In clinical trials, longitudinal data are commonly analyzed and compared between groups using a single summary statistic such as area under the outcome versus time curve (AUC). However, incomplete data, arising from censoring due to a limit of detection or missing data, can bias these analyses. In this article, we present a statistical test based on splines-based mixed-model accounting for both the censoring and missingness mechanisms in the AUC estimation. Inferential properties of the proposed method were evaluated and compared to ad hoc approaches and to a non-parametric method through a simulation study based on two-armed trial where trajectories and the proportion of missing data were varied. Simulation results highlight that our approach has significant advantages over the other methods. A real working example from two HIV therapeutic vaccine trials is presented to illustrate the applicability of our approach.
ScheffJDAlmonRRDuBoisDC, et al.
Assessment of pharmacologic area under the curve when baselines are variable.Pharmaceut Res2011;
28: 1081–1089.
2.
HeldensJWeststrateMVan den HovenR.Area under the curve calculations as a tool to compare the efficacy of equine influenza vaccines – a retrospective analysis of three independent field trials. J Immunol Methods2002;
264: 11–17.
3.
LydickEEpsteinRHimmelbergerD, et al.
Area under the curve: a metric for patient subjective responses in episodic diseases.Qual Life Res1995;
4: 41–45.
4.
NeoptolemosJPStockenDDFriessH, et al.
A randomized trial of chemoradiotherapy and chemotherapy after resection of pancreatic cancer.N Engl J Med2004;
350: 1200–1210.
5.
SchleyerEKühnSRührsH, et al.
Oral idarubicin pharmacokinetics – correlation of trough level with idarubicin area under curve.Leukemia1996;
10: 707–712.
6.
DuhMSLefebvrePFastenauJ, et al.
Assessing the clinical benefits of erythropoietic agents using area under the hemoglobin change curve.The Oncologist2005;
10: 438–448.
7.
HothornLA.Statistical analysis of in vivo anticancer experiments: tumor growth inhibition. Drug Inform J2006;
40: 229–238.
8.
WuJHoughtonPJ.Interval approach to assessing antitumor activity for tumor xenograft studies.Pharmaceut Stat2010;
9: 46–54.
9.
QianWParmarMSambrookR, et al.
Analysis of messy longitudinal data from a randomized clinical trial. Stat Med2000;
19: 2657–2674.
10.
ColeSRNapravnikSMugaveroMJ, et al.
Copy-years viremia as a measure of cumulative human immunodeficiency virus viral burden.Am J Epidemiol2010;
171: 198–205.
11.
RamosELMitchamJLKollerTD, et al.
Efficacy and safety of treatment with an anti-m2e monoclonal antibody in experimental human influenza.J Infect Dis2015;
211: 1038–1044.
12.
CalfeeDPPengAWCassLM, et al.
Safety and efficacy of intravenous zanamivir in preventing experimental human influenza a virus infection. Antimicrob Agents Chemother1999;
43: 1616–1620.
13.
AllisonDBPaultreFMaggioC, et al.
The use of areas under curves in diabetes research.Diabetes Care1995;
18: 245–250.
14.
VenterCSlabberMVorsterH.Labelling of foods for glycaemic index-advantages and problems. South African J Clin Nutr2003;
16: 118--126.
15.
PotteigerJJacobsenDDonnellyJ.A comparison of methods for analyzing glucose and insulin areas under the curve following nine months of exercise in overweight adults.Int J Obesity2002;
26: 87–89.
16.
WildingGEChandrasekharRHutsonAD.A new linear model-based approach for inferences about the mean area under the curve.Stat Med2012;
31: 3563–3578.
17.
BellMLKingMTFaircloughDL.Bias in area under the curve for longitudinal clinical trials with missing patient reported outcome data: summary measures versus summary statistics. SAGE Open2014;
4: 2158244014534858.
18.
ChandrasekharRShiYHutsonAD, et al.
Likelihood-based inferences about the mean area under a longitudinal curve in the presence of observations subject to limits of detection.Pharmaceut Stat2015;
14: 252–261.
19.
LittleRAnH.Robust likelihood-based analysis of multivariate data with missing values. Stat Sin2004;
14: 949–968.
20.
Jacqmin-GaddaHThiébautRChêneG, et al.
Analysis of left-censored longitudinal data with application to viral load in HIV infection.Biostatistics2000;
1: 355–368.
VaidaFLiuL.Fast implementation for normal mixed effects models with censored response.J Comput Graph Stat2009;
18: 797–817.
23.
SchistermanERotnitzkyA.Estimation of the mean of a k-sample u-statistic with missing outcomes and auxiliaries. Biometrika2001;
88: 713–725.
24.
SpritzlerJDeGruttolaVGPeiL.Two-sample tests of area-under-the-curve in the presence of missing data. Int J Biostat2008;
4.
25.
VardiYYingZZhangCH.Two-sample tests for growth curves under dependent right censoring. Biometrika2001;
88: 949–960.
26.
GarnerSARennieSAnanworanichJ, et al.
Interrupting antiretroviral treatment in HIV cure research: scientific and ethical considerations. J Virus Eradicat2017;
3: 82.
27.
LiJZEtemadBAhmedH, et al.
The size of the expressed HIV reservoir predicts timing of viral rebound after treatment interruption. AIDS2016;
30: 343.
28.
HendersonGEPeayHLKroonE, et al.
Ethics of treatment interruption trials in HIV cure research: addressing the conundrum of risk/benefit assessment. J Med Ethics2018;
44: 270–276.
29.
SnellerMCJustementJSGittensKR, et al.
A randomized controlled safety/efficacy trial of therapeutic vaccination in HIV-infected individuals who initiated antiretroviral therapy early in infection. Sci Transl Med2017;
9: eaan8848.
30.
SuedOAmbrosioniJNicolásD, et al.
Structured treatment interruptions and low doses of il-2 in patients with primary HIV infection. inflammatory, virological and immunological outcomes. PLoS One2015;
10: e0131651.
31.
LévyYThiébautRMontesM, et al.
Dendritic cell-based therapeutic vaccine elicits polyfunctional hiv-specific t-cell immunity associated with control of viral load.Eur J Immunol2014;
44: 2802–2810.
32.
PollardRBRockstrohJKPantaleoG, et al.
Safety and efficacy of the peptide-based therapeutic vaccine for hiv-1, vacc-4×: a phase 2 randomised, double-blind, placebo-controlled trial. Lancet Infect Dis2014;
14: 291–300.
33.
BarKJSnellerMCHarrisonLJ, et al.
Effect of HIV antibody vrc01 on viral rebound after treatment interruption. N Engl J Med2016;
375: 2037–2050.
34.
FagardCLe BrazMGünthardH, et al.
A controlled trial of granulocyte macrophage-colony stimulating factor during interruption of HAART. AIDS2003;
17: 1487–1492.
35.
PalichRGhosnJChaillonA, et al.
Viral rebound in semen after antiretroviral treatment interruption in an HIV therapeutic vaccine double-blind trial.AIDS2019;
33: 279–284.
36.
LévyYGahéry-SégardHDurierC, et al.
Immunological and virological efficacy of a therapeutic immunization combined with interleukin-2 in chronically hiv-1 infected patients. AIDS2005;
19: 279–286.
37.
BrundageTMVainoriusEChittickG, et al.
Brincidofovir decreases adenovirus viral burden, which is associated with improved mortality in pediatric allogeneic hematopoietic cell transplant recipients. Biol Blood Marrow Transplant2018;
24: S372.
38.
HillJAMayerBTXieH, et al.
Kinetics of double-stranded DNA viremia after allogeneic hematopoietic cell transplantation.Clin Infect Dis2018;
66: 368–375.
39.
ZeccaMWynnRDalleJH, et al.
Association between adenovirus viral load and mortality in pediatric allo-HCT recipients: the multinational advance study.Bone Marrow Transplant2019;
54: 1632–1642.
40.
KosulinKPichlerHLawitschkaA, et al.
Diagnostic parameters of adenoviremia in pediatric stem cell transplant recipients. Front Microbiol2019;
10: 414.
41.
SatterthwaiteFE.An approximate distribution of estimates of variance components. Biometr Bull1946;
2: 110–114.
42.
Hrong-Tai FaiACorneliusPL.Approximate f-tests of multiple degree of freedom hypotheses in generalized least squares analyses of unbalanced split-plot experiments. J Stat Comput Simulat1996;
54: 363–378.
43.
KenwardMGRogerJH.Small sample inference for fixed effects from restricted maximum likelihood. Biometrics1997;
53: 983–997.
44.
BailerAJ.Testing for the equality of area under the curves when using destructive measurement techniques.J Pharmacokinet Biopharmaceut1988;
16: 303–309.
VaidaFLiuL.lmec: linear mixed-effects models with censored responses, 2012. . R package version 1.0, https://CRAN.R-project.org/package=lmec (accessed 8 June 2021).
47.
HazraAGogtayN.Biostatistics series module 5: determining sample size. Ind J Dermatol2016;
61: 496.
48.
EilersPHMarxBD.Flexible smoothing with b-splines and penalties. Stat Sci1996;
11: 89–102.
49.
LauriniMPMouraM.Constrained smoothing b-splines for the term structure of interest rates. Insurance2010;
46: 339–350.
50.
PerelsonASRibeiroRM.Introduction to modeling viral infections and immunity.Immunol Rev2018;
285: 5.
