Edge-Only 3+ HER2 With a Failing On-Slide Control
A 47-year-old woman's 1.1 cm breast core biopsy is stained for HER2. The invasive carcinoma shows strong, complete membrane staining (3+) confined to the tumor at the tissue edge nearest the ink; the same tumor in the center of the section is 1+. On the same run the on-slide control pellet, which should show 2+ complete membrane staining, shows only faint, incomplete staining. ER shows 90% strong nuclear positivity in tumor with crisp internal control in normal ducts and stroma. The outside laboratory reported HER2 3+ on its own block two weeks earlier, and the oncologist has already discussed anti-HER2 therapy with the patient. The tumor is a 6 mm focus within the core.
1. Can this HER2 result be reported? Walk through the reasoning and state what is and is not interpretable on this slide.
Read the worked answer
No - the run is invalid, and that resolves the question before scoring even starts.
Reasoning:
- The on-slide control is the assay's proof that retrieval, antibody, detection, and chromogen all performed. A control pellet that is weaker than its known value means the run cannot validate any result on that slide - neither the central 1+ nor the edge 3+.
- The ER result is not a substitute. ER is a nuclear stain with a different retrieval and detection system, so its internal control says nothing about HER2 membrane staining.
- The edge-restricted pattern adds a second, independent concern: staining that hugs the tissue contour and stops abruptly often represents an edge/retraction artifact from uneven section adherence, thickness, drying, or incomplete deparaffinization, not biology.
Therefore: the slide is non-contributory. The correct action is to repeat the stain on a freshly cut section, with attention to section thickness (3-4 um), baking, and full deparaffinization, and to check that the control pellet is at its expected 2+.
Do not: report '3+' because the outside lab did, or report '1+' because most of the tumor is 1+. Either guess changes a treatment decision.
2. The repeat run has a valid control, and the same edge-predominant pattern recurs. How do you now distinguish artifact from true heterogeneity, and how do you score it?
Read the worked answer
Artifact versus heterogeneity - the discriminators:
| Favors artifact | Favors real heterogeneity |
|---|---|
| Staining follows the tissue contour exactly and stops at a sharp line | Staining follows tumor cell clusters, including internal nests |
| Present at every edge, including non-tumor edges and normal ducts | Confined to invasive tumor; normal ducts appropriate |
| No staining gradient with distance from the edge | 3+ clone adjacent to 1+ clone in a geographic pattern |
| Disappears on a re-cut or a differently oriented section | Reproducible across levels and blocks |
Practical steps: cut a new section a few micrometers deeper and stain it; examine the entire section including the non-tumor margins; and look for crush or cautery near the biopsy edge, which can produce nonspecific chromogen trapping.
Scoring once artifact is excluded: apply ASCO/CAP criteria to the whole invasive tumor. Estimate the proportion of tumor cells with complete, strong membrane staining. If the 3+ population exceeds the 10% threshold for heterogeneous positivity, the case is HER2 positive and the heterogeneity should be described in the comment. If the 3+ focus is at or below 10%, the case is equivocal and requires reflex in situ hybridization.
Key trap: 'the tumor is only 6 mm' does not lower the threshold. Size does not change the scoring rule, but it does mean you must be able to see enough tumor cells to make a proportion estimate - if the invasive focus is too small to quantify, say so and use ISH.
3. Reconcile the discrepancy with the outside 3+ result. Which preanalytical and analytical variables would you examine, and what would you recommend to the treating oncologist?
Read the worked answer
Analytical variables: different antibody clone or platform, different retrieval conditions, different detection system, and different scoring practice. These are checkable - request the outside laboratory's validation and the stain method.
Preanalytical variables, which matter more:
- Time to fixation and fixation duration. A thin core fixed for under an hour can show weak, unreliable membrane staining; over-fixation (a week in formalin) can produce false-negative HER2. Ask for the outside block's fixation record - often there is none, which is itself the answer.
- Delay to processing and block age. Cut sections older than about 4-6 weeks on a charged slide lose antigen.
- Section thickness and drying - the same edge artifact behaves differently between laboratories.
Recommendation:
- Obtain the outside block and stain it in your laboratory on the same run as your block, so the two are compared under one set of conditions. This is the single most informative experiment.
- If the discrepancy persists, in situ hybridization is the tiebreaker and is cheaper than an incorrect course of trastuzumab - HER2 ISH is less fixation-sensitive for gene copy number and gives a quantitative answer.
- Communicate to the oncologist that a single 3+ result on a 6 mm focus in a core, not confirmed by a valid in-house run, is not sufficient grounds to commit to therapy; state the plan and timeframe.
Document the discrepancy, the actions, and the final integrated result in the report.