Clinical Decompensation and Outcomes in Patients With Compensated Cirrhosis and a Hepatic Venous Pressure Gradient ≥20 mm Hg

Can Plasma Volume Determination in Cirrhosis Be Replaced by an Algorithm Using Body Weight and Hematocrit?

BACKGROUND

Patients with cirrhosis often develop hyperdynamic circulation with increased cardiac output, heart rate, and redistribution of the circulating volume with expanded plasma volume (PV). PV determination is part of the evaluation of patients with cirrhosis, but gold-standard methods are invasive, expensive, and time-consuming. Therefore, other estimations of PV would be preferable, and the aim of this study was therefore to study if PV, as assessed by a simplified algorithm based on hematocrit and weight, can replace the gold-standard method.

METHODS

We included 328 patients with cirrhosis who had their PV assessed by the indicator dilution technique as the gold-standard method (PVI-125). Actual PV was estimated as PVa = (1 - hematocrit)·(a + (b·body weight)). Ideal PV was estimated as PVi = c · body weight, where a, b, and c are constants.

RESULTS

PVI-125, PVa, and PVi were 3.99 ± 1.01, 3.09 ± 0.54, and 3.01 ± 0.65 (Mean ± SD), respectively. Although PVI-125 correlated significantly with PVa (r = 0.72, p < 0.001), a Bland-Altman plot revealed wide limits of confidence.

CONCLUSIONS

The use of simplified algorithms does not sufficiently estimate PV and cannot replace the indicator dilution technique.

Dietary salt in liver cirrhosis: With a pinch of salt!

Patients with liver cirrhosis are advised to limit their sodium consumption to control excessive fluid accumulation. Salt is the most common form in which sodium is consumed daily. Consequently, various recommendations urge patients to limit salt intake. However, there is a lack of consistency regarding salt restriction across the guidelines. Moreover, there is conflicting evidence regarding the efficacy of salt restriction in the treatment of ascites. Numerous studies have shown that there is no difference in ascites control between patients with restriction of salt intake and those without restriction. Moreover, patients with cirrhosis may have several negative effects from consuming too little salt, although there are no recommendations on the lower limit of salt intake. Sodium is necessary to maintain the extracellular fluid volume; hence, excessive salt restriction can result in volume contraction, which could negatively impact kidney function in a cirrhotic patient. Salt restriction in cirrhotic patients can also compromise nutrient intake, which can have a negative impact on the overall outcome. There is insufficient evidence to recommend restricted salt intake for all patients with cirrhosis, including those with severe hyponatremia. The existing guidelines on salt restriction do not consider the salt sensitivity of patients; their nutritional state, volume status and sodium storage sites; and the risk of hypochloremia. This opinion article aims to critically analyze the existing literature with regard to salt recommendations for patients with liver cirrhosis and identify potential knowledge gaps that call for further research.

Compensated liver cirrhosis: Natural course and disease-modifying strategies

Compensated liver cirrhosis (CLC) is defined as cirrhosis with one or more decompensating events, such as ascites, variceal haemorrhage, or hepatic encephalopathy. Patients with CLC are largely asymptomatic with preserved hepatic function. The transition from CLC to decompensated cirrhosis occurs as a result of a complex interaction between multiple predisposing and precipitating factors. The first decompensation event in CLC patients is considered a significant turning point in the progression of cirrhosis, as it signals a drastic decline in median survival rates from 10-12 years to only 1-2 years. Furthermore, early cirrhosis has the potential to regress as liver fibrosis is a dynamic condition. With the advent of effective non-invasive tools for detecting hepatic fibrosis, more and more patients with CLC are currently being recognised. This offers clinicians a unique opportunity to properly manage such patients in order to achieve cirrhosis regression or, at the very least, prevent its progression. There are numerous emerging approaches for preventing or delaying decompensation in CLC patients. A growing body of evidence indicates that treating the underlying cause can lead to cirrhosis regression, and the use of non-selective beta-blockers can prevent decompensation by lowering portal hypertension. Additionally, addressing various cofactors (such as obesity, diabetes, dyslipidaemia, and alcoholism) and precipitating factors (such as infection, viral hepatitis, and hepatotoxic drugs) that have a detrimental impact on the natural course of cirrhosis may benefit patients with CLC. However, high-quality data must be generated through well-designed and adequately powered randomised clinical trials to validate these disease-modifying techniques for CLC patients. This article discussed the natural history of CLC, risk factors for its progression, and therapeutic approaches that could alter the trajectory of CLC evolution and improve outcomes.

Effect of different infusion approaches on safety and efficacy of terlipressin: Current controversies

Terlipressin, a synthetic analog of vasopressin, has been widely used to treat acute variceal bleeding, hepatorenal syndrome, and other complications of cirrhosis. However, among different infusion approaches of terlipressin, its safety and efficacy are also heterogeneous. Previous studies have demonstrated that continuous infusion of terlipressin is more effective with a lower incidence of adverse events than intermittent intravenous infusion. This paper aims to review the relevant literature and summarize the data regarding the safety and effectiveness of different infusion approaches of terlipressin to guide clinical practice.

Portal Venous Remodeling Determines the Pattern of Cirrhosis Decompensation: A Systems Analysis

INTRODUCTION

As liver disease progresses, scarring results in worsening hemodynamics ultimately culminating in portal hypertension. This process has classically been quantified through the portosystemic pressure gradient (PSG), which is clinically estimated by hepatic venous pressure gradient (HVPG); however, PSG alone does not predict a given patient's clinical trajectory regarding the Baveno stage of cirrhosis. We hypothesize that a patient's PSG sensitivity to venous remodeling could explain disparate disease trajectories.

METHODS

We created a computational model of the portal system in the context of worsening liver disease informed by physiologic measurements from the field of portal hypertension. We simulated progression of clinical complications, HVPG, and transjugular intrahepatic portosystemic shunt placement while only varying a patient's likelihood of portal venous remodeling.

RESULTS

Our results unify hemodynamics, venous remodeling, and the clinical progression of liver disease into a mathematically consistent model of portal hypertension. We find that by varying how sensitive patients are to create venous collaterals with rising PSG we can explain variation in patterns of decompensation for patients with liver disease. Specifically, we find that patients who have higher proportions of portosystemic shunting earlier in disease have an attenuated rise in HVPG, delayed onset of ascites, and less hemodynamic shifting after transjugular intrahepatic portosystemic shunt placement.

DISCUSSION

This article builds a computational model of portal hypertension which supports that patient-level differences in venous remodeling may explain disparate clinical trajectories of disease.

Hepatic venous pressure gradient and rebleeding risk of patients with nonalcoholic steatohepatitis cirrhosis after variceal bleeding

Background and aims

Hepatic venous pressure gradient (HVPG) has a strong predictive value for variceal rebleeding in cirrhotic patients, but the accuracy of HVPG may be compromised in nonalcoholic steatohepatitis (NASH) cirrhosis. This study aimed to evaluate the accuracy of HVPG and portal pressure gradient (PPG) for predicting rebleeding in NASH cirrhosis after acute variceal bleeding.

Patients and methods

Thirty-eight NASH cirrhosis patients and 82 hepatitis B virus (HBV) cirrhosis patients with acute variceal bleeding were included in this study. All patients recived transjugular intrahepatic portalsystemic shunt (TIPS). The prognostic value of HVPG and PPG for variceal rebleeding was evaluated.

Results

Compared with HBV cirrhosis, NASH cirrhosis demonstrated a lower HVPG (15.3 ± 3.8 vs. 18.0 ± 4.8; p = 0.003) and lower PPG (18.0 ± 3.7 vs. 20.0 ± 3.4; p = 0.005). HVPG (AUC = 0.82; p = 0.002) and PPG (AUC = 0.72; p = 0.027) had promising prognostic value among NASH cirrhosis patients. The optimal threshold of HVPG and PPG for predicting rebleeding in NASH cirrhosis was 17 mmHg and 20 mmHg. At multivariate analysis, HVPG ≥17 mmHg was a significant predictor of variceal rebleeding (HR 9.40; 95% CI 1.85-47.70; p = 0.007).

Conclusion

In the patients with cirrhosis and vairceal bleeding, the levels of HVPG and PPG were found to be low in NASH cirrhosis than HBV cirrhosis. However, the prevalence of rebleeding was similar between two groups. HVPG measurement is still an accurate way to assess the risk of variceal rebleeding in NASH cirrhosis.

Hepatic Venous Pressure Gradient Predicts Further Decompensation in Cirrhosis Patients with Acute Esophageal Variceal Bleeding

BACKGROUND

The role of hepatic venous pressure gradient (HVPG) in predicting further decompensation in cirrhosis patients with acute variceal bleeding (AVB) is not known. We aimed to evaluate the role of HVPG in predicting further decompensation in cirrhosis patients with AVB Methods: In this prospective study, 145 patients with cirrhosis with esophageal or gastric AVB were included. HVPG was measured on the day of the AVB. Decompensating events occurring after 42-days of AVB were considered further decompensation.

RESULTS

The median age of the study cohort was 44 years; 88.3% males. The predominant etiology of cirrhosis was alcohol (46.2%). Overall, 40 (27.6%) patients developed further decompensation during median follow-up of 296 days following AVB. Gastro intestinal bleeding n = 27 (18.6%) and new-onset/worsening ascites n = 20 (13.8%) were the most common decompensating events. According to the multivariate model, HVPG was an independent predictor of any further decompensation in esophageal AVB patients but not in gastric variceal bleeding patients. HVPG cut-off of ≥16 mmHg predicted further decompensation in the esophageal AVB. However, HVPG was not an independent predictor of mortality.

CONCLUSION

HVPG measured during an episode of acute variceal hemorrhage from esophageal varices predicts further decompensating events in cirrhosis patients.

Clinical significance of measuring hepatic venous pressure gradient on transjugular liver biopsy for patients with pre-cirrhotic bridging fibrosis liver disease

PURPOSE

To demonstrate that patients with pre-cirrhotic bridging fibrosis (Meta-analysis of Histological Data in Viral Hepatitis, METAVIR stage F3) and clinically significant portal hypertension (CSPH, Hepatic Venous Pressure Gradient ≥10 mmHg) experience a higher rate of clinical decompensation than patients without CSPH.

MATERIALS AND METHODS

128 consecutive patients with pathology proven bridging fibrosis without cirrhosis between 2012 and 2019 were reviewed. Inclusion criteria were patients with HVPG measurement obtained during the same outpatient transjugular liver biopsy and clinical follow up of at least two years. Primary endpoint included rate of overall complication related to portal hypertension including evidence of either ascites, presence of varices on imaging or endoscopy, or evidence of hepatic encephalopathy.

RESULTS

Among 128 patients with bridging fibrosis (67 females and 61 males; average age 56 years), 42 (33%) and 86 (67%) were with and without CSPH (HVPG ≥10 mmHg). Median follow-up time was 4 years. Rate of overall complication (either ascites, varices or hepatic encephalopathy) in patients with and without CSPH was 36/42 (86%) and 39/86 (45%) (p < .001) respectively. Rate of developing ascites, varices and hepatic encephalopathy in patients with and without CSPH was 21/42 (50%) vs 26/86 (30%) (p = .034), 32/42 (76%) vs 26/86 (30%) (p ≤ .001) and 18/42 (43%) vs 12/86 (14%) (p = .001) respectively.

CONCLUSION

Patients with pre-cirrhotic bridging fibrosis and CSPH were associated with higher rates of developing ascites, varices and hepatic encephalopathy. Measuring HVPG during transjugular liver biopsy provides additional prognostic value in anticipating clinical decompensation in patients with pre-cirrhotic bridging fibrosis.

Low-dose continuous terlipressin infusion is effective and safer than intravenous bolus injections in reducing portal pressure and control of acute variceal bleeding

BACKGROUND AND AIMS

Continuous infusion of terlipressin is better tolerated, and equally effective at lower doses than intravenous boluses in type 1 hepatorenal syndrome. This approach in cirrhosis patients with acute esophageal variceal bleed was investigated by comparing the efficacy and adverse events of continuous versus bolus administration of terlipressin.

METHODS

One hundred ten consecutive cirrhosis patients with acute esophageal variceal bleed (AEVB) were randomized to receive either terlipressin as bolus (BOL, n = 55), 2 mg every 4 h, or, continuous infusion (CONI, n = 55), 4 mg/24 h for 5 days. Hepatic venous pressure gradient (HVPG) was measured at baseline, 12 and 24 h and response to terlipressin was defined as > 10% decline from baseline.

RESULTS

Baseline demographics, model for end-stage liver disease (MELD) and HVPG were comparable between groups. The primary objective of HVPG response at 24 h was achieved in significantly more patients in CONI than BOL group {47/55(85.4%) vs. 32/55(58.2%), p = 0.002}. Early HVPG response at 12 h was also higher in CONI group (71.5 vs. 49.1%, p < 0.01). Median dose of terlipressin was significantly lower {4.25 ± 1.26 mg vs. 7.42 ± 1.42 mg/24 h, p < 0.001)} and adverse events were fewer {20/55(36.3%) vs. 31/55(56.4%), p = 0.03} in the CONI than BOL group. Significantly higher incidence of very early rebleed was noted in BOL group {8/55 (14.5%) vs. 1/55, (1.8%), p = 0.03}. Baseline HVPG (OR 1.90, 95% CI = 1.25-2.89, p = 0.002) and MELD (OR 1.18, 95% CI = 0.99-1.41, p = 0.05) were predictors of rebleed.

CONCLUSION

"HVPG-tailored" continuous terlipressin infusion is more effective than bolus administration in reducing HVPG at a lower dose with fewer adverse events in cirrhotic patients.

CLINICAL TRIAL IDENTIFIER

NCT02695862.

Assessment of the Performance of Non-invasive Criteria for the Evaluation of Clinically Significant Portal Hypertension in Patients with Compensated Advanced Chronic Liver Disease

BACKGROUND AND AIMS

Assessment of clinically significant portal hypertension (CSPH) non-invasively using a combination of liver stiffness measurement (LSM) and platelet counts is proposed as an alternative to hepatic venous pressure gradient (HVPG) estimation. Utility of these criteria in compensated advanced chronic liver disease (cACLD) patients of different etiologies including nonalcoholic steatohepatitis (NASH) with BMI  >  30 kg/m² was studied in a large cohort.

METHODS

Consecutive patients of cACLD with available anthropometric and laboratory details, LSM, and HVPG were included in a retrospective analysis. A LSM of ≥ 25 kPa alone and LSM ≤ 15 kPa plus platelets ≥  150 × 10⁹/L were evaluated as non-invasive rule-in and rule-out criteria for CSPH, respectively. The NASH-ANTICPATE model (composite of BMI, platelets, and LSM) was evaluated in patients with obese NASH.

RESULTS

Patients with cACLD (n = 626) (mean age: 50.8 ± 12.4 years, 74.2% males) with alcohol (ALD, 30.3%), NASH (26.4%), hepatitis C (HCV, 16.6%), hepatitis B (HBV,10.2%) etiology were included. The prevalence of CSPH was  >  80% across all etiologies except in HBV (62.5%) and in obese non-NASH (71-72%). The rule-in criteria had a PPV  >  90% for all etiologies except in HBV (80.8%). The rule-out criteria had a negative predictive value (NPV) of 65%, 53%, and 40% in ALD, HCV, and NASH, respectively. The NASH-ANTCIPATE model had specificity of 100% and NPV of 33% to detect CSPH in obese NASH (n = 62).

CONCLUSIONS

LSM ≥ 25 kPa predicted CSPH in most etiologies except HBV. A significant proportion of patients have CSPH despite satisfying the rule-out criteria. The NASH-ANTICIPATE model is specific but fails to exclude CSPH in nearly two-third patients with obesity and NASH. There is a need for precise disease-specific non-invasive models for detecting CSPH.

The natural history of patients with compensated cirrhosis and elevated hepatic venous pressure gradient

Aims

Portal hypertension is a major complication of liver cirrhosis. Hepatic venous pressure gradient (HVPG) appears to be one of the best surrogates of clinical outcomes. However, the utility of elevated HVPG in predicting subsequent clinical decompensation is unclear.

Methods

We analyzed 410 patients who underwent HVPG assessment between 2014 and 2018. Of these, we identified and analyzed 20 patients with HVPG >12 mmHg without evidence of clinical decompensation (defined as ascites, non-bleeding esophageal varices or bleeding esophageal varices, hepatic encephalopathy, hepato-pulmonary syndrome, or hepatic hydrothorax). Additionally, we compared this group to 40 randomly selected cirrhotic patients with HVPG >12 mmHg with signs of clinical decompensation. Clinical events were subsequently assessed (mean = 33 months) after HVPG measurement.

Results

Patients with high HVPG without evidence of clinical decompensation had significantly lower model for end stage liver disease (MELD) scores (8 ± 4) compared to decompensated patients (13 ± 8, P = 0.05). HVPG measurements were similar in compensated (17 ± 6 mmHg) and decompensated (18 ± 4 mmHg) patients. Over follow-up for 33 months, 8/20 compensated patients had a decompensating event and neither MELD (8 and 8, respectively) nor HVPG (17 mmHg and 18 mmHg, respectively) differentiated patients who remained compensated vs. those that decompensated. Serum albumin at the time of HVPG measurement was significantly higher in patients who remained compensated than those with a decompensating event (3.5 g/dL vs. 2.6 g/dL, respectively, P = 0.05).

Conclusions

A small, unique, population of cirrhotic patients with substantially elevated HVPG appear to remain free of complications over long term follow-up.

Overview of Complications in Cirrhosis

BACKGROUND

Cirrhosis is the outcome of chronic liver disease of any etiology due to progressive liver injury and fibrosis. Consequently, cirrhosis leads to portal hypertension and liver dysfunction, progressing to complications like ascites, variceal bleeding, hepatic encephalopathy, hepatorenal syndrome, hepatopulmonary syndrome, cirrhotic cardiomyopathy, sarcopenia, hepatocellular carcinoma, and coagulation disorders. End-stage liver disease leads to an impaired quality of life, loss of social and economic productivity, and reduced survival.

METHODS

This narrative review explains the pathophysiology of complications of cirrhosis, the diagnostic approach and innovative management, with focus on data from India. A comprehensive literature search of the published data was performed in regard with the spectrum, diagnosis, and management of cirrhosis and its complications.

RESULTS

There is a change in the epidemiology of metabolic syndrome, lifestyle diseases, alcohol consumption and the spectrum of etiological diagnosis in patients with cirrhosis. With the advent of universal vaccination and efficacious long-term viral suppression agents for chronic hepatitis B, availability of direct-acting antiviral agents for chronic hepatitis C, and a booming liver transplantation programme across the country, the management of complications is essential. There are several updates in the standard of care in the management of complications of cirrhosis, such as hepatorenal syndrome, hepatocellular carcinoma, and hepatic encephalopathy, and new therapies that address supportive and palliative care in advanced cirrhosis.

CONCLUSION

Prevention, early diagnosis, appropriate management of complications, timely transplantation are cornerstones in the management protocol of cirrhosis and portal hypertension. India needs improved access to care, outreach of public health programmes for viral hepatitis care, health infrastructure, and disease registries for improved healthcare outcomes. Low-cost initiatives like immunization, alcohol cessation, awareness about liver diseases, viral hepatitis elimination, and patient focused decision-making algorithms are essential to manage liver disease in India.

Validation and comparison of non-invasive prediction models based on liver stiffness measurement to identify patients who could avoid gastroscopy

Several non‐invasive tests (NITs) based on liver stiffness measurement (LSM) have been developed to rule out varices needing treatment (VNT), including the Baveno VI criteria (B6C), the expanded Baveno VI criteria (EB6C), the LSM-spleen diameter to platelet ratio score (LSPS), and the VariScreen algorithm. We aimed to validate and compare those NITs in patients with compensated advanced chronic liver disease (cACLD). This retrospective study enrolled 354 patients with cACLD; LSM, platelet count (PLT), international normalized ratio (INR), gastroscopy and spleen diameter (SD) were collected. VNT prevalence was 28.5%. In comparison, patients with VNT included higher LSM, INR, and SD and lower PLT. Gastroscopies were spared for 27.7% of patients using the B6C with 1.0% VNT missed rate, 47.2% of patients using the EB6C with 5.9% VNT missed rate, 57.6% of patients using the LSPS with 9.9% VNT missed rate, and 45.5% of patients using the VariScreen algorithm with 3.0% VNT missed rate. Only the B6C and the VariScreen algorithm could safely avoid gastroscopies, and the VariScreen algorithm spared more gastroscopies than the B6C. The results were consistent with the previous when performed subgroup analysis. In conclusion, the VariScreen algorithm performed the best and can be used in clinical.

Impact of Intrahepatic Venovenous Shunt on Hepatic Venous Pressure Gradient Measurement

PURPOSE

To quantitatively analyze the impact of intrahepatic venovenous shunt (IHVS) on hepatic venous pressure gradient (HVPG) measurement.

MATERIALS AND METHODS

From 2015 to 2019, 222 HVPG measurements performed during transjugular intrahepatic portosystemic shunt creation were eligible for this study. Digital subtraction angiography (DSA) software color-coded each pixel of a two-dimensional DSA series by time-intensity curve to classify IHVS. Different degrees of IHVS were found in 36.5% of patients (81/222). Mild IHVS was found in 10.8% of patients (24/222), moderate IHVS was found in 10.8% of patients (24/222), and severe IVHS was found in 14.9% of patients (33/222).

RESULTS

Mean wedged hepatic vein pressure (WHVP) and HVPG were significantly lower in patients with IHVS compared with patients without IHVS (WHVP: 17.78 mm Hg ± 7.00 vs 24.89 mm Hg ± 8.69, P = .001; HVPG: 11.93 mm Hg ± 5.76 vs 18.6 mm Hg ± 6.85, P < .001). Mild IHVS had little effect on WHVP and HVPG. Mean WHVP and HVPG were 11 mm Hg lower in patients with moderate IHVS (WHVP: 20.38 mm Hg ± 8.38 vs 31.5 mm Hg ± 9.39, P = .026; HVPG: 13.88 mm Hg ± 6.33 vs 25.00 mm Hg ± 9.81, P < .001) and 15 mm Hg lower in patients with severe IHVS (WHVP: 13.45 mm Hg ± 5.28 vs 28.64 mm Hg ± 6.38, P = .017; HVPG: 8.27 mm Hg ± 3.85 vs 23.45 mm Hg ± 6.95, P < .001) than mean portal vein pressure and portal vein gradient.

CONCLUSIONS

For patients with moderate or severe IHVS, HVPG might greatly underestimate the actual value of portal vein pressure, and the portal vein should be catheterized to measure portal pressure.

Compensated Cirrhosis and 20 mm Hg: A Point of No Return?

This editorial comments on the findings from the study by Jindal et al. recently published in The American Journal of Gastroenterology. Weaknesses and strengths of the study are presented, and potential therapeutic strategies for compensated cirrhotic patients with an HVPG ≥20 mm Hg are proposed.

Thromboelastometry in patients with advanced chronic liver disease stratified by severity of portal hypertension

Background

Rotational thromboelastometry (ROTEM) has been studied in patients with advanced chronic liver disease (ACLD) without considering the impact of portal hypertension. We evaluated the influence of the hepatic venous pressure gradient (HVPG) on ROTEM results in patients with ACLD.

Methods

Cross-sectional study; ACLD patients undergoing HVPG measurement within the prospective Vienna Cirrhosis Study (NCT03267615) underwent concomitant ROTEM testing.

Results

Among 159 patients (68% male; Child–Pugh-A: 53%, Child–Pugh-B: 34%, Child–Pugh-C: 13%), 21 patients (13%) had a HVPG between 6 and 10 mmHg, 84 patients (53%) between 10 and 19 mmHg, and 54 patients (34%) ≥ 20 mmHg. Child–Pugh-C patients (vs. Child–Pugh-A and vs. Child–Pugh-B patients, respectively) showed longer clot formation time (CFT: median 187 s vs. 122 s vs. 122 s, p = 0.007) and lower maximum clot firmness (MCF: median: 45 mm vs. 56 mm vs. 56 mm, p = 0.002) in extrinsic thromboelastometry (EXTEM), while platelet counts were similar across Child–Pugh stages. In the overall cohort, ROTEM parameters did not differ by severity of portal hypertension. However, among compensated Child–Pugh-A patients, MCF decreased with increasing portal pressure, i.e. in higher HVPG strata (HVPG 9–10 mmHg: median MCF: 59 mm vs. HVPG 10–19 mmHg: 56 mm vs HVPG ≥ 20 mmHg: 54 mm, p = 0.023). Furthermore, patients with short CFT and high MCF in EXTEM had higher levels of lipopolysaccharide-binding protein, C-reactive protein, and procalcitonin, as well as higher leukocyte counts (all p < 0.05).

Conclusions

Portal hypertension seems to impact ROTEM results only in compensated Child–Pugh-A patients. Bacterial translocation and systemic inflammation may trigger a procoagulant state in patients with ACLD.

Electronic supplementary material

The online version of this article (10.1007/s12072-020-10093-3) contains supplementary material, which is available to authorized users.