Immunohistochemistry

Immunohistochemistry (IHC) uses labelled antibodies to find specific proteins in tissue sections. The antibody binds to its target in the tissue, and a chemical reaction produces a coloured deposit (usually brown with diaminobenzidine) at the site, which is seen under the light microscope and related to the cell structure.

It tells the pathologist what a cell is (its lineage), where it comes from, whether it carries a treatment target, and how fast it is dividing. It answers questions that routine staining cannot, and it adds to the morphology without replacing it.

Tissue is fixed and embedded in paraffin, sectioned, and then treated for antigen retrieval (heat or enzyme) to unmask epitopes. The primary antibody is applied, followed by a detection system, a chromogen and a counterstain. Automated staining platforms give reproducible results, and the test takes 1 to 2 days.

IHC is used in a large part of oncology diagnosis. A panel of antibodies, selected from the morphology and clinical information, is far more informative than a single marker, and no marker is perfectly sensitive and specific for any tumour.

Indications.

• Tumour classification: carcinoma versus lymphoma versus melanoma versus sarcoma (for example, cytokeratin, CD45, S100 and SOX10, desmin).
• Metastasis of unknown primary: patterns such as CK7, CK20, TTF-1, CDX2, GATA3, PAX8, PSA and others suggest the likely origin.
• Lung cancer subtyping (TTF-1, napsin A, p40, p63) and neuroendocrine markers (synaptophysin, chromogranin, INSM1).
• Lymphoma and leukaemia classification (CD3, CD20, CD30, CD15, CD10, BCL2, BCL6, cyclin D1, ALK, Ki-67).
• Predictive and prognostic markers: oestrogen and progesterone receptors, HER2, PD-L1, Ki-67, and mismatch repair proteins (MLH1, MSH2, MSH6, PMS2).
• Distinguishing in situ from invasive disease (p63 and smooth muscle markers around ducts).
• Sarcoma and soft tissue tumours (CD34, STAT6, SMA, desmin, MDM2, H3K27me3, β-catenin).
• Infection and viral markers: HHV-8 in Kaposi sarcoma, CMV, HSV, hepatitis B antigens, and some organisms.
• Deposits: amyloid typing, immunoglobulin light chains, IgG4.
• Dermatopathology, renal and neuropathology diagnosis.

Not indicated when morphology is diagnostic and extra tests would not change treatment, or when material is too small and needed for molecular tests.

Specimen requirements.

• Formalin fixed, paraffin embedded tissue or cell block, fixed within 1 hour of removal (cold ischaemia time) and for 6 to 72 hours for predictive markers.
• Unstained sections cut at 3 to 5 microm on charged slides.
• Decalcified tissue: strong acids damage antigens, so EDTA is preferred.
• Frozen tissue or special fixatives for some markers.
• Clinical information, morphological differential and the question to be answered.
• Appropriate positive and negative controls on each run.

Normal appearances.

Normal tissue shows the expected pattern for each antibody: for example, nuclear staining of oestrogen receptor in normal breast luminal cells, membranous staining of CD20 in B cells, and cytoplasmic cytokeratin in epithelial cells. Internal controls, such as the stroma and non-neoplastic cells, confirm that the stain worked.

Interpretation.

• Staining is scored by its location (nuclear, cytoplasmic, membranous), intensity (weak, moderate, strong) and proportion of positive cells.
• Hormone receptors: positive when 1% or more of tumour nuclei stain, with the proportion and intensity reported. Positive tumours are treated with endocrine therapy.
• HER2: scored 0 to 3+ by membrane staining, with 3+ positive, 2+ equivocal (needing in situ hybridisation), and 0 and 1+ negative.
• PD-L1: reported as a proportion score (tumour proportion score or combined positive score), with cut offs specific to the antibody clone, tumour type and drug.
• Mismatch repair proteins: loss of nuclear staining in tumour cells (with internal control staining) indicates mismatch repair deficiency, suggestive of Lynch syndrome or sporadic microsatellite instability.
• Ki-67: proliferation index as a percentage of positive nuclei, used in grading and prognosis.
• Lineage panels: the combination, for example CK7 positive and CK20 negative with TTF-1 positive suggests lung adenocarcinoma, and CK7 negative with CK20 and CDX2 positive suggests colorectal origin.
• A negative result is valid only when internal and external controls are positive.
• Aberrant results (unexpected positivity or negativity) are reconciled with morphology and further markers.

Critical values and findings needing immediate action.

• Findings that change treatment urgently (HER2 positivity, mismatch repair deficiency, PD-L1 high expression, CD30 or ALK positivity): reported at once to the team.
• A diagnosis of a lymphoma or an aggressive tumour in a patient with an emergency presentation (airway compression, cord compression): urgent contact with the clinician.
• Loss of mismatch repair proteins: referral for germline testing and counselling.

Limitations and artefacts.

• Pre-analytical factors: delayed or under fixation, prolonged fixation, and decalcification cause false negative or weak staining, especially for hormone receptors and HER2.
• Antigen retrieval, antibody clone, dilution and detection system change results, so validated procedures and controls are essential.
• False positives: endogenous biotin (liver, kidney), peroxidase activity, edge artefact, necrotic tissue, and non-specific binding.
• Cross-reactivity and aberrant expression: many markers are not tissue specific (for example, TTF-1 in thyroid and lung, CD117 in several tumours).
• Heterogeneous expression: a small biopsy can miss areas with a different staining pattern.
• Subjective scoring and interobserver variation, especially for HER2 2+, PD-L1 and Ki-67.
• Markers change after treatment, and a recurrent or metastatic tumour may differ from the primary.
• IHC supports a diagnosis, and a result that disagrees with the morphology is reviewed rather than accepted.
• Cost and turnaround limit its use to questions where it will help.

Special groups.

• Children: tumours with different markers (small round blue cell tumours need a panel), and molecular confirmation is often required.
• Patients with small biopsies and cell blocks: material is limited, so a stepwise panel is used to save tissue for molecular tests.
• Patients with metastasis of unknown primary: a stepwise panel linked to the clinical picture and imaging.
• Women with breast cancer: receptor testing on every invasive tumour, and retesting of metastases when treatment depends on it.
• Patients with suspected Lynch syndrome: mismatch repair IHC on colorectal and endometrial tumours.
• Regions: IHC is expensive and available in a few centres. Samples are sent to reference laboratories, with quality depending on fixation and transport. A limited panel for breast cancer, lymphoma and common tumours has the largest effect. Where IHC is unavailable, diagnosis relies on morphology and clinical findings.

Follow up tests.

• In situ hybridisation (HER2, MDM2, EBER) and fluorescence in situ hybridisation.
• Molecular and next generation sequencing tests.
• Flow cytometry and cytogenetics for haematological disease.
• Further antibodies in a second panel.
• Special stains and electron microscopy.
• Germline genetic testing and counselling.
• Second opinion review in a reference centre.
• Multidisciplinary meeting.