Vaccine clinical trials with active surveillance for infection often use the time to infection as the primary endpoint. A common method of analysis for such trials is to compare the times to infection between the vaccine and placebo groups using a Cox regression model. With new technology, we can sometimes additionally record the precise number of virions that cause infection rather than just the indicator that infection occurred. In this article, we develop a unified approach for vaccine trials that couples the time to infection with the number of infecting or founder viruses. We assume that the instantaneous risk of a potentially infectious exposure for individuals in the placebo and vaccine groups follows the same proportional intensity model. Following exposure, the number of founder viruses X* is assumed to be generated from some distribution on 0, 1, . . . , which is allowed to be different for the two groups. Exposures that result in X* = 0 are unobservable. We denote the placebo and vaccine means of X* by μ and μΔ so that 1 – Δ measures the proportion reduction in the mean number of infecting virions due to vaccination per exposure. We develop different semi-parametric methods of estimating Δ. We allow the distribution of X* to be Poisson or unspecified, and discuss how to incorporate covariates that impact the time to exposure and/or X*. Interestingly Δ, which is a ratio of untruncated means, can be reliably estimated using truncated data (X*
> 0), even if the placebo and vaccine distributions of X* are completely unspecified. Simulations of vaccine clinical trials show that the method can reliably recover Δ in realistic settings. We apply our methods to an HIV vaccine trial conducted in injecting drug users.
Burden of illness; Competing risks; Cox regression; Empirical process; Infectious disease; Marked process
Patients with Chronic Granulomatous Disease (CGD) suffer immunodeficiency due to defects in the phagocyte NADPH oxidase (NOX2) and concomitant reduction in reactive oxygen intermediates. This may result in a reduction in atherosclerotic injury.
Methods and Results
We prospectively assessed the prevalence of cardiovascular risk factors, biomarkers of inflammation and neutrophil activation, and the presence of MRI and CT quantified subclinical atherosclerosis in the carotid and coronary arteries of 41 CGD patients and 25 healthy controls in the same age range. Uni- and multivariable associations between risk factors, inflammatory markers and atherosclerosis burden were assessed. CGD patients had significant elevations in traditional risk factors and inflammatory markers compared with controls, including; hypertension, hsCRP, oxidized LDL, and low HDL. Despite this, CGD patients had a 22% lower internal carotid artery wall volume compared with controls (361.3 ± 76.4 mm3 vs. 463.5 ± 104.7 mm3, p<0.001). This difference was comparable in p47phox and gp91phox deficient subtypes of CGD, and independent of risk factors in multivariate regression analysis. In contrast, prevalence of coronary arterial calcification was similar between CGD patients and controls (14.6%, CGD, and 6.3%, controls, p=0.39).
The observation by MRI of reduced carotid but not coronary artery atherosclerosis in CGD patients despite the high prevalence of traditional risk factors raises questions about the role of NOX2 in the pathogenesis of clinically significant atherosclerosis. Additional high-resolution studies in multiple vascular beds are required to address the therapeutic potential of NOX-inhibition in cardiovascular diseases.
Clinical Trial Registration Information
clinicaltrials.gov. Identifier: NCT01063309.
atherosclerosis; inflammation; carotid arteries; coronary arteries; immune system
This paper proposes a unified framework to characterize the rate function of a recurrent event process through shape and size parameters. In contrast to the intensity function, which is the event occurrence rate conditional on the event history, the rate function is the occurrence rate unconditional on the event history, and thus it can be interpreted as a population-averaged count of events in unit time. In this paper, shape and size parameters are introduced and used to characterize the association between the rate function λ(·) and a random variable X. Measures of association between X and λ(·) are defined via shape- and size-based coefficients. Rate-independence of X and λ(·) is studied through tests of shape-independence and size-independence, where the shape-and size-based test statistics can be used separately or in combination. These tests can be applied when X is a covariable possibly correlated with the recurrent event process through λ(·) or, in the one-sample setting, when X is the censoring time at which the observation of N(·) is terminated. The proposed tests are shape- and size-based, so when a null hypothesis is rejected, the test results can serve to distinguish the source of violation.
Intensity function; Point process; Poisson process; Rate-independence; Rate function; Shape-independence; Size-independence
Malaria elimination efforts would benefit from vaccines that block transmission of Plasmodium falciparum gametocytes from humans to mosquitoes. A clear understanding of gametocyte-specific antibody responses in exposed populations could help determine whether transmission-blocking vaccines (TBV) would be boosted by natural gametocyte exposure, and also inform the development of serologic tools to monitor gametocyte exposure in populations targeted for malaria elimination. To this end, plasma was collected from Malian children and adults before and after the 6-month malaria season and probed against a microarray containing 1,204 P. falciparum proteins. Using publicly available proteomic data, we classified 91 proteins as gametocyte specific and 69 as proteins not expressed by gametocytes. The overall breadth and magnitude of gametocyte-specific IgG responses increased during the malaria season, although they were consistently lower than IgG responses to nongametocyte antigens. Notably, IgG specific for the TBV candidates Pfs48/45 and Pfs230 increased during the malaria season. In addition, IgGs specific for the gametocyte proteins Pfmdv1, Pfs16, PF3D7_1346400, and PF3D7_1024800 were detected in nearly all subjects, suggesting that seroconversion to these proteins may be a sensitive indicator of gametocyte exposure, although further studies are needed to determine the specificity and kinetics of these potential serologic markers. These findings suggest that TBV-induced immunity would be boosted through natural gametocyte exposure, and that antibody responses to particular antigens may reliably indicate gametocyte exposure.
Survival data from prevalent cases collected under a cross-sectional sampling scheme are subject to left-truncation. When fitting an additive hazards model to left-truncated data, the conditional estimating equation method (Lin & Ying, 1994), obtained by modifying the risk sets to account for left-truncation, can be very inefficient, as the marginal likelihood of the truncation times is not used in the estimation procedure. In this paper, we use a pairwise pseudolikelihood to eliminate nuisance parameters from the marginal likelihood and, by combining the marginal pairwise pseudo-score function and the conditional estimating function, propose an efficient estimator for the additive hazards model. The proposed estimator is shown to be consistent and asymptotically normally distributed with a sandwich-type covariance matrix that can be consistently estimated. Simulation studies show that the proposed estimator is more efficient than its competitors. A data analysis illustrates application of the method.
Canadian Study of Health and Aging; Composite likelihood; Estimating equation; Martingale; Prevalent sampling
Vaccine-induced immunity depends on long-lived plasma cells (LLPCs) that maintain antibody levels. A recent mouse study showed that Plasmodium chaubaudi infection reduced pre-existing influenza-specific antibodies—raising concerns that malaria may compromise pre-existing vaccine responses. We extended these findings to P. yoelii infection, observing decreases in antibodies to model antigens in inbred mice and to influenza in outbred mice, associated with LLPC depletion and increased susceptibility to influenza rechallenge. We investigated the implications of these findings in Malian children by measuring vaccine-specific IgG (tetanus, measles, hepatitis B) before and after the malaria-free 6-month dry season, 10 days after the first malaria episode of the malaria season, and after the subsequent dry season. On average, vaccine-specific IgG did not decrease following acute malaria. However, in some children malaria was associated with an accelerated decline in vaccine-specific IgG, underscoring the need to further investigate the impact of malaria on pre-existing vaccine-specific antibodies.
SCID is a syndrome characterized by profound T cell deficiency. BCG vaccine is contraindicated in SCID patients. Because most countries encourage BCG vaccination at birth, a high percent of SCID patients are vaccinated before their immune defect is detected.
To describe the complications and risks associated with BCG vaccination in SCID patients.
An extensive standardized questionnaire evaluating complications, therapeutics, and outcome regarding BCG in patients diagnosed with SCID was widely distributed. Summary statistics and association analysis was performed.
Data on 349 BCG vaccinated SCID patients from 28 centers in 17 countries was analyzed. Fifty-one percent of the patients developed BCG complications, 34% disseminated and 17% localized (a 33,000 and 400 fold increase, respectively, over the general population). Patients receiving early vaccination (≤ 1 month) showed an increased prevalence of complications (p=0.006) and death due to BCG complications (p<0.0001). The odds of experiencing complications among patients with T cells ≤ 250/uL at diagnosis was 2.1 times higher (95% CI, 1.4-3.4; p = 0.001) than among those with T cells > 250/uL. BCG complications were reported in 2/78 patients who received anti-mycobacterial therapy while asymptomatic and no deaths due to BCG complications occurred in this group. In contrast 46 BCG-associated deaths were reported among 160 patients treated with anti-mycobacterial therapy for a symptomatic BCG infection (p<0.0001).
BCG vaccine has a very high rate of complications in SCID patients, which increase morbidity and mortality rates. Until safer and more efficient anti-tuberculosis vaccines become available, delay in BCG vaccination should be considered to protect highly vulnerable populations from preventable complications.
primary immunodeficiency; SCID; vaccine; BCG; mycobacteria; new born screening; hematopoietic stem cell transplant; immune reconstitution syndrome
Background. Plasmodium falciparum reticulocyte-binding protein homologue 5 (PfRH5) is a blood-stage parasite protein essential for host erythrocyte invasion. PfRH5-specific antibodies raised in animals inhibit parasite growth in vitro, but the relevance of naturally acquired PfRH5-specific antibodies in humans is unclear.
Methods. We assessed pre–malaria season PfRH5-specific immunoglobulin G (IgG) levels in 357 Malian children and adults who were uninfected with Plasmodium. Subsequent P. falciparum infections were detected by polymerase chain reaction every 2 weeks and malaria episodes by weekly physical examination and self-referral for 7 months. The primary outcome was time between the first P. falciparum infection and the first febrile malaria episode. PfRH5-specific IgG was assayed for parasite growth-inhibitory activity.
Results. The presence of PfRH5-specific IgG at enrollment was associated with a longer time between the first blood-stage infection and the first malaria episode (PfRH5-seropositive median: 71 days, PfRH5-seronegative median: 18 days; P = .001). This association remained significant after adjustment for age and other factors associated with malaria risk/exposure (hazard ratio, .62; P = .02). Concentrated PfRH5-specific IgG purified from Malians inhibited P. falciparum growth in vitro.
Conclusions. Naturally acquired PfRH5-specific IgG inhibits parasite growth in vitro and predicts protection from malaria. These findings strongly support efforts to develop PfRH5 as an urgently needed blood-stage malaria vaccine.
Clinical Trials Registration NCT01322581.
RH5; blood-stage immunity; endemic population; Plasmodium falciparum; prospective cohort study; malaria
Severe Plasmodium falciparum malaria is a major cause of death in children. The contribution of the parasite burden to the pathogenesis of severe malaria has been controversial.
We documented P. falciparum infection and disease in Tanzanian children followed from birth for an average of 2 years and for as long as 4 years.
Of the 882 children in our study, 102 had severe malaria, but only 3 had more than two episodes. More than half of first episodes of severe malaria occurred after a second infection. Although parasite levels were higher on average when children had severe rather than mild disease, most children (67 of 102) had high-density infection (>2500 parasites per 200 white cells) with only mild symptoms before severe malaria, after severe malaria, or both. The incidence of severe malaria decreased considerably after infancy, whereas the incidence of high-density infection was similar among all age groups. Infections before and after episodes of severe malaria were associated with similar parasite densities. Nonuse of bed nets, placental malaria at the time of a woman’s second or subsequent delivery, high-transmission season, and absence of the sickle cell trait increased severe-malaria risk and parasite density during infections.
Resistance to severe malaria was not acquired after one or two mild infections. Although the parasite burden was higher on average during episodes of severe malaria, a high parasite burden was often insufficient to cause severe malaria even in children who later were susceptible. The diverging rates of severe disease and high-density infection after infancy, as well as the similar parasite burdens before and after severe malaria, indicate that naturally acquired resistance to severe malaria is not explained by improved control of parasite density. (Funded by the National Institute of Allergy and Infectious Diseases and others.)
Malaria and schistosomiasis often overlap in tropical and subtropical countries and impose tremendous disease burdens; however, the extent to which schistosomiasis modifies the risk of febrile malaria remains unclear.
We evaluated the effect of baseline S. haematobium mono-infection, baseline P. falciparum mono-infection, and co-infection with both parasites on the risk of febrile malaria in a prospective cohort study of 616 children and adults living in Kalifabougou, Mali. Individuals with S. haematobium were treated with praziquantel within 6 weeks of enrollment. Malaria episodes were detected by weekly physical examination and self-referral for 7 months. The primary outcome was time to first or only malaria episode defined as fever (≥37.5°C) and parasitemia (≥2500 asexual parasites/µl). Secondary definitions of malaria using different parasite densities were also explored.
After adjusting for age, anemia status, sickle cell trait, distance from home to river, residence within a cluster of high S. haematobium transmission, and housing type, baseline P. falciparum mono-infection (n = 254) and co-infection (n = 39) were significantly associated with protection from febrile malaria by Cox regression (hazard ratios 0.71 and 0.44; P = 0.01 and 0.02; reference group: uninfected at baseline). Baseline S. haematobium mono-infection (n = 23) did not associate with malaria protection in the adjusted analysis, but this may be due to lack of statistical power. Anemia significantly interacted with co-infection (P = 0.009), and the malaria-protective effect of co-infection was strongest in non-anemic individuals. Co-infection was an independent negative predictor of lower parasite density at the first febrile malaria episode.
Co-infection with S. haematobium and P. falciparum is significantly associated with reduced risk of febrile malaria in long-term asymptomatic carriers of P. falciparum. Future studies are needed to determine whether co-infection induces immunomodulatory mechanisms that protect against febrile malaria or whether genetic, behavioral, or environmental factors not accounted for here explain these findings.
The parasitic diseases malaria and schistosomiasis are tremendous public health burdens, each affecting over 200 million people worldwide with substantial geographic overlap in sub-Saharan Africa. Understanding how schistosomiasis influences the human immune response to Plasmodium, the agent of malaria, can be important for developing effective malaria vaccines. Past studies have tried to determine if infection with Schistosoma haematobium, which causes urinary schistosomiasis, affects the number of febrile attacks from malaria caused by Plasmodium falciparum in communities where the diseases overlap, but the findings have been inconsistent. Here, we examined 616 healthy people from a village in Mali for symptomless infections with S. haematobium and treated those with infections. We then followed them over a single malaria-transmission season of 7 months during which we diagnosed and treated all febrile malaria attacks. After the season, we examined archived blood collected at enrollment to look for occult P. falciparum infection. The study revealed that people who were infected with both parasites at the beginning of the season were better protected from the malaria attacks than those who were uninfected or infected with either parasite alone. Further studies are needed to confirm these findings and to determine the biological basis for this phenomenon.
Evaluating covariate effects on gap times between successive recurrent events is of interest in many medical and public health studies. While most existing methods for recurrent gap time analysis focus on modeling the hazard function of gap times, a direct interpretation of the covariate effects on the gap times is not available through these methods. In this article, we consider quantile regression that can provide direct assessment of covariate effects on the quantiles of the gap time distribution. Following the spirit of the weighted risk-set method by Luo and Huang (2011, Statistics in Medicine
30, 301–311), we extend the martingale-based estimating equation method considered by Peng and Huang (2008, Journal of the American Statistical Association
103, 637–649) for univariate survival data to analyze recurrent gap time data. The proposed estimation procedure can be easily implemented in existing software for univariate censored quantile regression. Uniform consistency and weak convergence of the proposed estimators are established. Monte Carlo studies demonstrate the effectiveness of the proposed method. An application to data from the Danish Psychiatric Central Register is presented to illustrate the methods developed in this article.
Clustered survival data; Data perturbation; Gap times; Quantile regression; Recurrent events; Within-cluster resampling
Chronic Granulomatous Disease (CGD) is an inherited defect in superoxide production leading to life-threatening infections, granulomas, and, possibly, abnormal immunoglobulin concentrations. We investigated whether factors controlling antibody production, such as B-cell activating factor (BAFF), were altered in CGD. CGD subjects had significantly increased mean (2.3-fold, p<0.0001) plasma concentrations of BAFF compared to healthy donors. Patients on IFN-γ treatment had significantly higher BAFF concentrations compared with CGD patients not taking IFN-γ (1.6-fold, p<0.005). Leukocytes from CGD subjects produced normal amounts of BAFF in response to IFN-γ or G-CSF in vitro. Expression of BAFF-R and TACI were significantly reduced on CGD B cells. Elevated BAFF in CGD correlated with CRP (R=0.44), ESR (R=0.49), and IgM (R=0.47) and increased rapidly in healthy subjects following intravenous endotoxin administration. These findings suggest that elevated BAFF in CGD subjects and healthy donors is a consequence of acute and chronic inflammation.
Chronic Granulomatous Disease; B-cell activating factor; inflammation
Influenza-like illnesses (ILI) are estimated to cause millions of deaths annually. Despite this disease burden, the etiologic causes of ILI are poorly described for many geographical regions.
Beginning in April 2010, we conducted an observational cohort study at five hospitals in Mexico City, enrolling subjects who met the criteria for ILI. Evaluations were conducted at enrollment and on day 28, with the collection of clinical data and a nasopharyngeal swab (or nasal aspirate in children). Swabs were tested by multiplex PCR for 15 viral pathogens and real-time PCR for influenza.
During the first year, 1065 subjects were enrolled in this study, 55% of whom were hospitalized; 24% of all subjects were children. One or more pathogens were detected by PCR in 64% of subjects, most commonly rhinovirus (25% of all isolates) and influenza (24% of isolates). Six percent of subjects died, and of those, 54% had no pathogen identified. Rhinovirus was the most common pathogen among those who died, although it did not have the highest case fatality rate.
Multiple respiratory viruses beyond influenza are associated with significant morbidity and mortality among adults and children in Mexico City. Detection of these agents could be useful for the adjustment of antibiotic treatment in severe cases.
Respiratory viral pathogens; Epidemiology; Hospital burden of disease; Influenza; Respiratory syncytial virus; Rhinovirus; Coronavirus
To investigate the risk of Plasmodium falciparum infection in a cohort of naturally exposed individuals, we assessed the time to first P. falciparum infection by polymerase chain reaction analysis of dried blood spots and found no age-related differences in infection risk.
Background In experimental models of human and mouse malaria, sterilizing liver stage immunity that blocks progression of Plasmodium infection to the symptomatic blood stage can be readily demonstrated. However, it remains unclear whether individuals in malaria-endemic areas acquire such immunity.
Methods In Mali, 251 healthy children and adults aged 4–25 years who were free of blood-stage Plasmodium infection by polymerase chain reaction (PCR) were enrolled in a longitudinal study just prior to an intense 6-month malaria season. Subsequent clinical malaria episodes were detected by weekly active surveillance and self-referral. Asymptomatic P. falciparum infections were detected by blood-smear microscopy and PCR analysis of dried blood spots that had been collected every 2 weeks for 7 months.
Results As expected, the risk of clinical malaria decreased with increasing age (log-rank test, P = .0038). However, analysis of PCR data showed no age-related differences in P. falciparum infection risk (log-rank test, P = .37).
Conclusions Despite years of exposure to intense P. falciparum transmission, there is no evidence of acquired, sterile immunity to P. falciparum infection in this population, even as clinical immunity to blood-stage malaria is clearly acquired. Understanding why repeated P. falciparum infections do not induce sterile protection may lead to insights for developing vaccines that target the liver stage in malaria-endemic populations.
preerythrocytic immunity; endemic population; Plasmodium falciparum; infection risk; malaria
Recent studies have shown high usage of the IGHV1-69 germline immunoglobulin gene for influenza hemagglutinin stem-directed broadly-neutralizing antibodies (HV1-69-sBnAbs). Here we show that a major structural solution for these HV1-69-sBnAbs is achieved through a critical triad comprising two CDR-H2 loop anchor residues (a hydrophobic residue at position 53 (Ile or Met) and Phe54), and CDR-H3-Tyr at positions 98±1; together with distinctive V-segment CDR amino acid substitutions that occur in positions sparse in AID/polymerase-η recognition motifs. A semi-synthetic IGHV1-69 phage-display library screen designed to investigate AID/polη restrictions resulted in the isolation of HV1-69-sBnAbs that featured a distinctive Ile52Ser mutation in the CDR-H2 loop, a universal CDR-H3 Tyr at position 98 or 99, and required as little as two additional substitutions for heterosubtypic neutralizing activity. The functional importance of the Ile52Ser mutation was confirmed by mutagenesis and by BCR studies. Structural modeling suggests that substitution of a small amino acid at position 52 (or 52a) facilitates the insertion of CDR-H2 Phe54 and CDR-H3-Tyr into adjacent pockets on the stem. These results support the concept that activation and expansion of a defined subset of IGHV1-69-encoded B cells to produce potent HV1-69-sBnAbs does not necessarily require a heavily diversified V-segment acquired through recycling/reentry into the germinal center; rather, the incorporation of distinctive amino acid substitutions by Phase 2 long-patch error-prone repair of AID-induced mutations or by random non-AID SHM events may be sufficient. We propose that these routes of B cell maturation should be further investigated and exploited as a pathway for HV1-69-sBnAb elicitation by vaccination.
The quest for universal influenza vaccine has gained wide interest with the discovery of human neutralizing antibodies that are able to variably cross neutralize and protect against different influenza strains, subtypes, groups and lineages. These antibodies, which bind to a highly conserved epitope in the hemagglutinin stem, are often encoded by rearranged IGHV1-69 germline genes that alone make contact with HA and prevent virus entry and emergence of escape mutants. Our study was undertaken to gain an understanding of what structural requirements enable a rearranged IGHV1-69 Ab to become a potent cross-neutralizing antibody. We found that in addition to a critical amino acid triad consisting of a pair of anchor residues in CDR-H2 and a properly positioned CDR-H3 Tyr, distinctive V-segment substitutions that arise in positions that are distinct from phase I AID somatic hypermutation (SHM) hotspot motifs are often required. As few as two V-segment SHM can fulfill this role which appears to facilitate the optimal binding of CDR-H2 Phe54 and CHR-H3-Tyr into adjacent hydrophobic pockets in the HA stem. These studies provide new information on the SHM requirements for IGHV1-69-encoded B cells to produce HV1-69-sBnAbs and suggest that there may exist alternative routes to their elicitation by vaccination.
Recurrent event data are commonly encountered in longitudinal follow-up studies related to biomedical science, econometrics, reliability, and demography. In many studies, recurrent events serve as important measurements for evaluating disease progression, health deterioration, or insurance risk. When analyzing recurrent event data, an independent censoring condition is typically required for the construction of statistical methods. In some situations, however, the terminating time for observing recurrent events could be correlated with the recurrent event process, thus violating the assumption of independent censoring. In this article, we consider joint modeling of a recurrent event process and a failure time in which a common subject-specific latent variable is used to model the association between the intensity of the recurrent event process and the hazard of the failure time. The proposed joint model is flexible in that no parametric assumptions on the distributions of censoring times and latent variables are made, and under the model, informative censoring is allowed for observing both the recurrent events and failure times. We propose a “borrow-strength estimation procedure” by first estimating the value of the latent variable from recurrent event data, then using the estimated value in the failure time model. Some interesting implications and trajectories of the proposed model are presented. Properties of the regression parameter estimates and the estimated baseline cumulative hazard functions are also studied.
Borrow-strength method; Frailty; Informative censoring; Joint model; Nonstationary Poisson process
Recurrent event data are largely characterized by the rate function but smoothing techniques for estimating the rate function have never been rigorously developed or studied in statistical literature. This paper considers the moment and least squares methods for estimating the rate function from recurrent event data. With an independent censoring assumption on the recurrent event process, we study statistical properties of the proposed estimators and propose bootstrap procedures for the bandwidth selection and for the approximation of confidence intervals in the estimation of the occurrence rate function. It is identified that the moment method without resmoothing via a smaller bandwidth will produce a curve with nicks occurring at the censoring times, whereas there is no such problem with the least squares method. Furthermore, the asymptotic variance of the least squares estimator is shown to be smaller under regularity conditions. However, in the implementation of the bootstrap procedures, the moment method is computationally more efficient than the least squares method because the former approach uses condensed bootstrap data. The performance of the proposed procedures is studied through Monte Carlo simulations and an epidemiological example on intravenous drug users.
bootstrap; independent censoring; intensity function; kernel estimator; Poisson process; rate function; recurrent events
The Canadian Study of Health and Aging (CSHA) employed a prevalent cohort design to study survival after onset of dementia, where patients with dementia were sampled and the onset time of dementia was determined retrospectively. The prevalent cohort sampling scheme favors individuals who survive longer. Thus, the observed survival times are subject to length bias. In recent years, there has been a rising interest in developing estimation procedures for prevalent cohort survival data that not only account for length bias but also actually exploit the incidence distribution of the disease to improve efficiency. This article considers semiparametric estimation of the Cox model for the time from dementia onset to death under a stationarity assumption with respect to the disease incidence. Under the stationarity condition, the semiparametric maximum likelihood estimation is expected to be fully efficient yet difficult to perform for statistical practitioners, as the likelihood depends on the baseline hazard function in a complicated way. Moreover, the asymptotic properties of the semiparametric maximum likelihood estimator are not well-studied. Motivated by the composite likelihood method (Besag 1974), we develop a composite partial likelihood method that retains the simplicity of the popular partial likelihood estimator and can be easily performed using standard statistical software. When applied to the CSHA data, the proposed method estimates a significant difference in survival between the vascular dementia group and the possible Alzheimer’s disease group, while the partial likelihood method for left-truncated and right-censored data yields a greater standard error and a 95% confidence interval covering 0, thus highlighting the practical value of employing a more efficient methodology. To check the assumption of stable disease for the CSHA data, we also present new graphical and numerical tests in the article. The R code used to obtain the maximum composite partial likelihood estimator for the CSHA data is available in the online Supplementary Material, posted on the journal web site.
Backward and forward recurrence time; Cross-sectional sampling; Random truncation; Renewal processes
In this paper, we study panel count data with informative observation times. We assume nonparametric and semiparametric proportional rate models for the underlying event process, where the form of the baseline rate function is left unspecified and a subject-specific frailty variable inflates or deflates the rate function multiplicatively. The proposed models allow the event processes and observation times to be correlated through their connections with the unobserved frailty; moreover, the distributions of both the frailty variable and observation times are considered as nuisance parameters. The baseline rate function and the regression parameters are estimated by maximising a conditional likelihood function of observed event counts and solving estimation equations. Large-sample properties of the proposed estimators are studied. Numerical studies demonstrate that the proposed estimation procedures perform well for moderate sample sizes. An application to a bladder tumour study is presented.
Dependent censoring; Frailty; Poisson process; Rate function; Serial events
The pretest–posttest study design is commonly used in medical and social science research to assess the effect of a treatment or an intervention. Recently, interest has been rising in developing inference procedures that improve efficiency while relaxing assumptions used in the pretest–posttest data analysis, especially when the posttest measurement might be missing. In this article we propose a semiparametric estimation procedure based on empirical likelihood (EL) that incorporates the common baseline covariate information to improve efficiency. The proposed method also yields an asymptotically unbiased estimate of the response distribution. Thus functions of the response distribution, such as the median, can be estimated straightforwardly, and the EL method can provide a more appealing estimate of the treatment effect for skewed data. We show that, compared with existing methods, the proposed EL estimator has appealing theoretical properties, especially when the working model for the underlying relationship between the pretest and posttest measurements is misspecified. A series of simulation studies demonstrates that the EL-based estimator outperforms its competitors when the working model is misspecified and the data are missing at random. We illustrate the methods by analyzing data from an AIDS clinical trial (ACTG 175).
Auxiliary information; Biased sampling; Causal inference; Observational study; Survey sampling
Severe malaria, including cerebral malaria (CM) and placental malaria (PM), have been recognized to have many of the features of uncontrolled inflammation. We recently showed that in mice genetic susceptibility to the lethal inflammatory autoimmune disease, systemic lupus erythematosus (SLE), conferred resistance to CM. Protection appeared to be mediated by immune mechanisms that allowed SLE-prone mice, prior to the onset of overt SLE symptoms, to better control their inflammatory response to Plasmodium infection. Here we extend these findings to ask does SLE susceptibility have 1) a cost to reproductive fitness and/or 2) an effect on PM in mice? The rates of conception for WT and SLE susceptible (SLEs) mice were similar as were the number and viability of fetuses in pregnant WT and SLEs mice indicating that SLE susceptibility does not have a reproductive cost. We found that Plasmodium chabaudi AS (Pc) infection disrupted early stages of pregnancy before the placenta was completely formed resulting in massive decidual necrosis 8 days after conception. Pc-infected pregnant SLEs mice had significantly more fetuses (∼1.8 fold) but SLE did not significantly affect fetal viability in infected animals. This was despite the fact that Pc-infected pregnant SLEs mice had more severe symptoms of malaria as compared to Pc-infected pregnant WT mice. Thus, although SLE susceptibility was not protective in PM in mice it also did not have a negative impact on reproductive fitness.
This paper considers semiparametric estimation of the Cox proportional hazards model for right-censored and length-biased data arising from prevalent sampling. To exploit the special structure of length-biased sampling, we propose a maximum pseudo-profile likelihood estimator, which can handle time-dependent covariates and is consistent under covariate-dependent censoring. Simulation studies show that the proposed estimator is more efficient than its competitors. A data analysis illustrates the methods and theory.
Approximate likelihood; Cross-sectional sampling; Product-limit estimator; Random truncation; Screening trials
H5N1 avian influenza represents an episodic zoonotic disease with potential to cause a pandemic, and resistance is of considerable concern. We sought to generate high titer H5N1 antibodies in healthy volunteers for the purpose of developing hyperimmune IVIG.
We conducted a dose escalating unblinded clinical trial involving 75 subjects between the ages of 18 and 59 years. Three cohorts of twenty-five subjects were enrolled sequentially receiving 90, 120, or 180 mcg of H5N1 A/Vietnam/1203/04 vaccine for four doses 28 days apart.
No significant dose related increases in the geometric mean titers (GMTs) of serum HAI were observed when comparing 90, 120, and 180 mcg. When analyzed together to determine the effect of additional vaccinations, HAI GMT after first, second, third and fourth vaccinations was 1:15.7, 1:22.2, 1:36.0, and 1:32.0, respectively (first vs baseline, P<0.0001; second vs first, P=0.02; third vs second, P<0.0001). The MN GMT after first, second, third, and fourth vaccinations was 1:17.5, 1:33.1, 1:55.7, and 1:68.4 respectively (all P<0.001).
Our study suggests that a third dose and fourth dose of the H5N1 A/Vietnam/1203/04 vaccine may result in a higher HA and MN GMT. There was no benefit to increasing the dose of the vaccine.
avian influenza; vaccination; passive immunotherapy; IVIG
Recurrent events are the natural outcome in many medical and epidemiology studies. To assess covariate effects on the gaps between consecutive recurrent events, the Cox proportional hazards model is frequently employed in data analysis. The validity of statistical inference, however, depends on the appropriateness of the Cox model. In this paper, we propose a class of graphical techniques and formal tests for checking the Cox model with recurrent gap time data. The building block of our model checking method is an averaged martingale-like process, based on which a class of multiparameter stochastic processes is proposed. This maneuver is very general and can be used to assess different aspects of model fit. Numerical simulations are conducted to examine finite-sample performance, and the proposed model checking techniques are illustrated with data from the Danish Psychiatric Central Register.
Correlated failure times; Induced-dependent censoring; Kaplan–Meier estimator; Renewal processes
Sera from H5N1 vaccinees and IVIG were analyzed for crossreactive Abs against HA of influenza A viruses. Data quantitatively shows the presence of two populations of heterosubtypic neutralizing Abs against either Group 1 only or Group 1 + 2 viruses.
(See the editorial commentary by Donis and Cox, on pages 1010–1012.)
Background. Lack of life-long immunity against influenza viruses represents a major global health care problem with profound medical and economic consequences. A greater understanding of the broad-spectrum “heterosubtypic” neutralizing human antibody (BnAb) response to influenza should bring us closer toward a universal influenza vaccine.
Methods. Serum samples obtained from 77 volunteers in an H5N1 vaccine study were analyzed for cross-reactive antibodies (Abs) against both subtype hemagglutinins (HAs) and a highly conserved pocket on the HA stem of Group 1 viruses. Cross-reactive Abs in commercial intravenous immunoglobulin were affinity purified using H5-coupled beads followed by step-wise monoclonal antibody competition or acid elution. Enzyme-linked immunosorbent assays were used to quantify cross-binding, and neutralization activity was determined with HA-pseudotyped viruses.
Results. Prevaccination serum samples have detectable levels of heterosubtypic HA binding activity to both Group 1 and 2 influenza A viruses, including subtypes H5 and H7, respectively, to which study subjects had not been vaccinated. Two different populations of Broadly neutralizing Abs (BnAbs) were purified from intravenous immunoglobulin by H5 beads: ∼0.01% of total immunoglobulin G can bind to HAs from both Group 1 and 2 and neutralize H1N1 and H5N1 viruses; ∼0.001% is F10-like Abs directed against the HA stem pocket on Group 1 viruses.
Conclusion. These data—to our knowledge, for the first time—quantitatively show the presence, albeit at low levels, of two populations of heterosubtypic BnAbs against influenza A in human serum. These observations warrant further investigation to determine their origin, host polymorphism(s) that may affect their expression levels and how to boost these BnAb responses by vaccination to reach sustainable protective levels.