Second Harmonic Generation Reveals Subtle Fibrosis Differences in Adult and Pediatric Nonalcoholic Fatty Liver Disease

Feng Liu, MD; Jing-Min Zhao, MD; Hui-Ying Rao, MD; Wei-Miao Yu, MD; Wei Zhang, MD; Neil D. Theise, MD; Aileen Wee, MBBS, FRCPath, FRCPA; and Lai Wei, MD


Am J Clin Pathol. 2017;148(6):502-512. 

In This Article

Abstract and Introduction


Objectives. Investigate subtle fibrosis similarities and differences in adult and pediatric nonalcoholic fatty liver disease (NAFLD) using second harmonic generation (SHG).

Methods. SHG/two-photon excitation fluorescence imaging quantified 100 collagen parameters and determined qFibrosis values by using the nonalcoholic steatohepatitis (NASH) Clinical Research Network (CRN) scoring system in 62 adult and 36 pediatric NAFLD liver specimens.

Results. Six distinct parameters identified differences among the NASH CRN stages with high accuracy (area under the curve, 0835–0.982 vs 0.885–0.981, adult and pediatric). All portal region parameters showed similar changes across early stages 0, 1C, and 2, in both groups. Parameter values decreased in adults with progression from stage 1A/B to 2 in the central vein region. In children, aggregated collagen parameters decreased, but nearly all distributed collagen parameters increased from stage 1A/B to 2.

Conclusions. SHG analysis accurately reproduces NASH CRN staging in NAFLD, as well as reveals differences and similarities between adult and pediatric collagen deposition not captured by currently available quantitative methods.


Nonalcoholic fatty liver disease (NAFLD) is a clinical syndrome predicted to be the next global epidemic affecting millions of people worldwide.[1] The prevalence of NAFLD in different regions of China ranges from 6.3% to 27.0% (median, 15%). The prevalence of NAFLD in children in China is 2.1%, although the prevalence increases to 68.2% among obese children.[2] Compared with adults, the prevalence of NAFLD doubled among US adolescents from 3.9% in 1988 to 1994 to 10.7% in 2007 to 2010.[3] The natural course of this disease, including its subtype, nonalcoholic steatohepatitis (NASH), is not clearly defined.[4] Complicating such definition is that while pediatric and adult populations are both affected, the disease is histologically distinctive in each, with differences in the degree of steatosis, qualities and location of histologic steatohepatitis, and patterns of steatofibrosis.[5]

As with the epidemics of hepatitis B and hepatitis C before it, liver biopsy plays an increasing role in documenting the severity of disease in affected individuals, elucidating mechanisms of injury, and evaluating responses to treatment.[5] In the absence of widely effective treatments, the careful monitoring for progression is likely to require, in some patients, repetitive biopsies over time requiring careful comparison. Moreover, evaluation of responses to therapy in clinical trials and, when therapies are standardized, in individual patients will require detailed analysis of histologic features in liver biopsy specimens (LBx), particularly the patterns and degree of scarring and possible features of regression of fibrosis.[6] For such purposes, semiquantitative staging systems have been developed,[7–11] all of which have pros and cons for any given clinical or investigational setting.[5] Some noninvasive technologies that assess stiffness and/or fibrosis have been developed, including transient elastography, magnetic resonance elastography, and acoustic radiation force impulse imaging, but standards in these tests were developed in adults and may not be applicable to pediatric NAFLD.[12] For some of these tests, steatosis itself is a confounding factor, increasing the stiffness of the liver even in the absence of scar.[13,14] It was found that children with portal-based NASH have more severe fibrosis.[15] Therefore, these noninvasive technologies are inaccurate for fibrosis quantification and insufficient for clinical trials.

Also, compared with patterns of scar in LBx of chronic viral hepatitis, there is a greater variety of patterns, distributions, and locations of scarring in NAFLD LBx. In viral hepatitis, there is a comparatively simple progression from portal fibrosis to portal-portal (sometimes including an intervening vein) fibrous septa, the emergence of parenchymal nodularity, and, finally, advanced/end-stage liver disease ("cirrhosis").[16] On the other hand, in NAFLD, there is considerably more variation: sclerosis of central veins (CVs), stellate fibrosis of portal tracts (PTs), perisinusoidal "chicken wire" fibrosis in varied regions of parenchyma (usually directional from acinus zone 3 to 1), compacted central-central and central-portal fibrous septa, and micronodular to macronodular cirrhosis;[6–10] moreover, the patterns in adults and children are usually different.[5] Thus, the semiquantitative staging systems for viral hepatitis had a somewhat simpler set of patterns to document; the complexity of NAFLD-associated scarring is not entirely captured by currently suggested methods.

Likewise, the relative complexity of NAFLD-associated scar probably reflects a more complex underlying pattern of cellular and molecular events. While most if not all scarring in chronic viral hepatitis depends on mechanisms and patterns of parenchymal extinction,[17] the much more finely fibrillar fibrosis of the classic "chicken wire" perisinusoidal fibrosis of NAFLD is a clearly different process. Furthermore, patterns of regression of fibrosis are also different. While advanced stage viral hepatitis shows a fairly limited set of features, described as the "hepatic repair complex,"[18] patterns and mechanisms of regression of perisinusoidal fibrosis are probably different. These have not yet been greatly detailed in NAFLD. This deficiency is likely to become decisively limiting when it comes to assessment of responses to antifibrotic agents for treatment of NAFLD. Different agents are likely to affect different kinds of scar, from the delicate deposition in the space of Disse to zonally spreading "chicken wire" fibrosis to established fibrous septa. Thus, there is a need for more detailed assessment of changes in scar than is provided by semiquantitative assessments.

Recently, a quantitative approach to detailed assessment of fibrosis has been applied to LBx in chronic viral hepatitis and primary biliary cholangitis.[19–21] Wang et al[22] reported that quantitative fibrosis parameters (q-FPs) could be valid for the assessment of progressing fibrosis in NAFLD by using the system of second harmonic generation/two-photon excitation fluorescence (SHG/TPEF) microscopy. However, the fine detail of collagen distribution in pediatric NAFLD remains unstudied. Therefore, we have applied similar techniques to investigate pediatric NAFLD and compared these findings with the features of adult NAFLD LBx. LBx in NAFLD are imaged by the system of SHG/TPEF microscopy established and adjusted as previously reported.[23] With input of imaging data from the liver sample, quantitative fibrosis (qFibrosis) can automatically compute the fully quantitative fibrosis scores based on the respective collagen architectural features. More important, it can perform in-depth analysis of features such as collagen distribution (eg, aggregated vs distributed), morphology (eg, width, length, perimeter, area, and orientation), and location (eg, portal vs central vs perisinusoidal). Such a strategy potentially overcomes some limitations of semiquantitative histologic fibrosis assessment and can thereby provide a level of detail necessary for fine-grained, thorough assessment of responses to antifibrotic therapies. Here we report the development of NAFLD-qFibrosis, verify its potential as a fibrosis assessment tool, and reveal differences between adult and pediatric patients with NAFLD in a Chinese cohort of patients.