Methodologies for the transfer of analytical methods: a review.

The transfer of a method from a laboratory to a production site is an important step in the development cycle of new pharmaceutical products. Method transfers are increasingly implemented due to the economical pressure coming from the rationalization of production sites, analytical subcontracting and fusion of pharmaceutical groups. However, no official guidance regarding study design, data analysis, or decision procedures is present neither in FDA documents nor in ICH documents for method transfers. The experiments performed in such a transfer and the methodology used to accept or reject it should be fitted for purpose. In order to provide to analysts a global view of the problematic of analytical method transfer, this paper reviews the documentation available in the scientific literature about the design of transfer studies and the required sample size. Special focus is also made on the statistical methodologies available for decision making with particular emphasis on risk management. Examples of transfer of pharmaceutical, bio-pharmaceutical and biological methods published in the literature are reviewed in order to illustrate the various possibilities among the strategies for methods transfer.

Harmonization of strategies for the validation of quantitative analytical procedures: a SFSTP proposal part IV. Examples of application.

A harmonized approach for the validation of analytical methods based on accuracy profile was introduced by a SFSTP commission on the validation of analytical procedure. This fourth and last document aims at illustrating this methodology and the statistics used. Therefore the validation of real case methods are proposed such as methods for the quality control of drugs, for the quantitation of impurities in drug substances, for bioanalysis or for the determination of nutriments. Furthermore, different types of analytical methods are used in order to demonstrate the applicability of the proposed approach to a wide range of methods such as liquid chromatography (LC-UV, LC-MS), spectrophotometry or ELISA.

Risk-based approach for the transfer of quantitative methods: bioanalytical applications.

The transfer of analytical methods from a sending laboratory to a receiving one requires to guarantee that this last laboratory will obtain accurate results. Undeniably method transfer is the ultimate step before routine implementation of the method at the receiving site. The conventional statistical approaches generally used in this domain which analyze separately the trueness and precision characteristics of the receiver do not achieve this. Therefore, this paper aims first at demonstrating the applicability of two recent statistical approaches using total error-based criterion and taking into account the uncertainty of the true value estimate of the sending laboratory, to the transfer of bioanalytical methods. To achieve this, they were successfully applied to the transfer of two fully automated liquid chromatographic method coupled on-line to solid-phase extraction. The first one was dedicated to the determination of three catecholamines in human urine using electrochemical detection, and the second one to the quantitation of N-methyl-laudanosine in plasma using fluorescence detection. Secondly, a risk-based evaluation is made in order to understand why classical statistical approaches are not sufficient to provide the guarantees that the analytical method will give most of the time accurate results during its routine use. Finally, some recommendations for the transfer studies are proposed.

Harmonization of strategies for the validation of quantitative analytical procedures. A SFSTP proposal--part III.

In the first two documents [Ph. Hubert, J.J. Nguyen-Huu, B. Boulanger, E. Chapuzet, P. Chiap, N. Cohen, P.A. Compagnon, W. Dewé, M. Feinberg, M. Lallier, M. Laurentie, N. Mercier, G. Muzard, C. Nivet, L. Valat, J. Pharm. Biomed. Anal. 36 (2004) 579-586; Ph. Hubert, J.J. Nguyen-Huu, B. Boulanger, E. Chapuzet, P. Chiap, N. Cohen, P.A. Compagnon, W. Dewé, M. Feinberg, M. Lallier, M. Laurentie, N. Mercier, G. Muzard, C. Nivet, L. Valat, E. Rozet, J. Pharm. Biomed. Anal., in press], a recent SFSTP Commission on the validation of analytical procedure has introduced a harmonized approach for the validation of analytical procedures. In order to complete this guide, the statistical methodology allowing to correctly conclude about the validity of a procedure is proposed in this third part of the guide. Indeed all the steps to obtain the decision tool namely the accuracy profile are described and illustrated step by step by a numerical example. This tool, based on the concept of total error (bias+standard deviation) build with a beta-expectation tolerance interval, allows to easily take the right decision and simultaneously minimizing the risk of the future use of the analytical procedure.

Comparison of two methods (left carotid artery and abdominal aorta) for surgical implantation of radiotelemetry devices in CD-1 mice.

The goal of this study was to compare two surgical methods, the left carotid (LC) and the abdominal aorta (AA), for mouse instrumentation with telemetry devices, to determine the best method for measuring cardiovascular (CV) parameters by radiotelemetry in freely moving mice. Surgery success rate, postsurgical recovery rate, clinical parameters, CV data (baseline and response to nicotine) and circadian rhythm measurements were compared between these techniques. Brains of LC-implanted mice were evaluated for potential ischaemia by direct observation of the Circle of Willis anatomy and histopathology. For this purpose, a total of 31 CD-1 male mice were instrumented with PA C20 devices (10 with LC and 21 with AA). Mortality, morbidity, physical examination, body weight (BW), water and food consumption (W/FC), mean blood pressure (MBP) and heart rate (HR) were monitored daily during the recovery period (10 days). CV baseline data were recorded continuously during two periods of four days, and finally, both LC- and AA-implanted mice received an acute subcutaneous administration of 1 mg/kg nicotine; BP and HR were recorded during 5 h after nicotine administration. Results showed that, in LC-implanted mice, 80% survived surgery and recovered well. In contrast, only 57% of mice implanted with the AA technique survived surgery and some presented lethal complications. Both techniques had similar recovery times for BW and W/FC, comparable return to normal circadian rhythm (day 6 post-surgery) and similar CV baseline values. No significant differences were observed in CV response to nicotine between both groups of implanted CD-1 mice. No histopathological changes suggestive of ischaemia were noted in the brain of mice implanted in the LC. Six out of the eight LC-implanted mice remained in good health and had good pressure signal for at least 100 days post-surgery, while most of the AA-implanted mice lost the signal pressure within 14-49 days post-surgery. In conclusion, we believe that LC implantation in mice is superior to the AA technique and is more appropriate for long-term telemetry studies, especially for smaller (transgenic) animals.

Harmonization of strategies for the validation of quantitative analytical procedures. A SFSTP proposal--part II.

As reported in a previous paper, the main objective of the new commission of the Société Française des Sciences et Techniques Pharmaceutiques (SFSTP) was the harmonisation of approaches for the validation of quantitative analytical procedures. In a series of meetings, members of this Commission have first tried to review the objectives of analytical methods and the objectives of validation methods and to recommend the use of two-sided beta-expectation tolerance intervals for total error of validation samples (accuracy profile) in the acceptance/rejection of analytical method in validation phase. In the context of the harmonization, the other objectives were: (i) to propose a consensus on the norms usually recognized, while widely incorporating the ISO terminology; (ii) to recommend to validate the analytical procedure accordingly to the way it will be used in routine; (iii) to elaborate a rational, practical and statistically reliable strategy to assure the quality of the analytical results generated. This strategy has been formalised in a guide and the three latter objectives made by the Commission are summarised in the present paper which is the second part of summary report of the SFSTP commission. The SFSTP guide has been produced to help analysts to validate their analytical methods. It is the result of a consensus between professionals having expertise in analytical and/or statistical fields. The suggestions presented in this paper should therefore help the analyst to design and perform the minimum number validation experiments needed to obtain all the required information to establish and demonstrate the reliability of its analytical procedure.

Improvement of the decision efficiency of the accuracy profile by means of a desirability function for analytical methods validation. Application to a diacetyl-monoxime colorimetric assay used for the determination of urea in transdermal iontophoretic extracts.

Validation of analytical methods is a widely used and regulated step for each analytical method. However, the classical approaches to demonstrate the ability to quantify of a method do not necessarily fulfill this objective. For this reason an innovative methodology was recently introduced by using the tolerance interval and accuracy profile, which guarantee that a pre-defined proportion of future measurements obtained with the method will be included within the acceptance limits. Accuracy profile is an effective decision tool to assess the validity of analytical methods. The methodology to build such a profile is detailed here. However, as for any visual tool it has a part of subjectivity. It was then necessary to make the decision process objective in order to quantify the degree of adequacy of an accuracy profile and to allow a thorough comparison between such profiles. To achieve this, we developed a global desirability index based on the three most important validation criteria: the trueness, the precision and the range. The global index allows the classification of the different accuracy profiles obtained according to their respective response functions. A diacetyl-monoxime colorimetric assay for the determination of urea in transdermal iontophoretic extracts was used to illustrate these improvements.

[Harmonisation of validation strategies and estimation of the associated uncertainty in the framework of assay laboratoires accreditation]. / Harmonisation des stratégies de validation et estimation de l'incertitude associée dans le cadre de l'accreditation des laboratoires d'essais.

In order to achieve the harmonization of validation strategies, the interpretation of several validation criteria is proposed. Furthermore, a simple and visual decision tool to assess the validity of an analytical procedure is described: the accuracy profile based on the estimation of the total error of the measurements. This profile build with beta-expectation tolerance intervals can also compute with efficiency the uncertainty related to the results of a laboratory, which is an essential parameter for the accreditation of laboratories under ISO 17025.

[The transfer of an automated method for the determination of noradrenaline in human urine: using total error as decision criterion]. / Le transfert d'une méthode de dosage automatisée de la noradrénaline dans l'urine humaine: utilisation de l'erreur totale comme critère de décision.

Two new statistical approaches based on the total error of measurements were applied to the transfer of an automated method for the determination of noradrenalin in human urine by LC-ECD coupled on-line to SPE. They showed that the receiving laboratory was mastering the analytical procedure allowing it to use the method in routine. Furthermore the risk based approach gave guarantee that the risk to have future measurements out of specification in the receiving laboratory was smaller than 5%, therefore being a risk management tool.

Comparison of FT-NIR transmission and UV-vis spectrophotometry to follow the mixing kinetics and to assay low-dose tablets containing riboflavin.

For several years, near-infrared spectroscopy (NIRS) has become an analytical technique of great interest for the pharmaceutical industry, particularly for the non-destructive analysis of dosage forms. The goal of this study is to show the capacity of this new technique to assay the active ingredient in low-dosage tablets. NIR spectroscopy is a rapid, non-destructive technique and does not need any sample preparation. As an example, a binary mixture of microcrystalline cellulose and riboflavin was used to prepare tablets of different weights by direct compression. A prediction model was built by using a partial least square regression fit method. The NIR assay was performed by transmission. The results obtained by NIR spectroscopy were compared with a conventional UV-vis spectrophotometry method. The study showed that tablets can be individually analysed by NIR with high accuracy. It was shown that the variability of this new technique is less important than that of the conventional method which is the UV-vis spectrophotometry.

The transfer of a LC-UV method for the determination of fenofibrate and fenofibric acid in Lidoses: use of total error as decision criterion.

Two new statistical approaches to assess the validity of the transfer of a LC-UV method for the determination of fenofibrate and fenofibric acid were investigated and compared to the conventional approaches generally used in this domain. These new approaches, namely the Tolerance Interval and the Risk approaches, are based on the simultaneous evaluation of the systematic (or trueness) and random (or precision) errors of the transfer into a single criterion called total error (or accuracy). The results of the transfer showed that only the total error based approaches fulfilled the objective of an analytical method transfer, i.e. to give guarantees that each future measurement made by the receiving laboratory will be close enough to the true value of the analyte in the sample. Furthermore the Risk approach was the most powerful one and allowed the estimation of the risk to have future measurements out of specification in the receiving laboratory, therefore being a risk management tool.

A new validation approach applied to the GC determination of impurities in organic solvents.

Organic solvents such as methanol, acetone, dichloromethane or toluene are frequently used in the pharmaceutical industry. The manufacturing of new active pharmaceutical ingredients (APIs) under GMP conditions commands to control adequately the quality of the different ingredients happening in the synthesis. Organic solvents have therefore to be controlled and their purity has to be determined before any GMP synthesis. A selective gas chromatography (GC) method has been developed to determine the purity of acetone, dichloromethane, methanol and toluene. Using this method, the main contaminants of each organic solvent can be quantified. Moreover, the developed method allows the simultaneous determination of ethanol, isopropanol, chloroform, benzene, acetone, dichloromethane, methanol and toluene. Propionitrile was used as the internal standard. The separation was obtained on a CP-SIL 8-CB low bleed/MS column (60 m x 0.32 mm i.d.x1.0 microm coating thickness). The GC method was fully validated using a new approach based on the accuracy profile as a decision tool. The determination of beta-expectation tolerance intervals for the estimation of total error - including both bias and precision - is used to better reflect the actual performances of the method, which is definitively the objective of the validation. The different validation criteria such as selectivity, response function, trueness, precision, accuracy, linearity or limits of detection and quantification were considered. The method was found to be able to quantitate with a good accuracy impurities around the 0.1% (v/v) concentration level for the different solvents.

Harmonization of strategies for the validation of quantitative analytical procedures. A SFSTP proposal--Part I.

This paper is the first part of a summary report of a new commission of the Société Française des Sciences et Techniques Pharmaceutiques (SFSTP). The main objective of this commission was the harmonization of approaches for the validation of quantitative analytical procedures. Indeed, the principle of the validation of theses procedures is today widely spread in all the domains of activities where measurements are made. Nevertheless, this simple question of acceptability or not of an analytical procedure for a given application, remains incompletely determined in several cases despite the various regulations relating to the good practices (GLP, GMP, ...) and other documents of normative character (ISO, ICH, FDA, ...). There are many official documents describing the criteria of validation to be tested, but they do not propose any experimental protocol and limit themselves most often to the general concepts. For those reasons, two previous SFSTP commissions elaborated validation guides to concretely help the industrial scientists in charge of drug development to apply those regulatory recommendations. If these two first guides widely contributed to the use and progress of analytical validations, they present, nevertheless, weaknesses regarding the conclusions of the performed statistical tests and the decisions to be made with respect to the acceptance limits defined by the use of an analytical procedure. The present paper proposes to review even the bases of the analytical validation for developing harmonized approach, by distinguishing notably the diagnosis rules and the decision rules. This latter rule is based on the use of the accuracy profile, uses the notion of total error and allows to simplify the approach of the validation of an analytical procedure while checking the associated risk to its usage. Thanks to this novel validation approach, it is possible to unambiguously demonstrate the fitness for purpose of a new method as stated in all regulatory documents.

Development of an antidepressant compliance questionnaire.

OBJECTIVE: The development of the Antidepressant Compliance Questionnaire (ADCQ), assessing patients' attitudes and beliefs on depression and antidepressants. METHOD: A 51-item questionnaire was applied to 85 psychiatric out-patients with a DSM-IV diagnosis of major depressive disorder (MDD). This data set was used to assess psychometric properties of the ADCQ. The questionnaire was also applied to 272 primary care out-patients with MDD. RESULTS: A principal component analysis revealed four dimensions with good internal consistency and acceptable test-retest reliability: 'perceived doctor-patient relationship', 'preserved autonomy', 'positive beliefs on antidepressants' and 'partner agreement', resulting in a final questionnaire comprising 33-items. Responses were independent from depression severity and patient age. The response patterns of both psychiatric and primary care patients are provided and illustrate the many erroneous beliefs on antidepressants. CONCLUSION: The ADCQ has good psychometric properties; further investigation should investigate whether this questionnaire is predictive of patient compliance.

An analysis of the SFSTP guide on validation of chromatographic bioanalytical methods: progress and limitations.

The Société Française des Sciences et Techniques Pharmaceutiques (SFSTP) published in 1997 a guide on the validation of chromatographic bio-analytical methods, which introduces new concepts in three different areas: stages of the validation, test of acceptability of a method and design of experiments to perform. In 'stages of validation', the SFSTP guide requires two phases to validate a method. The first phase, called 'prevalidation', is intended to (1) identify the model to use for the calibration curve; (2) evaluate the limits of quantitation; and (3) provide good estimates of the precision and bias of the method before designing the 'validation' phase per se. In the 'test of acceptability', the use of the interval hypotheses is envisaged by the SFSTP guide, not on the parameters of bias and precision, but on individual results by mixing mean bias and intermediate precision in a single test. The SFSTP guide also avoids the use of Satterthwaite's df for testing the acceptability. The reasons for those choices are discussed extensively. In 'design of experiments', much effort has been devoted to improving the quality of results by optimally designing and sizing the experiments to perform in validation. The rationale for using near D-optimal designs for the calibration curve is demonstrated and sample sizes are proposed to correctly size the validation experiments.

Compliance with antidepressants in a primary care setting, 1: Beyond lack of efficacy and adverse events.

BACKGROUND: Treatment guidelines recommend antidepressant treatment be continued for at least 6 months to ensure maximal improvement and to prevent relapse. Naturalistic studies show that the average length of treatment is shorter than 6 months and that dropout rates are high. Factors leading patients to discontinuation of therapy are not well understood. This study investigates when and why patients stop treatment and whether they inform their doctors. METHOD: Patients (N = 272) receiving antidepressant therapy due to an episode of major depressive disorder (DSM-IV) were asked to complete an antidepressant compliance questionnaire. Patients were then telephoned monthly while they continued on antidepressant therapy, up to 6 months. During each call, patients were asked standard questions. RESULTS: By endpoint, 53% of patients had discontinued antidepressant treatment. The most common reason given was "feeling better." However, different dropout reasons were prevalent at different times after initiation of therapy. Overall, 24% of the patients did not inform their physician about stopping the antidepressant medication. The likelihood of patients' informing their physicians differed according to the patients' reasons for discontinuation and according to the patients' perceptions of their relationship with their physicians. CONCLUSION: These results provide new guidelines for improving compliance. Strategy should be adapted to the stage of treatment, as patients' reasons for discontinuation vary as treatment progresses. The attitude of the physician and the information provided by the physician significantly influence whether patients inform the physician when they discontinue antidepressant therapy.

Compliance with antidepressants in a primary care setting, 2: the influence of gender and type of impairment.

BACKGROUND: DSM-IV diagnosis of major depressive disorder includes a requirement that symptoms result in significant clinical distress or impairment. This criterion is difficult to assess and is often overlooked. This study examines the use of the Sheehan Disability Scale as a possible method of assessing impairment, as well as the relationship between functioning and discontinuation of antidepressant medication. METHOD: Patients (N = 272) receiving antidepressant therapy due to an episode of major depressive disorder were asked to complete an antidepressant compliance questionnaire. Patients were telephoned monthly while they continued on antidepressant therapy, up to 6 months. During each call, the Sheehan Disability Scale was administered. RESULTS: Of patients referred to this study, 94.8% met DSM-IV criteria of at least 5 symptoms of major depressive disorder. Most patients had initial scores ranging from 5 to 8 on all 3 Sheehan disability subscales (occupational, social, and family functioning); 72% of patients had at least moderate impairment (scores > or = 4) on all 3 subscales. After 8 weeks of treatment, 42% of patients had scores < 4 on all 3 subscales (recovery); after 24 weeks, 64% of patients had scores < 4 on all 3 subscales. Dropout risk in men was related to improvement in occupational, social, and family functioning, whereas dropout risk in women was related only to improvement in family functioning. CONCLUSION: The Sheehan Disability Scale can be valuable in assessing impairment and thus in correctly diagnosing major depressive disorder. We suggest that scores of 4 or more (moderate impairment) on all 3 subscales indicate sufficient impairment for a strict diagnosis of major depressive disorder. Functional symptoms continued to improve for up to 24 weeks on antidepressant therapy, suggesting 6 months or more of therapy is necessary for maximum functional improvement. Premature discontinuation of antidepressant therapy is more likely to occur in women who experience significant improvement in family functioning or men who experience significant improvement in any functional area.

Adherence to treatment regimen in depressed patients treated with amitriptyline or fluoxetine.

OBJECTIVE: Non-compliance presents a constant challenge to effective therapy. Many studies only investigate early treatment discontinuation and not other measures like adherence to treatment regimen. We compared adherence in depressed patients using either a selective serotonin reuptake inhibitor (fluoxetine) or a tricyclic antidepressant (amitriptyline), and examined its clinical relevance through adverse events, drop-out rates, and outcome. Adherence was measured electronically with the MEMS (Medication Event Monitoring System). DESIGN: Nine-week double blind, randomized controlled trial. SETTING: Ambulatory psychiatric care. PATIENTS: Random sample of 66 depressed (DSM-III-R criteria) patients. INTERVENTION: Fluoxetine 20 mg or amitriptyline 150 mg. MAIN OUTCOME MEASURES: Time course of adherence and its relation to severe adverse events, drop-outs and outcome. RESULTS: Non-adherence to the treatment regimen occurred frequently in both treatment groups: 31% of patients had at least one 3-day drug holiday, and 34% of patients had at least one episode of three pills in a 24-h period. Over-consumption occurred more frequently during the early phases of treatment while under-consumption occurred more frequently during the later phases. Patients on amitriptyline (P=0.03) and patients with a higher pill intake (P=0.01) experienced more severe adverse events. Patients on amitriptyline (P=0.009) and patients with a lower adherence to the treatment regimen (P=0.004) discontinued from treatment more frequently. The final Hamilton score was significantly predicted by a longer duration of treatment and by a better adherence, but only in amitriptyline users. CONCLUSIONS: Non-adherence to the treatment regimen has important clinical consequences. Pharmacodynamics and human behavior predict risk for severe adverse events and drop-outs. Moreover, in amitriptyline users but not in fluoxetine users, better adherence predicts a better outcome.

A four-parameter index of marrow dysplasia has predictive value for survival in myelodysplastic syndromes.

Marrow dysplasia is a major characteristic of patients with myelodysplastic syndrome (MDS), along with marrow blastosis, cytopenia and cytogenetic anomalies. However, the impact of the degree of marrow dysplasia on survival has not been fully assessed. In this retrospective analysis of 111 patients selected according to the IPSS criteria of MDS diagnosis, the presence or absence of 21 dysplasia characteristics recognizable in bone marrow smears stained by the May-Grünwald-Giemsa method was correlated with patient survival. Using Cox proportional hazards regression analysis, megaloblastosis (MEGALO), neutrophil agranularity (AGRAN) and hypogranularity (HYPOGRAN) were highly significant predictors (p < 0.005), and Pelger-Huët anomaly (PELGHUET) a significant predictor (p = 0.05), of patient survival. The regression analysis yielded a dysplasia-based risk index (DI) where DI = 1.26 MEGALO + 0.82 AGRAN - 1.08 HYPOGRAN + 0.45 PELGHUET. The two subgroups of 60 and 47 patients with DI < or = 0 and > 0 showed highly significant differences in median survivals (2.6 vs 1.1 yrs; p <0.0001). Multivariate analysis further showed that DI offered additional predictive power that was independent of that provided by the IPSS (p=0.002 and 0.001 respectively). Analysis of survival curves stratified for IPSS and DI showed that the additional predictive power offered by inclusion of the DI essentially concerned the IPSS low/INT-1 risk categories. Further stratification for age did not improve survival prediction. The data indicate that a set of 4 dysplasia parameters can offer some prediction for survival of MDS patients in addition to that provided by the IPSS. Further multicenter studies should aim at including some form of evaluation of the degree of dysplasia in prognostic systems.

Comparison of rectal and infrared ear temperatures in older hospital inpatients.

OBJECTIVES: To assess the agreement between infrared emission detection (IRED) ear and rectal temperatures and to determine the validity of IRED ear thermometry in detecting rectal fever. DESIGN: Prospective, convenience sample, unblinded study. SETTING: An acute geriatric unit (teaching hospital) and a multidisciplinary intensive care unit. PARTICIPANTS: The study included 45 inpatients (26 women and 19 men), aged 78.3+/-6.9 years, admitted over a 4-month period. Twelve of the patients were definitely infected. MEASUREMENTS: Sequential rectal (RT) and ear temperature (ET) measurements were performed using mercury-in-glass and IRED ear thermometers, respectively. IRED ear temperatures were measured at both ears (unadjusted mode), with the highest of six ear temperatures considered the true value. RESULTS: Mean RT (37.39 degrees C +/- 0.52 degrees C) was significantly (P<.001) higher than mean ET (36.89 degrees C +/-0.59 degrees C). A highly significant positive correlation was found between RT and ET (slope = 0.69; 95% CI, 0.52-0.86; P<.001; r = 0.78). The mean bias (mean of the differences) between RT and ET was 0.50 degrees C +/-0.37 degrees C (95% CI, 0.41 degrees C-0.59 degrees C), and the 95% limits of agreement -0.22 degrees C and 1.23 degrees C (95% CI, -0.38 degrees C to 1.39 degrees C). According to the standard criterion (RT > or =37.6 degrees C), 14 patients were febrile. Using an optimum IRED ear fever threshold (37.2 degrees C), the sensitivity and specificity of IRED ear thermometry for predicting rectal fever were 86% and 89%, respectively (positive predictive value, 80%; negative predictive value, 93%). CONCLUSIONS: The degree of agreement between rectal temperature and the highest of six IRED ear temperatures was acceptable. Using an optimal IRED ear fever threshold of 37.2 degrees C (99 degrees F), IRED ear thermometry had acceptable sensitivity and specificity for predicting rectal fever.