📱 Rotate to landscape for the best experience
🔬
Vikram Deshpande · Pathology

IHC Tumor Workbench

BETA

Diagnostic marker-panel decision support with linked therapeutic markers. Set an IHC panel to rank candidate entities by relative support and see applicable treatment-relevant stains. Educational decision support — not a substitute for pathologist interpretation or guideline-based testing.

Theme
📋 How to Use This Site
① Find & set your markers
1. Type a marker name in the search box — e.g. CK7, TTF-1, SOX10.
2. Find the marker row in the list below.
3. Click P if the stain is Positive, or N if Negative.
② Add more markers
4. The Differential Diagnosis panel on the right updates instantly after each marker you set.
5. Search for your next marker and click P or N again — repeat for every stain in your panel.
③ Read the results
6. Diagnoses are ranked by Support — a score from 0 to 1.
7. A higher Support means your marker pattern better matches that diagnosis.
8. Support ≥ 0.8 = strong fit  ·  0.5–0.8 = partial  ·  < 0.3 = poor fit.
9. These are relative rankings, not diagnostic probabilities — always correlate with morphology and clinical context.
💡 Tip: Use Suggested Next Marker (bottom right) to find the single stain that best discriminates your top candidates. 🔄 Reset: Click Clear in the Marker Panel to start over. ⚗ Beta: Missing a marker? Use the Feedback tab to report it.

Marker panel

No markers set — click P / N below.
ADD MARKER
BROWSE LIST

Differential diagnosis

Relative support scores (0–1), not calibrated probabilities. Scores assume the case is one of the listed entities; an entity absent from the knowledge base cannot be ranked.
Set at least one marker to see a ranked differential.

Suggested next marker by expected information gain

—

1 · Pick organ system

2 · Pick tumour

Therapeutic / predictive / prognostic markers

In-depth IHC scoring · ISH algorithm · HER2-low · treatment implications
Each marker lists the drug(s), the actionable threshold, whether the IHC is an FDA companion diagnostic or a screen, and the confirmatory test required before treatment. Predictive claims are drug- and indication-specific — always use the validated assay for the intended therapy.

Tumour-of-unknown-primary immunoprofile grid

Click markers to set your panel — tumours rank by match in real time. Click once = POS, again = NEG, again = unset. ± counts as partial. Confirm with the last column.

IHC image gallery

46 reference images. Click any image to enlarge.

Gene fusions by tumour type

Recurrent somatic fusions (n≥2 cases) from TCGA. Therapeutically actionable fusions highlighted in purple.
Source: TCGA pan-cancer fusion analysis. Only fusions detected in ≥2 independent cases shown. Does not include cell-line or fusion-only datasets. Always correlate with clinical sequencing report.

Somatic mutations by tumour type

TCGA driver gene frequencies and recurrent hotspot mutations. Source: cBioPortal / TCGA pan-cancer.
Source: TCGA pan-cancer via cBioPortal. Driver gene frequencies = proportion of tumours with ≥1 somatic mutation in that gene. Hotspots = specific amino-acid substitutions recurring in ≥2 patients. For clinical reporting use a validated NGS panel.
Enter Alterations
Biomarker Status
Select Mutations / Alterations
About
Rule-based scoring across 20 primary sites using TCGA & ClinVar driver mutation data. Not a clinical diagnostic — use alongside morphology and IHC.
Primary Site Prediction
Add alterations on the left — top ranked sites appear here
⚗

IHC Tumor Workbench — Beta Feedback

Help us improve accuracy and completeness of this diagnostic reference

BETA v0.1
93 marker families
across 12 organ systems
Active development
Updated continuously
Your feedback matters
Curated by V. Deshpande
What are you reporting?

Submissions go to the workbench author via email — no account needed. ⚑ First submission triggers a one-time activation email to your inbox — click Activate to start receiving feedback.

How to use IHC Tumor Workbench
A step-by-step guide to building your IHC panel and reading the differential diagnosis.
1
Add markers to your panel
In the Differential Diagnosis tab, you'll see a list of all IHC markers. For each marker your case has been stained for:
P = Positive
N = Negative
Use the search box to quickly find markers by name — type "CK7", "TTF-1", "SOX10", "p40", etc. You can also filter by organ system using the category dropdown. Click P or N on as many markers as you have results for.
2
The differential diagnosis updates instantly
As soon as you set your first marker, the Differential Diagnosis panel on the right populates with a ranked list of candidate entities. Add more markers and the list re-ranks in real time — narrowing down to the most supported diagnoses.
Each entity row shows which of your markers support ↑ or contradict ↓ that diagnosis. The more markers you add, the more discriminating the result. Start with your most informative stains first (e.g. pankeratin, SOX10, TTF-1) then add reflex markers.
3
What do the scores mean?
The score next to each entity is a relative support score (0–1), not a probability. It tells you how consistent your panel is with that entity compared to all others in the knowledge base.
0.90+
Strong fit — your panel strongly supports this entity. Most expected markers are positive and few are contradicted.
0.50–0.89
Partial fit — some markers align but others are discordant. Consider additional stains to discriminate.
<0.30
Poor fit — multiple expected markers are absent or negative results are unexpected. Entity is unlikely given your panel.
Rank bar
The horizontal bar shows relative ranking strength — the top entity always fills 100% of its bar. Use it to see how separated the top candidates are.
⚠️ Important: Scores assume the case is one of the entities in the knowledge base. An entity not in the KB will score zero even if it is the correct diagnosis. This is a ranking tool, not a diagnostic probability calculator.
4
Check therapeutic / predictive markers
Once you have a leading diagnosis, switch to the Therapeutic / predictive markers tab. It shows all companion diagnostic and targeted therapy markers relevant to your marker panel — including predictive biomarkers, eligibility stains for clinical trials, and prognostic markers.
Examples: PD-L1 (immunotherapy), HER2/ERBB2 (trastuzumab), MMR/MSI (pembrolizumab), MDM2 amplification (MDM2 inhibitors), ALK/ROS1 (targeted lung therapy), NTRK (larotrectinib). Drug names and indications evolve — always confirm against current FDA labels and CAP/ASCO guidelines.
5
Use "Suggested next marker" to narrow faster
Below the differential panel, the Suggested Next Marker box recommends the single stain that would give you the most information to discriminate between your current top candidates — ranked by expected information gain. Order this stain next to most efficiently resolve the differential.
⚠️ Important caveats
This tool is educational decision support only — not a substitute for pathologist interpretation, clinical context, or guideline-based testing. Sensitivity and specificity values are approximate and sourced from WHO 5th edition Blue Books and peer-reviewed literature. Clone recommendations are typical but laboratory-validated protocols may differ. Morphology, clinical history, and molecular studies remain essential.
CLINICAL GUIDE · BREAST PATHOLOGY

HER2 Testing in Breast Cancer

ASCO/CAP 2018 Guidelines · HER2-Low · ISH Algorithm · Treatment Implications

IHC Scoring (ASCO/CAP 2018)

ScorePatternThresholdNext step
0 No staining, OR incomplete faint/barely perceptible staining Any % of cells HER2-negative. No reflex ISH.
1+ Incomplete membrane staining — faint/barely perceptible >10% invasive cells HER2-low — eligible for T-DXd. No reflex ISH required.
2+ Weak–moderate complete membrane staining, OR intense complete staining >10% invasive cells (weak–mod), OR ≤10% (intense) Equivocal → reflex ISH required
3+ Intense complete circumferential membrane staining >10% invasive cells HER2-positive — anti-HER2 therapy eligible
⚠ HER2-ultralow (2024): IHC >0 but <1+ (incomplete staining in ≤10% of cells). DESTINY-Breast06 showed T-DXd benefit — report and document carefully; do not call these IHC 0.

HER2-Low — The New Category

Definition: IHC 1+ OR IHC 2+ / ISH-negative (HER2/CEP17 <2.0 AND avg HER2 <4.0).
~55–60% of HR+ and ~35–40% of TNBC cases qualify as HER2-low. Re-testing of archival tissue or metastatic biopsy may be warranted given therapeutic implications.
KEY TRIAL
DESTINY-Breast04
T-DXd vs physician's choice chemo in HER2-low mBC. PFS 9.9 vs 5.1 mo (HR+ cohort); OS benefit confirmed.
KEY TRIAL
DESTINY-Breast06
T-DXd in HER2-low AND HER2-ultralow HR+ mBC. Extends benefit to ultralow category; supports meticulous IHC reporting.

ISH Algorithm — 5 Groups (ASCO/CAP 2018, Dual-Probe)

GroupHER2/CEP17 ratioAvg HER2/cellResult
1≥2.0≥4.0Positive
2≥2.0<4.0Positive (if IHC 3+) or Negative (if IHC 0/1+/2+) — requires IHC correlation
3<2.0≥6.0Positive (if IHC 3+/2+) or Negative (if IHC 0/1+)
4<2.0≥4.0 and <6.0Negative (unless IHC 3+)
5<2.0<4.0Negative
⚠ For groups 2–4: ISH result must be interpreted in conjunction with concurrent IHC result. If discordant, retest or use alternative ISH probe.

Treatment by HER2 Category

HER2-POSITIVE (IHC 3+ or 2+/ISH+)
Trastuzumab (Herceptin)PertuzumabT-DM1 (ado-trastuzumab emtansine)T-DXd (trastuzumab deruxtecan)Tucatinib + capecitabine + trastuzumabNeratinibLapatinib
HER2-LOW (IHC 1+ or 2+/ISH–)
T-DXd (trastuzumab deruxtecan) — DESTINY-Breast04
HR+/HER2-low: T-DXd after ≥1 line of endocrine therapy. TNBC/HER2-low: after ≥1 line of chemo.
HER2-ULTRALOW (>0 but <1+)
T-DXd showed PFS benefit in DESTINY-Breast06 (HR+ cohort). FDA approval pending — document carefully, do not score as 0.

Testing Workflow

1
Initial IHC on core biopsy or excision (invasive component only). Use validated antibody (4B5, SP3, CB11, or equivalent).
2
If 3+ → HER2-positive. Report and proceed to treatment planning. No ISH needed.
3
If 2+ → Equivocal. Reflex ISH on same block (dual-probe preferred). Apply 5-group algorithm.
4
If 1+ → HER2-low. No ISH required. Report as 1+ and flag HER2-low eligibility.
5
If 0 with ANY staining in ≤10% cells → Document HER2-ultralow carefully. Do not report as "0" without noting staining pattern.
6
Metastatic disease: Retest on metastatic biopsy when feasible. HER2 status can evolve (upregulation or loss). Prior anti-HER2 therapy does not preclude HER2-low re-classification.

Common Pitfalls & Reporting Tips

Avoid "HER2-negative" as sole report — distinguish 0, 1+ (HER2-low), and 2+/ISH– (HER2-low) explicitly. Each category has distinct therapeutic meaning.
Basolateral/lateral staining (not circumferential) should not be scored as 3+; seen in normal glands and may indicate non-specific staining.
Chromosome 17 polysomy: HER2/CEP17 ratio may be falsely low. Consider HER2 copy number and consult molecular correlates.
Post-neoadjuvant specimens: Residual disease HER2 status guides adjuvant T-DM1 (KATHERINE trial). Re-test residual invasive tumor after NAC.
Intratumoral heterogeneity: Report the % of cells with each staining pattern. For borderline 2+ vs 3+, err toward 2+ and reflex ISH. Neoadjuvant therapy may alter HER2 expression.
References: Wolff AC et al. J Clin Oncol 2018 (ASCO/CAP HER2 guidelines) · Modi S et al. NEJM 2022 (DESTINY-Breast04) · Curigliano G et al. NEJM 2024 (DESTINY-Breast06). Content is educational — always apply institutional protocols and current FDA labelling.