA,China (the donation was approved by the Ethics Committee of Shanghai, China, for medical research).All specimens were fixed with 10% (v/v) formaldehyde.There were three cirrhosis cases and three non-cirrhosis cases.Two male cases and one female case of cirrhosis,aged 55-65 years, died of gastrointestinal bleeding shock or hepatic coma, with no history of a HM episode;another two male cases and one female case of non-cirrhosis, aged 59-75 years, died of acute myocardial infarction, with no record of neurological diseases.

Methods

Gross anatomy and autopsy of the control and cirrhosis groupThe abdomen and spine were respectively anatomized,to expose the abdomen liver, stomach, intestine, portal vein, spinal canal, spinal cord, and its blood vessels.All tissues were calibrated with a caliper.

Histological detection of thoracic/lumbar spine -Nissl staining

The hepatic and spinal cord segments (T11-12, L3-4) were conventionally sliced for hematoxylin-eosin staining[21]and Nissl staining[22], respectively.Briefly, sections were stained with 1% (w/v) cresyl violet and 0.5% (w/v) toluidine blue solution, followed by alcohol dehydration, xylene washes, neutral gum mounting and observations under a light microscope (Olympus, Tokyo, Japan).

NF and Syn II immunofluorescence staining of spinal thoracic and lumbar anterior horn neurons and the lateral cord

Thoracic and lumbar cord slices were blocked with 3%(v/v) H2O2exogenous antigen, 10% (v/v) normal serum,and incubated with mouse anti-rat NF monoclonal antibody and rabbit anti-rat Syn II monoclonal antibody (1:500; Sigma, St.Louis, MO, USA) at 4oC overnight.Slices were then incubated with red and green fluorescent rabbit anti-mouse and goat anti-rabbit monoclonal antibody(1: 200; Sigma) and observed under a fluorescence microscope (Olympus, Tokyo, Japan).Sections were analyzed with Biosens Digiscope image analysis software(Shanghai Bio-Tech Co., Ltd., China).

Statistical analysis

Sample mean values of each group were analyzed with SAS 10.0 software (SAS Institute Inc., Cary, NC, USA)for one-way analysis of variance and pairwiset-test.Count data are expressed as mean±SD.AP＜0.05 was considered a statistically significant difference.

Author contributions:Aiqun Wu conceived and designed this study, and provided technical and funding support.Yu Lei analyzed the data and wrote the manuscript, while other authors assisted in the anatomy analysis and sampling.Huilong Huang,Yunheng Zhou and Suiliang Zhang were responsible for collection and verification of clinical cases.Aiqun Wu was responsible for this study.

Conflicts of interest:None declared.

Ethical approval:This study was approved by the Chinese Red Cross for voluntary donation used in medical research.

[1]Yin YH, Ma ZJ, Guan YH, et al.Clinical features of hepatic myelopathy in patients with chronic liver disease.Postgrad Med J.2009;85(1000):64-68.

[2]Panicked, SinhaS, Taly AB, et al.Hepatic myelopathy: a rare complication following extrahepatic portal vein occlusion and lienorenal shunt.Neurol India.2006;54(3):298-300.

[3]Leigh AD, Card WI.Hepatolenticular degeneration: a case associated with postero-lateral column degeneration.Neuropath Exp Neurol.1949;8(3):338-446.

[4]Zhang YX, Chen HH, Wu CL, et al.Clinical and neuroimaging characteristics and prognosis of patients with hepatic myelopathy and hepatic encephalopathy.Zhongfeng yu Shenjing Jibing Zazhi.2010;27(7):605-607.

[5]Koo JE, Lim YS, Myung SJ, et al.Hepatic myelopathy as a presenting neurological complication in patients with cirrhosis and spontaneous splenorenal shunt.Korean J Hepatol.2008;14(1):89-96.

[6]Nardone R, Buratti T, Oliviero A, et al.Corticospinal involvement in patients with a portosystemic shunt due to liver cirrhosis: a MEP study.J Neurol.2006;253(1):81-85.

[7]Lewis MB, MaeQuillan G, Bam ford JM, et al.Delayed myelopathic presentation of the acquired hepatocerebral degeneration syndrome.Neurology.2000;54(4):1011.

[8]Weng YB, Wang Y, Zhang ZT, et al.Pathological changes of motor neurons of brain and spinal cord after liver cirrhosis in rats.Zhonghua Gandan Waike Zazhi.2004;10(5):329-332.

[9]Tazawa K, Shimojima Y, Okano T, et al.An autopsy case with adult onset type II citrullinemia showing myelopathy.J Neurol Sci.2007;253(1-2):77-80.

[10]Giangaspero F, Dondi C, Scarani P, et al.Degeneration of the corticospinal tract following portosystemic shunt associated with spinal cord infarction.Virchows Arch.1985;406(4):475-481.

[11]Jiang WH.Neuroanatomy.Shanghai: Fudan University Press.2002:468-470.

[12]Ghia A, Ghia F, Coman L.The gastric mucosa in portal hypertension: structural and ultrastructural observations.Rom J Morphol Embryol.2010;51(2):271-275.

[13]Qiu FZ, Yang Z.Surgery in the treatment of portal hypertension status and trends.Waike Lilun yu Shijian.1999;4(2):65-67.

[14]Wang ZM, Yin RB, Luo DG.The clinical choice of portal hypertension elective surgery.Zhongguo Xiandai Yaowu Yingyong.2010;4(16):25-26.

[15]Khlynov IB, Chikunova MV, Lisovskaia TV.Effectiveness of nutritional support for the liver-cell deficiency in the liver cirrhosis.Eksp Klin Gastroenterol.2009;(2):39-43.

[16]Pagliaro L, D’Amico G, Pasta F.Portal hypertension in cirrhosis:natural history.In: Bosch J, Groszmann RJ, eds.Portal Hypertension: Pathophysiology and Treatment.London:Blackwell Scientific.1944:72-92.

[17]Qian NS, Dou KF.Clinical research of orthotropic liver transplantation in patients with hepatic myelopathy (2 cases analysis).Gan Dan Yi Waike Zazhi.2008;20(1):10-12.

[18]Lebovics E, DeMatteo RE, Schaffner F, et al.Portal-systemic myelopathy after portacaval shunt surgery.Arch Intern Med.1985;145(10):1921-1922.

[19]Pinarbasi B, Kaymakoglu S, Matur Z, et al.Are acquired hepatocerebral degeneration and hepatic myelopathy reversible.J Clin Gastroenterol.2009;43(2):176-181.

[20]Qu B, Liu C, Guo L, et al.The role of liver transplantation in the treatment of hepatic myelopathy: case report with review of the literature.Transplant Proc.2009;41(5):1987-1999.

[21]Yang DT, Erin BL, Xianlu Q, et al.Functional recovery following traumatic spinal cord injury mediated by a unique polymer scaffold seeded with neural stem cells.PNAS.2002;5(99):3024-3029.

[22]Yuan SZ.Cutting-edge Neuroscience Research Technical Manual.Beijing: Science Press.1997:36-37.