Copper in PSVD: When Having the Tool Doesn't Mean Necessarily Using It

Abstract

We have read with great interest the paper by Balcar L, Dominik N, et al., titled ‘Elevated Hepatic Copper Content in Porto-Sinusoidal Vascular Disorder (PSVD): Leading Down a Wrong Track.’ The study evaluates copper liver content in patients with PSVD and intriguingly reports elevated hepatic copper levels, which were associated with a higher degree of decompensation. However, the study sample was poorly characterised, and the rationale for ruling out Wilson disease (WD) as a differential diagnosis remains particularly unclear.

This work could be considered a long-term addendum into Peter Ferenci's legacy manuscript from 2005 [1] performed by the same group in Vienna 20 years ago. In that classic work, 359 liver biopsies from 114 patients WD, 219 non-cholestatic liver disease and 22 healthy controls had their intrahepatic copper measurement performed and compared, in an attempt to define intrahepatic copper value as a diagnostic tool for WD. Their main figure on that article could be complemented by the results from the present study (see adapted Figure 1), in which 92 patients with a known diagnosis of PSVD, had their intrahepatic copper content systematically evaluated as well. Despite the median copper values were low, as expected for a non-WD cohort (30 μg/g dry weight, IQR: 18–55), four patients (4.3%) presented with very increased intrahepatic copper levels (over the stablished cutoff for high suspicion of WD, ≥ 250 μg/g) and 29 (32%) with intermediate levels above normality (≥ 50 μg/g but < 250 μg/g), leading the authors to suggest PSVD should be included in the differential diagnosis whenever a patient presents with high copper levels in hepatic tissue. The opposite lecture would be that high levels of intrahepatic copper content would suggest PSVD in the absence of WD confirmation, being copper a negative predictor for worse prognosis. A third lecture would account for the magnified value of potential outliers in this scenario. And a forth lecture would really question the need for such a measurement, only explained by unlimited access to this tool irrespectively of clinical suspicion in a highly experienced centre for WD.

Adapted Figure 1 from Ferenci et al. [1] (figure 2 in the original manuscript).

A thorough clinical history would likely negate the need for measuring copper content in the majority of patients. In most cases, the diagnosis of WD could be excluded through ceruloplasmin levels and 24-h urinary copper measurements, at least in symptomatic patients with advanced liver disease [2, 3], like those from the present work. In fact, the need for a liver biopsy (LB) in this clinical scenario of WD suspicion would be minimal for different reasons: (1) patients with WD have no pathognomonic signs on regular histology and often show ‘common signs of cirrhosis’ whenever cirrhosis is clinically evident [4]—minimising the potential confusion between these two entities (WD and PSVD); (2) copper is heterogeneously deposited in the liver, and this is especially true among patients with severely architectural changes and higher degrees of fibrosis; (3) hepatic parenchymal copper evaluation has inherent limitations and requires adequate sizes of liver biopsy, which might be a problem among patients with more advanced liver disease; and (4) in the era of improved gene testing, ATP7B sequencing would be the highly required next step to ensure diagnosis. On a different perspective, whether copper should be routinely measured in all patients undergoing liver biopsy for diagnostic purposes (regardless of a WD suspicion or not) remains debatable, especially in regular clinical practice, where such tests may not be readily available in many centres worldwide. This underscores the enduring value of traditional medical approaches—relying on clinical manifestations, physical examination and basic laboratory tests—often sufficient for an accurate diagnosis.

After dietary absorption, copper accumulates in the liver whenever the excretion biliary pathways are impaired—either primarily (WD) or due to second hits (biliary obstructions). In this cohort of patients with PSVD, it is particularly intriguing that four patients exhibited very high copper levels (> 250 μg/g) in an unexpected manner, all of whom had severe portal hypertension. The lack of detailed clinical, imaging and laboratory data makes it difficult to assess the accuracy of the PSVD diagnosis, particularly in two of the four cases where PSVD was diagnosed on the basis of the presence of incomplete septal fibrosis in a biopsy—an indicator of PSVD that requires explant evaluation [5, 6]. This raises doubts about the diagnosis in at least these two patients. For the adequate interpretation of these increased hepatic copper levels, the authors excluded WD by ATP7B gene sequencing and apparently also cholestasis in a multi-step approach by measuring either serum bile acids, MRI, LB and/or GGT/AP levels. However, although clinically apparent cholestasis was excluded, both GGT and PA markers were significantly increased among patients with intrahepatic copper > 50 μg/g as a group compared to those with normal copper levels, and therefore, this exclusion approach seems too rudimentary in modern times. The authors assume the elevation of copper content in their patients might be related to some kind of ‘cholestatic effect’ associated with PSVD, either detectable or not with common evaluations. But this exclusion approach would also have considered the exclusion of other recognised genetic cholestatic diseases by exome sequencing (MRD3 deficiency, PFIC 1 or 2, LPAC, among others) [7, 8]. Altogether, whether copper content in these patients was increased as a cause, or a consequence, or due to the presence of one or two mutations of unrecognised genetic disorders, or on the context of other underlying abnormalities contributing to copper accumulation, is pure speculation due to the retrospective nature of this study, and remains to be clarified.

In the era of improved gene testing and easy access to gene evaluation, the latest recommendations launched by AASLD [2] and EASL (pending publication shortly) for WD state that invasive procedures such as LB were no longer routinely required for establishing a diagnosis. It can be helpful, however, in patients with high WD suspicion due to regular inconclusive copper abnormalities and/or unavailable results in ATP7B sequencing. Indeed, in the recent years, new accurate biomarkers (such as the circulating non-ceruloplasmin bound copper fraction) [9, 10] have been developed for WD recognition, avoiding even more the need for LB in this scenario and overcoming its limitations. On the contrary, diagnosis of PSVD is based on LB with predefined specific and unspecific characteristics to consider. Systematic evaluation of parenchymal copper content performed in all these LB without a previous high suspicion for WD could have led the authors to magnify these observations. Unfortunately, genetic exclusion of WD was performed only among patients with the highest copper levels, whereas patients with intermediate copper accumulation (> 50 μg/g but less than 250 μg/g) were not genetically assessed (even though ceruloplasmin levels were significantly decreased in those without cholestasis). Indeed, individuals carrying a single ATP7B mutation (estimated to affect 1 in 90 individuals worldwide) [11] might harbour some copper abnormalities and intermediate hepatic copper levels over a normal threshold for healthy individuals [12], without developing apparent WD.

PSVD is a heterogeneous group of disorders that commonly affect the porto-sinusoidal endothelium, with its pathophysiology remaining poorly understood. Whether copper accumulation directly causes toxicity to endothelial cells in the hepatic microcirculation, disrupting blood flow and leading to localised ischemia and compensatory hepatocyte hyperplasia, promoting NRH, remains to be demonstrated. Another speculative mechanism may involve limited or dysfunctional copper excretion through the bile, or impaired copper excretion due to changes in the microvasculature. These exciting questions open new avenues for future research, but well-characterised cohorts and adequate methodologies will be desirable if we attempt to evaluate pathogenesis and/ or prognostic effects.

ZM received speaker fees from Orphalan; consultancy fees from Orphalan, Alexion, Deep Genomics, Prime Medicine; grants from Gilead. VH-G received speaker fees from Cook Medical and Gore Medical.