K043576 · Roche Molecular Systems, Inc. · NTI · Jan 10, 2005 · Clinical Toxicology
Device Facts
Record ID
K043576
Device Name
ROCHE AMPLICHIP CYP450 TEST
Applicant
Roche Molecular Systems, Inc.
Product Code
NTI · Clinical Toxicology
Decision Date
Jan 10, 2005
Decision
SESE
Submission Type
Traditional
Regulation
21 CFR 862.3360
Device Class
Class 2
AI Performance
Output
Algorithm
Acceptance
Observed
Dev DS
Dev Readers
Test DS
Test Readers
CYP2C19 Genotype
Microarray-based hybridization pattern analysis with specific algorithms
—
99.6% agreement with sequencing
—
—
Method comparison study: 122 samples analyzed by bi-directional DNA sequencing; Reproducibility study: 809 samples tested at three sites.
—
Indications for Use
The Roche AmpliChip CYP450 test is intended to identify a patient's CYP2C19 genotype from genomic DNA extracted from a whole blood sample. Information about CYP2C19 genotype may be used as an aid to clinicians in determining therapeutic strategy and treatment dose for therapeutics that are metabolized by the CYP2C19 gene product.
Device Story
Microarray-based genotyping test; identifies CYP2C19 genotype from genomic DNA extracted from whole blood. Process: PCR amplification of target DNA; fragmentation; biotin labeling; hybridization to oligonucleotide probe microarray; staining with streptavidin-phycoerythrin; laser scanning. Software analyzes hybridization patterns to determine genotype (wild-type, mutant, heterozygous) based on probe intensity. Used in clinical laboratory settings by trained technicians. Output: Genotype report used by clinicians to predict enzymatic activity and adjust therapeutic strategy/dosing. Benefits: Personalized drug therapy based on metabolic phenotype.
Clinical Evidence
Bench testing only. Performance evaluated using 399 clinical samples compared against bi-directional DNA sequencing, PCR-RFLP, and allele-specific PCR. Overall genotype call rate and percent agreement were 99.7%. Specificity for wild-type samples was 100%. Reproducibility study across three sites using a six-member panel showed 99.8% call rate and 99.9% correct call rate. Interference studies confirmed no impact from elevated albumin, bilirubin, or lipids. Limit of detection established at 25 ng genomic DNA.
Technological Characteristics
Microarray-based genotyping; glass substrate with ~15,129 oligonucleotide probes (16-22 bases). Sensing via laser-induced fluorescence of phycoerythrin-labeled DNA. Connectivity: Affymetrix GeneChip platform (Fluidics Station 450Dx, Scanner 3000Dx). Software: GCOS v1.1 and AmpliChip CYP450 Data Analysis Software. PCR amplification using GeneAmp PCR System 9700. Single-use device.
Indications for Use
Indicated for patients requiring CYP2C19 genotyping to assist clinicians in determining therapeutic strategy and dosing for drugs metabolized by the CYP2C19 enzyme.
Regulatory Classification
Identification
A drug metabolizing enzyme genotyping system is a device intended for use in testing deoxyribonucleic acid (DNA) extracted from clinical samples to identify the presence or absence of human genotypic markers encoding a drug metabolizing enzyme. This device is used as an aid in determining treatment choice and individualizing treatment dose for therapeutics that are metabolized primarily by the specific enzyme about which the system provides genotypic information.
Special Controls
The special control is FDA's guidance document entitled "Class II Special Controls Guidance Document: Drug Metabolizing Enzyme Genotyping System."
*Classification.* Class II (special controls). The special control is FDA's guidance document entitled “Class II Special Controls Guidance Document: Drug Metabolizing Enzyme Genotyping Test System.” See § 862.1(d) for the availability of this guidance document.
{0}
1
# 510(k) SUBSTANTIAL EQUIVALENCE DETERMINATION DECISION SUMMARY ASSAY ONLY TEMPLATE
A. 510(k) Number:
k043576
B. Purpose for Submission:
Clearance of new device
C. Measurand:
Genotype of Cytochrome P450 2C19 (CYP2C19)
D. Type of Test:
Genotyping microarray
E. Applicant:
Roche Molecular Systems, Inc.
F. Proprietary and Established Names:
Roche AmpliChip CYP450 microarray
G. Regulatory Information:
1. Regulation section:
21 CFR §862.3360, drug metabolizing enzyme genotyping system
2. Classification:
Class II
3. Product code:
NTI, drug metabolizing enzyme genotyping system
4. Panel:
Toxicology (91)
H. Intended Use:
1. Intended use(s):
The Roche AmpliChip CYP450 test is intended to identify a patient's CYP2C19 genotype from genomic DNA extracted from a whole blood sample. Information
{1}
about CYP2C19 genotype may be used as an aid to clinicians in determining therapeutic strategy and treatment dose for therapeutics that are metabolized by the CYP2C19 gene product.
2. Indication(s) for use:
See intended use above.
3. Special conditions for use statement(s):
For professional use.
The information provided from this test may supplement therapeutic decision-making and should only be used in conjunction with routine monitoring by a physician. Because of the variability in the knowledge of clinical utility with specific drugs that are metabolized by CYP2C19, clinicians should use professional judgement in the interpretation of results from this type of test. Results from this type of assay should not be used to aid in predicting a patient's response to drugs for which, 1) the drug metabolizing enzyme activity of the allele, or 2) the drug metabolic pathway, has not been clearly established.
4. Special instrument requirements:
Affymetrix GeneChip® Microarray platform (GeneChip Fluidics Station 450Dx, GeneChip Scanner 3000Dx with Autoloader, Data station for the GeneChip Operating Software and AmpliChip CYP450 Data Analysis Software, GeneChip Operating Software (GCOS), version 1.1)
I. Device Description:
The Roche AmpliChip CYP450 2C19 test is a microarray-based genotyping test. The AmpliChip contains two distinct multiplexed drug metabolizing enzyme genotyping tests – one for CYP2D6 and one for CYP2C19. The CYP2D6 portion was reviewed and cleared under k042259.
The test is based on several processes: PCR amplification of purified genomic DNA, fragmentation and labeling of the amplified products, hybridization of the amplified products to a microarray, staining of the bound products, scanning of the microarray, and determination of the CYP450 genotype and predicted phenotype. The AmpliChip CYP4502C19 Test is designed to identify specific nucleic acid sequences and query for the presence of certain known sequence polymorphisms through analysis of the pattern of hybridization to a series of probes that are specifically complementary either to wild-type or mutant sequences.
Microarrays of oligonucleotide probes synthesized on a glass substrate are utilized for the analysis. Probe microarrays are manufactured in a series of cycles. The glass substrate is coated with linkers containing photolabile protecting groups. A mask is then applied that
2
{2}
exposes selected portions of the probe microarray. Illumination removes the photolabile protecting groups enabling selective nucleoside phosphoramidite addition only at the previously exposed sites. Next, a different mask is applied and the cycle of illumination and chemical coupling is performed again. By repeating this cycle, a specific set of oligonucleotide probes is synthesized, with each probe type in a known location. The AmpliChip CYP450 microarray consists of a square grid of 15,129 probes, each of which contains approximately $10^{7}$ copies of the specific oligonucleotide probe. Each probe sequence is 16 to 22 bases in length. A single Probe Set consists of four Probes, or Features, which have a fixed target except for at the substitution position where an A, C, G, and T are included to generate four unique probes. Of these four probes, one is designated the Perfect Match (PM) Probe based on the known genome sequence, and the other three are called Mismatch (MM) Probes. A Probe Set Pair consists of a Wild-type Probe Set and a Mutant Probe Set. Both Probe Sets are designed to hybridize to the same region of the target, but one is designed for the Wild-type allele and the other includes a known polymorphism. The assay is designed to distinguish 3 alleles of the CYP2C19 gene.
# J. Substantial Equivalence Information:
1. Predicate device name(s): Roche AmpliChip CYP450 microarray (CYP2D6)
2. Predicate 510(k) number(s): k042259
3. Comparison with predicate:
| Similarities | | |
| --- | --- | --- |
| Item | Device | Predicate |
| assay | AmpliChip contains probes for detecting both CYP2D6 and CYP2C19 genotype | AmpliChip contains probes for detecting both CYP2D6 and CYP2C19 genotype |
| technology | same | same |
| Manufacturing process | same | same |
| Differences | | |
| --- | --- | --- |
| Item | Device | Predicate |
| gene | CYP2C19 | CYP2D6 |
| Number of alleles | 3 | 27 |
# K. Standard/Guidance Document Referenced (if applicable):
NCCLS Guideline EP5-A, Evaluation of Precision Performance of Clinical Chemistry Devices
NCCLS Guideline EP7-A, Interference Testing in Clinical Chemistry
{3}
L. Test Principle:
The AmpliChip CYP450 Test permits the analysis of specific nucleic acid sequences and a query for the presence of known sequence polymorphisms through analysis of the pattern of hybridization to a series of probes that are specifically complementary either to wild-type or mutant sequences. Microarrays of oligonucleotide probes synthesized on a glass substrate are utilized for the analysis.
AmpliChip CYP450 probe microarrays are manufactured using technology that combines photolithographic methods and combinatorial chemistry. Over 15,000 different oligonucleotide probes are synthesized on a glass surface to analyze both sense and antisense strands of an amplified target DNA specimen. Within the $20 \times 20 \ \mu \mathrm{m}^2$ probe microarray, each probe type is located in a specific area called a probe cell, which contains approximately $10^6 - 10^7$ copies of a given probe. The AmpliChip CYP450 Microarray utilizes approximately 240 probes to detect each polymorphism.
The AmpliChip CYP450 Test amplifies the CYP450 genes in two separate reactions that are subsequently pooled after PCR amplification. The reaction containing CYP450 Primer Mix A uses primers that generate amplified product encompassing the promoter region and coding regions of the CYP2D6 gene and a CYP2D6 gene duplication-specific product, when present in the specimen or control. The reaction containing CYP450 Primer Mix B uses primers that generate amplified product encompassing a CYP2D6 gene deletion specific product, when present, in the specimen or control, as well as the 3 CYP2C19 alleles that this device is designed to detect. Amplification occurs only in the region of the CYP2C19 and gene between the primers; the entire genome is not amplified. Amplification is performed using the Applied Biosystems GeneAmp PCR System 9700 thermal cycler, utilizing a 35 cycle program.
The DNA amplicon from the two independent amplification reactions are pooled for each specimen and cleaved by incubation with a fragmentation mix (DNAse I and calf intestinal Alkaline Phosphatase) to generate small DNA fragments of an average size of 50 – 200 nucleotides. The fragmented DNA amplicon are subsequently labeled with biotin at their $3'$ termini by the action of Terminal Deoxynucleotide Transferase (TdT). The biotin-labeled CYP450 target DNA fragments are added to hybridization buffer containing a hybridization control. The mixture is hybridized to the oligonucleotides located on the AmpliChip CYP450 Microarray using the Affymetrix GeneChip Fluidics Station 450Dx and an AmpliChip CYP450-specific protocol. The hybridized AmpliChip CYP450 Microarray is washed and stained with a streptavidin-conjugated fluorescent dye (phycoerythrin). After staining, the CYP450 Microarray is scanned by an Affymetrix GCS3000Dx Scanner using a laser that excites the phycoerythrin moiety. The amount of emitted light is proportional to bound target DNA at each location on the probe microarray.
The image of each of the approximately 15,000 probe cells is stored in a data file and used for data analysis. Data analysis is performed by the GeneChip Operating Software (GCOS) and the AmpliChip CYP450 Data Analysis Software in several steps: (1) the
4
{4}
GeneChip Operating Software automatically places a grid over the image of the scanned microarray to demarcate the individual probe cells and to calculate the mean intensity of each probe cell, (2) the AmpliChip CYP450 Data Analysis Software uses CYP450-specific algorithms to analyze the intensity patterns and determine the genotype at each specified polymorphic site by analyzing the relative extent of hybridization to probes complementary to mutant and wild-type targets on the probe microarray, and (3) the AmpliChip CYP450 Data Analysis Software algorithm determines the genotype (wildtype, mutant, or heterozygous) at each specified polymorphism and compares this information to combinations of the genotype patterns from known alleles to identify the corresponding alleles for the test DNA. If the polymorphisms correspond to a defined allele of the respective gene, the allele is called with standard nomenclature. A report is generated that summarizes the genotype and lists the corresponding identified polymorphisms and alleles. The genotype information is used to predict an individual's CYP2C19 enzymatic activity based upon published studies.
## M. Performance Characteristics (if/when applicable):
### 1. Analytical performance:
#### a. Precision/Reproducibility:
To evaluate the reproducibility of the AmpliChip CYP450 Test a six-member panel was constructed from cell lines that represent all 3 CYP2C19 alleles on the array. The Reproducibility Panel samples were tested at a concentration of 50 ng DNA/PCR. The genotypes of the Reproducibility Panel samples are summarized below.
| CYP2C19 genotype |
| --- |
| *1 / *1 |
| *1 / *2 |
| *1 / *3 |
| *1 / *1 |
| *1 / *2 |
| *1 / *2 |
Testing was conducted at three sites; including two external sites and a laboratory at Roche Molecular Systems. The Reproducibility Panel was tested in triplicate for five runs by one operator at each of the three sites, using three lots of reagents.
Genotype calls for CYP2C19 were made for 807/809 (99.8%) samples. The 3 incorrect results included 2 samples that did not provide a genotype call (yielded a "no call" result) and 1 sample that incorrectly identified the genotype (genotype CYP2C19 *1/*1). The overall results were as follows: 806/809 samples called correctly (99.6%).
Twenty three system errors were encountered in the study. Twenty one system errors were related to scanner failures and two were related to fluidic station errors. Test results
{5}
were successfully obtained for twenty of the 21 scanner failures after re-scanning. Results for the two fluidic station errors were successfully obtained after repeat testing of samples from the amplification plate.
| CYP2C19 genotype | No. Tested | Genotype Calls N (%) | Correct Genotype Calls | Correct Call Rate Estimate (95% CI) |
| --- | --- | --- | --- | --- |
| *1 / *1 | 134 | 134 (100.0) | 133 | 0.99 (0.97) |
| *1 / *2 | 135 | 135 (100.0) | 135 | 1.00 (0.98) |
| *1 / *3 | 135 | 135 (100.0) | 135 | 1.00 (0.98) |
| *1 / *1 | 135 | 135 (100.0) | 135 | 1.00 (0.98) |
| *1 / *2 | 135 | 134 (99.3) | 134 | 1.00 (0.98) |
| *1 / *2 | 135 | 134 (99.3)) | 134 | 1.00 (0.98) |
| Total | 809 | 807 (99.8) | 806 | 1.00 (0.99) |
Note that in the above study, an additional sample was tested. The sample was identified as a $*2/*2$ sample based on testing by the Amplichip. During reproducibility testing, the Amplichip called the sample $*2/*2$ 100% of the time. Later sequencing of the sample revealed an extra polymorphism that is not tested by the chip making that sample a $*2/*10$ sample. This panel was removed from the reproducibility data analysis. The miscall is noted in the method comparison section below. The labeling also contains a limitation warning users that $*10$ alleles may be called by the chip as $*2$ alleles. Both $*2$ and $*10$ are predicted to be poor metabolizers.
The failure rate of the AmpliChip CYP450 Test system was evaluated by testing 100 replicates of genomic DNA purified from a whole blood specimen using a commercially available blood DNA extraction kit. The test solution contained approximately $50\mathrm{ng}$ DNA/PCR of the $*1/*1$ CYP2C19 genotype. There was one System Failure event where no result was obtained due to inability to scan the stained AmpliChip CYP450 Microarray resulting in a Whole System Failure rate of $1\%$ with a $95\%$ confidence interval from 94.55 - $99.97\%$ due to the instrument or the AmpliChip CYP450 Microarray. There was a $0\%$ Whole System Failure rate due to the AmpliChip CYP450 Test amplification and detection reagents. Of 100 valid replicates, one chip failed to scan the initial and subsequent attempts resulting in failure to produce a result.
# b. Linearity/assay reportable range:
Not applicable
c. Traceability, Stability, Expected values (controls, calibrators, or methods):
The Amplichip does not require calibration. It is a single use device. The sponsor recommends that the user follow the calibration and maintenance schedule recommended by the instrumentation manual.
{6}
The sponsor recommends that samples (anticoagulated whole blood) to be tested be stored at room temperature for up to 7 days, at 2-8 °C, for up to 1 month, or frozen at -20 °C for up to 7 weeks. Blood specimens can be subjected to up to five freeze thaw cycles. Fragmented amplicon can be stored at 2-8 °C for up to 18 hours. Protocols and acceptance criteria were described and found acceptable.
## d. Detection limit:
The limit of detection of the AmpliChip CYP450 Test was determined by analysis of dilutions of two genomic DNA samples to 2.5, 25 and 50 ng DNA/PCR. The concentration of the DNA samples was determined by use of a PicoGreen double stranded DNA Quantitation kit. The DNA samples were of genotype *2/*2 and *1/*1 for the CYP2C19 gene. The % positivity rate was determined from the number of correct genotype calls. The lowest level of genomic DNA at which a ≥ 95% positivity rate was obtained for correct detection of the CYP2C19 gene was 2.5 ng of input DNA. However, because of the limit of detection of the other gene on the chip (CYP2D6 - see K042259), the limit of detection for the system is 25ng of input DNA.
| DNA Amount (ng) | Number of Arrays | Number of Correct Calls | Positivity Rate | 95% Confidence Limit |
| --- | --- | --- | --- | --- |
| 50 | 144 | 144 | 100% | 97.5 – 100% |
| 25 | 144 | 144 | 100% | 97.5 – 100% |
| 2.5 | 144 | 144 | 100% | 97.5 – 100% |
## e. Analytical specificity:
Ten unique patient samples were tested with and without spiking of albumin, bilirubin and triglycerides to the following levels (approximately 10-fold greater than normal): Albumin - 6000 mg/dL; Bilirubin - 60 mg/dL; Triglycerides - 3000 mg/dL. The samples were extracted using a commercially available blood DNA extraction kit. Elevated levels of lipids, bilirubin and albumin in specimens did not interfere with the performance of the AmpliChip CYP450 Test.
This test has been validated for use with only human blood collected in EDTA anticoagulant. Testing of other specimen types may result in incorrect results or no results.
Potential carryover contamination was assessed with five runs of alternating two specimens of distinct genotype, along with the appropriate controls. The position of the specimens plus controls was varied between the runs. No carryover contamination was observed; the appropriate CYP2C19 genotype was obtained for all specimens.
## f. Assay cut-off: Not applicable
2. Comparison studies:
{7}
a. Method comparison with predicate device:
Method comparison studies were performed using bi-directional DNA sequencing as the comparator for the AmpliChip CYP450 test. DNA sequence analysis for genotype confirmation was performed for 122 samples with CYP2C19 genotype that had been previously analyzed by the AmpliChip CYP450 Test. One sample that was identified as CYP2C19 *2/*2 genotype by RFLP and by the AmpliChip CYP450 Test was shown by sequencing to be a CYP2C19 *2/*10 genotype. With this miscall, the agreement between the AmpliChip CYP450 Test and sequencing for CYP2C19 alleles was 99.6%. The AmpliChip CYP450 Test does not report CYP2C19 *10 alleles (see Procedural Limitations). The results are summarized below.
| CYP2C19 Allele | Number of Alleles Sequenced | AmpliChip Results | | | |
| --- | --- | --- | --- | --- | --- |
| | | Correct Calls | Miscalls | No Calls | Percent Agreement |
| *1 | 153 | 153 | 0 | 0 | 100.0% |
| *2 | 79 | 78 | 1 | 0 | 98.7% |
| *3 | 14 | 14 | 0 | 0 | 100.0% |
| Total | 246 | 245 | 1 | 0 | 99.6% |
The percent agreement for genotype detection of the AmpliChip CYP450 Test was calculated by determining the percentage of tested samples with the correct genotype assigned as compared to the total number of samples tested of that genotype. Genotype detection results for CYP2C19 were also calculated for the individual alleles and by sample. Genotype detection was evaluated using genomic DNA samples at approximately 50 ng/PCR. In addition to the sequencing confirmation presented above, additional samples were evaluated by PCR-RFLP.
| CYP2C19 Allele | Number of Unique Alleles Tested | Number of Correct Calls | Number of Miscalls | Number of No Calls | Percent Agreement | Number of Replicates |
| --- | --- | --- | --- | --- | --- | --- |
| *1 | 647 | 647 | 0 | 0 | 100% | 842 |
| *2 | 137 | 136 | 1 | 0 | 99.3% | 176 |
| *3 | 14 | 14 | 0 | 0 | 100% | 32 |
| Total | 798 | 797 | 1 | 0 | 99.9% | 1050 |
| CYP2C19 Genotype | Total Unique Samples | Number of Correct calls | Number of Miscalls | Number of No Calls | Percent Agreement | Genotype Call Rate |
| --- | --- | --- | --- | --- | --- | --- |
| *1/*1 | 270 | 270 | 0 | 0 | 100.0% | 100.0% |
| *1/*2 | 101 | 101 | 0 | 0 | 100.0% | 100.0% |
| *1/*3 | 6 | 6 | 0 | 0 | 100.0% | 100.0% |
{8}
| CYP2C19 Genotype | Total Unique Samples | Number of Correct calls | Number of Miscalls | Number of No Calls | Percent Agreement | Genotype Call Rate |
| --- | --- | --- | --- | --- | --- | --- |
| *2/*2 | 15 | 14 | 1^{1} | 0 | 93.3% | 93.3% |
| *2/*3 | 6 | 6 | 0 | 0 | 100.0% | 100.0% |
| *3/*3 | 1 | 1 | 0 | 0 | 100.0% | 100.0% |
| Total | 399 | 398 | 1 | 0 | 99.7% | 99.7% |
One sample, shown by sequencing to be CYP2C19 *2/*10*, was miscalled as a CYP2C19 *2/*2 genotype.
The genotype call rate and percent agreement for CYP2C19 was 99.7% for all 399 tested samples.
b. Matrix comparison:
Not applicable
3. Clinical studies:
a. Clinical Sensitivity:
Not applicable
b. Clinical specificity:
Not applicable
c. Other clinical supportive data (when a. and b. are not applicable):
The package provides information about predicted phenotypic activity of the alleles contained on the Amplichip test. The predicted phenotypes were identified from data in the literature for each allele. The literature references that were used to determine predicted phenotypes are listed below.
There are varying amounts of supportive data in these literature references to support phenotypic determinations for drugs that are metabolized by CYP2C19. In some cases, alleles have only been tested for phenotype in one or a few drugs and have not been verified for all drug classes that may be CYP2C19 substrates. Clinicians should use caution in predicting phenotype and adjusting treatment strategy for patients who express alleles that have not been investigated for activity in metabolizing a specific drug.
{9}
10
| Allele | Nucleotide Change | Predicted Enzyme Activity | Reference |
| --- | --- | --- | --- |
| *1 | None | Normal | Romkes et al. 1991
Richardson et al, 1997
Blaisdell et al, 2002 |
| *2 | 681G>A | None | de Morais et al, 1994a
Ibeanu et al, 1998b |
| *3 | 636G>A | None | de Morais et al, 1994b |
Some CYP450 alleles are not reported by the AmpliChip CYP450 Test. These alleles are listed below:
| CYP450 | Alleles Not Reported by AmpliChip CYP450 Test |
| --- | --- |
| CYP2C19 | 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16 |
New CYP450 alleles not identified at the time of release of AmpliChip CYP450 Data Analysis Software version 2.0 will not be correctly detected by the AmpliChip CYP450 Test. Most likely, a “No Call” result will be obtained for the relevant CYP450 gene.
Samples with CYP2C19 *2/*10 genotype may be miscalled by the AmpliChip CYP450 Test as CYP2C19 *2/*2 genotype. The expected phenotype of both of these genotypes is poor metabolizer.
4. Clinical cut-off:
Not applicable
5. Expected values/Reference range:
| Allele | Chinese | Black | Caucasian |
| --- | --- | --- | --- |
| *1 | 65% | 18% | 84% |
| *2 | 30% | 17% | 15% |
| *3 | 5% | <1% | <1% |
N. Proposed Labeling:
The labeling is sufficient and it satisfies the requirements of 21 CFR Part 809.10.
O. Conclusion:
The submitted information in this premarket notification is complete and supports a substantial equivalence decision.
Predicate graph will load when search results are available.
Embedding visualization will load when search results are available.
PDF viewer will load when search results are available.
Loading panels...
Select an item from Submissions
Click any panel, subpart, regulation, product code, or device to see details here.
Section Matches
Results will appear here.
Product Code Matches
Results will appear here.
Special Control Matches
Results will appear here.
Loading collections...
Loading
My Alerts
You will receive email notifications based on the filters and frequency you set for each alert.
Sort by:
Create Alert
Search Filters
Agent Token
Create a read-only bearer token for Claude, ChatGPT, or other agents that can call HTTP APIs.
Copy this now. It will not be shown again.
Connected apps
Apps you authorized through browser sign-in. Disconnecting revokes their access immediately.
Learn the FDA Browser
Two short videos show you everything — or skip straight to the written tutorial if you'd rather read. You can reopen this any time from the Tutorial button in the top bar.
Part 1 — Search, results, and everyday workflows 16 min
Part 2 — Embeddings: the galaxy map 3 min
1. Search: exact and fuzzy
Type a phrase like "coronary artery calcification" into the search box. You get two kinds of results. Exact results match the literal phrase — prefix searches work ("coronary artery calcificati") but suffix searches do not. Fuzzy results match on the meaning and intent of your phrase rather than the exact words, and are sorted by relevance score. Hover over the Exact or Fuzzy badge on any row to see exactly why it matched.
Use the checkboxes above the results to narrow: SaMD keeps only software-only devices, AI / ML keeps only devices with AI.
Exact vs. fuzzy search: what's the difference?
Exact matches on the literal phrase (prefix search works, suffix does not). Fuzzy matches on the meaning and intent of the phrase rather than the exact words. Hover over the badge on any row to see why it matched.
You search "coronary artery calcification" and want only software devices with AI. What two filters do you apply?
Narrow by SaMD (software-only devices), then narrow by AI/ML (devices with AI).
2. The results table
Scroll right in the results table. The intended use is extracted for you — no need to open the PDF. The device story gives a high-level snapshot of what the device does and how it's used. The AI Performance sub-table shows each output name, acceptance criteria, observed values, and development/test dataset descriptions — the same format Innolitics uses for regulatory strategy outputs, and the fastest high-level fingerprint of an AI device. It is AI-generated but has been very reliable in practice.
Where do you find a device's intended use without opening the PDF?
Scroll right in the search results table. The intended use column is extracted for you; no need to dig into the 510(k) summary PDF.
What does the AI Performance sub-table show, and why is it useful?
Output name, acceptance criteria, observed values, development dataset description, and test dataset description. It's the same format we use for regulatory strategy output and Fast 510(k) input, and the fastest high-level fingerprint of an AI device. AI-generated but reliable in practice.
3. Judging fuzzy relevance
Fuzzy results trail off in relevance as you scroll. Use three signals to decide how far down to go: the fuzzy badge explanations, the intended use column, and whether your target output (e.g., Cobb angle) still appears in the AI Performance sub-table. Once it stops appearing, you're past the relevant zone. A top hit with a low score (~0.4) and a stretched explanation is a hint the closest predicates are far away — the project may be headed for De Novo. Note the fuzzy search is a pattern match: it doesn't handle negation ("not") well, and hardware devices can appear — filter by SaMD/AI ML to cut them.
How do you judge how far down fuzzy search results to go?
Use the relevancy signals: the fuzzy badge explanations, the intended use column, and whether the target output (e.g., Cobb angle) still appears in the AI Performance sub-table. Once it stops appearing, results are trailing off in relevancy.
4. Device detail page: chat and citations
Click a device name to open its detail page: device facts on the left, a chat window on the right. Ask something like "Describe the training data". The answer carries little citation bubbles — click one to jump to the highlighted passage in the source PDF, so you can verify every AI answer against the document. There's also a Download PDF button for sharing.
How do you verify an AI chat answer on the device detail page?
Click the citation bubbles to jump to the relevant highlight in the source document.
Reading rule for every project: how many summaries do you read in full?
At least the three most relevant 510(k) or De Novo summaries, in full. After that, use targeted chat questions to confirm your memory quickly. The tool supports this professional habit — it doesn't replace it.
5. Side-by-side comparison
Select multiple rows in the results table (aim for under ~10), then open the PDF Viewer tab. Ask one question — it goes to all selected devices in parallel, each with citations. This is the fastest way to compare and contrast devices: training data, PCCP scope, how they handled adding new scanners, and so on.
What does the side-by-side PDF viewer mode do?
Select multiple devices, open the PDF viewer tab, and ask one question (e.g., "Describe the training data"). It queries all selected devices simultaneously with citations, so you can compare and contrast quickly.
6. Collections
With rows selected, go to the Collections tab and create a labeled collection (e.g., "Cobb Angle Project"). Reload that selection any time — before a client call, pull up the collection and ask questions across all of its devices at once.
How do you save a set of selected devices for later use?
Select the rows, go to the Collections tab, and create a labeled collection (e.g., "Cobb Angle Project"). You can reload the selection anytime and carry it into the PDF viewer and other tabs that support selections.
7. Product codes and the regulations tree
Click a product code in the results to jump to it in the regulations tree — identification text, sibling product codes, and devices you can open in a PDF viewer on the right. Click a regulation number to see its identification, special controls, and related product codes. You can also search by product code or regulation number at the top of the tree. Always read the special controls if any exist for your device — it broadens your search and sharpens pre-kickoff research.
What can you do from the regulations tree view?
Browse product codes and regulation numbers, read the identification text and special controls, browse sibling product codes, open device PDFs on the right, and search by product code or regulation number at the top of the tree.
8. Chart view
Click Show Chart and segment by regulation number (or product code) to see which regulations dominate your result set. Clicking a regulation takes you into the regulations tree. Great for spotting that most matches are, say, hardware laparoscopic devices — a cue to go back and filter.
How do you see which regulations dominate a search result set?
Click "Show Chart" and segment by Regulation Number. Clicking a regulation takes you to the regulations tree.
9. The predicate graph
Open the Predicates tab for a family-tree view of predicate relationships. Click a node to trace its parents and children; selections from search carry over pre-selected. Commonly predicated devices are worth reading — a lot of people predicated them for a reason. The visual lineage is also handy on client calls, e.g. to show how a predicate family evolved and justify why your predicate still holds.
In the predicate graph, why are commonly predicated devices worth reading?
A lot of people predicated them for a reason. Clicking a node traces parents and children, and selections from search carry over pre-selected.
10. Embeddings: the galaxy map
The Embeddings tab plots every matching document in a 2-D "galaxy map" where semantically similar devices cluster together. Hover or click clusters to explore, and let AI label the clusters for you. Embeddings beat product codes for grouping: two devices can carry different product codes (LLZ vs. QIH) yet do the same thing — the embedding captures the meaning of the intended use and device story. This is also exactly how retrieval-augmented generation (RAG) works under the hood, and it makes a great visual on client calls.
Try it yourself
Head to the search page and work through a few of these AI/ML fuzzy searches to build intuition: perivascular fat on CT · aortic valve calcification opportunistic screening on noncontrast CT · breast cancer prediction on digital pathology slides · autism detection · gestational age prediction · a hearing aid that can also detect a pulse · foundation model based analysis of ECG · large language models · penetration test. Watch how the relevance scores, intended use, and AI Performance tables tell you when results stop being meaningful.