Diagnostic Yield and Cost of Morphology-Guided Immunohistochemistry in Liver Tumors
Cheep Charoenlap1, Pakorn Arunsawat1, Tanaporn Prateepchaiboon2
1Department of Anatomical Pathology, Hatyai Hospital, Songkhla, Thailand.
Background:
In many routine pathology settings, particularly where resources are constrained, liver tumors are evaluated with an individualized, morphology-guided immunohistochemistry (IHC) strategy rather than fixed panels or molecular assays. Despite widespread use, the real-world diagnostic yield, stain utilization, and laboratory cost of this approach are not well defined.
Methods:
We retrospectively reviewed consecutive malignant liver tumor diagnoses at a university-affiliated hospital from January 2020 to December 2023. Cases were classified as primary liver cancers or metastatic tumors. IHC was performed using a morphology-guided, case-based sequence. Primary outcomes were histologic subtype assignment among primary liver cancers and determination of the primary site among liver metastases. Secondary outcomes were IHC stain utilization and laboratory cost per case.
Results:
Among 279 malignant liver tumors, 144 (51.6%) were primary and 135 (48.4%) were metastatic. Hepatocellular carcinoma accounted for 80.6% of primary liver cancers. Among metastases, a likely primary site was identified in most cases; colorectal (31.1%), lung (9.6%), and pancreatic (8.1%) primaries were most common, while cancers of unknown primary comprised 14.8%. IHC utilization was lower in primary tumors than in metastases (median 2 stains/case [IQR 2-4] vs 4 stains/case [IQR 2-7]; p<0.001). Correspondingly, median laboratory IHC cost per case was lower for primary tumors than for metastases (USD 29.90 [IQR 21.73-51.00] vs USD 46.92 [IQR 21.73-92.57]; p<0.001).
Conclusions:
An individualized, morphology-guided IHC strategy enabled reliable subtyping of primary liver cancers and identification of likely primary sites for most liver metastases while maintaining compact stain utilization and moderate laboratory costs. These data provide practical, real-world benchmarks for diagnostic performance and resource use and support targeted, hypothesis-driven IHC where broad fixed panels or molecular diagnostics are not routinely available.
