Patients
We retrospectively reviewed the data of 93 patients who were diagnosed with PA-HSOS and who had HVPG measurements at the affiliated Nanjing Drum Tower Hospital of Nanjing University School of Medicine between January 2016 and April 2020. All patients received regular anticoagulation-TIPS stepwise treatment. The inclusion criteria were as follows: 1. age between 18 and 80 years; 2. received a diagnosis of PA-HSOS according to the Nanjing criteria; the criteria included a confirmed history of PA-containing plant intake and the following three requirements: (1) abdominal distention and/or pain in the hepatic region, hepatomegaly and ascites; (2) elevation of serum total bilirubin or abnormal liver function testing; and (3) enhanced computed tomography (CT) or magnetic resonance imaging (MRI) features or pathological evidence that ruled out other known causes of liver injury7; 3. had anticoagulation-TIPS stepwise therapy after admission; and 4. had HVPG measurements taken after admission. The exclusion criteria were as follows: 1. patients with thrombosis of the portal vascular system (inaccurate HVPG result due to thrombosis); 2. patients with important data that was missing; and 3. patients with organ dysfunction or history of cancer. Ultimately, 76 patients were included in this study. The study flowchart is shown in Fig. 1.
Follow‑up
The patients had frequent follow-up at 1, 3, and 6 months after discharge from the hospital and then every 6 months thereafter. Every visit included a systematic examination for clinical signs and symptoms, routine blood tests, coagulation tests, liver function tests and blood ammonia levels, which were routinely rechecked. Ultrasound or CT scans could be used if necessary. The follow-up time was defined as the interval from admission to death or the last visit or the end of the study (01 September 2021).
DTSS estimates the severity of PA-HSOS
By evaluating the severity, the Drum Tower Severity Scoring (DTSS) method may accurately predict the outcome of supportive care and anticoagulation in PA-HSOS patients and may have potential significance for directing therapy16. The entire system contains AST, STB, Fib, and PPV, all of which may be collected by noninvasive procedures, making the model suitable for clinical application. According to the score, patients are then divided into three groups: mild, moderate and severe. DTSS proposes outpatient anticoagulation and follow-up every other week for mild patients; hospitalization for anticoagulation and weekly monitoring of pertinent parameters for intermediate patients; and direct TIPS for severe patients.
Instruments and testing methods
The HVPG method was as follows: under the Philips FD20 DSA instrument, the patient was placed in the supine position with the right mandibular angle set at 2 cm and the medial margin of the sternocleidomastoid muscle set as the puncture point, and routine disinfection, draping and local infiltration anesthesia with 2% lidocaine were performed. After the right internal jugular vein was punctured successfully, a guide wire was inserted into the inferior vena cava (IVC). Along the guide wire, a Cobra catheter was inserted into the IVC and superselected into the middle hepatic vein. A 5.5F thru-lumen embolectomy catheter (Fogaty, Edwards Lifesciences, USA) was used to fill the middle hepatic vein with balloon inflation, and then wedge hepatic venous pressure (WHVP) was measured at 3 points. The free hepatic venous pressure (FHVP), inferior vena cava pressure (IVCP) and right atrial pressure (RAP) were also measured when the balloon was completely closed; and HVPG (WHVP-FHVP) was calculated. If necessary, the end of the catheter was sent into the hepatic vein/inferior vena cava, and a high-pressure syringe was used for the injection. The hepatic vein and inferior vena cava were observed for dilatation, stenosis and contrast medium retention.
Transjugular liver biopsy (TJLB) was performed by the Charles Dotter's approach. A micropuncture device (Cook, Bloomington, IN, USA) was used to access the right internal jugular vein under ultrasound guidance, and a 7F transjugular sheath and 19-gauge 48 cm Quick-Core needle biopsy system (Cook) were placed into the middle hepatic vein under fluoroscopic guidance. A total of 63 individuals had TJLB examinations, with each having two to four core biopsies.
Treatment regimens
On admission, all patients with PA-HSOS were given appropriate supportive therapy, anticoagulation treatment was given immediately after active bleeding, and bleeding risk was ruled out. There are three categories of anticoagulation treatment. Monotherapy with low-molecular-weight heparin (4000 IU/bid) was utilized for individuals with a contraindication to warfarin or a baseline international normalized ratio (INR) > 1.5. When a patient's INR was less than 1.5 and their platelets (PLT) were greater than 50*10^9/L, the use of low-molecular-weight heparin combined with warfarin was then evaluated. The patients who had a platelet count of less than 50*10^9/L at baseline or who were hypersensitive to low-molecular-weight heparin were given warfarin alone. The starting dose for warfarin monotherapy was 1.5 mg daily and was subsequently adjusted to obtain a target INR of 2.0–3.09. If the PA-HSOS patients did not respond to the initial anticoagulant therapy, more aggressive treatments such as TIPS were then considered.
Definition
Nonresponse to anticoagulation therapy
If the patient meets the following criteria, then the treatment is called nonresponding, and TIPS is recommended7,17. The criteria were as follows: (1) after two weeks of anticoagulation treatment, either serum total bilirubin (STB) ≥ 5 mg/dl, or the peak portal vein velocity (PVV) < 10 cm/s, or (2) any two of the following five were met during 2 weeks of anticoagulation: no improvement in STB (< 5 mg/dl), an increment of less than 10% of the PVV at baseline, no significant relief of ascites, concurrent renal and coagulation deterioration (serum creatinine (Scr) higher than before treatment, prothrombin time (PT) prolongation >3 s or INR >3.0, or bleeding tendency of the skin and mucosa)16. The patients who achieved alleviation from ascites and had no worsening of the STB or PVV at baseline were considered anticoagulant responders and were advised to continue anticoagulation medication until complete resolution.
Bleeding area in the sinusoids
The typical manifestations of PA-HSOS are swelling, damage, and shedding of hepatic sinusoid endothelial cells in hepatic acinus zone III, so this study intended to use the area of bleeding in the sinusoids as a quantitative evaluation index for the severity of sinusoidal injury. Considering that the sampling method of liver biopsy in this study was mainly transjugular liver biopsy and the sampled tissue was slender, the bleeding area was defined as follows: among all the punctured liver tissues, the ratio of the area of subsinusoidal congestion to the total area of liver tissue was the bleeding area. This index was first evaluated by a liver pathologist with more than 3 years of work experience and was then reviewed by a senior liver pathologist with more than 10 years of work experience; if the results were inconsistent, the final results were determined by a consultation between the two.
Data analysis
Data on the measures with a normal distribution are reported as the mean±standard deviation, and an independent sample t test was used to compare the groups. Data on the measures with a nonnormal distribution are expressed as the medians and interquartile ranges, and the Mann‒Whitney U test was used to compare the groups. The Kruskal‒Wallis H test was used for comparisons between multiple groups. The chi-squared or Fisher's exact test was applied to count data, which was reported as the number of patients (%).
We discovered the variables related to nonresponse to anticoagulation using univariate logistic regression analysis (P<0.1), and subsequent multivariate logistic regression analysis (Enter method) was used to identify independent risk factors predicting nonresponse to anticoagulation (P<0.05). The cutoff value of independent risk factors (obtained at the maximum Youden index) for predicting nonresponse to anticoagulation was determined using receiver operating characteristic curve (ROC) analysis, and the area under the curve (AUC) of these variables in predicting ineffective anticoagulation was assessed. A model for predicting nonresponse to anticoagulation by multivariate logistic analysis was developed (forward stepwise LR methods), and the power of the model was similarly analyzed by ROC curves. Due to an insufficient sample size, we used bootstrapping logistic regression analysis to check the robustness of our predictors. Two hundred samples from 76 patients were chosen at random and replaced for this analysis; then, we performed logistic regression analysis for each sample and documented the frequency of each variable appearing in the final models18. Variables that appear in 50% or more of the models can be considered robust predictors of the outcome19.
In addition, Kaplan‒Meier survival curves were used to compare the prognostic survival of the patients in the different HVPG groups, and the differences were compared by the log-rank test. Linear regression was used when analyzing the relationship between HVPG and DTSS and bleeding area in the sinusoids.
Statistical analyses were performed using IBM SPSS Statistics 22.0 (Armonk, New York, USA), R studio version 4.2.1 and GraphPad Prism 9 (GraphPad Software, La Jolla, California, USA). A p value of < 0.05 indicated significant differences.
Ethics
Because the study was a retrospective study, informed consent was waived by the ethics committee. However, this study involved the survival outcome of patients followed up by telephone, so the Ethics Committee asked the investigator to ask whether the patient wanted to participate in this study during telephone follow-up, and the informed consent time and process were recorded after obtaining the patient 's consent.
The study complied with the relevant requirements of the Declaration of Helsinki and has been registered and approved by the Ethics Committee of Nanjing Drum Tower Hospital.