From Mutation to Clinical Decision
The disciplined journey from a single molecular finding to a defensible action: the seven questions that bridge the gap, and where the chain most often breaks.
Most interpretive errors in molecular oncology come from collapsing two separate things into one — seeing an alteration on a report and jumping straight to a drug.
“An alteration is present” and “this patient should receive this treatment now” is where clinical reasoning lives — and it is also where errors arise. This page deliberately walks that distance: the chain of questions that turns a single molecular finding into a defensible action, and the points where the chain most often breaks.
A mutation is not a decision. The distance between “an alteration is present” and “this patient should receive this treatment now” is where clinical reasoning lives.
The chain: seven questions between mutation and decision
A reported alteration becomes a treatment decision only if it survives every link in the chain below. If it fails at any one, the honest output is not silence, but a clear statement of why it stopped.
1Is the finding technically real?
Before anything clinical, the alteration must be reliable. Low variant allele frequency, low tumor purity, or coverage limitations can produce findings that are artifacts, or that underrepresent what is actually present. The first question is never “what does this mean?” — it is “do I believe this?” the question.
2Is this a driver or a passenger?
Tumors carry many alterations; only some of them drive the biology. Clonality and VAF help distinguish a clonal driver from a subclonal passenger or an emerging resistant clone. Targeting a passenger wastes a treatment line and exposes the patient to toxicity with no benefit. This link alone reframes a large share of “actionable” findings.
3Is there a matched therapy in this tumor type?
A drug matched to an alteration in one cancer does not automatically apply to another. The same alteration can be a strong target in one histology and clinically irrelevant in another.
On a report, “actionable” often means actionable somewhere — not here.
4What is the strength of the evidence?
Not all matches are equal. We grade each one using an established tiering system (AMP / ASCO / CAP), so that a well-established, tumor-matched indication never falls behind a weaker, off-label, or purely preclinical rationale. The evidence tier is stated explicitly, so the recommendation can be weighed, not merely accepted.
5Where is the patient in the course of disease?
The same finding leads to different decisions in a treatment-naive patient versus one who has progressed through multiple lines. Prior therapies, prior response, and current performance status determine whether an option is appropriate now, should be held in reserve, or is no longer relevant. A decision made without this context is a guess.
6Is the timing right?
Even a valid, well-matched, well-evidenced option may not belong to the current line. Sequencing matters: using an option too early can mean losing it for the moment when it would work better; using it too late can mean missing the window entirely. The question is not only “is this a good option?” but “is this the right moment for it?”
7What happens afterward?
A decision that ignores what comes next is incomplete. If a targeted approach is chosen, the likely resistance mechanism and monitoring plan are part of the decision itself, not a separate conversation for later. Thinking one step ahead is what turns a response into a strategy.
Where the chain breaks
Most interpretive errors are not exotic. They are predictable shortcuts:
- Reading “actionable” as “approved here.” The single most common error: treating a tumor-agnostic finding, or one belonging to a different tumor, as if it applied directly to the patient in front of you.
- Treating a biomarker with no matched drug as if it were a treatment. A finding can be biologically interesting and therapeutically empty. Documenting this honestly as “monitor — no matched therapy currently available” is not a failure; it is the correct answer.
- Targeting a subclone. Acting on a low-VAF, subclonal alteration as though it were a clonal driver.
- Ignoring co-occurring alterations. A second alteration can blunt or change the implication of the first; reading findings in isolation misses this.
- Off-label use without evidentiary discipline. Reaching for an off-label option without stating the evidence tier turns a reasonable idea into overreach.
Each of these is the result of skipping a link instead of walking the chain step by step.
Why two experts can still disagree.
Even when the chain is followed faithfully, two careful oncologists can reach different conclusions — usually because they weigh the same evidence tier, timing, or sequencing trade-off differently. This is not a flaw; it is the nature of decisions made under uncertainty.
The value of a structured second opinion is that it makes the reasoning visible — you can see which link is driving the conclusion and weigh it for yourself.
The real work is the journey
The work of precision oncology is not finding the mutation. Modern sequencing does that reliably.
The real work is the disciplined journey from that finding to a decision that fits this patient, this tumor, this moment — and being honest, at every link, about how solid the ground beneath it really is.
This content is for educational purposes only. It does not constitute individual medical advice and does not replace the judgment of the treating physician.
Submit a case for expert molecular evaluation
NGS Advisory reviews your existing NGS report against current guidelines, clarifies the exact variant and its actionability for your tumor type, and provides a structured, calibrated second opinion.