Comparison of Automated and Conventional IHC Visual Scoring Analysis for MHC Class I and Tapasin Expression in Cervical Carcinoma
Surasak Wanram PhD*, Jakkra Sombatwong BSc**, Anun Sakunpong BSc**, Prinya Prasongdee PhD**, Pawana Panomket PhD*, Parichart Wongsena MD*, Temduang Limpaiboon PhD**, Patcharee Jearanaikoon PhD**
Affiliation : * Biomedical Sciences Research Unit, College of Medicine and Public Health, Ubon Ratchathani University, Ubon Ratchathani, Thailand ** Centre for Medical Diagnostic Laboratory, Faculty of Associated Medical Sciences, Khon Kaen University, Khon Kaen, Thailand
Background : Cervical cancer (CXCA) is the second most common cancer among women in Thailand and worldwide.
Immune evasion caused by down-regulation of host immune responsive genes, such as MHC class I and loss of antigen
processing machinery (APM), presents a capability leading to cancer development. Immunohistochemical staining (IHC) is
regarded as a common technique for protein marker detection in clinical laboratories. At present, IHC automation has been
launched to facilitate the speed and feasibility to replace conventional IHC. However, evaluation of its use is still limited.
Objective : This study aimed to evaluate IHC scoring by automated visual analysis compared to conventional IHC analysis.
Material and Method: The paraffin-embedded tissues of 96 invasive CXCA were processed using a tissue microarray (TMA)
platform followed by automated IHC staining of the anti-MHC class I (heavy chain, β2M) and an APM-Tapasin expression.
Conventional IHC and automated slide scanning with scoring visual analysis were compared.
Results : The results showed significant association between conventional and automated IHC evaluation (p-value >0.05,
Chi-square) for MHC class I and Tapasin stated in percentage of positive cancer cells, whereas intensity was found (p-value
<0.05, Chi-square) with moderate agreement (p-value <0.001, kappa) 0.434-0.615 and 0.353-0.554, respectively. After
calculated values, the results showed significant association between conventional and automated IHC evaluation (p-value
>0.05, Chi-square) for MHC class I and Tapasin with the highest agreement level (p-value <0.001, kappa) of summation
0.595-0.755 and multiply scoring 0.633-0.689, respectively.
Conclusion and Discussion: The automation software for IHC scoring and interpretation can be used for the determination
of MHC class I and Tapasin in CXCA. In addition, an antigen presentation pattern must be included to allow an accurate
result for MHC class I in clinical use. An appropriate sample size and design of staging coverage as well as clinical prognosis
outcomes of progression should be used in further investigation.
Keywords : Automated immunohistochemistry, Automated visual analysis, Automated scoring, Cervical cancer, MHC class I, Tapasin
