# Immunohistochemistry: Principles and Panel Selection

## Overview

Immunohistochemistry (IHC) uses antigen-antibody reactions to detect specific proteins in tissue sections. It is the most widely used ancillary technique in surgical pathology, serving as an essential tool for tumor classification, prognostic and predictive biomarker assessment, and identification of infectious organisms. Understanding the principles underlying IHC allows the pathologist to design rational panels, interpret results in context, and recognize pitfalls that can lead to diagnostic error.

## Basic Principles

### Antigen-Antibody Reaction

The target of an IHC stain is an antigen -- specifically, an epitope on a protein expressed in tissue. This protein may be localized to the cytoplasm, nucleus, cell membrane, or extracellular matrix. The primary antibody is selected to bind specifically to this target antigen. Monoclonal antibodies recognize a single epitope, providing high specificity with minimal background staining. Polyclonal antibodies recognize multiple epitopes on the same antigen, which increases sensitivity but introduces potential cross-reactivity with unrelated proteins.

### Detection Systems

The direct method, in which the primary antibody itself is labeled with a chromogen, is rarely used in diagnostic practice due to low sensitivity. The indirect method uses an unlabeled primary antibody followed by a labeled secondary antibody directed against the species of the primary. Modern diagnostic laboratories use polymer-based detection systems as the current standard. These systems conjugate horseradish peroxidase (HRP) or alkaline phosphatase (AP) to a polymer backbone that carries the secondary antibody, eliminating the biotin-related background staining that plagued older avidin-biotin methods while providing higher sensitivity and cleaner signal. The most common chromogen is DAB (diaminobenzidine), which produces a brown precipitate at the site of the antigen. AEC and Fast Red are alternative red chromogens. Hematoxylin counterstain provides blue nuclear contrast against which the brown or red signal is visualized.

### Antigen Retrieval

Formalin fixation preserves tissue morphology but creates methylene cross-links that mask epitopes, potentially rendering antigens inaccessible to antibodies. Antigen retrieval reverses this masking. Heat-induced epitope retrieval (HIER) uses citrate buffer at pH 6 or EDTA buffer at pH 9 in a pressure cooker or microwave to break cross-links and re-expose epitopes. Enzyme-induced epitope retrieval (EIER) uses proteases such as proteinase K or trypsin to digest cross-linked proteins. The optimal retrieval method is antigen-specific and must be validated for each antibody in the laboratory's specific workflow.

## Pre-Analytic Variables

### Fixation Effects

Fixation is the single most important pre-analytic variable in IHC. Underfixation results in weak or absent staining, particularly for nuclear antigens that require good penetration of fixative. Overfixation causes excessive cross-linking that can mask epitopes even with aggressive retrieval protocols. The optimal fixation window for most antigens is 6 to 48 hours in 10% neutral buffered formalin. Cold ischemia time -- the interval between devascularization and fixation -- degrades phospho-proteins and labile biomarkers such as ER and PR, making documentation of this interval clinically important.

### Tissue Processing

Standard paraffin embedding is compatible with the vast majority of IHC markers. Decalcified tissue presents a challenge: EDTA-based decalcification preserves antigenicity, while strong acid decalcifiers destroy many antigens and should be avoided when IHC is anticipated. Previously frozen tissue can be fixed and processed for IHC but may show freezing artifact. Cytology cell blocks require separate validation from tissue sections because the processing and fixation conditions differ.

### Section Quality

Sections should be cut at 3 to 5 micrometers on charged (adhesive) slides to prevent tissue loss during the staining process. Unstained sections should not be stored for extended periods, as antigens degrade over weeks to months on exposed slides. Cutting fresh sections immediately before staining produces optimal results.

## Controls

### Positive Controls

External positive controls consist of known positive tissue run in parallel with the patient's specimen -- for example, tonsil for CD20 or breast carcinoma for ER. Internal positive controls are normal tissue elements within the patient's own specimen that are expected to stain positively, such as background lymphocytes for LCA or endothelial cells for CD31. Internal controls are preferred when available because they confirm that the entire staining process worked on that specific patient's tissue under the same fixation and processing conditions.

### Negative Controls

Reagent negative controls are prepared by omitting the primary antibody and substituting non-immune serum or antibody diluent. Tissue negative controls use known negative tissue to confirm specificity. Excessive background staining on negative controls indicates problems with blocking steps, antibody concentration, or retrieval conditions that must be resolved before results can be reported.

## Interpretation Principles

### Staining Pattern Assessment

Interpretation begins with identifying the subcellular localization of staining: nuclear (as seen with ER, PR, Ki-67, and p53), cytoplasmic (cytokeratins, S100), membranous (HER2, CD3), or combinations thereof. Staining intensity is graded as 0 (negative), 1+ (weak), 2+ (moderate), or 3+ (strong). The distribution is characterized as diffuse, focal, patchy, or heterogeneous. For certain biomarkers, the percentage of positive cells is critical -- ER uses a 1% positivity threshold, while Ki-67 may be assessed as a hot-spot value or global average depending on the clinical context.

### Common Pitfalls

Edge artifact produces false-positive staining at tissue edges due to crush or poor fixation. Necrotic tissue shows non-specific staining that should not be interpreted as true positivity. Liver, kidney, and placenta have high endogenous biotin, which was particularly problematic with older avidin-biotin detection systems. Entrapped normal cells -- such as benign breast epithelium or reactive lymphocytes within a tumor mass -- can be misinterpreted as tumor staining if the pathologist is not carefully correlating IHC with morphology. Endogenous pigments including melanin, hemosiderin, and lipofuscin can mimic the brown color of DAB chromogen.

## Rational Panel Construction

### Approach to Unknown Primary

The diagnostic approach begins with morphology. The pathologist first determines whether the tumor is epithelial, mesenchymal, melanocytic, lymphoid, or neuroendocrine in appearance. A first-line panel of broad-spectrum cytokeratin (AE1/AE3 or OSCAR), S100, LCA (CD45), and vimentin establishes the lineage. For cytokeratin-positive tumors, CK7/CK20 status narrows the differential, followed by organ-specific markers such as TTF-1, PAX8, GATA3, CDX2, NKX3.1, p40, and napsin A. For spindle cell neoplasms, useful markers include SMA, desmin, CD34, SOX10, STAT6, and ERG. For round blue cell tumors in the pediatric/young adult population, the panel includes CD99, NKX2.2, FLI1, desmin, myogenin, TLE1, and WT1.

### CK7/CK20 Expression Patterns

The CK7/CK20 profile provides a powerful first branch point in the workup of carcinoma of unknown primary. CK7-positive/CK20-negative tumors include lung, breast, thyroid, ovarian, endometrial, and cholangiocarcinoma. CK7-negative/CK20-positive tumors include colorectal carcinoma and Merkel cell carcinoma. Dual-positive tumors (CK7+/CK20+) include urothelial, pancreatic, gastric, and mucinous ovarian carcinomas. Dual-negative tumors (CK7-/CK20-) include hepatocellular carcinoma, renal cell carcinoma, prostate carcinoma, and adrenocortical carcinoma.

| CK7 | CK20 | Tumor Types |
|------|------|-------------|
| + | - | Lung, breast, thyroid, ovarian, endometrial, cholangiocarcinoma |
| - | + | Colorectal carcinoma, Merkel cell carcinoma |
| + | + | Urothelial, pancreatic, gastric, mucinous ovarian |
| - | - | Hepatocellular, renal cell, prostate, adrenocortical |

### Organ-Specific Markers

Lung adenocarcinoma is identified by TTF-1 and napsin A, while lung squamous carcinoma expresses p40 and CK5/6. Breast carcinoma markers include GATA3, mammaglobin, GCDFP-15, ER, and PR. Thyroid markers are TTF-1, PAX8, and thyroglobulin. Gastrointestinal markers include CDX2 and SATB2 for colorectal origin, and HepPar-1 and arginase for hepatocellular carcinoma. Renal markers include PAX8, CA-IX (for clear cell RCC), and CD117 (for chromophobe RCC). Prostate markers are NKX3.1, PSA, PSAP, and ERG. Urothelial carcinoma expresses GATA3, uroplakin, p63, and CK20. Gynecologic tumors express PAX8, WT1 (particularly serous carcinomas), ER, PR, and p16.

| Primary Site | Key IHC Markers |
|---|---|
| Lung adenocarcinoma | TTF-1, napsin A |
| Lung squamous | p40, CK5/6 |
| Breast | GATA3, mammaglobin, GCDFP-15, ER, PR |
| Thyroid | TTF-1, PAX8, thyroglobulin |
| Colorectal | CDX2, SATB2 |
| Hepatocellular | HepPar-1, arginase |
| Renal (clear cell) | PAX8, CA-IX |
| Renal (chromophobe) | PAX8, CD117 |
| Prostate | NKX3.1, PSA, PSAP, ERG |
| Urothelial | GATA3, uroplakin, p63, CK20 |
| Gynecologic (serous) | PAX8, WT1, ER, PR, p16 |

### Melanocytic Panel

S100 is the most sensitive melanocytic marker but lacks specificity. SOX10 is a nuclear stain that is sensitive for both melanoma and nerve sheath tumors. HMB-45 is cytoplasmic and more specific than S100 but less sensitive, with a characteristic pattern of decreasing expression toward the base of melanocytic lesions. Melan-A/MART-1 is sensitive for melanoma but also stains adrenal cortex and some steroid-producing tumors. MITF is a nuclear marker that can be positive in some carcinomas and macrophages, limiting its specificity.

### Neuroendocrine Markers

Synaptophysin is the most sensitive general neuroendocrine marker. Chromogranin A is more specific but less sensitive and may be negative in some high-grade neuroendocrine carcinomas that have lost their secretory granules. INSM1 is a nuclear marker that offers both sensitivity and specificity for neuroendocrine differentiation. CD56 is sensitive for neuroendocrine tumors but lacks specificity, staining many non-neuroendocrine neoplasms.

## Predictive and Prognostic Biomarkers

### Breast Cancer

ER and PR are assessed as nuclear staining and reported as positive if 1% or more of tumor nuclei show staining, using either the Allred scoring system or H-score. HER2 is scored based on membranous staining intensity and completeness according to ASCO/CAP guidelines: 0 indicates no staining or incomplete faint staining in 10% or fewer cells; 1+ shows incomplete faint staining in more than 10% (now recognized as "HER2-low"); 2+ shows weak-to-moderate complete staining in more than 10% or strong complete staining in 10% or fewer cells (equivocal, requiring reflex FISH testing); and 3+ shows strong complete membranous staining in more than 10% of cells (positive). Ki-67 is a proliferation marker used in scoring systems analogous to Oncotype DX, though the optimal cutoff remains debated, varying between 14% and 30% across different clinical contexts.

### Lung Cancer

PD-L1 assessment uses the tumor proportion score (TPS), with clinically significant cutoffs at 1% and 50%. The 22C3 clone on the Dako platform is validated for pembrolizumab, while the SP263 clone on the Ventana platform is validated for durvalumab. Different antibody clones are not interchangeable without formal laboratory validation, as they may produce discordant results.

### Mismatch Repair (MMR) / Microsatellite Instability

The MMR panel consists of four markers: MLH1, PMS2, MSH2, and MSH6. Loss of nuclear staining in tumor cells -- with retained staining in internal control cells such as lymphocytes and stromal cells -- indicates mismatch repair deficiency. Loss of MLH1 and PMS2 together may reflect either sporadic MLH1 promoter hypermethylation or germline mutation (Lynch syndrome), and further workup with MLH1 methylation testing or BRAF V600E testing is needed. Loss of MSH2 and MSH6 strongly suggests Lynch syndrome. Universal MMR screening is now recommended for all colorectal and endometrial carcinomas regardless of patient age or clinical features.

| Pattern of Loss | Interpretation | Next Step |
|---|---|---|
| MLH1 + PMS2 lost | Sporadic hypermethylation or Lynch syndrome | MLH1 methylation / BRAF V600E testing |
| MSH2 + MSH6 lost | Strongly suggests Lynch syndrome | Germline testing |
| PMS2 lost alone | PMS2 germline mutation | Germline testing |
| MSH6 lost alone | MSH6 germline mutation | Germline testing |

<image>A medical illustration showing the indirect immunohistochemistry detection method in a step-by-step diagram. Step 1: Tissue section on slide with target antigen on tumor cell. Step 2: Primary antibody (mouse anti-human) binding to the antigen. Step 3: Polymer-conjugated secondary antibody (anti-mouse) binding to the primary antibody, with multiple HRP enzyme molecules attached to the polymer backbone. Step 4: DAB substrate converted to brown precipitate by HRP at the antigen site. Step 5: Hematoxylin counterstain applied. Final result shows brown-stained target cells against a blue nuclear background. Each step is labeled with reagent names and incubation times.</image>

<image>A composite medical illustration showing four different IHC staining patterns. Panel A: Nuclear staining pattern (ER in breast carcinoma) with dark brown nuclei in tumor cells and clear cytoplasm. Panel B: Membranous staining pattern (HER2 3+ in breast carcinoma) showing crisp, complete brown rings around cell membranes. Panel C: Cytoplasmic staining pattern (chromogranin in a neuroendocrine tumor) with diffuse brown cytoplasmic granularity and unstained nuclei. Panel D: Combined nuclear and cytoplasmic staining (S100 in melanoma) showing both brown nuclei and cytoplasm. Each panel is clearly labeled with the marker name and staining pattern type.</image>

<image>A diagnostic algorithm flowchart for IHC workup of carcinoma of unknown primary. Starting with "Carcinoma confirmed by morphology and broad CK positivity," the chart branches based on CK7/CK20 status into four quadrants. Each quadrant leads to second-line markers: CK7+/CK20- branches to TTF-1, PAX8, GATA3, ER; CK7-/CK20+ branches to CDX2, SATB2; CK7+/CK20+ branches to GATA3, CDX2, PAX8; CK7-/CK20- branches to HepPar-1, NKX3.1, RCC marker, inhibin. Terminal boxes list the most likely primary sites for each marker combination.</image>

## Clinical Pearls

IHC must always be interpreted in the context of morphology -- it confirms or refines a morphologic differential diagnosis but never replaces careful histologic assessment. A rational panel of 4 to 6 markers is usually sufficient; avoid "shotgun" panels of 15 or more markers that increase cost, consume tissue, and complicate interpretation. Internal positive controls are more reliable than external controls because they confirm the stain worked on the patient's actual tissue under the same fixation conditions. Loss of staining can be as diagnostically important as positive staining, as seen with loss of INI1/SMARCB1, BAP1, ARID1A, or MMR proteins. For p53 interpretation, wild-type tissue shows variable weak-to-moderate staining; mutant patterns include either diffuse strong overexpression or complete absence (null pattern), and both are abnormal. HER2-low (1+ or 2+ with FISH non-amplification) is now clinically actionable with trastuzumab deruxtecan, making the distinction between a true 0 and a 1+ score critically important. GATA3 is positive in both breast and urothelial carcinoma, so additional markers such as mammaglobin or uroplakin are needed to distinguish between these two primaries. When IHC results are discordant with morphology, the stain should be repeated, controls checked, and pre-analytic variables reviewed before the result is accepted.

## References
- Dabbs DJ. *Diagnostic Immunohistochemistry: Theranostic and Genomic Applications*. 6th ed. Elsevier; 2021.
- Taylor CR, Rudbeck L. *Immunohistochemical Staining Methods*. 6th ed. Dako/Agilent; 2013.
- Wolff AC, et al. HER2 testing in breast cancer: ASCO/CAP clinical practice guideline focused update. *J Clin Oncol*. 2018;36(20):2105-2122.
- Allison KH, et al. Estrogen and progesterone receptor testing in breast cancer: ASCO/CAP guideline update. *J Clin Oncol*. 2020;38(12):1346-1366.
- Lin F, Prichard J. *Handbook of Practical Immunohistochemistry*. 3rd ed. Springer; 2023.