Swiss-Cheese Cores in an Overnight Run: Fixation or Processing?
A 61-year-old woman has a 1.3 cm spiculated mass in the right breast. Six 18-gauge core biopsies (each roughly 1 mm x 12 mm) are placed together in one cassette and processed overnight: 10% NBF 45 minutes, graded ethanols, three xylene stations, three paraffin stations at 62 C. Time in formalin before processing was 3 hours in a 60 mL container holding 20 mL of formalin. At microtomy the first two levels ribbon and float well, but levels three and four show a Swiss-cheese pattern, sharply demarcated clear spaces separating stromal fibers and epithelial clusters, and the tissue crumbles when the block is trimmed. A lymph node from the same patient, processed on the same run, sections cleanly. The pathologist needs ER, PR, and HER2 this week.
1. Decide whether this is a fixation problem or a processing problem. Which slide features and which laboratory records separate the two, and what observation argues most strongly against a fixation cause?
Read the worked answer
Sort the artifact by its signature first.
| Pattern | Looks like |
|---|---|
| Fixation (under) | Zonal: better at the edge, worse in the center; smudged chromatin; eosinophilic streaming |
| Processing/heat | Cells pulled apart with empty, sharply bordered spaces; crumbly, 'bubbly' stroma; worst in thin, fatty, or small tissue; unrelated to tissue edges |
| Microtomy | Uniform parallel thick/thin bands (chatter) or linear scratches running across tissue and background; not present here |
| Water-bath carryover | Extraneous fragments lying on the section, not gaps |
Records to pull: processor log (station times, paraffin oven and bath temperature, vacuum/pressure cycles), reagent age and ethanol concentrations by hydrometer, cloudiness of xylene, whether the cores were in a biopsy sponge, and the formalin volume-to-tissue ratio and total fixation time.
Do the arithmetic. Six cores of ~1 mm x 12 mm is roughly 0.06 mL of tissue in 20 mL of formalin - a ratio greater than 1:200, and 3 hours is generous for a 1 mm core. Fixation therefore cannot explain central under-fixation.
Strongest argument against fixation: the fixative time and volume were more than adequate and the artifact affects the interior of the tissue fragment rather than producing a gradient from the surface. The cleanly sectioned lymph node argues against a global reagent failure, but it is not exculpatory - large, collagen-rich tissue resists over-dehydration that destroys thin cores.
Conclusion: over-dehydration and thermal hardening in processing, to which 1 mm cores are the most vulnerable tissue in the run.
2. The oncologist needs interpretable ER/PR and HER2 slides from this block today. What will you do to the block, and what options would you reject?
Read the worked answer
Acceptable salvage, in order of least harm:
- Re-face the block, then soak the exposed face in warm water with a drop of liquid soap (or a commercial softening agent) for 10-20 minutes before cutting 3 um sections with a fresh blade; keep the block cold between passes.
- Cut at a slower speed with a smaller clearance angle; a disposable low-profile blade reduces chatter from hardened tissue.
- If the block is still friable, re-embed only after soaking, and accept that this is a single, documented re-embedding.
Reject:
- Cutting deeper and deeper hoping for better levels - the artifact is through the whole core, and deeper levels only consume tissue you may need for HER2 FISH.
- Repeated melting and re-embedding, which adds heat and further antigen loss.
- Reporting ER/PR as negative on a slide without a valid internal control. Heat damage from processing can produce a genuine false-negative, so the internal control (normal ducts and stroma) is the only thing standing between an artifact and a wrong treatment decision. If the internal control fails, the result is invalid, not negative.
Also do: document the event, and flag on the report that the specimen showed processing artifact if any nuclear detail remains suboptimal.
3. What quality-system action follows this case, and how would you detect the problem earlier next time?
Read the worked answer
Immediate: open a nonconformance, retain the H&E and the block, and notify the pathologist before results are released. Review every other case on that processor run - a reagent or temperature fault rarely affects one cassette.
Root-cause questions:
- Are thin biopsies (cores, endometrial curettings, cell-block fragments) run on the same schedule as large resection specimens? They should not be.
- Was the 62 C paraffin at or above set point, and were vacuum/pressure cycles lengthened recently?
- Are ethanol concentrations checked and recorded weekly?
Corrective actions: create a 'small/biopsy' processing program with shorter dehydration and lower paraffin temperature; place cores in biopsy sponges or lens paper to prevent loss and crush; add a first-pass QC check of one H&E from any new or changed program.
Monitoring: trend artifact-related re-cuts and pathologist complaints monthly; audit processor logs weekly. The measurable endpoint is the re-cut rate for small biopsies, not the presence of a policy.