Clinical sample planning by genotype
The Center for Medical Device Evaluation of China’s National Medical Products Administration published the 2026 revision of its registration review guidance for hepatitis C virus (HCV) genotyping reagents on October 13, 2026. The announcement is dated October 12 and specifies no separate effective date or transition period. The guidance applies to registration and change-registration applications and is advisory rather than a legally binding regulation.
A comparison with the 2017 guidance identifies a specific change in clinical sample planning. The earlier document specified at least 500 samples overall, statistically meaningful numbers for common genotypes, and possible reductions for rare types. The 2026 section instead recommends at least 120 cases for each common genotype or subtype, such as 1b and 2a, and at least 30 for each less common type or subtype, such as 3a, 3b and type 6.
For assays that do not distinguish 3a from 3b, both subtypes should be represented. Their combined sample size should meet statistical requirements, with at least 70 cases recommended. Combining the reporting category therefore does not remove the need to demonstrate representation of both subtypes.
The revised section does not retain the overall minimum of 500 samples. This does not establish a general reduction or increase in study size: the claimed detection range, subtype differentiation and statistical design determine the practical total. Planning should begin with the evidence needed for each claimed type.
For very rare types 4 and 5, the 2026 guidance retains the option of samples from HCV RNA-positive patients or scientifically validated serum panels. The instructions for use should describe the clinical validation information for those panels. Difficulty obtaining patient samples does not remove the need for appropriate validation.
Analytical coverage must support the claimed range
The guidance principally covers qualitative HCV genotyping assays based on real-time fluorescence PCR. Other qualitative methods may refer to it while assessing applicability to their own methodology. The intended clinical population consists of patients already confirmed as HCV RNA-positive; genotyping provides supporting clinical information.
Limits of detection must be established and verified separately for all claimed genotypes. For determination, each concentration level is tested at least 20 times, and the concentration yielding a 95% positive detection rate is used as the limit. Verification uses separately selected clinical samples or viral cultures, with at least 20 replicates at that concentration and a 95% positive detection rate. Coverage of all claimed genotypes was already present in the 2017 guidance and should not be described as a new 2026 requirement.
Analytical sensitivity and clinical agreement answer different questions. The former establishes detection at low concentrations for a particular genotype; the latter evaluates correct typing in the intended population. Results for a common genotype cannot simply establish the detection limit for a rare genotype. Similarly, pooled clinical agreement alone cannot describe performance across all claimed types. The revised guidance calls for genotype-specific agreement rates, Kappa statistics and their 95% confidence intervals.
Mixed infection and low-concentration detection
The 2026 document addresses microbial interference and competitive interference separately. Competitive interference concerns the effect of a high-concentration genotype on detection of a low-concentration genotype. It recommends combinations involving one genotype at its detection limit and another at a high concentration, taking account of the assay’s reaction design. Combinations within the same reaction system and those commonly encountered in mixed infection are recommended priorities.
This complements mixed-genotype studies. Simulated mixtures of clinical samples may be used, but their genotype composition and the concentration of each component should be established. A result from a mixture needs to be interpreted alongside the concentrations, ratios and genotype combinations tested.
The instructions for use should reflect these performance boundaries. The assay detects only the known genotypes included in its intended use; low viral concentrations may prevent successful genotyping. An inability to distinguish mixed-genotype samples because of the method’s design should be disclosed. Results are for clinical reference and should not be the sole basis for adjusting drug treatment.
Comparator coverage is part of study design
Clinical comparisons should use an approved, marketed comparator in China with a detection range covering the investigational assay and the ability to distinguish genotypes. Where marketed typing assays do not provide sufficient coverage, sequencing may also be incorporated. The sequencing method’s development, reagents, instruments and performance validation should be described, and its analytical performance should meet the evaluation’s needs.
Manufacturers can therefore assess claimed genotypes, planned enrollment and comparator coverage together. Obtaining samples for a rare genotype is only part of the task; an appropriate comparison method is also needed to interpret the results.
The guidance permits alternative methods that meet applicable regulatory requirements when supported by detailed research and validation. Existing projects should assess sample design and evidence coverage against their own claimed scope, product characteristics and statistical justification.
Sources
- Primary official source丙型肝炎病毒基因分型检测试剂注册审查指导原则（2026年修订版） (opens in a new tab)Center for Medical Device Evaluation, National Medical Products Administration
- Supporting source国家药监局器审中心关于发布《人表皮生长因子受体2基因扩增检测试剂（荧光原位杂交法）注册审查指导原则（2026年修订版）》等2项注册审查指导原则的通告（2026年第28号） (opens in a new tab)Center for Medical Device Evaluation, National Medical Products Administration
- Supporting source丙型肝炎病毒核酸基因分型检测试剂注册技术审查指导原则（2017年第209号附件4） (opens in a new tab)China Food and Drug Administration