生物等效性研究指导原则(英文版)

Technique Guideline for Human Bioavailability and BioequivalenceStudies on Chemical Drug ProductsContents(Ⅰ) Establishment and Validation for Biological Sample Analysis Methods (2)1. Common Analysis Methods (2)2. Method Validation (2)2.1 Specificity (2)2.2 Calibration Curve and Quantitative Scale (3)2.3 Lower Limit of Quantitation (LLOQ) (3)2.4 Precision and Accuracy (4)2.5 Sample Stability (4)2.6 Percent recovery of Extraction (4)2.7 Method Validation with microbiology and immunology (4)3. Methodology Quality Control (5)(Ⅱ) Design and Conduct of Studies (5)1. Cross-over Design (5)2. Selection of Subjects (6)2.1 Inclusion Criteria of Subjects: (6)2.2 Cases of Subjects (7)2.3 Division into Groups of the Subjects (7)3. Test and Reference Product, T and R (8)4. Sampling (8)(Ⅲ) Result Evaluation (9)(Ⅳ) Submission of the Contents of Clinical Study Reports (9)Technique Guideline for Human Bioavailability and BioequivalenceStudies on Chemical Drug ProductsSpecific Requirements for BA and BE Studies(Ⅰ) Establishment and Validation for Biological Sample Analysis MethodsBiological samples generally come from the whole blood, serum, plasma, urine or other tissues. These samples have the characteristics such as little quantity for sampling, low drug concentration, much interference from endogenous substances, and great discrepancies between individuals. Therefore, according to the structure, biological medium and prospective concentration scale of the analytes, it is necessary to establish the proper quantitative assay methods for biological samples and to validate such methods.1. Common Analysis MethodsCommonly used analysis methods at present are as follows: (1) Chromatography: Gas Chromatography(GS), High Performance Liquid Chromatography (HPLC), Chromatography－mass Spectrometry (LC-MS, LC-MS-MS, GC-MS, GC-MS-MS), and so on. All the methods above can be used in detecting most of drugs; (2) Immunology methods: radiate immune analysis, enzyme immune analysis, fluorescent immune analysis and so on, all these can detect protein and polypeptide; (3) Microbiological methods: used in detecting antibiotic drug.Feasible and sensitive methods should be selected for biologic sample analysis as far as possible.2. Method ValidationEstablishment of reliable and reproducible quantitative assay methods is one of the keys to bioequivalence study. In order to ensure the method reliable, it is necessary to validate the method entirely and the following aspects should be generally inspected:2.1 SpecificityIt is the ability that the analysis method has to detect the analytes exactly and exclusively, when interference ingredient exists in the sample. Evidences should be provided that the analytes are the primary forms or specific active metabolites of the test drugs. Endogenous instances, the relevant metabolites and degradation products in biologic samples should not interfere with the detection of samples. If there are several analytes, each should be ensured not to be interfered, and the optimal detecting conditions of the analysis method should be maintained. As for chromatography, at least 6 samples from different subjects, which include chromatogram of blank biological samples, chromatogram of blank biologic samples added control substance (concentration labeled) and chromatogram of biologic samples after the administration should beexamined to reflect the specificity of the analytical procedure. As for mass spectra (LC-MS andLC-MS-MS) based on soft ionization, the medium effect such as ion suppression should be considered during analytic process.2.2 Calibration Curve and Quantitative ScaleCalibration curve reflects the relationship between the analyte concentration and the equipment response value and it is usually evaluated by the regression equation obtained from regression analysis (such as the weighted least squares method). The linear equation and correlation coefficient of the calibration curve should be provided to illustrate the degree of their linear correlation. The concentration scale of calibration curves is the quantitative scale. The examined results of concentration in the quantitative scale should reach the required precision and accuracy in the experiment.Dispensing calibration samples should use the same biological medium as that for analyte, and the respective calibration curve should be prepared for different biological samples. The number of calibration concentration points for establishing calibration curve lies on the possible concentration scale of the analyte and on the properties of relationship of analyte/response value. At least 6 concentration points should be used to establish calibration curve, more concentration points are needed as for non-linear correlation. The quantitative scale should cover the whole concentration scale of biological samples and should not use extrapolation out of the quantitative scale to calculate concentrations of the analyte. Calibration curve establishment should be accompanied with blank biologic samples. But this point is only for evaluating interference and not used for calculating. When the warp* between the measured value and the labeled value of each concentration point on the calibration curve is within the acceptable scale, the curve is determined to be eligible. The acceptable scale is usually prescribed that the warp of minimum concentration point is within ±20% while others within ±15%. Only the eligible calibration curve can be carried out for the quantitative calculation of clinical samples. When linear scale is somewhat broad, the weighted method is recommended to calculate the calibration curve in order to obtain a more exact value for low concentration points. ( *: warp=[(measured value － labeled value)/labeled value]×100%)2.3 Lower Limit of Quantitation (LLOQ)Lower limit of quntitation is the lowest concentration point on the calibration curve, indicating the lowest drug concentration in the tested sample, which meets the requirements of accuracy and precision. LLOQ should be able to detect drug concentrations of samples in 3~5 eliminationhalf-life or detect the drug concentration which is 1/10~/20 of the C max. The accuracy of the detection should be within 80~120% of the real concentration and its RSD should be less than 20%. The conclusions should be validated by the results from at least 5 standard samples.2.4 Precision and AccuracyPrecision is, under the specific analysis conditions, the dispersive degree of a series of the detection data from the samples with the same concentration and in the same medium. Usually, the RSD from inter- or intra- batches of the quality control samples is applied to examine the precision of the method. Generally speaking, the RSD should be less than 15% and that around LLOQ should be less than 20%. Accuracy is the contiguous degree between the tested and the real concentrations of the biological samples (namely, the warp between the tested and the real concentrations of the quality-controlled samples). The accuracy can be obtained by repeatedly detecting the analysis samples of known concentration which should be within 85~115% and which around LLOQ should be within 80~120%.Generally, 3 quality-control samples with high, middle and low concentrations are selected for validating the precision and accuracy of the method. The low concentration is chosen within three times of LLOQ, the high one is close to the upper limit of the calibration curve, and the middle concentration is within the low and the high ones. When the precision of the intra-batches is detected, each concentration should be prepared and detected at least 5 samples. In order to obtain the precision of inter-batches, at least 3 qualified analytical batches, 45 samples should be consecutively prepared and detected in different days.2.5 Sample StabilityAccording to specific instances, as for biological samples containing drugs, their stabilities should be examined under different conditions such as the room temperature, freezing, thaw and at different preservation time, in order to ensure the suitable store conditions and preservation times. Another thing that should be paid attention to is that the stabilities of the stock solution and the analyte in the solution after being treated with, should also be examined to ensure the accuracy and reproducibility of the test results.2.6 Percent recovery of ExtractionThe recovery of extraction is the ratio between the responsive value of the analytes recovered from the biological samples and that of the standard, which has the same meaning as the ratio of the analytes extracted from the biologic samples to be analyzed. The recovery of extraction of the 3 concentrations at high, middle and low should be examined and their results should be precise and reproduceable.2.7 Method Validation with microbiology and immunologyThe analysis method validation above mainly aims at chromatography, with many parameters and principles also applicable for microbiological and immunological analysis. However, some special aspects should be considered in the method validation. The calibration curve of the microbiological and immunological analysis is non-linear essentially, so more concentration pointsshould be used to construct the calibration curve than the chemical analysis. The accuracy of the results is the key factor and if repetitive detection can improve the accuracy, the same procedures should be taken in the method validation and the unknown sample detection.3. Methodology Quality ControlThe unknown samples are detected only after the method validation for analysis of biological samples has been completed. The quality control should be carried out during the concentration detection of the biological samples in order to ensure the reliability of the method in the practical application. It is recommended to assess the method by preparing quality-control samples of different concentrations by isolated individuals.Each unknown sample is usually detected for only one time and redetected if necessary. In the bioequivalence experiments, biological samples from the same individual had better to be detected in the same batch. The new calibration curve should be established when detecting biological samples of each analysis batch and high, middle and low concentrations of the quality-control samples should be detected at the same time. Each concentration should at least have two samples and should be equally distributed in the detection sequence of the unknown samples. When there are a large number of unknown samples in one analysis batch, the number of the quality-control samples at different concentrations should be increased to make the quality-control samples exceed 5% of the unknown sample population. The warp of detection result from the quality-control samples should usually be less than 15%, while the warp of the low concentration point should be less than 20% and at most 1/3 results of the quality-control samples at different concentrations are allowed to exceed the limit. If the detection results of the quality-control samples do not accord with the above requirements, the detection results of the samples in this analysis batch should be blanked out.The samples with concentrations higher than the upper quantitation limit should be detected once more using corresponding diluted blank medium. As for those samples with concentrations lower than the lower quantitation limit, during pharmacokinetics analysis, those sampled before reaching C max should be calculated as zero while those after C max should be calculated as ND (Not detectable), so as to decrease the effect of the zero value on the AUG calculation.(Ⅱ) Design and Conduct of Studies1. Cross-over DesignCurrently, the crossover design is the most wildly applied method in the BE study. As for the drug absorption and clearance, there is a transparent variation among individuals. Therefore, the coefficient of variability among individuals is far greater than that of the individual himself. That is why the bioequivalence study is generally required to be designed on the principle of self crossover control. Subjects are randomly divided into several groups and treated in sequence, of whichsubjects in one group take the test products first and then the reference product, while subjects in the other take the reference products first and then the test products. A long enough interval is essential between the two sequences, which is called Wash-out period. In this way, every subject has been treated twice or more times sequentially, which is equal to self-control. Therefore, the influence of drug products on drug absorption can be discriminated from the others, and the effect of various test periods and individual difference on the results can be eliminated.Two-sequence crossover design, three-sequence crossover design are adopted respectively according to the amount of the test product. If two varieties of drug products are to be compared, the two-treatment, two-period or two-sequence crossover design will be a preferable choice. When three varieties of products (two test products and one reference product) are included, thethree-formulation, three-period and double 3×3 Latin square design will be the suitable choice. And a long enough wash-out period is required among respective periods.Wash-out period is set on purpose to eliminate the mutual disturbance of the two varieties of drug products and avoid the treatment in the prior period from affecting that of the next period. And the wash-out period is generally longer than or equal to 7 elimination half lives.While the half-lives of some drugs or their active metabolites are too long, it is not suitable to apply the crossover design. Under this circumstance, parallel design is adopted, but the sample size should be enlarged.However, as for some highly variable drugs, except for increase of the subjects, repetitive cross-over design can be applied, to test possibly existing difference in individual when receive the same preparation twice.2. Selection of Subjects2.1 Inclusion Criteria of Subjects:The difference among individuals of the subjects should be minimized so that the difference of the drug products can be detected. The inclusion criteria and exclusion criteria should be noted in the trial scheme.Male healthy subjects are recruited generally. And as to the drugs of special purpose, proper subjects are recruited according to specific conditions. If female healthy subjects are recruited, the possibility of gestation should be avoided. If the drugs to be tested have some known adverse effects, which may do harm to the subjects, patients can also be included as the subjects.Age: 18~ 40 years old generally. The difference in age of the subjects in one batch should not be more than 10 years.Body weight: not less than 50kg as to normal subjects. Body Mass Index (BMI), which is equal to body weight (kg)/ body height 2 (m2), is generally required to be in the range of standard body weight. For the subjects in one batch, the taken dosage is the same, the range of the bodyweight, therefore, should not have great disparity.The subjects should receive the overall physical examination and be proved healthy. There is not medical history of heart, kidney, digestive tract, nervous system, mental anomaly, metabolism dysfunction, and so on. The physical examination has revealed normal blood pressure, heart rate, electrocardiogram, and respiratory rate. Laboratory data have revealed normal hepatic function, renal function and blood function. Those examinations are essential to prevent the metabolism of drugs in vivo from being interfered by the diseases. According to the classification and safety of drugs, special items examinations are required before, during and after the test, such as the blood glucose examination, which is required in the drug trial of hypoglyceimic agents.In order to avoid the interference by other drugs, no administration of other drugs is allowed from two weeks before and till the end of the test. Moreover, the cigarette, wine,beverage with caffeine, or some fruit juice that may affect the metabolism of the drug, is forbidden during the trial period also. The subjects had better have no appetite of cigarette and wine. Possible effects of the cigarette-addicted history should not be neglected in the discussion of results.Due to the metabolism variance resulted by known genetic polymorphism of drugs, the safety factor which may be effected by the slow metabolism speed of drugs should be considered.2.2 Cases of SubjectsThe cases of the subjects should meet the statistic requirement. And according to the current statistical methods, 18~24 cases are enough for most drugs to meet the requirement of sample size. But as to some drugs of high variability, more cases may be required correspondingly.The cases of a clinic trial are determined by three fundamental factors: （1）Significance level: namely, the value of α, for which value 0.05 or 5% is often adopted;（2）Power of a test: namely, the value of 1-β. β is the index that represents the probability of the type error, which is also theⅡprobability of misjudging the actually efficacy drugs as inefficient drugs, and value not less than 80% is commonly stated; （3）Coefficient of variance（CV%）and Difference（θ）: In the equivalence test of two drugs, the greater CV% and θ of the test indexes are, the more cases are required. The CV% and θ are unknown before the trial and can only be estimated by the above parameters of the owned reference products or running the preliminary test. Moreover, when a BA test has been finished, the value of N can be calculated according to the parameters such as θ, CV% and 1-β and then compared with the cases adopted in the finished BA test to determine whether the cases are reasonable or not.2.3 Division into Groups of the SubjectsThe subjects should be randomly divided into different comparable groups. The cases of the two groups should guarantee the best comparability.3. Test and Reference Product, T and RThe quality of the reference products directly affect the results reliability of BE trial. Generally, the domestic innovator products of the same dosage form which has been approval to be on sale are commonly selected. If it failed in acquiring the innovator products, the key product on the market can also be chosen as the reference product and the related quality certifications (such as the test results of the assay and dissolution) and the reasons for option should be provided. When it comes to the drug study of specific purpose, other on-sale dosage forms which are of the same kind and similar with pharmaceutics properties are selected as the reference products and those reference products should be already on sale and qualified in quality. The difference in assay between the test product and reference product should not exceed 5%.The test product should be the scale-up product or manufacture scale product, which is consistent with the quality standards for clinical application. And the indexes such as the in vitro dissolution, stability, content or valence assay, consistency reports between batches should be provided to the test unit for reference. As for some drugs, the data of polymorphs and optical isomers should be offered additionally. The test and reference product should be noted with the advanced development unit, batch number, specification, storage conditions and expiry date.For future reference, the test and reference product should be kept long enough after the trialtill the product is approved to be on sale.4. SamplingThere is a significant sense in designing the sampling point to guarantee both the reliability of the trial results and the rationality of calculating the pharmacokinetics parameters. Commonly, there should be preliminary tests or the pharmacokinetics literatures at home and abroad served as the evidences of designing the reasonable sampling points. When the blood-drug concentration assay is performed, the absorption phase, balance phase and clearance phase should be considered overall. There must be enough sampling points in every phase of the C-T curve and around the T max. The concentration curve, therefore, can fully reflect the entire procedure of the drugs distribution in vivo. And the blank blood samples are taken before the administration. Then at least 2~3 points are sampled in the absorption phase, at least 3points are sampled near the C max and 3-5 points in the clearance phase. Try to avoid that the first point gets the C max, and running the preliminary test may avoid this. When the continuously-sampling results show that the drugs’ primary forms or the active metabolites are at the point of 3～5 half- lives or the blood drug concentration is 1/10～1/20 ofC max, the values of AUC0-t/AUC0-∞are generally bigger than 80% .For the terminal clearance item doesn’t affect the evaluation of the products’ absorption process much, as to the long half-life drugs, the sampling periods should be continued long enough, so that the whole absorption process can be compared and analyzed. In the multiple administration study, the BA of some drugs is known to beaffected by the circadian rhythm, samples of which should be taken 24 hours continuously if possible.When the BA of the test drugs can’t be determined by detecting the blood-drug concentration, if the primary forms and the active metabolites of the test drugs are mainly be excreted in urine (more than 70% of the dosage), the BA assay may be performed by detecting the urine drug concentration, which is the test of the accumulated excretion quantity of drugs in urine to reflect the intake of drugs. The test products and trial scheme should accord with the demands of BA assay. The urine samples should be collected at intervals, and the collection frequency and intervals of which should meet the demands of evaluating the excretion degree of the primary forms and the active metabolites of the test products in urine. However this method cannot reflect the absorption speed of the drugs and gets many error factors, it is not recommended generally.Some drugs metabolize so rapidly in vivo that it is impossible to detect the primary forms in biological samples. Under these circumstances, the method determining the concentration of corresponding active metabolites in biological samples is adopted to perform the BA and BE studies.(Ⅲ) Result EvaluationAt present, the weighting function of AUC on drug absorption degree is comparatively affirmed, while C max and T max sometimes are not sensitive and seemly enough for weighting the absorption speed due to their dependence on the arrangement of sampling time, and they are therefore not suitable for drug products with multi-peak phenomena and for experiments with large individual variation. During the evaluation, if there are some special instances of inequivalence, a specific analysis should be performed for specific problems.As for AUC，the 90% confidence interval is generally required within the scope of 80%~125%. As for the drugs with narrow treatment spectrum, the above scope should likely be appropriately reduced. While in a few instances, having been validated to be reasonable, the scope can also be increased. So does C max. And as for T max, statistical evaluation is required only when its release speed is closely correlated to clinical therapeutic effects and safety, the equivalence scope of which can be ascertained according to the clinical requirements.When bioavailability ratio of test products is higher than that of reference products, which is called suprabioavailability, the following two instances can be considered: 1). Whether the reference product itself is a product with low bioavailability, which results in the improvement of the test drug's bioavailability; 2). The quality of the reference product meets the requirement, and the test drug really has higher bioavailability.(Ⅳ) Submission of the Contents of Clinical Study ReportsIn order to satisfy the demand of evaluation, a clinical report of bioequivalence study shouldinclude the following contents: (1)Experiment subjective;(2) Establishment of analysis methods for bioavailability samples and data of inspection, as well as provision of the essential chromatograms;(3) Detailed experiment design and operation methods , including data of all the subjects，sample cases，reference products，given dosage，usage and arrangement of sampling time;(4) All data about original measurement of unknown sample concentrations，pharmacokinetics parameters and drug-time curve of each subjects;(5) Data handling procedure and statistical analysis methods as well as detailed procedure and results of statistics;(6) Observation results of clinical adverse reactions after taking medicine，midway exit and out of record of subjects and the reasons;(7) Result analysis and necessary discussion on bioavailability or bioequivalence; (8) References. A brief abstract is required before the main body; at the end of the main body, names of the experiment unit, chief persons of the study and experiment personnel should be signed to take the responsibility for the results of the study.。