Knowledge · when each test is used

When each test is used

Nine situations, the test that fits, and the reasoning. Each case states what the test does not answer as plainly as what it does.

Illustrative clinical situations, not real patients.

Compare every test Book a free consultation

How to read these

Choosing a test is choosing a question

The right test is the one that answers the question in front of you, not the largest one available.

Nine situations, each with the test that fits it, the reasoning behind that choice, and the three things we do once it is chosen: what gets planned, what it gets matched to, and what access work follows. Every case also states what the test does not answer, because that is the half people are rarely told and the half that causes disappointment.

The cases

Nine situations, and the test that fits

Arranged roughly in the order they arise across a journey, from the first decision through monitoring to inherited risk.

Case 01

The biopsy is too small to profile

Advanced non-squamous lung cancer. Diagnostic biopsy largely used by earlier tests.

Can we get a molecular result before first-line treatment starts?

Which test: Signature STb O or STb O+, from blood
Why that one: Tissue exhaustion is the most common reason profiling fails in lung cancer. Plasma is run in parallel with any tissue attempt rather than after one fails, because sequential attempts cost weeks that the first treatment decision does not have.

Planned

Plasma drawn the same week, run in parallel with any remaining tissue attempt, so a failed tissue result costs nothing in time. Board-signed decision timed to reach the first-line discussion.

Matched

Any driver found is matched to licensed therapy and, separately, to trials filtered by line of treatment, prior exposure and travel distance.

Access

Assistance and biosimilar mapping against whatever is matched, and a navigator who calls the hospital pathology team directly rather than sending the family back to ask.

What it answers

Whether a targetable driver is present
Whether the panel read fusions, which several lung drivers are
Enough to inform the first-line discussion rather than the second

What it does not answer

A negative plasma result does not exclude a driver, because shedding varies
It does not replace tissue confirmation where plasma is uninformative
It does not report PD-L1, which is immunohistochemistry on tissue

Signature STb O Lung cancer entry

Case 02

A genomic report already exists, but no plan came with it

Metastatic colorectal cancer. A panel from another laboratory, several months old.

Does this need repeating, or interpreting?

Which test: Decipher, or Decipher Plus if there is a gap
Why that one: If sequencing was done properly, repeating it adds cost and delay rather than information. The useful question is what the earlier panel could and could not detect. In colorectal cancer the common gap is HER2 amplification, which a point-mutation-only panel will not report.

Planned

The existing report is read against what its assay could physically detect, and the gap is named specifically. A repeat is recommended only where one would add something.

Matched

Re-annotation against current evidence, then matching to therapy and to trials that did not exist or did not recruit when the first report was written.

Access

Only the missing assay is ordered, not a second full panel. Records from every hospital involved are consolidated into one timeline the treating team can file.

What it answers

A ranked, evidence-tiered interpretation of data that already exists
Whether the earlier panel had a scope gap worth filling
Tumour board sign-out without a new sample

What it does not answer

It cannot recover an alteration class the original assay never read
It does not tell you the first laboratory was wrong
It does not substitute for new testing where the disease has changed materially

Decipher Decipher Plus

Case 03

Treatment is running and the only feedback is a scan every three months

Advanced disease on matched targeted therapy, with a baseline profile on file.

Is it still working, and would we know sooner?

Which test: Sentinel, at two, four or six draws a year
Why that one: Serial plasma measured against an existing baseline produces a trend rather than a single value. The clinically important question is settled first: what would a rising result actually cause the treating team to do?

Planned

Before the first draw, the escalation is agreed in writing: what a rise would cause the treating oncologist to do, and at what point. Monitoring that nobody would act on is declined.

Matched

Matched to a draw schedule keyed to the existing baseline and the treatment cycles, with the resistance plan naming in advance what gets tested if the trend turns.

Access

Draws arranged at home or locally on that schedule, and every result routed to the treating oncologist rather than to the patient alone.

What it answers

A trend line between scans rather than a gap
An earlier signal of change than imaging typically gives
Resistance alterations, where they are detectable in plasma

What it does not answer

An earlier signal is not the same as a better outcome, and trials are still testing whether acting earlier helps
It does not replace imaging or clinical assessment
It is uninformative where the tumour sheds little DNA

Sentinel Clear

Case 04

A family pattern, and nobody has been tested

An unaffected relative asking about inherited risk. Two affected relatives, no genetic testing in the family.

Who should be tested first?

Which test: Heritage Core for the affected relative, then Heritage Adhoc for the family
Why that one: Testing an unaffected person first is far less informative than it sounds. If no variant is found, that may mean the family alteration is elsewhere rather than absent. Testing an affected relative as the index case is what makes every later result in the family interpretable.

Planned

A testing order rather than a test: the affected relative identified as the index case, then a cascade sequence naming who is tested next and what each result settles.

Matched

Once a family variant is known, each relative is matched to a focused single-variant test rather than another broad panel, and to the surveillance pathway that variant calls for.

Access

Counselling arranged for relatives wherever they live, including those who have not decided whether they want to know, with the paperwork handled centrally.

What it answers

Whether a specific inherited alteration is present in the family
Once identified, a clear yes or no for each relative
A surveillance discussion that can begin before any cancer develops

What it does not answer

It does not say whether someone will develop cancer, only how risk changes
A negative result in an untested family is not reassurance
It does not decide what anyone should tell their children, which is what counselling is for

Heritage Core Heritage Adhoc

Case 05

Acute leukaemia, and treatment starts this week

Newly presenting acute myeloid leukaemia. Treatment cannot wait for a full panel.

Which results change what happens now, and which can follow?

Which test: Signature BML+, with staged reporting
Why that one: A result that arrives after the first cycle answers a question already decided. So testing is staged: acute promyelocytic leukaemia is excluded immediately, the findings that change the first cycle are reported within days, and the fuller risk picture follows to inform the transplant discussion.

Planned

Staged reporting agreed at the point of ordering: the findings that change the first cycle are reported ahead of the full panel rather than held back with it.

Matched

Urgent targetable findings matched to therapy within the first reporting stage; the full risk picture then feeds the transplant discussion.

Access

Sample logistics handled from the ward, and the report delivered into the haematology team’s hands rather than posted to a patient mid-admission.

What it answers

Urgent targetable findings, first
Formal risk classification, combining cytogenetics with molecular findings
A baseline for measurable residual disease monitoring

What it does not answer

It does not diagnose leukaemia. Morphology, flow cytometry and cytogenetics remain essential
It does not replace immediate testing for acute promyelocytic leukaemia
It does not remove the need to re-profile at relapse, when the picture commonly changes

Signature BML+AML entry

Case 06

The lineage is not clean, and a sibling is being considered as a donor

Ambiguous-lineage leukaemia. Transplant under discussion, with a sibling as potential donor.

One test or two, and is the donor suitable?

Which test: Signature BML+, which carries the lymphoid genes too
Why that one: A panel built for one lineage can mischaracterise a mixed-phenotype leukaemia, a therapy-related neoplasm or a transformation. Reading both in one run avoids that. It also flags germline predisposition, which matters urgently here: an affected sibling would be an unsuitable donor, and that is far better established before transplant than after.

Planned

One comprehensive panel covering both lineages, so an ambiguous presentation is characterised once. Germline predisposition is assessed before a donor is selected, not after.

Matched

Matched to therapy on the lineage the molecular picture actually supports, and matched against donor suitability where a related donor is in question.

Access

A germline finding triggers the cascade pathway for relatives at no separate consultation, so the sibling question and the family question are answered together.

What it answers

Myeloid and lymphoid drivers in one run
MDS and MPN overlap genes, where the origin is unclear
Germline predisposition flags, before a related donor is chosen

What it does not answer

A germline flag on a tumour panel is not a germline diagnosis. Confirming it needs a separate test
It does not choose the donor or the conditioning regimen
It does not replace expert haematopathology review

Signature BML+MDS and MPN entry

Case 07

The diagnosis itself depends on the molecular result

Adult glioma. Classification cannot be completed on histology alone.

Which assays does the classification actually require?

Which test: Signature Custom, designed around the classification
Why that one: Adult glioma cannot be classified without IDH status and 1p/19q codeletion, and MGMT promoter methylation is a separate epigenetic test that a mutation panel does not perform. This is not profiling added to a diagnosis; it is part of making one, so the panel has to carry every assay the classification requires.

Planned

The panel is designed around the classification criteria the diagnosis depends on, so every marker required to assign an entity is read in one run.

Matched

The assigned classification matched to the treatment protocol and the trial eligibility criteria that reference it, since both are written in terms of the entity, not the gene.

Access

The report is written so it can be attached to a trial screening form or a protocol referral without translation.

What it answers

IDH status, 1p/19q codeletion and MGMT methylation together
Whether molecular features change the grade assigned on histology
A design matched to the classification rather than to a catalogue

What it does not answer

Design and validation take longer than a catalogue tier. If the decision is next week, that matters
It does not substitute for neuropathology review
Plasma is much less informative in brain tumours, because the blood-brain barrier limits shedding

Signature Custom Glioma entry

Case 08

The gene name is not enough to act on

Gastrointestinal stromal tumour. A report naming the gene but not the exon.

Does the report actually say what we need?

Which test: Signature Custom, or a tier reporting at exon level
Why that one: In GIST, which exon of KIT is altered changes the discussion, and the PDGFRA D842V variant behaves differently from every other alteration in that gene. A report saying only “PDGFRA mutated” is not actionable, and the difference is not a detail.

Planned

Reporting specified at exon and variant level at the point of design, because the gene name alone does not determine sensitivity or resistance.

Matched

Matched at the level the evidence is written at: the specific alteration to the specific agent, including where an alteration in the same gene predicts the opposite response.

Access

Where the matched agent is not routinely available, assistance, biosimilar and named-patient routes are mapped alongside the recommendation.

What it answers

Exon-level resolution, reported explicitly
Whether a PDGFRA alteration is specifically D842V
Wild-type status, which triggers a different workup including SDH and a germline question

What it does not answer

It does not change what the alteration is, only whether the report tells you
It does not replace the site, size and mitotic rate assessment that establishes recurrence risk
It does not predict which resistance mutation will emerge later

Signature Custom GIST entry

Case 09

Treatment finished, and every appointment feels like an exam

After curative-intent treatment. No evidence of disease.

Is molecular surveillance worth doing in this cancer?

Which test: Clear, where the evidence supports it
Why that one: Molecular residual disease detection is more established in some cancers than others. The honest first step is checking where the evidence stands in this particular disease, and saying so when it does not yet support routine use.

Planned

Eligibility is checked first, and surveillance is declined where the evidence does not support it. Where it proceeds, the schedule and the stopping point are set in advance.

Matched

Matched to a tumour-informed assay built on the original profile, so a signal is measured against the disease that was actually treated.

Access

Draws arranged locally on the agreed schedule, with results going to the treating team and a named person to call between appointments.

What it answers

Whether residual disease is detectable below imaging
A schedule matched to the cancer type and the follow-up plan
Where the evidence currently stands in that specific disease

What it does not answer

In several cancers, whether acting on an early signal improves outcomes is still an open question
It is not reassurance, and a negative result does not exclude disease
It does not replace scheduled clinical follow-up

Clear Colorectal entry

The tenth case

If none of these looks like your situation

Most situations do not match a written case exactly. That is what the free consultation is for, and it is also where we say that no test would change anything, when that is the honest answer.

Book a free consultation Compare every service

Take the next step

Three ways forward. Pick the one that fits today.

Book a free consultation

A no-obligation conversation with our care team, arranged through KPCIRC.

Book a consultation

Begin Blueprint Care

Commission your decision report and a dedicated clinical team.

Explore Blueprint Care

Ask on WhatsApp, free

Signal replies to your first question within four hours, at no cost.

Start with Signal