Federal regulations · Through 2026-10-07

21 CFR 866.6095: High throughput DNA sequencing for hereditary cancer predisposition assessment test system.

(a) Identification. A high throughput DNA sequencing for hereditary cancer predisposition assessment test system is a qualitative in vitro diagnostic (IVD) system intended for analysis of human DNA extracted from human specimens to detect germline mutations in a panel of targeted cancer-related genes. It is intended to aid in hereditary cancer predisposition assessment by qualified health care professionals in accordance with professional guidelines. The device is not intended for screening, prenatal testing, or as a stand-alone diagnostic test. The device is for prescription use only.
(b) Classification. Class II (special controls). The special controls for this device are:
(1) Design verification and validation must include:
(i) A description of genomic coverage that includes:
(A) A list of all genes, variant types, and target regions within each gene that the device detects;
(B) Summary information regarding the clinical significance of each gene, including references;
(C) A description of the genes with high clinical significance that are detected by the device, defined as genes for which the test result(s) may lead to prophylactic screening, confirmatory procedures, or treatment that may incur morbidity or mortality; and
(D) A description of any within-gene targeted regions that cannot be reported.
(ii) Specifications for specimen requirements, including any specimen collection devices, handling, and storage.
(iii) Specifications of the DNA extraction method and criteria for DNA quality and quantity that are prerequisite to performing the assay.
(iv) Detailed documentation of the methodology and protocols for each step of the test, including reagents, instrumentation, and software required. The documentation must include the analysis algorithms used for mutation detection and annotation, and specify the quality metrics, variant calling thresholds, and filters at each step of the test, including the criteria for run failures, batch failures, specimen failures, invalid calls (e.g., failed quality control), and “no calls” (i.e., absence of a result), as applicable.
(v) Description of required instrumentation and equipment, and any ancillary reagents, instrumentation, or equipment.
(vi) Detailed documentation of device software, including software applications and hardware-based devices that incorporate software. The documentation must include verification, validation, hazard analysis, and risk assessment activities.
(vii) Documentation of internal and external controls that are recommended or provided and control procedures. The documentation must identify those control elements that are incorporated into the testing procedure.
(viii) Detailed documentation pertaining to the probability of test failure based on data from clinical samples, description of scenarios in which a test can fail, and any risk mitigations, including follow-up actions to be taken.
(ix) Detailed documentation of the rules, procedures, tools, and criteria used for establishing mutation-hereditary disease relationships and mutation annotation, evaluation, and classification (e.g., pathogenic, likely pathogenic, variant of unknown significance, benign, and likely benign).
(x) Detailed documentation of any internal or external database(s) or decision rules used for mutation annotation, including:
(A) The protocol(s) used for variant interpretation, including training of personnel, monitoring accuracy of decision, resolution of discordant interpretations, and updating interpretations;
(B) Detailed documentation of the basis for interpretation, including the use of alternate databases, literature, and guidelines, and the basis for risk reporting;
(C) Methods for data preservation and security; and
(D) Data formats and nomenclature.
(xi) Information that demonstrates the performance characteristics of the device, evaluated either specifically for each gene/mutation or, when determined to be acceptable and appropriate by FDA, using a representative approach based on other mutations of the same type, including:
(A) Data that adequately support the intended specimen type(s) (e.g., whole blood), specimen handling protocol, and DNA extraction method.
(B) A summary of the evidence that demonstrates how the analytical quality metrics and thresholds used to determine the acceptability of reporting support the minimum accuracy requirements.
(C) Data to adequately support device accuracy using clinical specimens representing all indicated specimen types, mutation types, and size ranges intended to be detected and reported by the device. Accuracy data must fulfill the following:
(1) Accuracy of the device must be evaluated with clinical specimens collected in accordance with the device labeling and selected without bias, or well-characterized human cell line samples, when determined to be acceptable and appropriate by FDA.
(2) Accuracy must be evaluated by comparison to bidirectional Sanger sequencing or other orthogonal methods identified as appropriate by FDA. Performance criteria for both the comparator method(s) and the device must be predefined and appropriate to the device's intended use. Detailed study protocols must be documented.
(3) A sufficient number of specimens must be tested. For BRCA1 and BRCA2 genes, a minimum of 120 variant positive specimens must be tested. For other genes with high clinical significance, at least 40 variant-positive specimens must be tested per gene, including representative specimens for each indicated variant type based on a justification determined to be appropriate and acceptable by FDA. For remaining genes detected by the device, testing must include variant positive specimens representing each variant type, unless the variant type has a prevalence of less than 0.01 percent. Specimen selection must be prioritized based on clinical significance. The selected specimens must be representative of zygosity and challenging genomic context (e.g., guanine-cytosine content, near tandem repeats and homopolymer stretches, pseudogene), and must cover the range of sizes (for insertions, deletions, copy number variant (CNV) amplifications and CNV deletions) intended to be detected and reported by the device.
(4) Except as permitted by FDA under paragraph (b)(1)(xi)(C)(5) of this section, tested specimens must be selected based on results obtained from the orthogonal method. Positive percent agreement (PPA) and negative percent agreement (NPA) must be calculated and demonstrated for each variant type detected and reported by the device, as well as for clinically relevant variants. PPA is calculated as the number of variants that are tested positive by both the device and the orthogonal method (true “positives” (TP)) divided by the number of variants tested positive as determined by the orthogonal method (TP plus false negatives (FN) by the device). NPA is calculated as the number of variants that are tested negative (wild type) by both the device and the orthogonal method (true “negatives” (TN)) divided by the number of variants tested negative (wild type) by the orthogonal method (TN plus false positives (FP) by the device). Point estimates for PPA and NPA must be calculated along with 95 percent two-sided confidence intervals (CI). Uncertainty of the point estimate must be within an acceptable range, as identified by FDA, and must be demonstrated using the 95 percent CI.
(5) When it is determined by FDA to be appropriate and acceptable to select samples based on the results obtained with the device, accuracy must be presented as technical positive predictive value (TPPV) and technical negative predictive value (TNPV). TPPV relates to the likelihood that a variant call is a true positive and reflects the number of false positives per test. TPPV is calculated as the number of variants that are tested positive by both the device and the orthogonal method (TP) divided by the number of variants tested positive by the device (TP plus FP). TNPV relates to the likelihood that a variant call is a true negative and reflects the number of false negatives per test. TNPV is calculated as the number of variants that are tested negative by both the device and the orthogonal method (TN) divided by the number of variants tested negative by the device (TN plus FN).
(6) Any “no calls” or invalid calls in the study must be reported separately. The percent of final “no calls” or invalid calls must be clinically justifiable.
(7) Accuracy as a function of each performance metric (e.g., coverage depth, base quality scores) must be documented to provide evidence of the accuracy of the overall run.
(8) Detailed documentation for accuracy of the device must include information and results for the overall study, each mutation type, and each gene. The accuracy must further be described based upon stratification within each mutation type by zygosity, genomic context, and size (for indels and CNVs). Overall accuracy for reporting of substitutions must be ≥99.0 percent; insertions and deletions, ≥99.0 percent; CNVs, ≥99.0 percent for positive agreement (PPA, TPPV); and ≥99.9 percent for negative agreement (NPA, TNPV).
(D) Documentation of the data to adequately support device precision using clinical specimens representing all specimen types, mutation types, and sizes intended to be detected and reported by the device. The precision study must fulfill the following:
(1) The study must be performed using multiple instruments and multiple operators, on multiple non-consecutive days, and using multiple reagent lots. If the device is to be performed at more than one site, different sites must be included and reproducibility across sites must be evaluated.
(2) Representative clinical specimens of each mutation type must be tested (both positive and negative), considering clinical significance, prevalence, zygosity, genomic context, and size (for indels and CNVs). The precision for CNV detection must be demonstrated on the gene level for genes with high clinical significance. Alternatively, a justification for why such data are not needed must be found acceptable and appropriate by FDA.
(3) The study must assess the performance of all steps, including DNA extraction, unless a separate extraction study is performed.
(4) The study must use predefined performance criteria. Agreement estimates such as PPA/NPA and average positive agreement/average negative agreement must be provided, including point estimates and 95 percent confidence intervals. Documentation from the precision study must be demonstrated for the overall precision study, in addition to each mutation type, each gene, and each sample. Precision must further be demonstrated upon stratification within each mutation type by zygosity, genomic context, and size (for indels and CNVs). The overall precision point estimates for each variant type must be >99.0 percent.
(5) Any “no calls” or invalid calls in the study must be included in precision study results and reported separately. The percent of “no calls” or invalid calls and key quality control metrics parameters (e.g., coverage, sequencing score) must be summarized and based on stratification in the same way as the precision estimates.
(E) Documentation of the nucleic acid assay input range and the evidence to adequately support the range.
(F) Detailed documentation of additional analytical validation studies, including endogenous and exogenous interfering substances, specimen and reagent stability, cross-reactivity, carryover and cross-contamination, guard-banding, and index misassignment, as applicable. If specimens are pooled, index cross-contamination must be evaluated and demonstrate that pooling does not negatively impact test performance.
(G) Specimen type and matrix comparison data must be generated if more than one specimen type or anticoagulant can be tested with the device, including failure rates for the different specimen types.
(xii) Information that adequately supports the variant annotation and clinical interpretation of the test must include:
(A) A summary documenting the clinical significance for each gene on the test panel, including the associated conditions/cancers, the most prevalent and representative mutations, and summary of clinical evidence with references, including expected frequency in the general population and different ethnicities, and risks of developing the disease in relevant ethnic populations and the general population.
(B) Detailed documentation of the data to adequately support the performance of the variant annotation algorithms (e.g., concordance studies between the device generated variant classifications and externally established variant classifications, manual classifications by medical professionals, or classifications generated from clinical test reports).
(C) Documentation of any procedures or protocols for incorporation of any updates of valid scientific evidence into variant classification algorithms.
(2) The labeling required under § 809.10 of this chapter must include the following, as applicable:
(i) The intended use must include a description of the intended specimen type(s) and matrix (e.g., whole blood), the validated germline mutation types (e.g., single nucleotide variant, insertion, deletion, CNV), and a statement that the test is for hereditary cancer predisposition assessment and to aid in identifying hereditary genetic variants potentially associated with a diagnosed cancer.
(ii) The name of the testing facility or facilities (e.g., for single-site assays).
(iii) A summary of device description in accordance with paragraphs (b)(1)(i) through (b)(1)(viii) of this section.
(iv) A section that provides summary information on how the test works, how to interpret the results of the test, and an explanation on the database(s) used for mutation annotation.
(v) A summary of the information that demonstrates the performance characteristics of the device as required under paragraph (b)(1)(xi) of this section.
(vi) The following limiting statements:
(A) A statement that the test is not intended for use as a stand-alone diagnostic to diagnose cancer or other health conditions, and is not intended for use for prenatal testing nor as a cancer screening test.
(B) A statement that the test is specifically designed for heritable germline mutations and is not appropriate for the detection of somatic mutations.
(C) A description of the intended test population.
(D) A statement that the risk of cancer or disease for an individual cannot be predicted.
(E) A statement that: test results should be interpreted in the context of clinical findings, family history, lifestyle, environment, and other factors; molecular testing may not detect all possible mutations leading to cancer predisposition; a negative result does not rule out the possibility that the individual has an unidentified variant leading to cancer; and for more information, physicians ordering the test may wish to consult with a clinical medical geneticist or genetic counselor.
(F) A statement that other factors, such as ethnicity, may affect whether the test results are relevant for a particular patient and may also affect how their genetic health results are interpreted.
(G) A statement describing the situations that a patient should not receive the test (e.g., a patient with bone marrow transplant).
(H) A statement describing the challenging genomic contexts that may have reduced performance, where results should be interpreted with care.
(I) A statement disclosing the genetic coverage of the test, including any gaps in coverage.
(J) Statements describing testing conditions that were identified to cause test failures (e.g., low specimen volume, poor DNA quality).
(vii) For variants detected and reported by the device under the category of “variants of uncertain significance” or equivalent designation, a limiting statement that the clinical significance has not been demonstrated with adequate clinical evidence in accordance with established guidelines (e.g., professional guidelines).
(viii) For variants detected and reported by the device under the category of “variants with evidence of clinical significance” or equivalent designation, reference(s) for physicians to access internal or external information concerning decision rules or conclusions about the level of evidence for clinical significance.

Where this section sits in the code

Title 21—Food and Drugs
CHAPTER I—FOOD AND DRUG ADMINISTRATION, DEPARTMENT OF HEALTH AND HUMAN SERVICES
SUBCHAPTER H—MEDICAL DEVICES
PART 866—IMMUNOLOGY AND MICROBIOLOGY DEVICES
Subpart G—Tumor Associated Antigen immunological Test Systems

(b) Classification. Class II (special controls). The special controls for this device are:

(1) Design verification and validation must include:

(i) A description of genomic coverage that includes:

(A) A list of all genes, variant types, and target regions within each gene that the device detects;

(B) Summary information regarding the clinical significance of each gene, including references;

(C) A description of the genes with high clinical significance that are detected by the device, defined as genes for which the test result(s) may lead to prophylactic screening, confirmatory procedures, or treatment that may incur morbidity or mortality; and

(D) A description of any within-gene targeted regions that cannot be reported.

(ii) Specifications for specimen requirements, including any specimen collection devices, handling, and storage.

(iii) Specifications of the DNA extraction method and criteria for DNA quality and quantity that are prerequisite to performing the assay.

(iv) Detailed documentation of the methodology and protocols for each step of the test, including reagents, instrumentation, and software required. The documentation must include the analysis algorithms used for mutation detection and annotation, and specify the quality metrics, variant calling thresholds, and filters at each step of the test, including the criteria for run failures, batch failures, specimen failures, invalid calls (e.g., failed quality control), and “no calls” (i.e., absence of a result), as applicable.

(v) Description of required instrumentation and equipment, and any ancillary reagents, instrumentation, or equipment.

(vi) Detailed documentation of device software, including software applications and hardware-based devices that incorporate software. The documentation must include verification, validation, hazard analysis, and risk assessment activities.

(vii) Documentation of internal and external controls that are recommended or provided and control procedures. The documentation must identify those control elements that are incorporated into the testing procedure.

(viii) Detailed documentation pertaining to the probability of test failure based on data from clinical samples, description of scenarios in which a test can fail, and any risk mitigations, including follow-up actions to be taken.

(ix) Detailed documentation of the rules, procedures, tools, and criteria used for establishing mutation-hereditary disease relationships and mutation annotation, evaluation, and classification (e.g., pathogenic, likely pathogenic, variant of unknown significance, benign, and likely benign).

(x) Detailed documentation of any internal or external database(s) or decision rules used for mutation annotation, including:

(A) The protocol(s) used for variant interpretation, including training of personnel, monitoring accuracy of decision, resolution of discordant interpretations, and updating interpretations;

(B) Detailed documentation of the basis for interpretation, including the use of alternate databases, literature, and guidelines, and the basis for risk reporting;

(D) Data formats and nomenclature.

(xi) Information that demonstrates the performance characteristics of the device, evaluated either specifically for each gene/mutation or, when determined to be acceptable and appropriate by FDA, using a representative approach based on other mutations of the same type, including:

(A) Data that adequately support the intended specimen type(s) (e.g., whole blood), specimen handling protocol, and DNA extraction method.

(B) A summary of the evidence that demonstrates how the analytical quality metrics and thresholds used to determine the acceptability of reporting support the minimum accuracy requirements.

(C) Data to adequately support device accuracy using clinical specimens representing all indicated specimen types, mutation types, and size ranges intended to be detected and reported by the device. Accuracy data must fulfill the following:

(1) Accuracy of the device must be evaluated with clinical specimens collected in accordance with the device labeling and selected without bias, or well-characterized human cell line samples, when determined to be acceptable and appropriate by FDA.

(2) Accuracy must be evaluated by comparison to bidirectional Sanger sequencing or other orthogonal methods identified as appropriate by FDA. Performance criteria for both the comparator method(s) and the device must be predefined and appropriate to the device's intended use. Detailed study protocols must be documented.

(3) A sufficient number of specimens must be tested. For BRCA1 and BRCA2 genes, a minimum of 120 variant positive specimens must be tested. For other genes with high clinical significance, at least 40 variant-positive specimens must be tested per gene, including representative specimens for each indicated variant type based on a justification determined to be appropriate and acceptable by FDA. For remaining genes detected by the device, testing must include variant positive specimens representing each variant type, unless the variant type has a prevalence of less than 0.01 percent. Specimen selection must be prioritized based on clinical significance. The selected specimens must be representative of zygosity and challenging genomic context (e.g., guanine-cytosine content, near tandem repeats and homopolymer stretches, pseudogene), and must cover the range of sizes (for insertions, deletions, copy number variant (CNV) amplifications and CNV deletions) intended to be detected and reported by the device.

(4) Except as permitted by FDA under paragraph (b)(1)(xi)(C)(5) of this section, tested specimens must be selected based on results obtained from the orthogonal method. Positive percent agreement (PPA) and negative percent agreement (NPA) must be calculated and demonstrated for each variant type detected and reported by the device, as well as for clinically relevant variants. PPA is calculated as the number of variants that are tested positive by both the device and the orthogonal method (true “positives” (TP)) divided by the number of variants tested positive as determined by the orthogonal method (TP plus false negatives (FN) by the device). NPA is calculated as the number of variants that are tested negative (wild type) by both the device and the orthogonal method (true “negatives” (TN)) divided by the number of variants tested negative (wild type) by the orthogonal method (TN plus false positives (FP) by the device). Point estimates for PPA and NPA must be calculated along with 95 percent two-sided confidence intervals (CI). Uncertainty of the point estimate must be within an acceptable range, as identified by FDA, and must be demonstrated using the 95 percent CI.

(5) When it is determined by FDA to be appropriate and acceptable to select samples based on the results obtained with the device, accuracy must be presented as technical positive predictive value (TPPV) and technical negative predictive value (TNPV). TPPV relates to the likelihood that a variant call is a true positive and reflects the number of false positives per test. TPPV is calculated as the number of variants that are tested positive by both the device and the orthogonal method (TP) divided by the number of variants tested positive by the device (TP plus FP). TNPV relates to the likelihood that a variant call is a true negative and reflects the number of false negatives per test. TNPV is calculated as the number of variants that are tested negative by both the device and the orthogonal method (TN) divided by the number of variants tested negative by the device (TN plus FN).

(6) Any “no calls” or invalid calls in the study must be reported separately. The percent of final “no calls” or invalid calls must be clinically justifiable.

(7) Accuracy as a function of each performance metric (e.g., coverage depth, base quality scores) must be documented to provide evidence of the accuracy of the overall run.

(8) Detailed documentation for accuracy of the device must include information and results for the overall study, each mutation type, and each gene. The accuracy must further be described based upon stratification within each mutation type by zygosity, genomic context, and size (for indels and CNVs). Overall accuracy for reporting of substitutions must be ≥99.0 percent; insertions and deletions, ≥99.0 percent; CNVs, ≥99.0 percent for positive agreement (PPA, TPPV); and ≥99.9 percent for negative agreement (NPA, TNPV).

(D) Documentation of the data to adequately support device precision using clinical specimens representing all specimen types, mutation types, and sizes intended to be detected and reported by the device. The precision study must fulfill the following:

(1) The study must be performed using multiple instruments and multiple operators, on multiple non-consecutive days, and using multiple reagent lots. If the device is to be performed at more than one site, different sites must be included and reproducibility across sites must be evaluated.

(2) Representative clinical specimens of each mutation type must be tested (both positive and negative), considering clinical significance, prevalence, zygosity, genomic context, and size (for indels and CNVs). The precision for CNV detection must be demonstrated on the gene level for genes with high clinical significance. Alternatively, a justification for why such data are not needed must be found acceptable and appropriate by FDA.

(3) The study must assess the performance of all steps, including DNA extraction, unless a separate extraction study is performed.

(4) The study must use predefined performance criteria. Agreement estimates such as PPA/NPA and average positive agreement/average negative agreement must be provided, including point estimates and 95 percent confidence intervals. Documentation from the precision study must be demonstrated for the overall precision study, in addition to each mutation type, each gene, and each sample. Precision must further be demonstrated upon stratification within each mutation type by zygosity, genomic context, and size (for indels and CNVs). The overall precision point estimates for each variant type must be >99.0 percent.

(5) Any “no calls” or invalid calls in the study must be included in precision study results and reported separately. The percent of “no calls” or invalid calls and key quality control metrics parameters (e.g., coverage, sequencing score) must be summarized and based on stratification in the same way as the precision estimates.

(E) Documentation of the nucleic acid assay input range and the evidence to adequately support the range.

(F) Detailed documentation of additional analytical validation studies, including endogenous and exogenous interfering substances, specimen and reagent stability, cross-reactivity, carryover and cross-contamination, guard-banding, and index misassignment, as applicable. If specimens are pooled, index cross-contamination must be evaluated and demonstrate that pooling does not negatively impact test performance.

(G) Specimen type and matrix comparison data must be generated if more than one specimen type or anticoagulant can be tested with the device, including failure rates for the different specimen types.

(xii) Information that adequately supports the variant annotation and clinical interpretation of the test must include:

(A) A summary documenting the clinical significance for each gene on the test panel, including the associated conditions/cancers, the most prevalent and representative mutations, and summary of clinical evidence with references, including expected frequency in the general population and different ethnicities, and risks of developing the disease in relevant ethnic populations and the general population.

(B) Detailed documentation of the data to adequately support the performance of the variant annotation algorithms (e.g., concordance studies between the device generated variant classifications and externally established variant classifications, manual classifications by medical professionals, or classifications generated from clinical test reports).

(C) Documentation of any procedures or protocols for incorporation of any updates of valid scientific evidence into variant classification algorithms.

(2) The labeling required under § 809.10 of this chapter must include the following, as applicable:

(i) The intended use must include a description of the intended specimen type(s) and matrix (e.g., whole blood), the validated germline mutation types (e.g., single nucleotide variant, insertion, deletion, CNV), and a statement that the test is for hereditary cancer predisposition assessment and to aid in identifying hereditary genetic variants potentially associated with a diagnosed cancer.

(ii) The name of the testing facility or facilities (e.g., for single-site assays).

(iii) A summary of device description in accordance with paragraphs (b)(1)(i) through (b)(1)(viii) of this section.

(iv) A section that provides summary information on how the test works, how to interpret the results of the test, and an explanation on the database(s) used for mutation annotation.

(v) A summary of the information that demonstrates the performance characteristics of the device as required under paragraph (b)(1)(xi) of this section.

(vi) The following limiting statements:

(A) A statement that the test is not intended for use as a stand-alone diagnostic to diagnose cancer or other health conditions, and is not intended for use for prenatal testing nor as a cancer screening test.

(B) A statement that the test is specifically designed for heritable germline mutations and is not appropriate for the detection of somatic mutations.

(D) A statement that the risk of cancer or disease for an individual cannot be predicted.

(E) A statement that: test results should be interpreted in the context of clinical findings, family history, lifestyle, environment, and other factors; molecular testing may not detect all possible mutations leading to cancer predisposition; a negative result does not rule out the possibility that the individual has an unidentified variant leading to cancer; and for more information, physicians ordering the test may wish to consult with a clinical medical geneticist or genetic counselor.

(F) A statement that other factors, such as ethnicity, may affect whether the test results are relevant for a particular patient and may also affect how their genetic health results are interpreted.

(G) A statement describing the situations that a patient should not receive the test (e.g., a patient with bone marrow transplant).

(H) A statement describing the challenging genomic contexts that may have reduced performance, where results should be interpreted with care.

(I) A statement disclosing the genetic coverage of the test, including any gaps in coverage.

(J) Statements describing testing conditions that were identified to cause test failures (e.g., low specimen volume, poor DNA quality).

(vii) For variants detected and reported by the device under the category of “variants of uncertain significance” or equivalent designation, a limiting statement that the clinical significance has not been demonstrated with adequate clinical evidence in accordance with established guidelines (e.g., professional guidelines).

(viii) For variants detected and reported by the device under the category of “variants with evidence of clinical significance” or equivalent designation, reference(s) for physicians to access internal or external information concerning decision rules or conclusions about the level of evidence for clinical significance.

Collected 2026-10-09T05:45:50Z. Source file · JSON

Browse this collection