MCPModPack-package: Design and analysis of dose-finding trials

Description Details Author(s) References Examples

Description

The MCPModPack package facilitates the design and analysis of dose-finding clinical trials with normally distributed, binary and count endpoints using the MCPMod methodology.

Details

Package: MCPModPack
Type: Package
Version: 0.3
Date: 2020-08-01
License: GPL-3

Key functions included in the package:

The package comes with three example data sets:

Author(s)

Alex Dmitrienko <admitrienko@medianainc.com>

References

Bornkamp, B., Bezlyak, V., Bretz, F. (2015). Implementing the MCP-Mod procedure for dose-response testing and estimation. Modern Approaches to Clinical Trials Using SAS. Menon, S., Zink, R. (editors). SAS Press: Cary, NC.

Bretz, F., Pinheiro, J.C., Branson, M. (2005). Combining multiple comparisons and modeling techniques in dose response studies. Biometrics. 61, 738-748.

Bretz, F., Tamhane, A.C., Pinheiro, J. (2009). Multiple testing in dose response problems. Multiple Testing Problems in Pharmaceutical Statistics. Dmitrienko, A., Tamhane, A.C., Bretz, F. (editors). New York: Chapman and Hall/CRC Press.

Nandakumar, S., Dmitrienko, A., Lipkovich, I. (2017). Dose-finding methods. Analysis of Clinical Trials Using SAS: A Practical Guide (Second Edition). Dmitrienko, A., Koch, G.G. (editors). SAS Press: Cary, NC.

Pinheiro, J. C., Bornkamp, B., Bretz, F. (2006). Design and analysis of dose finding studies combining multiple comparisons and modeling procedures. Journal of Biopharmaceutical Statistics. 16, 639-656.

Pinheiro J., Bornkamp B., Glimm E., Bretz F. (2013). Model-based dose finding under model uncertainty using general parametric models. Statistics in Medicine. 33, 1646-1661.

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
    
# MCPMod-based analysis of a dose-finding trial with a normally distributed endpoint

# Select the candidate dose-response models and initial values 
# of the non-linear model parameters (linear, quadratic, exponential, 
# emax, logistic and sigemax)
models = list(linear = NA, 
              quadratic = -0.5, 
              exponential = 0.3, 
              emax = 0.3, 
              logistic = c(0.5, 0.1), 
              sigemax = c(0.5, 5))

# One-sided Type I error rate
alpha = 0.025

# Direction of the dose-response relationship
direction = "increasing"

# Model selection criterion
model_selection = "AIC"

# The treatment effect for identifying the target dose 
# (this effect is defined relative to the placebo effect)
Delta = 0.5

# Perform an MCPMod-based analysis of the trial's data
# The data set normal is included in the package
results = MCPModAnalysis(endpoint_type = "Normal", 
                     models = models, 
                     dose = normal$dose, 
                     resp = normal$resp, 
                     alpha = alpha, 
                     direction = direction, 
                     model_selection = model_selection, 
                     Delta = Delta)

# Simple summary of the MCPMod analysis results
results

# Detailed summary of the MCPMod analysis results (remove tempfile)
AnalysisReport(results, 
  "MCPMod analysis summary (Normally distributed endpoint)", 
  tempfile("MCPMod analysis summary (Normally distributed endpoint).docx", fileext=".docx")) 
  

Example output

Loading required package: mvtnorm
***************************************

Descriptive statistics

***************************************

 Dose  n  Mean         95% CI    SE
    0 20 0.345   (0.02, 0.67) 0.166
 0.05 20 0.457 (0.132, 0.782) 0.166
  0.2 20  0.81 (0.485, 1.135) 0.166
  0.6 20 0.934  (0.609, 1.26) 0.166
    1 20 0.949 (0.624, 1.274) 0.166

***************************************

Hypothesis testing and model selection

***************************************

Model-specific dose-response contrasts

 Dose Linear Quadratic Exponential   Emax Logistic SigEmax
    0 -0.437    -0.517      -0.299 -0.601   -0.374  -0.362
 0.05 -0.378    -0.408      -0.291 -0.384   -0.369  -0.362
  0.2 -0.201    -0.112      -0.258  0.006   -0.330  -0.351
  0.6  0.271     0.428      -0.023  0.411    0.397   0.400
    1  0.743     0.608       0.871  0.567    0.676   0.674

 Contrast correlation matrix

      Models Linear Quadratic Exponential  Emax Logistic SigEmax
      Linear  1.000     0.971       0.934 0.940    0.979   0.975
   Quadratic  0.971     1.000       0.823 0.988    0.962   0.956
 Exponential  0.934     0.823       1.000 0.775    0.885   0.882
        Emax  0.940     0.988       0.775 1.000    0.911   0.901
    Logistic  0.979     0.962       0.885 0.911    1.000   1.000
     SigEmax  0.975     0.956       0.882 0.901    1.000   1.000

 Model-specific contrast tests

       Model Test statistic Adjusted p-value Significant contrast
      Linear          2.972           0.0040                  Yes
   Quadratic          3.279           0.0014                  Yes
 Exponential          2.260           0.0248                  Yes
        Emax          3.422           0.0011                  Yes
    Logistic          2.808           0.0061                  Yes
     SigEmax          2.758           0.0070                  Yes

Adjusted critical value: 2.254

***************************************

Dose-response modeling

***************************************

Dose-response model: Linear 
Parameter estimates
   e0 delta 
0.492 0.559 

***************************************

Dose-response model: Quadratic 
Parameter estimates
    e0 delta1 delta2 
 0.390  1.768 -1.232 

***************************************

Dose-response model: Exponential 
Parameter estimates
   e0    e1 delta 
0.524 0.378 1.148 

***************************************

Dose-response model: Emax 
Parameter estimates
   e0  eMax  ed50 
0.322 0.746 0.142 

***************************************

Dose-response model: Logistic 
Parameter estimates
   e0  eMax  ed50 delta 
0.169 0.773 0.087 0.071 

***************************************

Dose-response model: SigEmax 
Parameter estimates
   e0  eMax  ed50     h 
0.345 0.612 0.110 1.912 

***************************************

Dose selection

***************************************

Model selection criteria

       Model Significant contrast    AIC Test statistic Model weight
      Linear                  Yes 220.50          2.972        0.157
   Quadratic                  Yes 219.72          3.279        0.232
 Exponential                  Yes 223.60          2.260        0.033
        Emax                  Yes 219.14          3.422        0.310
    Logistic                  Yes 220.83          2.808        0.133
     SigEmax                  Yes 220.82          2.758        0.134

Selected model (based on the smallest AIC): Emax

***************************************

Model-specific estimated target doses (based on Delta = 0.5)

       Model Target dose
      Linear       0.895
   Quadratic       0.387
 Exponential       0.967
        Emax       0.289
    Logistic       0.226
     SigEmax       0.239

Selected target dose (based on the smallest AIC): 0.289 

MCPModPack documentation built on Jan. 13, 2021, 6:37 p.m.