Abstract
Regression of liver fibrosis could be confirmed in patients with virologic response to IFN even in patients with liver cirrhosis. Our quantitative evaluation provided information superior to that obtained by the classical fibrosis staging score system.
Introduction
In clinical hepatology, the "gold standard" for assessing liver damage is the fibrosis staging system (F0 to F4). While useful, this system is inherently limited: it is a discrete, subjective 5-point scale based on a pathologist's visual assessment of tissue architecture. The authors identify a critical gap: the F-scale lacks the granularity to detect subtle changes in the density or volume of fibrous tissue. If a patient’s fibrosis reduces by 20%, they might still appear as an F4 (cirrhosis) to a human eye, leading to the false conclusion that treatment had no effect. This study develops a precise, objective method using digital image analysis to track the actual "fibrosis area" over time.
Methods
The researchers performed paired liver biopsies (before and after treatment) on 25 patients with chronic Hepatitis C. The cohort was divided into a Complete Response (CR) group (20 patients who eradicated the virus via Interferon therapy) and a Conservative Therapy (CT) group (5 patients who did not receive Interferon).
The technical pipeline for quantification involved:
- Staining: Tissue samples were treated with Azan-Mallory (AM) stain to color collagen blue and hepatocytes red.
- Digitization: Samples were captured via light microscope and stitched into whole-tissue images.
- Segmentation: Using Adobe Photoshop, images were converted into binary "two-toned" images to isolate blue-stained fibrous areas.
- Quantification: NIH Image software calculated the "relative area of fibrosis"—the ratio of scar tissue pixels to total sample pixels.
Key Results
The study utilized 50 total liver biopsies with an average interval of 37–57 months between samples. The quantitative results revealed a stark contrast between the treatment groups:
- CR Group (Successful Treatment): The mean fibrosis area dropped significantly from 7.2% ± 1.5% to 2.7% ± 0.5% (p=0.0001). Notably, 19 out of 20 patients showed a decrease in fibrosis area.
- CT Group (Untreated): The mean fibrosis area nearly doubled, increasing from 8.4% ± 4.3% to 15.9% ± 7.7%.
- Sensitivity: In the CR group, only 11 of 20 patients showed improvement in their traditional F-score, yet digital analysis confirmed regression in 19 of 20, proving the quantitative method's superior sensitivity.
Discussion and Insights
The most significant insight is that digital area measurement provides a "higher resolution" lens than traditional staging. In several cases where a pathologist saw "no change" in the F-score, digital analysis revealed a significant reduction in the actual density of the fibrous septa.
Furthermore, contrary to older medical dogmas, the study found that even patients at the F4 (cirrhosis) stage showed significant regression of scar tissue once the viral trigger (HCV) was removed. This suggests that liver cirrhosis—long considered an irreversible progression—can actually regress with successful virologic clearance.
Limitations and Future Work
The authors acknowledge that with only 25 patients, this study is a small-scale proof of concept. Additionally, the manual steps in Photoshop for area selection introduce potential human bias, and "sampling error" remains a risk inherent to liver biopsies. Future work should focus on more automated, objective criteria to further refine the digital analysis of liver architecture and evaluate new anti-fibrotic drugs.
Summary updated