Tree-temporal scan statistics for safety signal detection in vaccine clinical trials.

The evaluation of safety is critical in all clinical trials. However, the quantitative analysis of safety data in clinical trials poses statistical difficulties because of multiple potentially overlapping endpoints. Tree-temporal scan statistic approaches address this issue and have been widely employed in other data sources, but not to date in clinical trials. We evaluated the performance of three complementary scan statistical methods for routine quantitative safety signal detection: the self-controlled tree-temporal scan (SCTTS), a tree-temporal scan based on group comparison (BGTTS), and a log-rank based tree-temporal scan (LgRTTS). Each method was evaluated using data from two phase III clinical trials, and simulated data (simulation study). In the case study, the reference set was adverse events (AEs) in the Reference Safety Information of the evaluated vaccine. The SCTTS method had higher sensitivity than other methods, and after dose 1 detected 80 true positives (TP) with a positive predictive value (PPV) of 60%. The LgRTTS detected 49 TPs with 69% PPV. The BGTTS had 90% of PPV with 38 TPs. In the simulation study, with simulated reference sets of AEs, the SCTTS method had good sensitivity to detect transient effects. The LgRTTS method showed the best performance for the detection of persistent effects, with high sensitivity and expected probability of type I error. These three methods provide complementary approaches to safety signal detection in clinical trials or across clinical development programmes. All three methods formally adjust for multiple testing of large numbers of overlapping endpoints without being excessively conservative.

Assurance in vaccine efficacy clinical trial design based on immunological responses.

The assurance of a future clinical trial is a key quantitative tool for decision-making in drug development. It is derived from prior knowledge (Bayesian approach) about the clinical endpoint of interest, typically from previous clinical trials. In this paper, we examine assurance in the specific context of vaccine development, where early development (Phase 2) is often based on immunological endpoints (e.g., antibody levels), while the confirmatory trial (Phase 3) is based on the clinical endpoint (very large sample sizes and long follow-up). Our proposal is to use the Phase 2 vaccine efficacy predicted by the immunological endpoint (using a model estimated from epidemiological studies) as prior information for the calculation of the assurance.

Assessing correlates of protection in vaccine trials: statistical solutions in the context of high vaccine efficacy.

BACKGROUND: The use of correlates of protection (CoPs) in vaccination trials offers significant advantages as useful clinical endpoint substitutes. Vaccines with very high vaccine efficacy (VE) are documented in the literature (VE ≥95%). The rare events (number of infections) observed in the vaccinated groups of these trials posed challenges when applying conventionally-used statistical methods for CoP assessment. In this paper, we describe the nature of these challenges, and propose easy-to-implement and uniquely-tailored statistical solutions for the assessment of CoPs in the specific context of high VE. METHODS: The Prentice criteria and meta-analytic frameworks are standard statistical methods for assessing vaccine CoPs, but can be problematic in high VE cases due to the rare events data available. As a result, lack of fit and the problem of infinite estimates may arise, in the former and latter methods respectively. The use of flexible models within the Prentice framework, and penalized-likelihood methods to solve the issue of infinite estimates can improve the performance of both methods in high VE settings. RESULTS: We have 1) devised flexible non-linear models to counteract the Prentice framework lack of fit, providing sufficient statistical power to the method, and 2) proposed the use of penalised likelihood approaches to make the meta-analytic framework applicable on randomized subgroups, such as regions. The performance of the proposed methods for high VE cases was evaluated by running simulations. CONCLUSIONS: As vaccines with high efficacy are documented in the literature, there is a need to identify effective statistical solutions to assess CoPs. Our proposed adaptations are straight-forward and improve the performance of conventional statistical methods for high VE data, leading to more reliable CoP assessments in the context of high VE settings.

Statistical methods for the analysis of clinical trials data containing many zeros: An application in vaccine development.

In recent years, many vaccines have been developed for the prevention of a variety of diseases. Many of these vaccines, like the one for herpes zoster, are supposed to act in a multilevel way. Ideally, they completely prevent expression of the virus, but failing that they help to reduce the severity of the disease. A simple approach to analyze these data is the so-called burden-of-illness test. The method assigns a score, say W, equal to 0 for the uninfected and a post-infection outcome X > 0 for the infected individuals. One of the limitations of this test is the potential low power when the vaccine efficacy is close to 0. To overcome this limitation, we propose a Fisher adjusted test where we combine a statistic for infection with a statistic for post-infection outcome adjusted for selection bias. The advantages and disadvantages of different methods proposed in the literature are discussed. We compared the methods via simulations in herpes zoster, HIV, and malaria vaccine trial settings. In addition, we applied these methods to published data on HIV vaccine. The paper ends with some recommendations and conclusions.

Hemagglutination Inhibition Antibody Titers as a Correlate of Protection Against Seasonal A/H3N2 Influenza Disease.

Background. To investigate the relationship between hemagglutinin-inhibition (HI) antibody levels to the risk of influenza disease, we conducted a correlate of protection analysis using pooled data from previously published randomized trials. Methods. Data on the occurrence of laboratory-confirmed influenza and HI levels pre- and postvaccination were analyzed from 4 datasets: 3 datasets included subjects aged <65 years who received inactivated trivalent influenza vaccine (TIV) or placebo, and 1 dataset included subjects aged ≥65 years who received AS03-adjuvanted TIV (AS03-TIV) or TIV. A logistic model was used to evaluate the relationship between the postvaccination titer of A/H3N2 HI antibodies and occurrence of A/H3N2 disease. We then built a receiver-operating characteristic curve to identify a potential cutoff titer between protection and no protection. Results. The baseline odds ratio of A/H3N2 disease was higher for subjects aged ≥65 years than <65 years and higher in seasons of strong epidemic intensity than moderate or low intensity. Including age and epidemic intensity as covariates, a 4-fold increase in titer was associated with a 2-fold decrease in the risk of A/H3N2 disease. Conclusions. The modeling exercise confirmed a relationship between A/H3N2 disease and HI responses, but it did not allow an evaluation of the predictive power of the HI response.

Effects of varying antigens and adjuvant systems on the immunogenicity and safety of investigational tetravalent human oncogenic papillomavirus vaccines: results from two randomized trials.

BACKGROUND: A prophylactic human papillomavirus (HPV) vaccine targeting oncogenic HPV types in addition to HPV-16 and -18 may broaden protection against cervical cancer. Two Phase I/II, randomized, controlled studies were conducted to compare the immunogenicity and safety of investigational tetravalent HPV L1 virus-like particle (VLP) vaccines, containing VLPs from two additional oncogenic genotypes, with the licensed HPV-16/18 AS04-adjuvanted vaccine (control) in healthy 18-25 year-old women. METHODS: In one trial (NCT00231413), subjects received control or one of 6 tetravalent HPV-16/18/31/45 AS04 vaccine formulations at months (M) 0,1,6. In a second trial (NCT00478621), subjects received control or one of 5 tetravalent HPV-16/18/33/58 vaccines formulated with different adjuvant systems (AS04, AS01 or AS02), administered on different schedules (M0,1,6 or M0,3 or M0,6). RESULTS: One month after the third injection (Month 7), there was a consistent trend for lower anti-HPV-16 and -18 geometric mean antibody titers (GMTs) for tetravalent AS04-adjuvanted vaccines compared with control. GMTs were statistically significantly lower for an HPV-16/18/31/45 AS04 vaccine containing 20/20/10/10 µg VLPs for both anti-HPV-16 and anti-HPV-18 antibodies, and for an HPV-16/18/33/58 AS04 vaccine containing 20/20/20/20 µg VLPs for anti-HPV-16 antibodies. There was also a trend for lower HPV-16 and -18-specific memory B-cell responses for tetravalent AS04 vaccines versus control. No such trends were observed for CD4(+) T-cell responses. Immune interference could not always be overcome by increasing the dose of HPV-16/18 L1 VLPs or by using a different adjuvant system. All formulations had acceptable reactogenicity and safety profiles. Reactogenicity in the 7-day post-vaccination period tended to increase with the introduction of additional VLPs, especially for formulations containing AS01. CONCLUSIONS: HPV-16 and -18 antibody responses were lower when additional HPV L1 VLPs were added to the HPV-16/18 AS04-adjuvanted vaccine. Immune interference is a complex phenomenon that cannot always be overcome by changing the antigen dose or adjuvant system.

Model-based estimates of long-term persistence of induced HPV antibodies: a flexible subject-specific approach.

In infectious diseases, it is important to predict the long-term persistence of vaccine-induced antibodies and to estimate the time points where the individual titers are below the threshold value for protection. This article focuses on HPV-16/18, and uses a so-called fractional-polynomial model to this effect, derived in a data-driven fashion. Initially, model selection was done from among the second- and first-order fractional polynomials on the one hand and from the linear mixed model on the other. According to a functional selection procedure, the first-order fractional polynomial was selected. Apart from the fractional polynomial model, we also fitted a power-law model, which is a special case of the fractional polynomial model. Both models were compared using Akaike's information criterion. Over the observation period, the fractional polynomials fitted the data better than the power-law model; this, of course, does not imply that it fits best over the long run, and hence, caution ought to be used when prediction is of interest. Therefore, we point out that the persistence of the anti-HPV responses induced by these vaccines can only be ascertained empirically by long-term follow-up analysis.

Long-term persistence of anti-HPV-16 and -18 antibodies induced by vaccination with the AS04-adjuvanted cervical cancer vaccine: modeling of sustained antibody responses.

OBJECTIVES: Strong and sustained HPV-16 and -18 antibody responses have been observed in previously unexposed women aged 15-25 years vaccinated with the AS04-adjuvanted HPV-16/18 L1 virus-like particle vaccine. While awaiting the extended results of ongoing trials, our objective was to predict the long-term persistence of anti-HPV-16/18 antibodies in vaccinees by applying three statistical models using immunogenicity data from vaccinated women with serum samples collected up to 6.4 years after first vaccination. Two different data lock-points (up to 5.5 years and up to 6.4 years) were used to assess the robustness of the models. METHODS: Three statistical models were applied to estimate the long-term persistence of anti-HPV-16/18 antibodies in 393 women vaccinated with the AS04-adjuvanted HPV-16/18 vaccine. Individual antibody levels for each study participant at each timepoint up to 6.4 years were input to previously published power-law and modified power-law models. The power-law model estimates antibody decay over time. The modified power-law model takes into account both antibody persistence over time and immune memory. A third model, the piece-wise model, fits the data based on three different non-overlapping intervals (between Months 7 and 12, Months 12 and 21, and over 21 months), corresponding to the observed decay of vaccine-induced antibodies. RESULTS: HPV-16 and -18 antibodies peaked at Month 7 and gradually plateaued at Months 18-24 and remained stable through 6.4 years. Mean antibody levels at the last timepoint were several fold higher than those associated with natural infection. All three models predict that HPV-16 and -18 mean antibody levels will remain well above those associated with natural infection for at least 20 years, when using data from 5.5 as well as 6.4 years' follow-up. Predictions are similar for the modified power-law model and improve with longer follow-up for both the power-law and the piece-wise models. CONCLUSIONS: Vaccination with the AS04-adjuvanted HPV-16/18 vaccine is predicted to provide long-term persistence for both HPV-16 and -18 antibodies, independent of the statistical model applied. Model predictions are based on conservative mathematical assumptions. Since the input of longer term data of up to 6.4 years showed an improved profile compared with that for data up to 5.5 years, the predictions of antibody persistence based on population means are conservative when predicting that antibody levels will remain well above levels induced by natural infection for 20 years.

AIR inhaled insulin versus subcutaneous insulin: pharmacokinetics, glucodynamics, and pulmonary function in asthma.

OBJECTIVE: This study evaluated pharmacokinetic and glucodynamic responses to AIR inhaled insulin relative to subcutaneous insulin lispro, safety, pulmonary function, and effects of salbutamol coadministration. RESEARCH DESIGN AND METHODS: Healthy, mildly asthmatic, and moderately asthmatic subjects (n = 13/group, aged 19-58 years, nonsmoking, and nondiabetic) completed this phase I, open-label, randomized, crossover euglycemic clamp study. Subjects received 12 units equivalent AIR insulin or 12 units subcutaneous insulin lispro or salbutamol plus AIR insulin (moderate asthma group only) before the clamp. RESULTS: AIR insulin exposure was reduced 34 and 41% (both P < 0.01) in asthmatic subjects (area under the curve(0-t'), 24.0 and 21.1 nmol x min x l(-1) in mild and moderate asthma subjects, respectively) compared with healthy subjects (35.2 nmol x min x l(-1)), respectively. Glucodynamic (G) effects were similar in healthy and mildly asthmatic subjects (G(tot) = 38.7 and 23.4 g, respectively; P = 0.16) and were reduced in moderately asthmatic subjects (G(tot) = 10.7 g). Salbutamol pretreatment (moderately asthmatic subjects) improved bioavailability. AIR insulin had no discernable effect on pulmonary function. AIR insulin adverse events (cough, headache, and dizziness) were mild to moderate in intensity and have been previously reported or are typical of studies involving glucose clamp procedures. CONCLUSIONS: This study suggests that pulmonary disease severity and asthma treatment status influence the metabolic effect of AIR insulin in individuals with asthma but do not affect AIR insulin pulmonary safety or tolerability. In view of the potential interactions between diabetes treatment and pulmonary status, it is prudent to await the results of ongoing clinical trials in diabetic patients with comorbid lung disease before considering the use of inhaled insulin in such patients.

Accounting for variability in individual hierarchical clinical trial data.

Meta-analytical approaches have been extensively used to analyze medical data. In most cases, the data come from different studies or independent trials with similar characteristics. However, these methods can be applied in a broader sense. In this paper, we show how existing meta-analytic techniques can also be used as well when dealing with parameters estimated from individual hierarchical data. Specifically, we propose to apply statistical methods that account for the variances (and possibly covariances) of such measures. The estimated parameters together with their estimated variances can be incorporated into a general linear mixed model framework. We illustrate the methodology by using data from a first-in-man study and a simulated data set. The analysis was implemented with the SAS procedure MIXED and example code is offered.

Conditional mixed models with crossed random effects.

The analysis of continuous hierarchical data such as repeated measures or data from meta-analyses can be carried out by means of the linear mixed-effects model. However, in some situations this model, in its standard form, does pose computational problems. For example, when dealing with crossed random-effects models, the estimation of the variance components becomes a non-trivial task if only one observation is available for each cross-classified level. Pseudolikelihood ideas have been used in the context of binary data with standard generalized linear multilevel models. However, even in this case the problem of the estimation of the variance remains non-trivial. In this paper, we first propose a method to fit a crossed random-effects model with two levels and continuous outcomes, borrowing ideas from conditional linear mixed-effects model theory. We also propose a crossed random-effects model for binary data combining ideas of conditional logistic regression with pseudolikelihood estimation. We apply this method to a case study with data coming from the field of psychometrics and study a series of items (responses) crossed with participants. A simulation study assesses the operational characteristics of the method.

AIR inhaled insulin in subjects with chronic obstructive pulmonary disease: pharmacokinetics, glucodynamics, safety, and tolerability.

OBJECTIVE: In this open-label, randomized, crossover study, pharmacokinetic and glucodynamic responses were compared in healthy subjects versus subjects with moderate chronic obstructive pulmonary disease (COPD), following administration of 12 units equivalent AIR inhaled insulin versus 12 units subcutaneous insulin lispro. RESEARCH DESIGN AND METHODS: Three nonsmoking groups (n = 15 each)--healthy subjects (baseline mean +/- SD age 38 +/- 13 years, forced expiratory volume in 1 s [FEV1] 4.06 +/- 1.04 l), subjects with chronic bronchitis (aged 53 +/- 9 years, FEV1 2.14 +/- 0.60 l), and subjects with pulmonary emphysema (aged 58 +/- 6 years, FEV1 1.67 +/- 0.61 l)--were randomly assigned to one of three treatment sequences. Three euglycemic glucose clamp procedures were performed. RESULTS: In subjects with chronic bronchitis and emphysema, AIR inhaled insulin administration resulted in reduced insulin exposure (area under the serum insulin concentration curve from time zero until time of return to baseline [AUC(0-t')]) (55.7%, P = 0.13 and 78.5%, P < 0.001, respectively) and reduced total insulin effect (total glucose infusion rate) (60.4%, P < 0.01 and 67.1%, P < 0.01, respectively) relative to healthy subjects. Subcutaneous insulin lispro administration resulted in similar responses across study groups for insulin exposure and metabolic effect. Intrasubject pharmacokinetic and glucodynamic variability ranged from 17 to 52% across groups. No significant differences were shown for pre- and postclamp pulmonary function tests. During clamps, FEV1 and forced vital capacity declined modestly in both COPD groups, with no difference between AIR insulin and subcutaneous insulin lispro. CONCLUSIONS: Short-term exposure to AIR inhaled insulin was well tolerated by COPD subjects, showing similar time-exposure and time-action profiles, but with reduced insulin absorption and metabolic effect compared with healthy subjects. Further clinical evaluation is warranted in patients with comorbid diabetes and COPD.

Prentice's approach and the meta-analytic paradigm: a reflection on the role of statistics in the evaluation of surrogate endpoints.

We put a perspective on the strengths and limitations of statistical methods for the evaluation of surrogate endpoints. Whereas using several trials overcomes some of the limitations of a single-trial framework (Prentice, 1989, Statistics in Medicine 8, 431-440), arguably the evaluation of surrogate endpoints can never be done using only statistical evidence but such evidence should be seen as but one component in a decision-making process that involves, among others, a number of clinical and biological considerations. We briefly present a hierarchical framework that incorporates ideas from Prentice's work and is uniformly applicable to different types of surrogate and true clinical outcomes.

Modelling associations between time-to-event responses in pilot cancer clinical trials using a Plackett-Dale model.

This work was motivated by the need to find surrogate endpoints for survival of patients in oncology studies. The goal of this article is to determine associations between five time-to-event outcomes coming from three clinical trials for non-small cell lung cancer. To this end, we propose to use the multivariate Dale model for time-to-event data introduced by Tibaldi et al. (Stat. Med. 2003). We fit the model to these data, using a pseudo-likelihood approach to estimate the model parameters. We evaluate and compare the performance of different dimensional models and we relate the Dale model association parameter, i.e. the odds ratio, to well-known quantities such as Kendall's tau and Spearman's rho. Finally, the results are discussed with a perspective on surrogate marker validation. Some suggestions are made regarding further studies in this field.

Pseudo-likelihood estimation for a marginal multivariate survival model.

In this paper, we propose a multivariate Plackett-Dale model for survival outcomes. A pseudo-likelihood method for the estimation of the parameters is proposed and these ideas are applied to two case studies. The modelling approach is similar in spirit but different from Parner's approach. The first study is in AIDS, where the overall survival time and different opportunistic infections in HIV-infected patients are studied. The second study is on adoption data where the association of the survival times within families is modelled, illustrating the use of the proposed methodology for the context of population genetics.

Psychomotor development in Argentinean children aged 0-5 years.

In Argentina, there is no information on ages of attainment of developmental milestones and very few data about environmental factors that influence them. A national survey on the psychomotor development of children under 6 years of age was carried out with the help of 129 paediatricians. Logistic regression was applied to a final sample of 3573 healthy, normal children in order to estimate selected centiles (25th, 50th, 75th and 90th), together with their respective confidence intervals, of the ages of attainment of 78 developmental items belonging to the following areas: personal-social (18 items), fine motor (19), language (18) and gross motor (23). The 50th centile obtained for each of the 43 comparable items was compared with those obtained in previously standardised tests: DDST, Denver II, Bayley and Chilean scales. Neither significant nor systematic differences were found between our results and those described in the tests used for comparison. Multiple logistic regressions showed that social class, maternal education and sex (female) were associated with earlier attainment of some selected developmental items, achieved at ages later than 1 year. Selected items achieved before the first year of life were not affected by any of the independent environmental variables studied. The information is useful in helping paediatricians in their daily practice for surveillance of development, as baseline information for epidemiological studies on development in our country and for cross-cultural analysis.

Organización de un estudio nacional para la evaluación del desarrollo psicomotor infantil, con un programa de entrenamiento para pediatras / The organization of a national survey for evaluating child psychomotor development in Argentina, with a training program for pediatricians

Doscientos once pediatras fueron seleccionados e invitados a participar en una encuesta nacional destinada a evaluar la edad de cumplimiento de pautas de desarrollo en niños de 0 a 5 años. Después de un estudio piloto y del proceso de entrenamiento en cascada, el 61,1 por ciento de los pediatras completó satisfactoriamente la recolección de la información sobre 229 pautas de desarrollo. En el estudio piloto hubo significativamente mayor proporción de pruebas de desarrollo no realizadas por los pediatras en el grupo de niños mayores de 1 año, lo que confirma la impresión de que los pediatras están más familiarizados con la evaluación del desarrollo en lactantes que en niños mayores. El impacto del entrenamiento fue significativo en el sentido que el porcentaje de las pruebas de desarrollo no realizadas por los pediatras descendió del 3,26 por ciento en el estudio piloto al 1,88 por ciento después del entrenamiento (p<0,01). La muestra final obtenida de 3.573 niños representó el 0,11 por ciento de la población nacional del mismo rango etario, el índice de masculinidad fue de 1,01. No hubo diferencias significativas en la distribución geográfica de la muestra en relación a la población nacional. El nivel socioeconómico evaluado a partir del nivel educacional materno esta sesgado hacia niveles educacionales altos en relación a la población nacional. La media más menos error/estándar de los puntajes estandarizados Z para peso y estatura derivados de los estándares nacionales variaron entre 0,06 más menos 0,03 y 0,03 más menos 0,06 en niños, y 0,04 más menos 0,06 y 0,08 más menos 0,04 en niñas, lo cual representó una diferencia mínima con los estándares nacionales,sin significancia biológica. la muestra es comparable con otras muestras de carácter internacional. Además de haberse obtenido satisfactoriamente una muestra representativa para el análisis de la edad de alcance de pautas madurativas en niños sanos de nuestro país.

Edad de cumprimiento de pautas de desarrollo en niños argentinos sanos menores de seis años / Age of development milestones by healthy argentinians under six

En la Argetina no hay información sobre la media y la variación individual de la edad en que los niños sanos de nuestro país alcanzan determinadas pautas del desarrollo. Sobre estas bases se llevó a cabo el Programa Colaborativo Nacional de Evaluación del Desarrollo en Niños Menores de Seis Años. En el marco de este estudio transversal, se obtuvo una muestra nacional y geográficamente proporcional de 3.573 niñas y niños sanos menores de seis años. Los niños fueron evaluados por 129 pediatras previamente entrenados y de acurdo a un instructivo normalizado. Sobre esta muestra, y con el método de regresión logística, se estimaron los centilos 25§, 50§, 75§ y 90§ y sus respectivos intervalos de confianza, de la edad de cumprimiento de 78 pautas de desarrollo, agrupadas por las siguientes áreas: 18,19, 18 y 23 en las áreas personal social, motriz fina, lenguaje y motriz gruesa respectivamente. Se comparó el centilo 50§ de nuestros resultados en cada pauta, con el centilo 50§ descripto por las pruebas de Denver, Denver II, Bayley y Rodríguez...

Comparación de la información obtenida mediante una ficha anestésica manual vs. una automática computarizada: análisis de los datos y su implicancia médico-legal / Comparison between the information obtained by anesthesia record handwritten with the one recorded by an automatic computarized anesthesia record system: analysis of data and its medico legal implications

Objetivos: Comparar la información encontrada en las fichas anestésicas tradicionales confeccionadas de puño y letra con la registrada con el sistema automático de registro anestésico utilizado por nuestro Servicio. Lugar: Hospital de Gastroenterología de Buenos Aires "Dr. Carlos Bonorino Udaondo". Material y Métodos: Se seleccionaron en forma aleatoria para su análisis 100 fichas manuales y 100 automáticas confeccionadas durante el período 1994 - 1995. Se compararon todos los datos consignados contenidos en ambos tipos de fichas y se estudiaron las variables fisiológicas de los pacientes anestesiados. Resultados: Los parámetros registrados en forma automática superan ampliamente los standards de mínima de seguridad intraoperatoria. Es mayor el número de eventos, drogas y datos del paciente documentados en el sistema automático que en las fichas manuales, no existiendo en las primeras el problema de la ilegilibilidad. Conclusión: La ficha anestésica automática resulta una herramienta de gran utilidad en la cabecera del paciente, aporta mayor información sobre lo ocurrido durante el intraoperatorio y se comporta como un elemento de documentación que facilita la defensa médico-legal del anestesiólogo involucrado en una demanda judicial.