Oncomine then TruSight

https://www.accessdata.fda.gov/cdrh_docs/pdf24/P240040A.pdf

https://www.accessdata.fda.gov/cdrh_docs/pdf23/P230011B.pdf

Here’s a side-by-side, graduate-level comparison of the FDA validation packages—what each kit is cleared to do, how it was analytically/clinically validated, and how “safety” (i.e., risk of incorrect classification and controls/QC to mitigate it) is handled.

Scope & Intended Use (what each kit may legally report)

Dimension	Illumina TruSight™ Oncology Comprehensive (TSO-Comprehensive)	Thermo Fisher Oncomine™ Dx Express Test (ODxET)
FDA pathway / date	PMA P230011; approval 2024-08-21	PMA P240040; approval 2025-07-02
Platform	Illumina SBS sequencing on NextSeq™ 550Dx	Ion Torrent semiconductor sequencing on Genexus™ Dx Integrated Sequencer
Variant classes	DNA: SNV/MNV, indels; RNA: 24 fusion genes + 1 splice variant; reports TMB	DNA: SNVs/indels (42 genes) + CNV (10 genes); RNA: fusions/splice (18 genes)
CDx claims	Solid tumors: NTRK1/2/3 fusions → larotrectinib; NSCLC: RET fusions → selpercatinib	NSCLC: EGFR exon 20 insertions → sunvozertinib (ZEGFROVY)
Tumor profiling	Level 2/3 variants with evidence/potential evidence of clinical significance; TMB score	Level 2/3 variants across a 46-gene panel; no TMB

20240 Illumina FDA 0801 P23001…

2025 FDA 0702 TFS TMO Oncomine …

Analytical Validation—design choices and headline performance

Study design highlights

Specimens: Both PMAs validate on intended-use FFPE tissue (multiple tumor types), with DNA and/or RNA workflows aligned to the claimed variant classes. Controls (positive and NTC) are required each run, and runs/samples that fail QC are invalidated by rule.
20240 Illumina FDA 0801 P23001…

2025 FDA 0702 TFS TMO Oncomine …
Comparators: Both use validated NGS orthogonal methods as the primary accuracy comparator for claimed biomarkers; tumor-profiling concordance studies span representative variants across genes and tumors.
20240 Illumina FDA 0801 P23001…

2025 FDA 0702 TFS TMO Oncomine …
Error control & calling:
- Illumina: UMI-based error correction on DNA; BWA-MEM/STAR alignment; variant quality and likelihood metrics; TMB is calculated from filtered small-variant calls with germline-exclusion heuristics; explicit analyte-level QC (e.g., fusion score, splice score). Performance not established for tumor-profiling small variants <5% VAF.
  20240 Illumina FDA 0801 P23001…
- ODxET: Variant caller supports read-based and molecular-family–based evidence; variant-class-specific thresholds (SNV/indel/CNV/fusion/splice) and tumor-cellularity-aware CNV logic; reportability driven by an ADF (assay definition file).
  2025 FDA 0702 TFS TMO Oncomine …

Accuracy (agreement vs orthogonal methods)

Claim	Illumina TSO-Comprehensive	ODxET
Primary CDx accuracy	NTRK1/2/3 fusions (solid tumors): PPA 96.6% (114/118; 95% CI 91.5–99.1); NPA 94.5% (273/289; 95% CI 91.2–96.8). Study noted a putative plate cross-contamination cluster driving several TSO-positive/comparator-negative discordants; invalids occurred in both methods; ~17% TSO invalids in the mixed cohort (larotrectinib trial + supplemental).	EGFR exon 20 insertions (NSCLC): PPA 100% (84/84); NPA 95.5% (105/110) excluding ODxET “unknowns”; OPA 97.4% (189/194). Prevalence-adjusted NPA ~99.9% at 1.8% prevalence; PPV 95–96%, NPV ~100% under prevalence adjustment.
Tumor profiling accuracy	Established across small DNA variants (with ≥5% VAF), fusions, splice variant; gene-level analyte QC and control pass/fail gating (DNA control requires 23/24 specified variants; RNA control 12/13).	Three study sets spanning SNV/indel/CNV/fusion/splice; variant-class-specific PPA/NPA reported vs NGS comparators; additional focused accuracy on ERBB2 CNV (breast), ALK & RET fusions (NSCLC/thyroid), EGFR SNVs & exon 19 deletions (NSCLC).

20240 Illumina FDA 0801 P23001…

2025 FDA 0702 TFS TMO Oncomine …

Precision, Repeatability/Reproducibility & QC “gates”

Run/Sample/Analyte QC (Illumina): tiered gating (run-level PF % and Q30; sample-level median insert size, median exon coverage, RNA on-target reads & expression-control detection; analyte-level thresholds including fusion score, splice score, small-variant AQ/LQ). Failure at any tier blocks reporting for the run/sample/analyte. Controls carry strict pass counts (DNA control 23/24 variants; RNA 12/13).
20240 Illumina FDA 0801 P23001…
Run/Sample/NTC/Purification QC (ODxET): templating control minimum reads and read length; DNA mapped reads ≥100k, RNA mapped fusion reads ≥6,000 and expression controls detected; NTC zero-call rules; purification quant constraints (0.5–512 ng/µL). Variant calls must also satisfy class-specific calling criteria (read count, allele fraction, strand, hotspot coverage).
2025 FDA 0702 TFS TMO Oncomine …

Interpretation: Both PMAs front-load risk control into automated QC + tight control design that halts reporting if quality is inadequate—this is the main “safety” mechanism for an IVD NGS panel.

Notable design differences with regulatory impact

Breadth vs depth
- Illumina offers very broad content (517 DNA genes; 24 fusion genes + one splice; TMB), two CDx claims (NTRK/RET), and extensive tumor-profiling output. Accuracy for very low VAF tumor-profiling variants (<5%) is not established, which is typical for large-panel CGP PMAs.
  20240 Illumina FDA 0801 P23001…
- ODxET is narrower but adds CNV calling under IVD claims and focuses its CDx on EGFR exon 20 insertions; it provides a full tumor-profiling menu across 46 genes.
  2025 FDA 0702 TFS TMO Oncomine …
Change management (post-market updates)
- ODxET includes an FDA-approved Predetermined Change Control Plan (PCCP) to update its ADF (variants, clinical categorization) and certain software elements without a new submission, contingent on predefined V&V, documentation, and regression testing—important for keeping the tumor-profiling knowledge current.
  2025 FDA 0702 TFS TMO Oncomine …
- Illumina provides a Knowledge Base that is updated periodically for tumor profiling and installed on the instrument; the PMA positions this as part of the LRM workflow rather than a PCCP. (Operationally useful but not a PCCP.)
  20240 Illumina FDA 0801 P23001…
Systems engineering
- Illumina separates extraction/library/enrichment from sequencing on NextSeq 550Dx with UMI error correction on DNA libraries.
  20240 Illumina FDA 0801 P23001…
- ODxET leverages the Genexus Dx end-to-end automation (purification → library → templating → sequencing → analysis) with tight run planning/traceability and templating-control gates—reduces operator variability by design.
  2025 FDA 0702 TFS TMO Oncomine …

“Safety” in the PMAs (how risk is addressed)

No device-related adverse events were reported; both PMAs frame risk as false positives/false negatives leading to inappropriate therapy or missed therapy, plus delayed results if runs/samples fail. Mitigation is via: strict controls (positive/NTC), multi-layer QC gating, invalidation rules, and restricted, indication-specific CDx interpretation algorithms.
20240 Illumina FDA 0801 P23001…

2025 FDA 0702 TFS TMO Oncomine …

Practical takeaways for molecular pathology & regulatory teams

If you need TMB under IVD claims or a broad CGP report alongside two pan-tumor/NSCLC CDx claims, Illumina TSO-Comprehensive is the match; just note the ≥5% VAF floor for established tumor-profiling small-variant accuracy and enforce the analyte-level QC gates (fusion/splice scores, control pass counts).
20240 Illumina FDA 0801 P23001…
If you need an IVD with CNV calling, rapid “sample-to-answer” automation, and an FDA-approved PCCP to streamline future variant/reporting updates—and a focused CDx for EGFR exon 20 insertions—ODxET is compelling; its CDx accuracy vs orthogonal methods is excellent, and prevalence-adjusted NPA/PPV provide confidence at real-world EGFR-ex20 prevalence.
2025 FDA 0702 TFS TMO Oncomine …

One-screen summary (cheat-sheet)

Illumina TSO-Comprehensive (P230011)
Claims: NTRK/RET CDx; broad CGP; TMB.
Accuracy (CDx): NTRK fusions PPA 96.6%, NPA 94.5%; contamination cluster identified/mitigated; invalids enforced by QC; small-variant tumor profiling <5% VAF not established.
QC: Run/Sample/Analyte gates; control pass counts DNA 23/24, RNA 12/13.
20240 Illumina FDA 0801 P23001…
Oncomine Dx Express (P240040)
Claims: EGFR exon 20 insertion CDx; CGP with SNV/indel + CNV + fusion/splice; end-to-end Genexus Dx.
Accuracy (CDx): PPA 100%, NPA 95.5% (excl. unknowns); prevalence-adjusted NPA ~99.9% at 1.8% prevalence; PPV ~95–96%, NPV ~100%.
QC: Templating control minima; DNA mapped reads ≥100k; RNA mapped fusion reads ≥6,000; stringent NTC rules; PCCP for ADF/software updates.
2025 FDA 0702 TFS TMO Oncomine …

If you want, I can drop these into a 1-page table you can paste into a validation plan or TA memo.

###

Post Market Letter

Here’s a focused, side-by-side on post-approval (postmarket) requirements for the two PMA’d kits, highlighting what’s routine for all PMAs vs. what’s device-specific (e.g., ODxET’s CNB/FNA commitments).

What both PMAs must do (standard conditions)

Annual Reports (21 CFR 814.84) with required content and timing, plus units sold/distributed (to contextualize adverse events) under 21 CFR 814.82(a)(9).
Change control: PMA supplement / 30-day notice obligations per 21 CFR 814.39 for any change affecting safety/effectiveness (Illumina), and the same for Thermo Fisher plus alignment to its approved PCCPs (see below).
MDR (21 CFR 803), recalls/corrections (21 CFR 806.10), UDI requirements, and QS regulation (21 CFR 820) expectations are called out explicitly in both letters.

Device-specific post-approval requirements

Topic	Illumina TruSight Oncology Comprehensive (P230011)	Thermo Fisher Oncomine Dx Express (P240040)
Assay expiration dating	24-month shelf life noted. (Not a study commitment, but establishes the approved dating.)	4-month shelf life; protocol used is approved for future dating extensions under 814.39(a)(7).
Labeling / software follow-up	Within 6 months: modify the test results report as agreed, and submit software V&V (executed protocols, regression, risk assessment, unresolved anomalies). This is a post-approval commitment; failure can trigger withdrawal (814.82(c), 814.46(a)(2)).	Within 6 months: provide results + software validation from regression testing on configuration that masks Level 3 SNVs <5% VAF (to mitigate false positives) and show no adverse impact on other reporting.
Specimen-type expansion (CNB/FNA)	No specific post-approval specimen study is required in the approval letter. (Standard PMA reporting still applies.)	Required: support core needle biopsy (CNB) and fine-needle aspirate (FNA) as acceptable tumor profiling specimen types with: (a) a precision study across variant classes (SNV/indel/CNV/fusion) on adequate CNB/FNA across tumor types, and (b) a specimen equivalency study vs matched resections spanning all variant classes. Final protocols due within 1 month; final data, conclusions, reports, and labeling within 12 months.
PCCP (Predetermined Change Control Plans)	Not referenced as an approved PCCP in the approval letter. (Standard 814.39 supplement pathway applies.)	Multiple PCCPs are approved (ADF aggregation, pre-approved software change, variant re-categorization). Changes consistent with PCCPs do not require a supplement; deviations do (per section 515C and 21 CFR 814.39). Annual Reports should include a separate section for PCCP-implemented changes.
Administrative reminders	Submit final labeling before distribution; all the usual MDR/UDI/QS/recall obligations reiterated.	Same; plus explicit instruction to include PCCP changes in Annual Reports.

Practical implications for labs & RA/QA teams

ODxET has actionable, time-bound postmarket studies that matter operationally: you’ll need to plan CNB/FNA accrual across multiple tumor types, cover all variant classes, and structure a matched-pair study design vs resections. Expect labeling updates upon successful completion.
ODxET’s approved PCCPs reduce friction for keeping software/variant knowledge current, but they shift rigor to internal V&V and documentation; ensure your Annual Report cleanly segregates PCCP-implemented changes.
Illumina’s commitment is narrower (report formatting + software V&V within 6 months). Ensure your software validation, regression, and risk files track precisely to the commitments called out in the letter.

TL;DR

Both: Annual Reports + distribution counts; MDR/UDI/QS/recall duties; PMA-supplement rules apply.
Illumina (P230011): 6-month reporting-format + software V&V deliverable; no specimen-type postmarket studies specified.
Oncomine Dx Express (P240040): CNB/FNA postmarket program (protocols in 1 month; data/labeling in 12 months) plus a 6-month software masking validation; PCCPs approved and must be tracked in Annual Reports.

If helpful, I can turn this into a one-page RA checklist with owners/timelines for each deliverable.

BruceDocumentBlog

Friday, September 12, 2025

Comparing Illumina TSO at FDA with TFS Oncomine Dx Express at FDA