Thẩm định phương pháp

phân tích theo ICH

Mục tiêu:
•Trình bày được khái niệm chung về thẩm
định phương pháp phân tích
•Trình bày được một số chỉ tiêu cần xác
định khi thẩm định phương pháp phân
tích theo ICH: định nghĩa, cách tiến hành
và báo cáo kết quả
Definition
• Method validation is the process by which it is
established, through laboratory studies, that the
performance characteristics of the method meet
the requirements for its intended purpose.
When?
Analytical methods need to be validated, verified,
or revalidated:
•Before initial use in routine testing
•When transferred to another laboratory
•Whenever the conditions or method parameters
for which the method has been validated change
(e.g., instrument with different characteristics or
samples with a different matrix) and the change is
outside the original scope of the method.
ICH
(International Conference on Harmonisation)
• https://www.ich.org/

The International Council for Harmonisation of

Technical Requirements for Pharmaceuticals for Human
Use (ICH)
INTRODUCTION – ICH Q2 R1
• The tripartite harmonized ICH Guideline on Text
(previously coded Q2A) was finalized in October
1994.
• The Guideline on Methodology (Q2B, May 1997)
has been incorporated into the Guideline on Text
in November 2005 and then renamed Q2(R1),
without any changes in the contents of the two
Guidelines.
SPECIFICITY
• Specificity is the ability to measure accurately
and specifically the analyte of interest in the
presence of other components that may be
expected to be present in the sample matrix
• Identification tests
to ensures the identity of the analyte of interest.
• Purity tests
to ensures an accurate statement of the impurity
content (that is, in related substances tests, heavy
metals and organic volatile impurity limits)
• Assays
to provide an exact result for a determination of the
content or potency of the analyte.
Methodology
• Identification (qualitative analyses)
to discriminate between compounds of closely
related structures, or by comparison to known
reference materials.

• Assays
using spiked samples to show that the method
results are unaffected by the presence of
impurities or excipients.
 Impurity tests

Impurities available
by spiking the drug substance or product with the
appropriate levels of impurities and determining them
with the appropriate accuracy and precision.
Impurities not available
Compare results to a second well-characterized
procedure.
Include samples stored under relevant stress
conditions, (for example, light, heat, humidity,
acid/base hydrolysis, and oxidation). For assay,
the two results are compared. For impurity tests,
the impurity profiles are compared head-to-head.
Documentation
• For chromatographic procedures, representative
chromatograms with peaks labeled should be
included. Resolution, plate count (efficiency),
and tailing factor should be measured and
documented.
• Peak purity tests using advanced detection such
as photodiode array or mass spectrometry
should be used to show that the response is not
due to more than one component.
• Specificity vs. Sensitivity
PRECISION
• Precision is the degree of agreement among
individual test results when an analytical method
is used repeatedly to multiple samplings of a
homogeneous sample.
• Repeatability
Results of the method operating over a short time
interval under the same conditions (inter-assay
precision).
• Intermediate precision (formerly ruggedness)
Results from within-laboratory variations due to
random events such as different days, analysts,
equipments, etc.
Experimental design should be employed so that the
effects (if any) of the individual variables can be
monitored.
• Reproducibility
Results of collaborative studies between
laboratories.
Methodology
• to calculate statistically significant estimates of
standard deviation or relative standard deviation
(coefficient of variation). Assays should be of
samples that have all gone through the entire
analytical procedure from sample preparation
through final analysis.
• A minimum of 9 determinations covering the
specified range of the procedure (for example, 3
levels, 3 repetitions each) or a minimum of 6
determinations at 100% of the test or target
concentration is recommended.
Documentation
• expressed as the standard deviation or the
relative standard deviation (coefficient of
variation) for a statistically significant number of
measurements and confidence interval.
Statistical tables, bar charts, and other types of
graphs are commonly used to document
precision.
Association of Official Analytical Chemists (AOAC)
DATA PRESENTATION

Precision, standard deviation and uncertainty

Precision: variability of results obtained under
different circumstances (e.g. repeatability and
reproducibility); expressed by SD
Uncertainty: lack knowledge of the true value
The significant figures (or significant digits)
of a number are the digits that are known with
some degree of confidence
Rules for significant figures:
All non zero digits are significant.
E.g. 325 has 3 significant figures

Zeros between non zero digits are significant.

E.g. 1009 has 4 significant figures

Leading zeros are insignificant.

E.g. 0.0005 has 1 significant figures

Trailing zeros in a number containing a decimal point is significant.

E.g. 25.00 has 4 significant figures
Trailing zeros in a number not containing a decimal point can be
either significant or insignificant. To indicate which zeros are
significant, place a bar over the zeros which are significant or
indicate the number of significant figures in the brackets.

E.g. 2500 has 2 significant figures, 35000 has 3 significant figures,

12000 has 4 significant figures, 800 (2 sf) has 2 significant figures.
writing an uncertainty, how many significant figures?

SD: 2 significant figures.

The measurement result can then be written to the same
number of decimal places. For example: (1.123±0.032) M.

uncertainty reasonable?
NO simple answer
The more repeats done, the smaller the uncertainty
ACCURACY
• Accuracy is the closeness of test results to the
true value.
 Methodology
Drug substance
• Comparison of the results with the analysis of a
standard reference material.
• Comparison to a second, well-characterized
method.
Drug product
• Evaluate by analyzing synthetic mixtures of
known amounts or samples spiked with known
quantities of components.
• Comparison to a second, well-characterized
method.
Quantitation of impurities
• Analyze samples (drug substance or drug
product) spiked with known amounts of
impurities. (If impurities are not available, see
specificity.)

Data from a minimum of 9 determinations over a

minimum of 3 concentration levels covering the
specified range (for example, 3 concentrations, 3
replicates of each concentration).
Matrix effects vs. Recovery
Documentation
• Reported as the percent recovery of the known,
added amount, or as the difference between the
mean and true value with confidence intervals.
Association of Official Analytical Chemists (AOAC)
Examples
• The amount of calcium in different samples of milk powder
(in mg of calcium per g of milk powder) were analyzed by
two methods, one employing extraction followed by
analysis using atomic absorption spectroscopy, the other
using a complexometric titration method
LIMIT OF DETECTION (LOD)
• Characteristic of limit tests, the LOD is defined as
the lowest concentration of an analyte in a
sample that can be detected, not quantitated. It
is a limit test that specifies whether or not an
analyte is above or below a certain value.
Methodology
Noninstrumental methods
• by analyzing samples at known concentrations
and establishing the minimum level at which the
analyte can be reliably detected.
• Instrumental methods
• determined as a signal to noise ratio, usually 2:1
or 3:1, Or,
• according to the formula: LOD = 3.3(SD/S)
(SD) standard deviation of the response based on
either the standard deviation of the blank, the
residual standard deviation of the regression line,
or the standard deviation of y-intercepts of
regression lines.
(S) slope of the calibration curve
Documentation
• Express the LOD as the concentration of the
analyte.
• Document and support the method used to
determine LOD.
• An appropriate number of samples should be
analyzed at the limit to validate the level.
LIMIT OF QUANTITATION (LOQ)
• LOQ is the lowest concentration of an analyte in
a sample that can be determined (quantitated)
with acceptable precision and accuracy under
the stated operational conditions of the method.
Methodology
Noninstrumental methods
• Determine LOQ by analyzing samples at known
concentrations and establishing the minimum
level at which the analyte can be reliably
quantitated.
Instrumental methods
determined as a signal to noise ratio, usually 10:1, Or,
according to the formula: LOQ = 10(SD/S).
• (SD) standard deviation of the response based on
either the standard deviation of the blank, the
residual standard deviation of the regression line, or
the standard deviation of y-intercepts of regression
lines.
• (S) slope of the calibration curve
Documentation
• Express LOQ as a concentration, with the
precision and accuracy of the measurement.
• Document and support the method used to
determine LOD.
• An appropriate number of samples should be
analyzed at the limit to validate the level.
 SYSTEM SUITABILITY
• System suitability is the checking of a system to
ensure system performance before or during the
analysis of unknowns.
• System suitability tests are an integral part of
chromatographic methods.
• System suitability parameters are established as a
direct result of robustness studies.
Methodology
• Plate count (N), tailing factor (T), resolution (Rs) and
repeatability (%RSD) are determined from replicate
injections of a standard (an analyte peak and an
internal standard, related compound, excipient,
and/or impurity, etc.) compared against method
specifications.
• If RSD < 2.0%, five replicates are used.
• If RSD > 2.0%, six replicates are used.
• System suitability must be demonstrated at
appropriate intervals before, during, and after
the analysis of unknown samples, or whenever
there is a significant change in instrumentation,
or in a critical reagent.
Documentation
• accomplished by summarizing data on
repeatability, efficiency, tailing and resolution for
the replicate injections.
• No sample analysis is acceptable unless system
suitability specifications have been met
ROBUSTNESS
• Robustness is the capacity of a method to
remain unaffected by small, deliberate variations
in method parameters; a measure of the
reliability of a method.
• Robustness should be evaluated in late
development, or early in the method validation
process. If the results of a method or other
measurements are susceptible to variations in
method parameters, these parameters should be
adequately controlled and a precautionary
statement included in the method documentation.
• Robustness can be used to establish system
suitability parameters.
• Normally, after implementing a validated
method, it can be adjusted within the confines
of the robustness study without triggering a
revalidation. However, method changes, outside
the range of parameters validated, would
require at least some revalidation to show
equivalency of results.
Methodology
• Purposely vary method parameters over a known
range, and determining the effect (if any) on the
method results.
• Multivariate statistical experimental design can be
used to control method variables (for example,
Factorial, Fractional Factorial, or Plackett-Burman
designs).
• Theoretical modeling software can also be used to
predict robustness and then verified experimentally.
Documentation
• Robustness can be illustrated by many different
means, using summary tables, bar, and control
charts, effect and probability plots, and other
means of result comparisons.
DATA ELEMENTS REQUIRED FOR
ASSAY VALIDATION
• Category 1: Analytical methods for the
quantitation of major components of bulk drug
substances or active ingredients in finished
pharmaceutical products.
• Category 2: Analytical methods for the
determination of impurities in bulk drug
substances or degradation compounds in
finished pharmaceutical products, including
quantitative assays and limit tests.
•Category 3: Analytical methods for the
determination of performance
characteristics (for example, dissolution,
drug release).
•Category 4: Identification tests.
 LINEARITY AND RANGE
• Linearity
The ability of the method to elicit test results that are
directly, or by a well-defined mathematical
transformation, proportional to analyte concentration
within a given range.
• Range
The interval between the upper and lower levels of
analyte (inclusive) that have been demonstrated to be
determined with a suitable level of precision, accuracy,
and linearity using the method as written.
Methodology
• Linearity
Demonstrate across the entire range of the
analytical procedure.
A minimum of five concentrations is
recommended.
• Range
Verify that the method provides acceptable
precision, accuracy, and linearity when applied to
samples at the extreme as well as within the
range.
 Recommended minimum Ranges:
• Assay of Drug Substance or Finished Product
From 80–120% of the test concentration.
• Determination of an Impurity
From 50–120% of the specification.
• Content Uniformity
A minimum of 70–130% of the test concentration
unless a wider or more appropriate range is
justified based upon the dosage form.
• Dissolution Testing
+/-20% over the specified range of the dissolution
test.
Documentation
• The report should include:
The slope of the regression line.
The correlation coefficient.
y-intercept.
The residual sum of squares.
Fitting Straight Lines - Simple Regression
r always lies between −1 and +1.
 METHOD OF LEAST SQUARES

the sum of the squares of the residuals

Setting up and evaluating a statistical model for regression

1.Perform statistical tests to check whether the

assumptions made in setting up the model are
reasonable

2. If the model does not have to be rejected, it can be

used to make estimates of quantities of interest with
associated confidence intervals.
 Diagnostic Plots

If the model is correct,

the sampling distribution of the residuals will
Partitioning the corrected sum of squares of the responses
F is about 1
EXCEL
 Normal Probability Plot X Variable 1 Residual Plot
3 0.5
2

Residuals
Y

1 0
0 5 10 15 20 25
0
0 20 40 60 80 100 120
-0.5
Sample Percentile X Variable 1
 Regression Statistics
Multiple R 0.4167546
R Square 0.1736844
Adjusted R Square 0.13433603 R2
Standard Error 0.21669556 Sy/x
Observations 23

ANOVA
df SS MS F Significance F
Regression 1 0.207268917 0.207269 4.414018433 0.047895911
Residual 21 0.986096301 0.046957
Total 22 1.193365217

Coefficients Standard Error t Stat P-value Lower 95% Upper 95% Lower 95.0% Upper 95.0%
Intercept (a) 1.09781488 0.117481184 9.344602 6.26348E-09 0.853499382 1.3421304 0.8534994 1.342130374
X Variable 1 (b) 0.02196252 0.010453582 2.100957 0.047895911 0.000223108 0.0437019 0.0002231 0.043701937
ESTIMATION AND CONFIDENCE INTERVALS
NOTES:
LINEAR CALIBRATION

y = a + bx

the limits for the 95% prediction interval for y

 Common Mistakes Using Regression
Analysis
• Lacking an awareness of the assumptions of least-squares
regression
• Knowing how to evaluate the assumptions of least-squares
regression
• Knowing what the alternatives to least-squares regression
are if a particular assumption is violated
• Using a regression model without knowledge of the subject
matter
• Predicting Y outside the relevant range of X
 TO AVOID THE COMMON MISTAKES:
• Always start with a scatter plot (to observe the possible
relationship between X and Y).
• Check the assumptions of regression (after the
regression model has been fit, before using the results of
the model).
• Plot the residuals versus the independent variable. (to
determine whether the model fit to the data is an
appropriate one and to check visually for violations of
the equal variation assumption).
• Use a histogram, box-and-whisker plot, or normal
probability plot of the residuals (to graphically evaluate
whether the normality assumption has been seriously
violated).
• If the assumptions violated, use alternative methods to
least-squares regression or alternative least-squares
models.
• If the assumptions not violated, then undertake the
inferential aspects of the regression analysis. A test for the
significance of the slope and a confidence interval
estimate of the slope can be carried out.
SUMMARY
The standard deviation of the estimate of
concentration from m measurements of an unknown
Wine of unknown glucose concentration and calibration
solutions of glucose were treated in the same way as
follows: Wine (200 mL) or glucose solution (200 mL) was
diluted to 5mL in a volumetric flask by the addition of
enzyme solution and buffer. Triplicate measurements of
the absorbance of the treated wine solution were 0.253,
0.243, 0.238.
FAQ on calibration
• Which type of balance do I need for the preparation of calibration
standards?
The US Pharmacopeia defines the minimum permissible weight of a
balance as a load that will give a relative uncertainty of less than 0.1%.
• Is serial dilution of a standard solution for the preparation
of calibration standards acceptable?
• Which type of volumetric glassware may I use for the
preparation of calibration standards?

Class A glassware according to ISO standard 1042:1983

(e.g. ± 0.04 ml for a 25 ml flask)
• How many points do I need for a calibration curve?
• How many replicates per calibration point?
The ISO standard 11095:1996 demands at least two replicate analyses
per calibration level and recommends as many as possible.
• Why shall the concentration levels of the calibration
standards be equidistant?
• When do I need to prepare matrix matched
calibration standards?
A matrix matched calibration is needed in those cases,
where the matrix (even after clean-up procedures) has
an influence on the signal obtained for the analyte
during measurement.

In which sequence shall I measure the calibration

standards?
Generally the sequence in which the calibration
standards are measured should be random.
• Does a correlation coefficient (r) of 0.99 indicate
linearity of calibration?
No! The correlation coefficient is a measure of how
much of the variability of y can be predicted by x.
• Why is calibration curve always plotted with the instrument signals
on the vertical (Y) axis and the standard concentrations on the
horizontal (x) axis?
assume that all the errors are in the y-values and that the standard
concentrations (x-values) are error-free