Objectives: After liver transplant, veno-occlusive disease and infectious complications may result from subclinical pulmonary hypertension. In this retrospective study, we investigated whether our preemptive bundle therapy was effective for subclinical pulmonary hypertension and extrasinusoidal platelet aggregation after liver transplant.
Materials and Methods: After January 2014, nutrition therapy with glutamine, synbiotics, phosphodiesterase 3 inhibitors, prostaglandin E1, prostaglandin I2, closed-loop artificial pancreas, and sivelestat has been used to reduce bacterial translocation, vascular endothelial cell damage, and extrasinusoidal platelet aggregation, which is administered as preemptive bundle therapy for all liver transplant recipients. In this study, we evaluated the prognosis of 84 liver transplant recipients who underwent liver transplants through 2018. Subclinical pulmonary hypertension was evaluated in 49 adult liver transplant recipients with an evaluable main pulmonary artery trunk cross-sectional area using enhanced computed tomography in the acute phase after transplant, with 14 of these patients receiving preemptive bundle therapy.
Results: Subclinical pulmonary hypertension was reduced in the preemptive bundle therapy group (n = 14) compared with the nontherapy group (n = 35). The preemptive bundle therapy group showed more rapid recovery of platelet, prothrombin time, and bilirubin levels after liver transplant compared with the nontherapy group. The prognosis of patients in the preemptive bundle therapy group was significantly better than in the nontherapy group. Extrasinusoidal platelet aggregation was significantly lower in the preemptive bundle therapy group than in the nontherapy group.
Conclusions: Preemptive bundle therapy reduced sinusoidal endothelial cell injury, extrasinusoidal platelet aggregation, and subclinical pulmonary hypertension after liver transplant, resulting in good posttransplant recovery.
Key words : Extrasinusoidal platelet aggregation, Liver transplantation, Phosphodiesterase 3 inhibitor, Veno-occlusive disease
Portopulmonary hypertension (PoPH)1 is considered the direct effect of bacterial translocation (BT) from the gut and vasoconstrictors that are present in the lungs as a result of a portosystemic shunt.2,3 Additionally, PoPH is worsened by capillarization of the hepatic sinusoid, by increasing intrahepatic production of vasoconstrictors, and by low production of vasodilators such as prostacyclin (prostaglandin I2, ie, PGI2) and nitric oxide.2,3 After liver transplant (LT), the patient may develop cor pulmonale, such as expansion of pulmonary artery, massive bilateral pleural effusion, and expansion of the right ventricle, and this is considered a subclinical pulmonary hypertension (PH) state equivalent to PoPH. Our studies have shown that extrasinusoidal platelet aggregation (EPA) resulting from sinusoidal endothelial cell injury and subsequent veno-occlusive disease (VOD) also cause PH4-10 after LT. Additionally, BT has been shown to induce sinusoidal endothelial cell damage and platelet activation by neutrophil extracellular traps (NETs) in sinusoids, causing EPA and affecting PH formation after LT.7,9,11 Therefore, we hypothesized that perioperative enhancement of intestinal immunity may foster suppression of BT and reduction of EPA, resulting in a reduction of post-LT subclinical PH.
Perioperative administration of a phosphodi-esterase 3 (PDE3) inhibitor, which has antiplatelet and protective effects on sinusoidal endothelial cells and which has been shown to reduce ischemia-reperfusion injury (IRI) during LT,12-15 has also been shown to reduce EPA following IRI. Thus, PDE3 inhibitors may suppress subclinical PH after LT.16 Intensive insulin therapy with a closed-loop artificial pancreas, which can strictly control the blood glucose level without hypoglycemia, also has a protective effect on sinusoidal endothelial cells, and thus, it was introduced in the LT perioperative period.17 Additionally, prostaglandin E1 (PGE1), which has antiplatelet, vasodilation, and anti-inflammatory effects,18 and PGI2, which has antiplatelet, vasodilation, and vascular endothelial protective effects,19,20 have been shown to improve insulin resistance21 and are already in use for PoPH.22
Based on Japanese public insurance requirements, PGE1 and PGI2 are administered during the LT perioperative period to reduce subclinical PH by reduction of EPA. Additionally, for recipients who undergo artificial ventilation after transplant, sivelestat is administered to control intrasinusoidal and intrapulmonary NETs caused by BT from the gut.23,24 In related living donor LT, appropriate graft selection is performed to avoid severe small-for-size graft syndrome, and splenectomy is actively performed to control portal venous pressure.
Materials and Methods
This retrospective study was approved by the Kanazawa University Ethics Committee (approval No. 2018-007), and informed consent was obtained from all living recipients who were included in this study. The study protocol for the artificial pancreas was approved by the institutional review board of Kanazawa University Hospital. The study was conducted in accordance with the Declaration of Helsinki (Clinical Trial ID: UMIN000016451; https://rctportal.niph.go.jp/). The need to acquire written informed consent from deceased recipients was waived because of the retrospective nature of this study.
Eighty-four recipients who underwent LT at Kanazawa University Hospital through 2018 were enrolled (Table 1). There were 5 deceased donor LT recipients and 79 related living donor LT recipients. The total number of living donors was 80 because there was 1 patient with retransplant; all living donors have remained in good health after donation. The living donors’ relationships with the recipients were 25 spouses, 12 parents, 31 children, 10 siblings, and 2 nephews. All 14 recipients who underwent LT after 2014 received preemptive bundle therapy, whereas those who underwent LT before 2013 did not receive this therapy. Preemptive bundle therapy was performed as a regular treatment within Japanese public insurance coverage parameters.
In 49 recipients who underwent a detailed evaluation of the thorax and abdomen in the acute phase after LT with enhanced multidetector row computed tomography (MDCT), the pulmonary artery cross-sectional area and blood data (platelets count, prothrombin time, and total bilirubin) were evaluated. The group without preemptive bundle therapy before 2013 was defined as nontherapy group A, and the group with preemptive bundle therapy after 2014 was defined as preemptive bundle therapy group B.
In 28 recipients who underwent graft liver biopsy in the acute phase after LT, immunohistochemical staining for glycoprotein Ib? (GpIb?, ie, CD42b) of platelet surface antigens was performed with liver biopsy specimens to evaluate state of EPA and VOD.
Preemptive bundle therapy for extrasinusoidal platelet aggregation and subclinical pulmonary hypertension
Beyond nutrition therapy
Synbiotics such as Elental (EA Pharma), GFO (Otsuka Chemical Industrial), Lactobacillus casei (Yakult Honsha), bifidobacteria (Yakult Honsha), and Clostridium butyricum (Miyarisan Pharmaceutical) containing glutamine and branched-chain amino acids were administered orally or via enteral feeding tube to perioperative LT recipients for enhancement of intestinal immunity.
Phosphodiesterase 3 inhibitor
The perfusate (University of Wisconsin solution; Astellas Pharma) was mixed with 10 mg of PDE3 inhibitor (Milrinone; Astellas Pharma) and administered intravenously (0.1-0.5 ?g/kg/min) when oral intake was not possible, including during surgery. Cilostazol was orally administered (100-200 mg/d) when oral ingestion was possible.
Prostaglandin E1 analogue
Prostandin (Maruishi Pharmaceutical), a PGE1 analogue, was continuously administered (0.01 ?g/kg/min) as a preemptive therapy to prevent intraoperative hypertension.
To evaluate the cardiovascular function, a pulmonary artery catheter was inserted just before surgery, after introduction of general anesthesia, and was left in place until the patient was removed from the ventilator after surgery. Then, in patients with subclinical PH who had a pulmonary artery wedge pressure ?15 mm Hg and a mean pulmonary artery pressure ?25 mm Hg, beraprost (PGI2 analogue; Kaken Pharmaceutical) was administered orally at 30 to 60 ?g/d.
Postoperatively, in patients who were on artificial ventilation because of respiratory distress syndrome associated with systemic inflammatory response syndrome, we expected to suppress NETs introduced by BT from the gut with 4.8 mg/kg/d sivelestat (Maruishi Pharmaceutical), which was administered based on public insurance parameters.
Intensive insulin therapy by closed-loop artificial pancreas
Detailed mechanisms of the closed-loop artificial pancreas (STG-55; Nikkiso) have been reported in previous studies.17,25 Venous blood was continuously sampled, and the blood glucose levels were continuously monitored. The artificial pancreas can automat-ically adjust the blood glucose level to the target blood glucose level, which comprises the closed-loop system. A target blood glucose level was set at 80 to 110 mg/dL for tight glycemic control. The artificial pancreas system was used to control blood glucose until 24 hours before the start of surgery. The accuracy and reliability of an artificial pancreas to continuously monitor blood glucose have been established previously.17,25
Immunohistochemistry for extrasinusoidal platelet aggregation evaluation
Platelet activation and aggregation in the sinusoid, the perisinusoidal space (space of Disse), and the parenchyma of the graft liver were evaluated by immunohistochemical staining for platelet surface receptor CD42b.4-6 The expression of CD42b was examined immunohistochemically with the respective primary antibodies with the EnVision+ system (Dako). Dewaxed 4-?m sections were incubated with 1:50 with protein blocking serum for 10 minutes to block nonspecific binding proteins, and immunostaining was performed with the EnVision+ system (Dako). Briefly, the slides were incubated with each primary antibody (1:50) at 4 °C overnight. After wash, the EnVision+ polymer solution was applied for 1 hour. The reaction products were visualized with 3,3?-diaminobenizidine tetrahydrochloride (Sigma Chemical) and H2O2. The specimens were then lightly counterstained with hematoxylin and examined under a fluorescence microscope. Primary antibodies used for immunostaining were CD42b rabbit antihuman polyclonal antibody (Atlas Antibodies). A similar dilution of control rabbit immunoglobulin G (Dako) was applied instead of the primary antibody, as a negative control. Positive and negative controls were routinely included.
Extrasinusoidal platelet aggregation was classified into 4 grades (0 to 4) based on the EPA-positive zone 3 rate (in %, as EPA-positive zone 3 vs all zone 3). Grade 0 was defined as no EPA-positive staining in zone 3, whereas grade 1 was 0 to 33% staining, grade 2 was 33% to 66% staining, and grade 3 was >66% EPA-positive staining in zone 3.
Posttransplant contrast-enhanced MDCT images of the chest and abdomen were available for 49 recipients. Slice thickness for each examination ranged from 2.0 to 5.0 mm in all cases.
Image analysis with Ziostation2 analysis software
The cross-sectional area of the pulmonary artery main trunk was calculated with Ziostation2 analysis software (Ziosoft) with the minimum cross-sectional area of the pulmonary artery main trunk. Briefly, the pulmonary artery was extracted from the enhanced MDCT image and automatically converted into a 3-dimensional (3D) image. The pulmonary artery cross-sectional area was continuously measured according to this 3D image, and its minimum value was used as the cross-sectional area. Figure 1 shows the time course of enhanced MDCT images in the acute phase after LT for recipients who underwent preemptive bundle therapy, including axial and sagittal images of chest-enhanced computed tomography images on days 13, 26, and 45 after LT. The cross-sectional area of the pulmonary artery main trunk and the right ventricle volume also decreased, and the bilateral pleural effusion also gradually decreased with the reduction of the cross-sectional area of the pulmonary artery main trunk.
Numerical data are presented as the mean ± SD and compared by the Student’s t test or Welch’s t test. Correlations between the 2 factors were evaluated with the chi-square test. Overall survival was compared by the log-rank test with the Kaplan-Meier method. SPSS software was used for the analyses. P < .05 was considered statistically significant.
Table 1 shows the details of the all-nontherapy group (n = 70), nontherapy group A (n = 35), and preemptive bundle therapy group B (n = 14). Nontherapy group A comprised the recipients who received enhanced MDCT during the acute phase after LT in the all-nontherapy group. Group B had no mortality, which was a significantly better prognosis compared with the all-nontherapy group and group A. In group B, the splenectomy frequency was significantly higher compared with group A to control the portal venous pressure after graft reperfusion. The cold ischemic time was significantly longer in group B compared with the all-nontherapy group and group A, reflecting the high frequency of deceased donor LT.
Table 2 shows the results of the analyses of the pulmonary artery cross-sectional area (S) ratios and the blood data after surgery between groups A and B. Subclinical PH was evaluated from the ratio of the pulmonary artery cross-sectional area of each recipient versus the normal pulmonary artery cross-sectional area. Enhanced MDCT images after posttransplant recovery were used for surviving patients, whereas for recipients who died in the early phase after transplant, the image data we used were from the period when the general condition was relatively stable before transplant (Child-Turcotte-Pugh status A). The values for the pulmonary artery cross-sectional area ratio (pulmonary artery S ratio) after LT in the nontherapy group A and preemptive bundle therapy group B were compared, and group B showed a significantly lower pulmonary artery S ratio up to 3 weeks after LT. At week 4, results for group B also tended to be lower compared with group A. Thus, subclinical PH was reduced in group B by preemptive bundle therapy compared with group A.
Group B showed a more rapid platelet level recovery after LT compared with group A. Similarly, group B had better recovery of prothrombin time and international normalized ratio compared with group A. Additionally, total bilirubin in group A was better compared with group B, but the recovery after LT in group B was better compared with group B. Thus, introduction of preemptive bundle therapy suppressed EPA and subclinical PH after LT, and platelet count, hepatic synthesis ability, and hepatobiliary excretion ability also showed good recovery.
The prognosis of patients in preemptive bundle therapy group B was significantly better compared with the all-nontherapy group and nontherapy group A, and all group B recipients recovered (Figure 2).
In 28 patients who underwent graft liver biopsy during the acute phase after LT in groups A and B, liver biopsy specimens were immunohistochemically evaluated for EPA. As shown in Figure 3, the degree of EPA was significantly lower in the preemptive bundle therapy group B compared with group A, and 80% of group B recipients were EPA grade 0, which indicated that there was no EPA. This result confirms that group B had good platelet count recovery and that the recovery after transplant for this group was quicker compared with group A.
A PDE3 inhibitor could be a key drug in preemptive bundle therapy. The efficacy and safety of a PDE3 inhibitor preadministration for IRI, EPA, and VOD have been extensively shown, including in our research.12,13,16,26 However, clinical application of PDE3 inhibitor has not been reported; here, we have described its clinical use in the present study.
The monocrotaline (MCT)-induced hepatic EPA and VOD animal model that we used in previous studies26,27 also induces PH,28,29 which suggests that EPA is a trigger for PH. Previously, we had used clinical data and the MCT-induced VOD rat model to show that the pathophysiology of VOD involves sinusoidal endothelial cell damage and subsequent EPA.26,27 We have also demonstrated that EPA and VOD are significantly reduced by either oral preadministration or portal vein preinjection of a PDE3 inhibitor, which protects and strengthens sinusoidal endothelial cells, with the MCT-induced VOD rat model.26,30 Additionally, we found that PH was also reduced (unpublished data).
We have also observed no effect of postadministration PDE3 inhibitor in MCT-induced EPA and VOD models.26 Preemptive therapy is critical to avoid serious EPA, VOD, and PH. Therefore, we injected perfusate (University of Wisconsin solution) that included a PDE3 inhibitor and continued intravenous administration for recipients when oral intake was not possible, including during surgery. When oral intake was possible, the medication was taken orally.
Hyperglycemia pathophysiology may be related to sinusoidal endothelial dysfunction, inflammatory changes, and poor immune function.31,32 In 2008, Van den Berghe proposed accurate and continuous blood glucose monitoring and a computer-assisted blood glucose control system, which is the closed-loop system, in the intensive care unit to avoid hypoglycemia.33 In response to this proposal, Hanazaki and colleagues reported the safety and effectiveness of intensive insulin treatment using the closed-loop glycemic control system and an artificial pancreas in a clinical trial that enrolled patients for digestive surgery.25 Based on their results, we introduced the closed-system artificial pancreas for strict blood glucose control in highly invasive hepatobiliary and pancreatic surgery such as LT during the perioperative period. Strict glycemic control using a closed-loop artificial pancreas was shown to be safe with no hypoglycemia in LT recipients, and it was useful for infection control and maintained a mortality rate of zero.17 Recipients undergoing LT cannot control their blood glucose level under 150 mg/dL without hypoglycemia; however, with the closed-loop artificial pancreas, it was possible to strictly control the blood glucose level below 150 mg/dL without hypoglycemia. In our experience,17 insulin has the effect to reduce IRI during transplant and protect endothelial cells; therefore, intensive insulin therapy with closed-loop artificial pancreas is deemed useful.34,35,36
Because PGE1 has been shown to reduce IRI in sinusoidal endothelial cells,18 PGE1 was continuously administered for preemptive therapy and to control intraoperative hypertension. Because PGE1 has an antiplatelet aggregation effect, fibrinolytic activity, vasodilation activity, an anti-inflammatory effect including inhibition of tumor necrosis factor ? release from Kupffer cells,18 adhesion molecule expression, and neutrophil adherence to endothelial cells, PGE1 is expected to protect vascular endothelial cells including sinusoids and reduce EPA and NETs, thereby reducing transmission of death mediators and vasoconstrictors from the liver to the lungs. Similarly, PGI2, an anti-PH drug, has an antiplatelet effect, a vasodilation effect, a protective effect on vascular endothelial cells, and a cytoprotective effect on liver cells.19,20 Therefore, hepatoprotection including sinusoidal endothelial cells and pulmonary arterial pressure reduction effects are expected. Additionally, PGI2 enhances endothelial nitric oxide synthase activation by stimulating insulin in vascular endothelial cells and improves insulin resistance by restoring the capillary dilatation ability and insulin transfer to the skeletal muscle stroma.21 Therefore, a synergistic effect with intensive insulin therapy using a closed-loop artificial pancreas is expected. Prostacyclin also increases hepatic venous blood flow without increasing portal venous pressure,22 so it is highly likely to reduce pulmonary arterial pressure, which is useful for PH.
Prevention of BT by preemptive bundle therapy suppresses NET, further reduces sinusoidal endothelial cell injury, and reduces EPA and VOD and thereby reduces subclinical PH and leads to rapid recipient recovery. Recovery of the blood platelet count is a good barometer of recipient recovery after LT.37 However, in this study of 28 recipients with graft liver biopsy, an inverse correlation was found between platelet level recovery and EPA grade, indicating that EPA was a main cause of transfusion-resistant thrombocytopenia after LT. However, the rapid recovery of platelets after LT in group B may also be the result of the high frequency of splenectomy.
Based on the results of pulmonary artery catheterization, the pulmonary artery diameter on the MDCT coronal image or the pulmonary artery diameter-to-aorta diameter ratio is useful as a simple evaluation method for PoPH.37-41 However, this is not always an objective evaluation method, because the cross-sectional area is evaluated with only a single axis and corrections for individual differences such as sex and body type are not included. Our evaluation method used Ziostation2 analysis software, which can continuously evaluate the true cross-sectional area and automatically and accurately construct 3D images, and this is considered a better and more objective evaluation method. We used the ratio of the minimum continuous measurement value of the accurate pulmonary artery cross-sectional area to the minimum value of the healthy state in 3D MDCT images for simple evaluation of subclinical PA, for which corrections for individual differences and objective numerical values are included. Additionally, pulmonary artery catheteri-zation and echocardiographic evaluation are considered useful for objective evaluation of PH, but enhanced MDCT is also better for subclinical PH evaluation for several reasons, as follows.
First, enhanced MDCT is less invasive than pulmonary artery catheterization, and analysis software (Ziostation2) can be used to perform accurate 3D automatic image analysis. Additionally, echocardiography has poor reproducibility. However, because the correlation between the pulmonary artery cross-sectional area (S) ratio and mean pulmonary artery pressure, pulmonary artery wedge pressure, and pulmonary vascular resistance measured by pulmonary artery catheterization was not specifically investigated in detail, this remains a simple evaluation method.
This study has some limitations including the small number of patients and the retrospective study design. For all recipients of LT after 2014 (group B), enhanced MDCT was performed frequently in the acute phase after LT; however, in earlier LT, MDCT was performed only in recipients for whom computed tomography was deemed necessary. Additionally, bias in group A patients cannot be ruled out. To show the usefulness of preemptive bundle therapy, a multicenter prospective randomized control study is required. However, it is not ethical to conduct this double-blind study in Japan, where living donor LT is relied upon, because there is a risk that the recipients may be disadvantaged by not receiving preemptive bundle therapy. Therefore, in the future, we will conduct a multicenter prospective single-arm study.
The preemptive bundle therapy described here suppressed sinusoidal endothelial cell injury, EPA, BT, and subclinical PH after LT and contributed to the rapid recovery of LT recipients.
Volume : 19
Issue : 11
Pages : 1173 - 1181
DOI : 10.6002/ect.2021.0176
From the 1Department of General and Digestive Surgery, Kanazawa Medical University, Kahoku, Ishikawa, Japan; the 2Department of Gastroenterologic Surgery, Kanazawa University, Kanazawa, Ishikawa, Japan; and the 3Department of Surgery, Toyama Prefectural Central Hospital, Toyama City, Toyama, Japan
Acknowledgements: The authors have not received any funding or grants in support of the presented research or for the preparation of this work and have no declarations of potential conflicts of interest. The authors are grateful to Nagata, who is affiliated with the Radiology Center of Kanazawa Medical University for measuring the cross-sectional area of the pulmonary artery with Ziostation2 analysis software, and are grateful to the staff at Kanazawa University Hospital (Kanazawa, Japan). We also thank Jodi Smith, PhD, from Edanz Group for editing a draft of this manuscript.
Corresponding author: Hiroyuki Takamura, Dept. of General and Digestive Surgery, Kanazawa Medical University, 1-1 Daigaku, Uchinada-Machi, Kahoku, Ishikawa 920-0293, Japan
Phone: +81 76 286 2211
Table 1. Comparative Study of Clinicopathological Factors in Each Liver Transplant Group
Figure 1. Transition of Pulmonary Artery Cross-Sectional Area and Bilateral Pleural Effusion in Recipient After Liver Transplant
Table 2. Comparative Study of Pulmonary Artery S Ratio and Laboratory Data Between Preemptive Bundle Therapy Group B and Nontherapy Group A
Figure 2. Overall Survival Rate of Recipients After Liver Transplant Was Evaluated Separately for Patients With and Without Preemptive Bundle Therapy
Figure 3. Preemptive Bundle Therapeutic Effect on Extrasinusoidal Platelet Aggregation in Adult Liver Transplant Recipients