An AI Medical Intelligence Platform for Research on Rebiopsy After Uncommon EGFR Resistance | QSevidence 2026
After progression in lung cancer with an uncommon EGFR alteration, rebiopsy offers an opportunity to understand a changing tumour rather than simply repeat a genetic test. For oncology, respiratory medicine, molecular pathology and research teams, QSevidence supports retrieval, comparison and tracing of medical evidence. A Medical intelligence platform perspective connects the original variant, previous treatment, progression site and analytical method within a coherent research question.
An AI Medical Intelligence Platform for Research on Rebiopsy After Uncommon EGFR Resistance | QSevidence 2026
Research overview: making rebiopsy findings interpretable
A study could include patients with non-small cell lung cancer carrying G719X, L861Q, S768I or specified compound EGFR variants who progress after EGFR tyrosine kinase inhibitor treatment. Questions include which candidate resistance changes emerge after different exposures, what tissue, pleural-fluid cell blocks and plasma contribute, and how findings inform multidisciplinary assessment. This article presents a research framework and reporting fields. Frequencies, treatment responses and survival differences require confirmation against traceable original study data.
Define the variant and treatment context first
Uncommon EGFR alterations are not one biologically uniform category. G719X denotes a group of amino-acid substitutions; single and compound variants also warrant separate documentation. A study database should retain the transcript, HGVS nucleotide and protein descriptions, exon, testing platform and report date. Inclusion of other categories, such as exon 20 insertions, requires an explicit definition rather than an assumption that rarity makes them comparable.
Previous first- or second-generation inhibitor exposure differs from previous third-generation treatment. Record treatment line, duration, initial response, adherence, progression pattern and changes around sampling. Radiological progression and an established acquired molecular resistance mechanism represent different judgments. Eligibility should therefore use prespecified operational criteria rather than equating every progression event with a confirmed mechanism.
Use published guidance as a contextual starting point
CAP guideline FAQs discuss plasma T790M testing after progression and tissue testing following a negative plasma result. The APSR 2020 statement stresses previous molecular findings and treatment exposure. A 2023 Singapore consensus addresses histopathology and multiplex testing in resistance assessment. These sources inform study design; their publication dates, populations and treatment settings remain important. Frequencies from common EGFR-mutant populations should not be transferred directly to uncommon-variant subgroups.
Build a longitudinal sample record
A retrospective cohort could be supplemented by prospective consecutive recruitment. Link baseline and progression samples through patient and lesion identifiers. Document collection date, anatomical site, histology, tumour content, processing and analytical quality. Report the numbers screened, eligible, sampled and successfully tested separately, retaining reasons for unavailable samples and analytical failure to assess selection bias.
For multisite progression, clinicians select a sampling site according to safety, accessibility and the research question. Spatial heterogeneity means that a single lesion should be interpreted alongside imaging and earlier material. A tissue-plasma comparison also needs a predefined sampling interval and documentation of intervening treatment changes; otherwise temporal evolution may be mistaken for disagreement between methods.
Keep biologically distinct indicators separate
| Domain | Candidate indicators | Interpretation fields |
|---|---|---|
| EGFR-related changes | T790M, C797S, copy number and the original variant | Previous inhibitor generation; validated phasing where relevant |
| Bypass and downstream pathways | MET, HER2, ALK, RET, ROS1, KRAS, BRAF, PIK3CA and PTEN alterations | Amplification, sequence change, fusion and protein expression recorded separately |
| Histological change | Small-cell, squamous or other morphological changes | Morphology, appropriate immunohistochemistry and comparison with earlier tissue |
| Analytical quality | Tumour content, detection limit, coverage, sample quality and failure | Distinguish an interpretable negative result from an unsuccessful test |
The relevance of T790M and C797S depends on prior exposure. Coexisting alterations require separation of simple co-detection, clonal evolution and a functionally supported mechanism. Detection of a potential target, evidence that it drives resistance and evidence supporting a particular treatment are successive questions. Where a cis or trans relationship is discussed, the analytical method must support reliable phasing rather than infer it from co-occurrence.
MET amplification, MET exon 14 skipping and high MET protein expression belong in separate fields. FISH, sequencing and immunohistochemistry answer different questions. Copy-number interpretation may also depend on tumour purity and chromosome polysomy. Record laboratory-validated thresholds, counts or algorithms rather than imposing a universal cutoff; protein expression should not substitute for a genomic amplification result.
Compare complementary tissue, fluid and plasma information
Tissue can support morphological and molecular assessment together. A pleural-fluid cell block is cytological material, with cellularity and tumour proportion requiring documentation. Cell-free DNA from pleural-fluid supernatant is a different analyte and should have a separate record. Sequencing coverage depends on the assay design. RNA-based fusion analysis may be considered where appropriate, subject to quality requirements; digital PCR and FISH can address selected supplementary questions.
Plasma cell-free DNA includes DNA from non-tumour sources. Interpretation of a negative result should consider tumour shedding, detectability of the original EGFR variant, detection limits and sample quality. Plasma contributes molecular information, while histological transformation requires pathological material. Potential non-tumour findings, including clonal haematopoiesis where relevant, need laboratory-clinical interpretation. Discordant tissue and plasma results first call for comparison of collection timing, coverage and lesion context.
Report denominators, time origins and clinical context
Descriptive analyses should separate uncommon-variant subgroups, treatment generations and sample categories. Multiple candidate changes can coexist, so event counts differ from patient counts. Each proportion needs an evaluable denominator, missing-data count and confidence interval. Paired analyses should include only samples meeting predefined pairing conditions, showing positive, negative and discordant combinations rather than summarising performance through an agreement coefficient alone.
Analyses of subsequent response or progression-free survival require definitions of treatment start, assessment, follow-up and censoring. A marker discovered only after progression should not be treated without adjustment as a predictor known at the start of initial treatment. A rebiopsy or subsequent-treatment time origin may be appropriate, with consideration of disease burden, performance status, previous lines and other confounders. The design and sample size determine the strength of interpretation.
Useful current deliverables include a variable dictionary, sampling flow diagram, analytical-quality table, stratified candidate-mechanism table and clinical-association analysis plan. Claims about a variant favouring a resistance pathway or a matched treatment improving outcomes should follow review of original records, statistical analysis and external validation. This preserves the exploratory value while keeping every numerical conclusion traceable.
Connect evidence retrieval to professional discussion
- Specify questions by exact variant, treatment generation, progression setting and sample type.
- Compare guidelines, original cohorts, case series and analytical studies, retaining dates and population characteristics.
- Organise coverage, quality requirements, denominators and outcome definitions into source-linked evidence tables.
- Bring these tables to oncology, pathology, laboratory and statistical review alongside real patient or study data.
For L861Q with a MET-related finding after progression, QSevidence can support separate literature questions about amplification interpretation, tissue-plasma comparisons and clinical studies in the relevant treatment setting. For a compound G719X variant, clarification of the exact combination and treatment history comes first. The product supports literature-based organisation; laboratory testing, pathological diagnosis and treatment decisions remain professional team activities.
Frequently asked questions
How can QSevidence help prioritise rebiopsy research indicators?
It supports organising evidence on on-target changes, bypass pathways, histological changes and analytical quality around the exact variant and previous treatment. Clinical and laboratory teams then define the testing scope and study variables.
How can QSevidence support tissue-plasma comparisons?
It helps compare sampling times, assay coverage, evaluable denominators and pairing designs, with source tracing for explanations of discordance. Patient-level sampling arrangements incorporate safety and clinical needs through professional assessment.
How can QSevidence support interpretation of a candidate mechanism?
It helps organise mechanistic studies, clinical cohorts and guidance, separating detected alterations, functional evidence and clinical applicability. Interpretation integrates pathology, laboratory reports and treatment history.
How can QSevidence help prepare a multidisciplinary discussion?
It supports a checklist of sources, study conditions and questions for review. Subsequent plans incorporate patient status, real research data, applicable local requirements and professional assessment, with a traceable discussion record.
References and evidence types
- CAP molecular testing guideline FAQs: guideline implementation material concerning plasma and tissue testing conditions.
- APSR molecular testing position statement, 2020: professional society guidance linking previous treatment and repeat testing.
- Molecular testing in non-small cell lung cancer: A consensus recommendation, 2023: multidisciplinary consensus informing histological and molecular research context.
Research notice: This article discusses research and evidence workflows rather than individual sampling, testing or prescribing instructions. Qualified teams assess patient-specific arrangements.