Frozen Section — performance, reporting and limitations

Definition / Overview

Frozen section (FS) is an intraoperative histological technique in which fresh tissue is rapidly frozen, cryostat-sectioned, stained, and interpreted to provide real-time diagnostic information to the operating surgeon. The result directly influences intraoperative decision-making: extent of resection, margin clearance, lymph node status, and tissue identification.

FS is a consultation service, not a definitive diagnostic service. The pathologist's role is to answer a specific, clinically framed question within a constrained timeframe, with the understanding that the paraffin-embedded permanent section remains the gold standard for final diagnosis.

Indications for Frozen Section

Appropriate indications

• Margin assessment: Surgical margins in carcinoma resections (skin, head and neck, breast, lung, colorectal, gynaecological). The question must be framed as "is this margin involved?" not "what is the tumour type?"
• Lymph node status: Sentinel lymph node biopsy (breast, melanoma), regional nodal sampling in thyroid and lung surgery.
• Tissue identification: Confirming parathyroid tissue during thyroid/parathyroid surgery; identifying adrenal cortex; confirming the nature of an unexpected intraoperative finding.
• Adequacy of tissue for diagnosis: Ensuring a biopsy contains diagnostic material before the patient leaves the operating theatre, particularly for lymphoma workup or culture.
• Tumour characterisation: Distinguishing benign from malignant when the result will change the operative approach (e.g. ovarian mass: benign cyst vs. borderline vs. carcinoma).

Situations where frozen section should be deferred or declined

• Lesions $\leq 5\,\text{mm}$ where the entire specimen may be consumed by FS, precluding permanent section diagnosis.
• Calcified or heavily ossified tissue (decalcification required; FS is unreliable).
• Fatty tissue (adipose does not section well when frozen).
• Situations where the result will not change intraoperative management.
• Lesions where diagnosis requires ancillary studies (e.g. lymphoma subtyping, soft tissue tumour with complex molecular requirements).
• Infectious or prion-risk specimens (biosafety considerations; prion-contaminated cryostats require specialist decontamination).

Technical Performance of Frozen Section

Specimen handling

1. Fresh, unfixed tissue is transported immediately to the pathologist. Delay degrades morphology.
2. The pathologist (or designated scientist) grosses the specimen, selecting the most representative area: the closest margin, the interface of tumour and normal tissue, or the area of clinical concern.
3. Tissue is embedded in optimal cutting temperature (OCT) compound on a cryostat chuck.
4. Rapid freezing is achieved using liquid nitrogen spray, isopentane cooled in liquid nitrogen, or the cryostat's built-in freezing platform. Isopentane provides more uniform freezing and fewer ice crystal artifacts than direct liquid nitrogen immersion.
5. Sections are cut at $5\text{-}8\,\mu\text{m}$ on the cryostat, typically at $-20\,^\circ\text{C}$ to $-25\,^\circ\text{C}$.
6. Sections are transferred to glass slides (room-temperature slides promote adherence), fixed briefly in cold acetone or 95% ethanol, and stained.

Staining

• Haematoxylin and eosin (H&E): Standard stain; rapid protocol (total staining time 60-90 seconds).
• Toluidine blue: Faster than H&E; useful for rapid nuclear detail assessment, particularly in neuropathology intraoperative consultations.
• Oil Red O: Can be performed on frozen sections for lipid identification (not possible on paraffin sections after processing).
• Immunohistochemistry on frozen sections is feasible but rarely performed intraoperatively due to time constraints; it is more commonly used on frozen tissue retrieved from the cryostat archive.

Turnaround time

• Target: result communicated to surgeon within 20 minutes of specimen receipt.
• Complex specimens (multiple margins, large resections) require clear pre-operative communication about expected turnaround.

Reporting Frozen Section Results

Communication

• Results are communicated verbally to the surgeon, documented contemporaneously in the pathology information system or a dedicated FS log.
• The verbal report must be unambiguous. Use clear language: "The margin is free of tumour," "Carcinoma is present at the inked margin," "This tissue is consistent with parathyroid."
• Avoid hedging language that the surgeon cannot act upon. If genuinely uncertain, state: "I cannot exclude malignancy on frozen section; I recommend deferral to permanent section."
• The final written FS report is incorporated into the permanent section report, with concordance or discordance noted.

Diagnostic categories for frozen section reporting

Concordance and discordance

• Concordance rate between FS and permanent section should exceed 97-98% in a quality programme.
• Discordance is classified as:
• False negative: FS reported benign, permanent section shows malignancy. Most commonly due to sampling error.
• False positive: FS reported malignant, permanent section benign. Most commonly due to misinterpretation of artifact or reactive change.
• Deferred: Not a discordance; appropriate deferral is a quality indicator, not a failure.
• All discordant cases must be reviewed at departmental audit and the surgeon notified promptly when a permanent section result changes the FS diagnosis.

Limitations of Frozen Section

Morphological limitations

• Nuclear detail is less crisp than in paraffin sections; chromatin texture and nucleolar prominence are harder to assess.
• Cytoplasmic detail is reduced; mucin, keratinisation, and glandular differentiation may be difficult to evaluate.
• Tissue architecture is partially preserved but compression and freeze artifact distort spatial relationships.
• Lipid-rich tissue (breast, adrenal cortex, adipose) sections poorly and yields unreliable morphology.
• Small lesions risk complete consumption, leaving no tissue for permanent section.

Sampling limitations

• FS samples only the submitted tissue; a negative margin on FS does not guarantee the entire circumferential margin is clear.
• Discontinuous tumour spread (e.g. lobular carcinoma of the breast, perineural invasion) may be missed on a single level.
• Sentinel lymph node FS has lower sensitivity for micrometastases ($\leq 2\,\text{mm}$) and isolated tumour cells than permanent section with serial levels and cytokeratin immunohistochemistry.

Ancillary study limitations

• Receptor immunohistochemistry (ER, PR, HER2) cannot be reliably performed or interpreted on frozen tissue; formalin fixation is required.
• Molecular studies (FISH, NGS, PCR-based assays) require fixed tissue or fresh tissue handled under specific protocols; FS material is generally unsuitable.
• Special stains for organisms, amyloid typing, and enzyme histochemistry are possible on frozen sections but are not routine intraoperative applications.

Artifacts in Frozen Section

Artifact recognition is critical to avoiding false-positive and false-negative diagnoses.

Ice crystal artifact

• Mechanism: Slow or uneven freezing allows intracellular water to form ice crystals, which expand and rupture cell membranes.
• Appearance: Irregular clear vacuoles within the cytoplasm and nucleus; "Swiss cheese" or honeycomb pattern within cells; loss of nuclear detail.
• Mitigation: Rapid freezing technique; isopentane preferred over direct liquid nitrogen immersion for small specimens; avoid saline-soaked tissue.
• Diagnostic impact: Can mimic clear cell change, vacuolar degeneration, or lipid accumulation; nuclei appear artifactually vesicular.

Freeze-thaw artifact

• Tissue that has been partially thawed and refrozen shows exaggerated ice crystal damage and nuclear smearing.

Compression / crush artifact

• Excessive pressure during sectioning or tissue handling compresses cells, producing elongated, hyperchromatic nuclei that can mimic high-grade malignancy.
• Particularly problematic in lymphoid tissue and small cell tumours.

Sectioning artifacts

• Chatter: Vibration during cutting produces parallel lines across the section; seen with hard or incompletely frozen tissue.
• Thickness variation: Uneven sections produce areas of apparent hyperchromatism in thicker zones, mimicking increased nuclear density.
• Folding and tearing: Sections fold over themselves, creating apparent multilayering or pseudostratification.

Staining artifacts

• Rapid H&E protocols produce less crisp nuclear staining than standard paraffin H&E.
• Gomori trichrome on frozen sections can show irregular red-stained areas that interfere with interpretation, particularly in muscle biopsies.
• Nuclear vacuolisation (peripherally marginated chromatin with a clear centre) is an artifact seen in frozen sections of some tissues and should not be mistaken for viral cytopathic effect or intranuclear inclusions.

OCT compound contamination

• OCT compound on the tissue surface produces a pale, amorphous, non-staining material that can obscure the margin or mimic mucin.
• Careful grossing and embedding technique minimises this.

Tissue-specific artifacts

Specific Clinical Scenarios

Thyroid surgery: parathyroid identification

• Parathyroid tissue on FS shows chief cells with pale cytoplasm, small uniform nuclei, and a delicate vascular stroma; fat cells are present in normal parathyroid.
• Thyroid follicular tissue, lymph node, and thymic tissue are the main differentials.
• Touch imprint cytology can supplement FS for rapid identification.

Ovarian mass

• FS is used to distinguish benign, borderline (low malignant potential), and malignant epithelial tumours.
• Sensitivity for borderline tumours is lower than for invasive carcinoma; microinvasion and micropapillary patterns may be missed on FS.
• Mucinous tumours require extensive sampling; a single FS section from a large mucinous tumour is unreliable for excluding borderline or malignant areas.

Sentinel lymph node (breast)

• FS is appropriate when a positive result will lead to immediate axillary lymph node dissection.
• Sensitivity of FS for macrometastases ($> 2\,\text{mm}$): approximately 70-80%; for micrometastases: substantially lower.
• Many centres have moved to intraoperative molecular assays (e.g. one-step nucleic acid amplification, OSNA) as an adjunct or alternative.

Brain tumour intraoperative consultation

• Smear preparation (squash preparation) is the preferred technique in neuropathology: a small fragment of tissue is smeared between two slides, fixed, and stained with toluidine blue or H&E.
• Smear preparations show better cytological detail than FS in brain tissue and avoid freeze artifact.
• FS is used for tissue adequacy and broad categorisation (glioma vs. metastasis vs. lymphoma vs. non-neoplastic).

Quality, Safety, and Governance

Laboratory quality indicators

• FS-to-permanent section concordance rate (target $\geq 97\%$).
• Turnaround time (target $\leq 20$ minutes from receipt to verbal report).
• Rate of appropriate deferral (a marker of appropriate clinical judgement, not failure).
• Discordance case review at departmental morbidity and mortality or quality meetings.

Biosafety

• Cryostats are a potential source of aerosolised infectious material. Universal precautions apply.
• Prion-risk specimens (Creutzfeldt-Jakob disease, suspected prion disease) must not be processed in the standard cryostat. Dedicated equipment and specialist decontamination protocols are required; FS should be declined or performed only with appropriate containment.
• Tuberculosis-risk specimens: FS should be deferred; fresh tissue should be sent for microbiological culture and the specimen handled in a biosafety cabinet.

Documentation

• Every FS must be documented: specimen identity, time of receipt, time of verbal report, result communicated, name of surgeon and pathologist.
• Discordant cases require an addendum to the permanent section report explaining the discordance and, where relevant, the clinical implications.

Exam-Focused Summary: Key Points for RCPA Part 1 and Part 2

• Know the indications and contraindications for FS; the examiner will present a scenario and ask whether FS is appropriate.
• Artifact recognition on the slide: Ice crystal artifact, crush artifact, and OCT contamination are the most commonly tested. Know their appearance and how to distinguish them from genuine pathology.
• Thyroid nuclear features on FS are unreliable: Nuclear clearing, grooves, and pseudoinclusions of papillary thyroid carcinoma are exaggerated by freeze artifact; do not over-diagnose PTC on FS alone.
• Concordance and discordance: Understand the classification of discordance, its causes, and the governance response.
• Communication is part of the competency: The RCPA expects trainees to demonstrate clear, unambiguous verbal reporting and appropriate escalation when a diagnosis cannot be made on FS.
• Smear preparations in neuropathology are superior to FS for cytological detail in brain tissue; know when to use each technique.
• Sentinel lymph node FS limitations: Sensitivity is imperfect, particularly for micrometastases; know the role of OSNA and serial permanent sections with cytokeratin IHC.