Cancer Precision Treatment Navigator: Biomarkers, Targeted Therapies, Immunotherapy and Resistance
Editorial review date: October 15, 2026 | Publisher: SmartCancer.org Editorial Team
Cancer precision medicine seeks to identify treatments that are more likely to benefit an individual patient by combining the cancer's histology, stage, molecular alterations, protein expression, immune characteristics and clinical context. The goal is not simply to find a mutation and choose a drug. It is to determine whether a biomarker is reliable, whether the proposed treatment has demonstrated meaningful clinical benefit, whether the patient is eligible, and how the treatment's effectiveness and resistance will be evaluated.
This guide provides a structured pathway for understanding molecular testing, targeted therapies, immune checkpoint inhibitors, treatment resistance, liquid biopsies and clinical trial matching. It is intended for patients, caregivers, clinicians and cancer researchers who want to make better-informed treatment decisions.
A molecular finding is not automatically a treatment recommendation. The same alteration can have different implications across cancer types, and a biomarker-matched treatment may fail despite apparently appropriate testing. Treatment decisions must account for clinical evidence, the exact cancer diagnosis, disease stage, prior treatment, competing options, safety and patient preferences.
Contents
- What is precision oncology?
- The precision treatment decision pathway
- Types of cancer biomarkers
- Choosing the right biomarker tests
- Biomarker and treatment map by cancer type
- Targeted therapy: matching a drug to a cancer dependency
- Immunotherapy biomarkers and treatment selection
- Targeted therapy versus immunotherapy versus chemotherapy
- Why cancer treatments stop working
- Resistance testing and treatment after progression
- Liquid biopsies and circulating tumor DNA
- AI, multi-omics and emerging precision oncology
- How to assess evidence and actionability
- Questions to ask the oncology team
- Frequently asked questions
1. What Is Precision Oncology?
Precision oncology is an approach to cancer care that combines information about the tumor's biological characteristics with clinical information about the patient to guide treatment decisions.
.png)
Traditional cancer classification remains essential. The organ where a tumor began, its histological subtype, its stage and whether it has spread all influence treatment. Precision oncology adds another layer: the biological features that may explain why a cancer grows, spreads, responds to treatment or develops resistance.
These features can include:
- Genomic alterations: mutations, gene fusions, amplifications, deletions and other changes in DNA or RNA.
- Protein expression: biomarkers such as estrogen receptor, progesterone receptor, HER2 and PD-L1.
- DNA repair characteristics: including mismatch repair deficiency and homologous recombination repair alterations.
- Immune-related characteristics: including microsatellite instability, tumor mutational burden and features of the tumor immune microenvironment.
- Clinical characteristics: cancer stage, sites of metastasis, performance status, organ function, previous treatments and coexisting conditions.
Biomarker testing can identify treatment opportunities, help exclude ineffective treatments, inform prognosis in selected settings and identify relevant clinical trials. Nevertheless, precision medicine is not a replacement for established pathology, staging or evidence-based standard care.
For a foundational explanation, see the National Cancer Institute's guide to biomarker testing for cancer treatment.
2. The Precision Treatment Decision Pathway
A reliable precision oncology workflow links each testing result to a clear clinical question. Testing is useful when it can change a decision, improve the characterization of the disease or support a meaningful clinical trial option.
Primary cancer site, histology, grade and relevant pathology findings.
Localized versus advanced disease, curative versus disease-control goals, and urgency of treatment.
Use disease-specific molecular panels, immunohistochemistry, in situ hybridization and other indicated tests.
Confirm the alteration, assay limitations, biological significance and level of clinical evidence.
Approved biomarker-directed treatment, established non-targeted treatment, combination strategies or appropriate clinical trials.
Clinical goals, imaging, laboratory tests, toxicity assessment and any validated molecular monitoring.
If disease progresses, consider repeat molecular assessment and determine the next evidence-supported treatment.
Why the sequence matters
A highly informative molecular result may be clinically irrelevant if the patient cannot safely receive the matched drug, if the treatment is unlikely to act quickly enough, or if another treatment offers a better-established benefit. Conversely, an actionable finding can be missed when testing is incomplete, tissue is inadequate or the assay does not cover the relevant alteration.
In some cancers, molecular results should be available before a treatment decision is finalized. In urgent situations, however, the oncology team may need to begin appropriate treatment while testing continues. The decision depends on the cancer type, disease burden, clinical stability and available alternatives.
3. Types of Cancer Biomarkers
| Biomarker category | What it tells us | Typical clinical use | Important limitation |
|---|---|---|---|
| Predictive | Whether a treatment is more or less likely to provide benefit in a defined setting. | EGFR alterations, selected HER2 findings, PD-L1 expression and mismatch repair status. | Prediction is not certainty; performance depends on the treatment and cancer context. |
| Prognostic | Information about the likely course or outcome of a disease, irrespective of a particular treatment effect. | Risk assessment and outcome estimation in validated disease-specific settings. | A prognostic association does not prove that a specific treatment improves outcome. |
| Diagnostic or classification | Helps identify a tumor subtype or distinguish diseases that require different management. | Pathology markers, selected gene fusions and defining molecular alterations. | Must be interpreted alongside morphology, history and other diagnostic evidence. |
| Pharmacodynamic or response-related | May indicate a biological effect of treatment or a change in disease activity. | Selected serial blood tests and molecular response studies. | A change in a biomarker does not always establish improved survival or justify changing treatment. |
| Hereditary or germline | Identifies inherited variants that may influence cancer risk or treatment selection. | BRCA1/2 and other cancer-predisposition or DNA-repair genes. | May have implications for relatives and require genetic counseling. |
Somatic versus germline testing
Somatic alterations arise in cancer cells or develop during a person's lifetime. These can help identify tumor-specific treatment opportunities.
Germline alterations are inherited and are generally assessed using non-tumor tissue, often blood or saliva, with appropriate clinical interpretation. A finding on a tumor sequencing report may raise suspicion of an inherited variant, but tumor testing alone does not necessarily establish a germline diagnosis.
This distinction matters because an alteration can influence both treatment selection and family cancer-risk assessment. Patients with potentially hereditary findings should discuss confirmatory testing and genetic counseling with their care team.
4. Choosing the Right Biomarker Tests
4.1 Tissue-based next-generation sequencing
Next-generation sequencing (NGS) can assess multiple genomic alterations in one test. Depending on the assay, it may identify single-nucleotide variants, insertions and deletions, gene amplifications, deletions and selected gene fusions.
A broad panel can be particularly useful in advanced cancers when multiple molecular alterations have established treatment implications or may qualify a patient for a clinical trial. But broad sequencing is not automatically the best test for every situation.
Before ordering a panel, the team should establish whether it adequately assesses the alteration types relevant to the cancer. For example, a DNA-only assay may not detect all clinically important RNA-defined fusions, and a genomic panel may not replace immunohistochemistry or another validated protein-based assay.
4.2 Immunohistochemistry and in situ hybridization
Immunohistochemistry (IHC) measures protein expression in tissue. It is important for biomarkers such as estrogen receptor, progesterone receptor, HER2 and PD-L1.
Fluorescence in situ hybridization (FISH) and other in situ hybridization techniques can help assess selected gene amplifications, rearrangements and other structural changes.
These approaches answer different biological questions from DNA sequencing. They should be selected according to the relevant disease-specific testing recommendations rather than treated as interchangeable methods.
4.3 MSI, mismatch repair and tumor mutational burden
Microsatellite instability-high (MSI-H) and mismatch repair deficiency (dMMR) are important biomarkers for identifying patients who may benefit from particular immune checkpoint inhibitors in defined clinical settings.
Tumor mutational burden (TMB) estimates the number of mutations in tumor DNA. High TMB may help identify candidates for certain immunotherapies under specific treatment indications and test criteria. However, TMB thresholds, assays and predictive performance are not interchangeable across every tumor type.
A report should specify the assay, the result, any relevant threshold and the treatment context to which the result applies.
4.4 Germline testing
Germline testing may be indicated according to cancer type, personal history, age at diagnosis, family history or the presence of particular tumor findings. Some inherited alterations influence the choice of targeted treatment, while others primarily change prevention, surveillance and family counseling.
4.5 How to choose between a focused test and a broad panel
| Clinical situation | Potential approach | Key question |
|---|---|---|
| A specific biomarker is required for a treatment decision | Validated focused assay or an appropriately comprehensive panel. | Will the test reliably answer the treatment question in time? |
| Advanced cancer with multiple possible targets | Consider a multigene panel that covers clinically relevant alterations. | Does the panel detect substitutions, indels, copy-number changes and relevant fusions? |
| Protein-expression marker drives the decision | IHC or another validated protein assay. | Is the scoring method appropriate for the specific drug and indication? |
| Tumor tissue is insufficient or unsafe to obtain | Consider validated plasma ctDNA testing where clinically appropriate. | Would a negative result need tissue-based follow-up? |
| Possible inherited cancer predisposition | Appropriate germline assessment and genetic counseling. | Does the finding require confirmation in a non-tumor sample? |
| Progression following targeted treatment | Consider repeat tissue and/or plasma molecular profiling. | Has the tumor acquired a resistance mechanism that changes treatment options? |
Testing strategies should follow disease-specific evidence, clinical guidelines and the intended use of the assay. The FDA's list of authorized companion diagnostic devices is a useful reference for specific biomarker–drug relationships in the United States. Regulatory authorization and treatment availability vary by jurisdiction.
5. Biomarker and Treatment Map by Cancer Type
The following table is a navigation aid, not a complete testing protocol. It highlights representative biomarkers that can influence treatment selection in specific clinical settings. A listed biomarker does not mean that every patient with that finding should receive a particular drug.
| Cancer type | Representative biomarkers | Potential treatment relevance | Important interpretation issue |
|---|---|---|---|
| Non-small cell lung cancer | EGFR, ALK, ROS1, BRAF, KRAS, MET, RET, ERBB2/HER2 and NTRK alterations; PD-L1 expression. | Driver-directed therapies, selected antibody-based treatments and immune checkpoint inhibitors. | Test scope matters. A comprehensive assessment may need both DNA and RNA analysis. Treatment sequence depends on the exact driver and clinical context. |
| Breast cancer | ER, PR, HER2; selected PIK3CA, ESR1, BRCA1/2 and other pathway alterations. | Endocrine therapy, HER2-directed treatment, selected pathway inhibitors and PARP-targeted approaches. | Receptor status, prior therapies and disease progression can change the treatment strategy. Some biomarkers may need reassessment in advanced disease. |
| Colorectal cancer | KRAS/NRAS, BRAF V600E, MSI/dMMR and selected HER2 or NTRK findings. | Anti-EGFR treatment selection, selected BRAF or HER2-directed approaches, and immunotherapy in indicated biomarker-defined settings. | RAS status and other molecular findings can determine whether targeted treatment is appropriate. Primary tumor site and treatment history also matter. |
| Ovarian cancer | BRCA1/2, homologous recombination deficiency (HRD) where validated, and selected protein-expression biomarkers such as folate receptor alpha (FRα). | PARP-inhibitor strategies in indicated settings, treatment-selection decisions and selected antibody–drug conjugates. | Eligibility depends on the specific disease setting, prior treatment, assay result and current product label. |
| Prostate cancer | BRCA1/2 and other homologous recombination repair alterations, MSI/dMMR in relevant cases, and imaging-defined PSMA expression. | Selected PARP-based approaches, biomarker-directed treatment and radioligand therapy in eligible patients. | Somatic and germline findings are distinct. Genomic markers and imaging biomarkers answer different questions. |
| Gastric and gastroesophageal junction cancer | HER2, PD-L1 combined positive score (CPS), MSI/dMMR and CLDN18.2 expression in appropriate settings. | Selected HER2-directed therapy, immunotherapy and CLDN18.2-directed treatment. | Protein-expression thresholds, testing methods and treatment indications must match the drug being considered. |
| Melanoma | BRAF V600 and other relevant molecular and clinical markers. | BRAF/MEK-targeted combinations in eligible patients and immunotherapy. | Both targeted treatment and immunotherapy may be appropriate; sequencing depends on the specific clinical circumstances. |
| Cholangiocarcinoma | FGFR2 fusions or rearrangements, IDH1 mutations and selected other alterations. | Targeted therapies in biomarker-defined settings, often particularly relevant in advanced disease. | Confirm the exact alteration and whether the matched agent is supported for the patient's treatment setting. |
| Endometrial cancer | MSI/dMMR, selected HER2 expression or amplification, and other histology-specific molecular markers. | Immunotherapy and other biomarker-informed approaches in defined clinical settings. | Histology, molecular subtype and stage can substantially change the management pathway. |
| Thyroid cancer | Selected RET, NTRK and BRAF alterations, among others. | Targeted approaches in eligible tumor subtypes and clinical settings. | Some alterations are much more relevant in certain histologies than others. |
How to use this table: begin with the cancer type and pathological subtype, identify the alterations or protein markers relevant to that disease, and then confirm the treatment evidence and regulatory indication. The table should not be used to select medication without an oncology assessment.
Tissue-agnostic treatment: when a biomarker can matter across cancer types
Some therapies are approved for particular molecular features rather than only one organ of origin. This is known as a tissue-agnostic approach.
Examples include selected treatments for tumors with NTRK gene fusions, MSI-H/dMMR tumors and certain other biomarker-defined populations. However, the exact eligibility criteria, prior-treatment requirements, tumor exclusions and evidence base differ by drug and indication.
A tissue-agnostic biomarker does not mean that every cancer carrying that alteration will respond equally well. The tumor's lineage, biological context, available alternative treatments and clinical trial evidence remain relevant.
6. Targeted Therapy: Matching a Drug to a Cancer Dependency
Targeted cancer therapies interfere with specific proteins, signaling pathways or biological processes that contribute to tumor growth or survival. They can include small-molecule inhibitors, monoclonal antibodies, antibody–drug conjugates and other treatment platforms.
The central question is whether the alteration identified by testing is a clinically meaningful treatment target or a reliable marker for benefit from a particular therapy.
Common classes of targeted treatment
| Target or pathway | Representative biomarker | Therapeutic principle | What must be established |
|---|---|---|---|
| Receptor tyrosine kinases | EGFR, ALK, ROS1, RET or MET alterations in relevant cancers. | Block oncogenic signaling through the altered receptor or pathway. | The precise alteration, tumor type and treatment-specific eligibility. |
| BRAF–MEK signaling | Selected BRAF alterations, particularly BRAF V600 in established settings. | Suppress a growth-promoting signaling pathway, sometimes with combination treatment. | That the variant is one for which the proposed drug or combination has evidence. |
| HER2 | HER2 amplification, overexpression or selected mutations depending on cancer type. | Block HER2 signaling or deliver cytotoxic payloads through HER2-directed treatment. | Whether the assay measures the specific HER2 feature required by the treatment. |
| DNA repair | Selected BRCA1/2 and other homologous recombination repair alterations. | Exploit vulnerabilities in DNA repair through treatments such as PARP inhibitors in defined settings. | Gene-specific evidence, disease setting, treatment history and eligibility criteria. |
| Antibody–drug conjugates | Selected protein targets, including HER2 and FRα in appropriate settings. | Use an antibody to direct a linked cytotoxic payload toward cells expressing the target. | Target expression or other required biomarker criteria and the relevant treatment indication. |
What makes a genomic alteration actionable?
An actionable alteration has a sufficiently credible relationship to a treatment decision in a defined clinical setting. Actionability should not be confused with the presence of a mutation that appears biologically interesting.
For each finding, the oncology team should determine:
- Is the alteration pathogenic or likely oncogenic, or is its significance uncertain?
- Is it a recognized driver, a passenger alteration, a resistance mechanism or an incidental finding?
- Has a treatment been shown to benefit patients with this exact alteration in this cancer type or another applicable setting?
- Is that treatment approved or otherwise supported by credible clinical evidence for this patient?
- Does the patient's clinical condition permit treatment, and is it preferable to the available alternatives?
A variant of uncertain significance (VUS) should not ordinarily be used by itself to justify a cancer treatment. Additional evidence may clarify its significance over time, but uncertainty is not equivalent to an actionable target.
Potential benefits and limitations of targeted therapy
For an appropriately selected patient, targeted therapy can provide substantial disease control and may avoid some toxicities associated with other treatment strategies. It can also produce substantial adverse effects, including organ-specific toxicities, drug interactions and treatment-emergent complications. Its suitability depends on the drug and the patient, not just on the molecular report.
Targeted treatment can fail for several reasons: the alteration may not be the dominant driver, tumor cells may have different molecular profiles, drug exposure may be insufficient, or the tumor may evolve under treatment pressure. Resistance assessment is therefore an essential part of precision oncology rather than an afterthought.
7. Immunotherapy Biomarkers and Treatment Selection
Cancer immunotherapy aims to help the immune system recognize or attack malignant cells. Immune checkpoint inhibitors targeting PD-1, PD-L1 or CTLA-4 are established treatments in multiple cancers, but they do not work equally well for every patient.
Immunotherapy selection may depend on the cancer type, treatment setting, PD-L1 expression, MSI/dMMR status, TMB in selected indications and other clinical or molecular features.
7.1 PD-L1 expression
PD-L1 is a protein involved in immune checkpoint signaling. Some tumors use this pathway to reduce immune attack. PD-L1 testing can help determine eligibility for particular treatments in specific cancers.
PD-L1 results require careful interpretation. Different assays and scoring systems may be used, including tumor proportion score (TPS) and combined positive score (CPS). Their thresholds and indications are not interchangeable.
A high PD-L1 result does not guarantee response. A low or negative result does not exclude benefit in every cancer or treatment regimen.
7.2 MSI-H and dMMR
Mismatch repair deficiency can cause microsatellite instability and the accumulation of DNA errors. MSI-H/dMMR status can identify patients who may benefit from certain immune checkpoint inhibitors in approved or guideline-supported settings.
These are among the important tumor biomarkers in precision immuno-oncology. Even so, they are not perfect predictors, and response varies across tumor types and clinical situations.
7.3 Tumor mutational burden
TMB measures the number of mutations in a tumor using a defined assay. A higher mutation burden may increase the number of abnormal proteins that the immune system can recognize, but this relationship does not translate into a universal prediction of treatment benefit.
Technical differences among assays, cutoffs, tumor biology and other immune features all affect interpretation. TMB should be used only in the clinical contexts where the assay and treatment criteria are applicable.
7.4 Other emerging immunotherapy biomarkers
Research is examining gene-expression signatures, immune-cell infiltration, interferon signaling, antigen presentation, microbiome-related factors and other characteristics of the tumor microenvironment.
These approaches may refine patient selection in the future. Many remain investigational or lack sufficient evidence to guide routine treatment independently. Their presence on a research report should not be confused with a validated companion diagnostic.
Biomarkers change the estimated probability of benefit; they do not determine an individual's outcome with certainty. Biomarker results must be combined with cancer type, stage, treatment regimen, clinical status, other available therapies and the known risks of immune-related toxicity.
For a concise overview of established immune biomarkers and their limitations, see The Jackson Laboratory's May 2026 immunotherapy biomarker resource.
8. Targeted Therapy vs Immunotherapy vs Chemotherapy
These approaches are not universally competing alternatives. They may be used in different sequences, in combinations or in entirely different clinical circumstances.
| Feature | Targeted therapy | Immunotherapy | Chemotherapy |
|---|---|---|---|
| Primary mechanism | Interferes with a specific molecular target or pathway. | Modifies or supports immune-mediated tumor control. | Damages cells or disrupts processes required for cell division and survival. |
| Biomarker role | Often central when the drug is designed for a particular driver alteration. | Important for some treatments and diseases; less decisive or not required in others. | May depend more on cancer type, stage and established regimen than on a single target. |
| Common limitations | Acquired resistance, target heterogeneity and drug-specific toxicity. | Primary resistance, acquired resistance and immune-mediated adverse events. | Non-selective effects on normal dividing cells and drug-specific toxicity. |
| Monitoring | Clinical assessment, imaging, relevant laboratory tests and selected molecular testing. | Clinical assessment, imaging and monitoring for immune-related adverse events; selected biomarkers where useful. | Clinical assessment, imaging, blood counts and other regimen-specific safety monitoring. |
| Best choice | Depends on target validity, clinical evidence, patient eligibility and competing options. | Depends on treatment indication, expected benefit, toxicity and the overall disease context. | Depends on the disease setting and evidence for the regimen, including whether combination treatment is appropriate. |
Why sequencing matters
In some cancers, a biomarker-directed therapy provides an established treatment opportunity. In others, immunotherapy or a combined approach may be preferred. The order can influence the chance of disease control, cumulative toxicity and which treatments remain available later.
For example, a high PD-L1 score in a driver-positive lung cancer does not automatically settle the choice between immunotherapy and targeted treatment. The exact oncogenic alteration, disease setting, approved treatment options, expected benefit and safety implications of the sequence must be reviewed together.
There is no single best sequence for all cancers. A rational treatment plan should specify why a particular option is preferred, what alternatives exist and what findings would prompt a change.
9. Why Cancer Treatments Stop Working
Cancer resistance can be present before treatment starts or emerge after an initial response. Resistance can affect targeted therapy, immunotherapy, chemotherapy and combination regimens.
9.1 Primary resistance
Primary resistance occurs when a tumor does not respond adequately to a treatment from the outset. Reasons include:
- The identified alteration is not the dominant driver of tumor growth.
- The tumor has additional pathways that maintain growth despite the treatment.
- Different tumor regions contain different molecular populations.
- The drug cannot achieve adequate exposure at a relevant disease site.
- The tumor's immune environment or antigen-presentation machinery limits an immunotherapy response.
9.2 Acquired resistance
Acquired resistance develops during or after an initial period of benefit. Treatment can select for cell populations that are less sensitive to the drug, while cancer cells may also acquire new alterations or change their biological state.
| Resistance mechanism | Biological explanation | Potential clinical implication |
|---|---|---|
| On-target mutation | A new change in the drug target reduces treatment binding or effectiveness. | May support consideration of another inhibitor if a validated option exists. |
| Bypass signaling | Alternative pathways maintain growth despite inhibition of the original target. | May justify further profiling or a combination strategy in a supported setting. |
| Target amplification or activation | Increased target abundance or pathway activity helps the tumor overcome inhibition. | Requires interpretation in the context of the assay and available treatments. |
| Lineage or cell-state change | Tumor cells acquire different biological characteristics that alter drug sensitivity. | Repeat pathology, sometimes with molecular profiling, may be important. |
| Tumor heterogeneity | Some cell populations respond while others survive and expand. | A single biopsy or plasma sample may not capture every resistant population. |
| Immune escape | Tumor cells or the microenvironment evade immune recognition or suppress antitumor activity. | May explain immunotherapy failure, although actionable clinical options vary. |
9.3 Resistance to immunotherapy
Failure of immune checkpoint inhibition may reflect insufficient immune recognition, impaired antigen presentation, suppressive immune cells, changes in interferon signaling or other tumor-specific and host-related factors.
These mechanisms are biologically plausible and actively studied, but not every mechanism can currently be detected reliably or matched to an effective alternative treatment. A mechanistic explanation is not, by itself, evidence that a proposed rescue treatment will improve clinical outcomes.
9.4 Resistance is not the same as treatment failure caused by every other factor
Before concluding that the tumor has developed molecular resistance, the clinical team should also assess adherence, dosing, drug interactions, treatment interruptions, toxicity-related dose changes, imaging findings and the possibility of a new or unrelated condition.
Resistance is one explanation for progression, not a substitute for a complete clinical assessment.
For further scientific context, see the 2025 review Acquired resistance in cancer: towards targeted therapeutic strategies in Nature Reviews Cancer.
10. Resistance Testing and Treatment After Progression
When cancer progresses during treatment, the next treatment should not be chosen solely by repeating the original strategy or by selecting an experimental drug based on a single newly detected mutation.
The team should establish whether the progression is confirmed, whether it is limited or widespread, and whether additional testing is likely to change management.
Step 1: Confirm progression and clinical stability
Assess symptoms, physical examination, imaging, relevant laboratory results and the pace of progression. The immediate clinical risk can determine whether there is time for an extensive molecular investigation.
Step 2: Review the original molecular findings
Compare the original pathology and sequencing report with the treatment history. Identify the original drug target, expected resistance patterns and whether the cancer had a documented response.
Step 3: Consider new tissue or plasma testing
Repeat tissue biopsy may identify a new histological pattern, altered protein expression or molecular resistance mechanism. Plasma ctDNA testing may identify emerging genomic alterations when a validated test is appropriate.
These approaches are complementary rather than universally interchangeable. If a blood-based assay is negative or non-informative, it may not exclude an alteration that is present in the tumor.
Step 4: Distinguish a validated resistance target from an exploratory finding
Some acquired alterations have established treatment implications. Others are supported mainly by laboratory studies, small case series or early-phase trials. A report that labels an alteration as potentially resistant should be checked against the treatment history and the strength of the clinical evidence.
Step 5: Compare the next treatment options
Options may include another targeted agent, a different established regimen, an evidence-supported combination, immunotherapy where appropriate, local treatment for selected sites of progression, a clinical trial or symptom-focused supportive care.
Not all resistance mechanisms have a proven rescue treatment. When a matched option is not established, the priority remains choosing the best available evidence-based approach rather than assuming that every molecular finding is treatable.
Step 6: Define the next reassessment point
Before a new treatment begins, document how response will be assessed, what toxicity requires intervention, when imaging will be repeated and which findings would warrant treatment modification.
At progression, ask three separate questions: Is the cancer truly progressing? Has its biology changed in a clinically meaningful way? Is there a proven treatment that acts on the newly identified change? These questions should not be treated as equivalent.
11. Liquid Biopsies and Circulating Tumor DNA
Circulating tumor DNA (ctDNA) consists of small DNA fragments released into the bloodstream by tumor cells. Plasma-based molecular testing can help identify genomic alterations without requiring an immediate tissue biopsy.
Potential uses
- Finding treatment-relevant alterations when tissue testing is difficult or a biopsy poses unacceptable risk.
- Obtaining potentially actionable results more quickly when tissue results would arrive too late for a treatment decision.
- Assessing selected genomic resistance mechanisms in advanced cancer.
- Supporting research into treatment response and molecular residual disease.
Important limitations
Not every tumor releases enough ctDNA to be detected. Some cancers and disease sites shed very little DNA into the bloodstream. A plasma test may also have limitations in detecting certain fusions, copy-number changes or other alteration types, depending on the assay.
A negative result therefore does not necessarily mean that the tumor lacks an actionable alteration. Tissue testing should be pursued where clinically appropriate if a negative, inconclusive or discordant result could otherwise lead to a missed treatment option.
What the 2026 ASCO guideline means for practice
The June 18, 2026, ASCO guideline on ctDNA testing in solid tumors and lymphoma provides several important guardrails:
- ctDNA testing for tumor genetic alterations may be appropriate when tissue testing is difficult, unsafe, unavailable in time or allowed by the relevant drug indication.
- A negative or inconclusive ctDNA result should prompt consideration of tissue-based testing when clinically appropriate.
- ctDNA should not be used as a blanket replacement for other testing methods.
- Outside clinical trials, testing should generally be offered when a result can inform a clinical decision.
- Routine treatment changes based solely on molecular residual disease findings or nonspecific ctDNA concentration measures are not justified by this general guideline without adequate evidence of clinical utility.
For details, read the full ASCO guideline: Circulating Tumor DNA Testing in Solid Tumors and Lymphoma.
Molecular residual disease is a distinct question
Molecular residual disease (MRD) testing seeks to detect cancer-associated molecular signals after treatment, sometimes when imaging shows no obvious disease. In selected cancers, MRD can provide prognostic information and help estimate recurrence risk.
However, detecting a molecular signal earlier than conventional assessment does not automatically prove that starting treatment earlier improves survival. The clinical utility depends on the cancer, test, threshold, evidence and resulting intervention. Decisions to escalate, de-escalate or start treatment based on MRD results should follow disease-specific evidence and applicable guidelines.
12. AI, Multi-Omics and Emerging Precision Oncology
AI and computational methods may help connect the large amount of information generated by modern cancer diagnostics. Their greatest value is as decision-support tools that organize evidence and identify patterns for clinical review, rather than as independent authorities that prescribe treatment.
12.1 Multi-omics integration
Multi-omics approaches combine different types of biological data, potentially including:
- Genomics: mutations, fusions, amplifications and other DNA changes.
- Transcriptomics: RNA expression and selected evidence of active biological pathways.
- Proteomics: protein abundance, signaling and protein-level changes.
- Epigenomics: regulatory changes that influence gene activity.
- Spatial and single-cell analysis: information about tumor-cell populations and their local environments.
Integrated analysis may improve biological understanding and identify candidates for further investigation. Yet a predicted pathway dependency is not necessarily a clinically actionable vulnerability. A proposed drug match must still be assessed against human clinical evidence and treatment-specific eligibility.
12.2 AI-assisted treatment matching
An AI-supported precision oncology system could connect a patient's verified diagnosis and biomarkers to:
- Relevant approved treatment indications and clinical guidelines.
- Published clinical trials and the patient characteristics used to select trial participants.
- Known resistance mechanisms and associated evidence.
- Drug interactions, contraindications and safety considerations requiring clinician review.
- Evidence gaps, uncertainty and alternative treatment pathways.
A reliable system should identify the source of every recommendation, display the date of the evidence review and make uncertainty visible. It should not treat a computational association or a preclinical result as equivalent to a successful randomized clinical trial.
12.3 Patient-derived organoids and functional testing
Researchers are investigating patient-derived tumor models, including organoids and other experimental systems, to test how cancer cells respond to candidate treatments. These approaches could eventually complement genomic matching in selected situations.
However, assay standardization, representativeness of the sampled tumor, turnaround time and evidence that model-guided decisions improve patient outcomes remain essential considerations. Functional testing should not be presented as a universally validated replacement for standard treatment selection.
12.4 Emerging treatment platforms
Personalized cancer vaccines, cellular therapies, radiopharmaceuticals, novel antibody–drug conjugates and other investigational platforms expand the range of potential precision treatments. Their availability, eligibility criteria and evidence vary considerably.
A treatment can be scientifically promising without being proven to improve meaningful outcomes in a given cancer. Emerging approaches should be classified by the maturity of the evidence and, where appropriate, offered through ethically conducted clinical trials.
13. How to Assess Evidence and Actionability
One of the most common problems in precision oncology is confusing biological plausibility with demonstrated clinical benefit. A molecular tumor board or treating oncologist should assess both the strength of the biomarker evidence and the evidence for the proposed intervention.
A practical actionability framework
The following framework is an editorial navigation tool, not a replacement for formal clinical guidelines or an official regulatory classification.
| Evidence tier | What it means | How to interpret the finding |
|---|---|---|
| Tier 1: Established clinical actionability | A biomarker–treatment relationship is supported by an applicable regulatory indication or established clinical guideline. | Check the exact indication, assay, eligibility, treatment sequence and patient-specific suitability. |
| Tier 2: Supported but context-dependent | Credible human clinical evidence supports a treatment in a defined setting, but its applicability or regulatory status may differ. | Clarify the strength of the evidence, the relevant population and whether expert review is needed. |
| Tier 3: Investigational clinical evidence | Early clinical studies suggest possible benefit, but efficacy or the appropriate patient group remains uncertain. | Consider a relevant clinical trial; do not assume the treatment is established. |
| Tier 4: Preclinical or mechanistic hypothesis | The proposed match is primarily supported by laboratory experiments, computational predictions or biological reasoning. | Treat as hypothesis-generating unless stronger clinical evidence exists. |
| Tier 5: Unknown or non-actionable | The finding has uncertain significance or no credible treatment link has been established for the current decision. | Do not use the result alone to justify treatment. Seek clarification or reassessment where appropriate. |
Evidence quality and the outcome that matters
For a proposed treatment, evaluate the study population, study design, comparator, sample size, follow-up, response durability, progression-free survival, overall survival, quality of life and adverse effects, as relevant.
Objective response rate can be important, particularly in advanced disease, but a response rate alone does not establish longer survival or better quality of life. The magnitude and durability of benefit should be evaluated against appropriate alternatives.
Single-patient reports and retrospective series can generate hypotheses, especially in rare cancers. They generally cannot establish comparative treatment efficacy because of selection bias, confounding, publication bias and the absence of an appropriate comparator.
How to evaluate a molecular report
For each potentially actionable result, record the following:
- Finding: exact gene, variant, fusion, amplification or protein-expression result.
- Assay: testing method, specimen source and relevant quality limitations.
- Classification: pathogenic, likely pathogenic, oncogenic, uncertain or other laboratory designation as applicable.
- Tumor context: cancer type, histology, stage and prior treatments.
- Treatment evidence: approved indication, guideline support, human clinical data or preclinical evidence.
- Expected benefit and harm: outcomes from comparable patients, relevant uncertainty and treatment-specific toxicity.
- Next action: treatment discussion, confirmatory testing, expert review, clinical trial evaluation or no immediate action.
The objective is not to maximize the number of drugs suggested by a report. It is to identify the options that are most defensible for the individual patient.
14. Questions to Ask the Oncology Team
Patients and caregivers can use the following questions to make a molecular treatment discussion more productive.
About the diagnosis and testing
- Has the cancer diagnosis and histological subtype been confirmed?
- Which biomarkers are required before the next treatment decision?
- Was testing performed on tumor tissue, blood or another specimen?
- Did the test cover relevant mutations, fusions, amplifications and protein-expression markers?
- Could inadequate tissue or a negative plasma result have missed an important finding?
About treatment selection
- Which finding supports the proposed treatment, and how strong is the evidence?
- Is the treatment approved for this cancer and clinical setting in our jurisdiction?
- How does it compare with other established options?
- What are the realistic goals: cure, recurrence reduction, tumor shrinkage, disease control or symptom relief?
- What are the main risks, treatment interactions and monitoring requirements?
About resistance and next steps
- What findings will determine whether treatment is working?
- When will response be reassessed?
- If progression occurs, would repeat biopsy or ctDNA testing change the treatment decision?
- Is there a relevant clinical trial, and what evidence supports it?
- When would a second opinion from a disease-specific specialist or molecular tumor board be valuable?
15. Clinical Trials and Additional Resources
Clinical trials can provide access to investigational treatment approaches, including biomarker-selected therapies that are not yet standard care. A trial match should be evaluated using the exact eligibility criteria rather than the presence of a promising biomarker alone.
- ClinicalTrials.gov — search recruiting and ongoing trials by cancer type, intervention, biomarker and location.
- National Cancer Institute: Biomarker Testing for Cancer Treatment — patient-focused explanation of testing, interpretation and limitations.
- ASCO: Molecular Testing and Biomarkers Guidelines — professional guidance on molecular testing and biomarker-related decisions.
- FDA: Authorized Companion Diagnostic Devices — information on diagnostic tests linked to particular treatments in the United States.
- National Cancer Institute — authoritative educational material on cancer diagnosis, treatment and research.
16. Frequently Asked Questions
What is the difference between precision oncology and personalized cancer treatment?
The terms overlap. Precision oncology emphasizes using measurable tumor and patient characteristics to guide treatment decisions. Personalized cancer care is broader and includes clinical circumstances, goals, preferences, supportive care and other individual needs.
Does every patient with advanced cancer need comprehensive genomic profiling?
Not necessarily. Testing is recommended in many advanced cancers when the results may influence treatment, and it is routine in selected disease settings. The appropriate test depends on the diagnosis, available therapies, likely clinical utility, specimen quality, timing and cost. The oncology team should explain which results could change management.
Does finding an actionable mutation mean targeted therapy will work?
No. A validated biomarker can increase the likelihood of benefit in a defined setting, but response is not guaranteed. Tumor heterogeneity, coexisting alterations, resistance mechanisms, clinical factors and drug-specific limitations can all affect the outcome.
Can immunotherapy work when PD-L1 is negative?
In some cancers and treatment settings, yes. PD-L1 is not a universal predictor, and negative results do not exclude benefit from every immune checkpoint inhibitor or regimen. The decision must follow disease-specific evidence and applicable treatment criteria.
What is the difference between MSI-H, dMMR and TMB-high?
MSI-H describes a pattern of microsatellite instability. dMMR describes deficient mismatch repair function, often assessed through expression of mismatch repair proteins. TMB-high refers to a high estimated number of tumor mutations using a defined assay. These markers are related in some tumors but are not interchangeable.
Can a blood liquid biopsy replace a tissue biopsy?
Not universally. Plasma ctDNA can be useful when tissue is unavailable, risky to obtain or too slow for a clinically important decision. However, a negative or inconclusive plasma result may require tissue testing, depending on the circumstances and the biomarker sought.
Should treatment be changed when ctDNA or a molecular residual disease test becomes positive?
Not automatically. The clinical utility of acting on a result depends on the cancer, assay and evidence that the resulting intervention improves outcomes. Outside an appropriately designed clinical trial or an established disease-specific pathway, a molecular result alone may not justify changing treatment.
What should happen when targeted therapy stops working?
The clinical team should confirm progression, review the treatment history and consider whether repeat tissue or plasma testing can identify a clinically meaningful resistance mechanism. The next treatment should be selected from evidence-supported options, considering disease behavior, patient condition, safety and clinical trial eligibility.
Can AI choose the best cancer treatment?
AI can help organize molecular findings, identify relevant studies, summarize evidence and generate treatment hypotheses. Its recommendations require validation against current clinical evidence and independent clinical review. A predicted drug match is not proof of efficacy.
When should a molecular tumor board or second opinion be considered?
It can be particularly useful when a case has multiple potentially actionable alterations, conflicting test results, uncommon histology, a possible resistance mechanism, a proposed off-label treatment or several competing clinical trial options. Expert review does not guarantee a better outcome, but it can help clarify evidence and uncertainty.
Conclusion: Precision Is More Than Finding a Mutation
Precision oncology works best when every step connects to the next: a reliable diagnosis, an appropriate biomarker test, an accurate interpretation, a treatment supported by clinical evidence, and a plan to monitor benefit, harm and resistance.
The central objective is not to find the largest number of molecular abnormalities or the longest list of possible drugs. It is to determine which findings are clinically meaningful, which treatment is most defensible for the patient, and what evidence should trigger reassessment.
As genomic profiling, ctDNA testing, AI and multi-omics continue to develop, the same principle remains essential: biological plausibility, predictive accuracy and demonstrated clinical benefit are different levels of evidence. Keeping them separate is fundamental to trustworthy precision cancer care.
This guide is for education and informed discussion with qualified healthcare professionals. It does not diagnose cancer, recommend an individual treatment or replace oncology consultation. Treatment indications and regulatory approvals differ across countries and can change as new evidence emerges. Do not delay or discontinue prescribed cancer treatment based on a biomarker report, an online resource or an AI-generated recommendation.
References and Evidence Sources
- National Cancer Institute. Biomarker Testing for Cancer Treatment.
- Lockwood CM, et al. Circulating Tumor DNA Testing in Solid Tumors and Lymphoma: ASCO Guideline. Journal of Clinical Oncology Oncology Practice. Published June 18, 2026.
- US Food and Drug Administration. List of FDA-Authorized Companion Diagnostic Devices.
- The Jackson Laboratory. Immunotherapy Biomarkers. Clinical Education resource, May 2026.
- Soragni A, et al. Acquired resistance in cancer: towards targeted therapeutic strategies. Nature Reviews Cancer. Published June 3, 2025.
- Chakravarty D, et al. Somatic Genomic Testing in Patients With Metastatic or Advanced Cancer: ASCO Provisional Clinical Opinion.
- US Food and Drug Administration. Tissue-Agnostic Drug Development in Oncology.
- ClinicalTrials.gov. Clinical trial registry.
Editorial method: This guide distinguishes established clinical uses from exploratory research and emphasizes the importance of disease-specific guidelines, validated diagnostic methods and regulatory labels. Biomarker–drug examples are representative rather than exhaustive. Clinicians should verify current indications, companion diagnostic requirements and applicable local guidance before making treatment decisions.
.png)
.png)
.png)
Comments
Post a Comment