Looking at patients with less than 3 months of survival

Number of metastases

Most patients with less than 3 months already have more than 15

##    nummets  N
## 1:     >15 22
## 2:    6-15  1

Survival based on number of mets

## Loading required package: ggplot2

Number of affected organs

##    Anz..Befalener.Organe.0.weniger.gleich.2..1..mehr.als.2  N
## 1:                                                       0 51
## 2:                                                       1 19

Number of affected organs in patients with less than 3 months

##    Anz..Befalener.Organe.0.weniger.gleich.2..1..mehr.als.2  N
## 1:                                                       0  6
## 2:                                                       1 17

Survival plot for affected organs

Looking at N0M1 vs N1-3M1

##    N.Status  N
## 1:       1a  3
## 2:        3 31
## 3:        0 12
## 4:        1  9
## 5:       2c  3
## 6:       1b  8
## 7:        x  3
## 8:        2  1

only less than 3 months survival

##    N.Status  N
## 1:       1a  1
## 2:        3 15
## 3:        0  2
## 4:        1  2
## 5:       2c  1
## 6:       1b  2

Survival plot for N0 vs N1-3

##    N0  N
## 1:  1 21
## 2:  0  2

Looking at M1a,b vs M1c

##    X0.M1a.M1b..1.M1c  N
## 1:                 1 39
## 2:                 0 31

Only in 3 Months survival patients, most are M1c

##    X0.M1a.M1b..1.M1c  N
## 1:                 1 21
## 2:                 0  2

Survival curve for M1a,b M1c

Most patients have more than one met, so its hard to separate them with just lung or liver or brain

Time from Stage I till Stage IV

Descriptive statsistics for study and control group

median age

##    studygrp   V1
## 1:     TRUE 60.8
## 2:    FALSE 51.2

Gender

##                      studygrp
## Sex..1.female.2.male. FALSE TRUE
##                     1    23    7
##                     2    24   16

Mutation

##                            studygrp
## Mutationstatus.BRAF.NRAS.WT FALSE TRUE
##                           1    26    9
##                           2     6    1
##                           3     9    6
##                           5     1    4
##                           6     5    3

Breslow

##                studygrp
## Breslow.codiert FALSE TRUE
##               1     5    3
##               2     6    3
##               3     8    7
##               4    13    7
##               5    15    3

Treatment

##                   studygrp
## Group.Stat         FALSE TRUE
##                        4    8
##   Immuntherapie       11    4
##   IT + TT             17    1
##   Referenten           3    6
##   Targeted therapy    12    4
##                                     studygrp
## Group                                FALSE TRUE
##   Allovectin                             1    0
##   BMS CA209-067                          2    2
##   CMEK162X2102 (RAF265+MEKi)             3    0
##   CMEK162X2110 (BRAF/MEKi)               0    1
##   CMEK162X2201 (MEKi)                    1    0
##   COLUMBUS Studie (Vemu)                 0    1
##   CSOM230X2404                           0    1
##   DTIC                                   1    2
##   DTIC 2010                              1    3
##   DTIC 2011                              1    1
##   GSK MEK116513 (Combi V)/BRAFi          1    0
##   GSK MEK116513 (Combi V)/BRAFi+MEKi     3    0
##   Ipi 2008-11                            7    0
##   Ipi 2012                               3    1
##   Ipi 2013                               2    0
##   Ipi 2014                               1    1
##   METRIC GSK MEK 114267(MEKi)            3    0
##   Nexavar/DTIC                           0    1
##   Nexavar/DTIC 2010                      0    1
##   Nexavar/DTIC/VP                        0    1
##   NY-ESO01                               1    0
##   Pazopanib                              0    1
##   Philogen                               1    1
##   Temodal                                0    1
##   Temodal 2011                           0    1
##   Temodal Chur                           1    0
##   Vemu 2008-11                          10    1
##   Vemu 2012                              3    1
##   Vemu 2013                              0    1
##   Vemu 2014                              1    0

Number of mets

##        studygrp
## nummets FALSE TRUE
##    6-15     4    1
##    1-5     36    0
##    >15      7   22

Affected organs

##                                                        studygrp
## Anz..Befalener.Organe.0.weniger.gleich.2..1..mehr.als.2 FALSE TRUE
##                                                       0    45    6
##                                                       1     2   17

N stage

##         studygrp
## N.Status FALSE TRUE
##       0     10    2
##       1      7    2
##       1a     2    1
##       1b     6    2
##       2      1    0
##       2c     2    1
##       3     16   15
##       x      3    0

M stage

##          studygrp
## M.Stadium FALSE TRUE
##         1    14    0
##         2    15    2
##         3    18   21

Ulceration

##                                          studygrp
## Ulceration..0.no..1.yes..3..leer.unknown. FALSE TRUE
##                                         0     8    6
##                                         1     7    6
##                                         3    32   11

LDH

##             studygrp
## LDH.elevated FALSE TRUE
##            1     1   17
##            2    21    6
##            3    25    0

S100

##             studygrp
## S100elevated FALSE TRUE
##            0    17    2
##            1    11   20
##            3    19    1

survival curve study group and control group

logistic regression on all factors for short and long survivors in univariate analysis

looking at age with regression and Mann-Whitney

## 
##  Wilcoxon rank sum test with continuity correction
## 
## data:  Age.at.first.diagnosis by studygrp
## W = 366, p-value = 0.02956
## alternative hypothesis: true location shift is not equal to 0
##                     term    estimate  std.error statistic     p.value
## 1            (Intercept) -3.62448569 1.34092382 -2.702977 0.006872159
## 2 Age.at.first.diagnosis  0.05252568 0.02319687  2.264343 0.023553014
## Waiting for profiling to be done...
##                      Coef         OR     2.5 %    97.5 %     p.value
## 1:            (Intercept) 0.02666281 0.0015349 0.3108016 0.006872159
## 2: Age.at.first.diagnosis 1.05392962 1.0093997 1.1065374 0.023553014

looking at sex, regression and fisher’s exact

## 
##  Fisher's Exact Test for Count Data
## 
## data:  lea.sex
## p-value = 0.1995
## alternative hypothesis: true odds ratio is not equal to 1
## 95 percent confidence interval:
##  0.6858928 7.4466131
## sample estimates:
## odds ratio 
##   2.166248
##                             term   estimate std.error statistic
## 1 factor(Sex..1.female.2.male.)1 -1.1895841 0.4316655 -2.755801
## 2 factor(Sex..1.female.2.male.)2 -0.4054651 0.3227486 -1.256288
##       p.value
## 1 0.005854869
## 2 0.209011681
## Waiting for profiling to be done...
##                              Coef        OR     2.5 %   97.5 %     p.value
## 1: factor(Sex..1.female.2.male.)1 0.3043478 0.1207115 0.674077 0.005854869
## 2: factor(Sex..1.female.2.male.)2 0.6666667 0.3475319 1.244832 0.209011681

looking at breslow, regression and fisher’s exact

## 
##  Fisher's Exact Test for Count Data
## 
## data:  lea.bres
## p-value = 0.4454
## alternative hypothesis: two.sided
##                       term   estimate std.error  statistic    p.value
## 1 factor(Breslow.codiert)1 -0.5108256 0.7302967 -0.6994768 0.48425410
## 2 factor(Breslow.codiert)2 -0.6931472 0.7071067 -0.9802583 0.32695862
## 3 factor(Breslow.codiert)3 -0.1335314 0.5175492 -0.2580072 0.79640138
## 4 factor(Breslow.codiert)4 -0.6190392 0.4688072 -1.3204559 0.18668286
## 5 factor(Breslow.codiert)5 -1.6094379 0.6324555 -2.5447448 0.01093576
## Waiting for profiling to be done...
##                        Coef        OR      2.5 %    97.5 %    p.value
## 1: factor(Breslow.codiert)1 0.6000000 0.12305677 2.4453025 0.48425410
## 2: factor(Breslow.codiert)2 0.5000000 0.10551288 1.8951666 0.32695862
## 3: factor(Breslow.codiert)3 0.8750000 0.30678669 2.4378421 0.79640138
## 4: factor(Breslow.codiert)4 0.5384615 0.20234470 1.3145020 0.18668286
## 5: factor(Breslow.codiert)5 0.2000000 0.04632242 0.6060404 0.01093576

looking at ulceration, regression and fisher’s exact

## 
##  Fisher's Exact Test for Count Data
## 
## data:  lea.ulc
## p-value = 0.2485
## alternative hypothesis: two.sided
##                                                 term   estimate std.error
## 1 factor(Ulceration..0.no..1.yes..3..leer.unknown.)0 -0.2876821 0.5400617
## 2 factor(Ulceration..0.no..1.yes..3..leer.unknown.)1 -0.1541507 0.5563486
## 3 factor(Ulceration..0.no..1.yes..3..leer.unknown.)3 -1.0678406 0.3495123
##    statistic     p.value
## 1 -0.5326837 0.594252558
## 2 -0.2770757 0.781721993
## 3 -3.0552301 0.002248879
## Waiting for profiling to be done...
##                                                  Coef        OR     2.5 %
## 1: factor(Ulceration..0.no..1.yes..3..leer.unknown.)0 0.7500000 0.2469277
## 2: factor(Ulceration..0.no..1.yes..3..leer.unknown.)1 0.8571429 0.2759429
## 3: factor(Ulceration..0.no..1.yes..3..leer.unknown.)3 0.3437500 0.1654959
##       97.5 %     p.value
## 1: 2.1565676 0.594252558
## 2: 2.5803710 0.781721993
## 3: 0.6608769 0.002248879

looking at braf status, regression and fisher’s exact

## 
##  Fisher's Exact Test for Count Data
## 
## data:  lea.braf
## p-value = 0.1344
## alternative hypothesis: two.sided
##                                   term   estimate std.error  statistic
## 1 factor(Mutationstatus.BRAF.NRAS.WT)1 -1.0608720 0.3867459 -2.7430723
## 2 factor(Mutationstatus.BRAF.NRAS.WT)2 -1.7917595 1.0801233 -1.6588472
## 3 factor(Mutationstatus.BRAF.NRAS.WT)3 -0.4054651 0.5270463 -0.7693160
## 4 factor(Mutationstatus.BRAF.NRAS.WT)5  1.3862944 1.1180337  1.2399397
## 5 factor(Mutationstatus.BRAF.NRAS.WT)6 -0.5108256 0.7302967 -0.6994768
##       p.value
## 1 0.006086729
## 2 0.097146595
## 3 0.441705767
## 4 0.214997707
## 5 0.484254099
## Waiting for profiling to be done...
##                                    Coef        OR       2.5 %     97.5 %
## 1: factor(Mutationstatus.BRAF.NRAS.WT)1 0.3461538 0.153209706  0.7116237
## 2: factor(Mutationstatus.BRAF.NRAS.WT)2 0.1666667 0.008824972  0.9756398
## 3: factor(Mutationstatus.BRAF.NRAS.WT)3 0.6666667 0.223479458  1.8487217
## 4: factor(Mutationstatus.BRAF.NRAS.WT)5 4.0000000 0.591735449 78.2485115
## 5: factor(Mutationstatus.BRAF.NRAS.WT)6 0.6000000 0.123056766  2.4453025
##        p.value
## 1: 0.006086729
## 2: 0.097146595
## 3: 0.441705767
## 4: 0.214997707
## 5: 0.484254099

looking at LDH, regression and fisher’s exact

## 
##  Fisher's Exact Test for Count Data
## 
## data:  lea.ldh
## p-value = 1.177e-11
## alternative hypothesis: two.sided
##                    term   estimate   std.error    statistic     p.value
## 1 factor(LDH.elevated)1   2.833213    1.028992  2.753388453 0.005898187
## 2 factor(LDH.elevated)2  -1.252763    0.462910 -2.706277318 0.006804220
## 3 factor(LDH.elevated)3 -19.566069 2150.802594 -0.009097101 0.992741664
## Waiting for profiling to be done...
##                     Coef           OR         2.5 %       97.5 %
## 1: factor(LDH.elevated)1 1.700000e+01  3.493428e+00 3.063244e+02
## 2: factor(LDH.elevated)2 2.857143e-01  1.048588e-01 6.663129e-01
## 3: factor(LDH.elevated)3 3.181005e-09 2.071408e-157 2.207380e+41
##        p.value
## 1: 0.005898187
## 2: 0.006804220
## 3: 0.992741664

looking at s100, regression and fisher’s exact

## 
##  Fisher's Exact Test for Count Data
## 
## data:  lea.s100
## p-value = 1.733e-06
## alternative hypothesis: two.sided
##                    term  estimate std.error statistic     p.value
## 1 factor(S100elevated)0 -2.140066 0.7475450 -2.862792 0.004199256
## 2 factor(S100elevated)1  0.597837 0.3753786  1.592624 0.111244543
## 3 factor(S100elevated)3 -2.944439 1.0259255 -2.870032 0.004104303
## Waiting for profiling to be done...
##                     Coef         OR       2.5 %    97.5 %     p.value
## 1: factor(S100elevated)0 0.11764706 0.018651379 0.4102095 0.004199256
## 2: factor(S100elevated)1 1.81818182 0.887002998 3.9301129 0.111244543
## 3: factor(S100elevated)3 0.05263158 0.002929061 0.2534472 0.004104303

looking at number of organ mets, regression and fhisher’s exact

## 
##  Fisher's Exact Test for Count Data
## 
## data:  lea.org
## p-value = 1.744e-09
## alternative hypothesis: true odds ratio is not equal to 1
## 95 percent confidence interval:
##   10.1720 621.4708
## sample estimates:
## odds ratio 
##   57.29433
##                                                               term
## 1 factor(Anz..Befalener.Organe.0.weniger.gleich.2..1..mehr.als.2)0
## 2 factor(Anz..Befalener.Organe.0.weniger.gleich.2..1..mehr.als.2)1
##    estimate std.error statistic      p.value
## 1 -2.014903 0.4346088 -4.636130 3.549924e-06
## 2  2.140066 0.7475099  2.862927 4.197475e-03
## Waiting for profiling to be done...
##                                                                Coef
## 1: factor(Anz..Befalener.Organe.0.weniger.gleich.2..1..mehr.als.2)0
## 2: factor(Anz..Befalener.Organe.0.weniger.gleich.2..1..mehr.als.2)1
##           OR      2.5 %     97.5 %      p.value
## 1: 0.1333333 0.05098486  0.2888725 3.549924e-06
## 2: 8.5000000 2.43777767 53.6153441 4.197475e-03

looking at N stage status, regression and fisher’s exact

## 
##  Fisher's Exact Test for Count Data
## 
## data:  lea.n
## p-value = 0.4127
## alternative hypothesis: two.sided
##         term     estimate    std.error    statistic    p.value
## 1  N.Status0  -1.60943791    0.7745967 -2.077775411 0.03773005
## 2  N.Status1  -1.25276297    0.8017837 -1.562469938 0.11817732
## 3 N.Status1a  -0.69314718    1.2247449 -0.565952303 0.57142620
## 4 N.Status1b  -1.09861229    0.8164966 -1.345519766 0.17845744
## 5  N.Status2 -17.56606849 3956.1803280 -0.004440159 0.99645728
## 6 N.Status2c  -0.69314718    1.2247449 -0.565952303 0.57142620
## 7  N.Status3  -0.06453852    0.3593976 -0.179574135 0.85748691
## 8  N.Statusx -17.56606849 2284.1017773 -0.007690580 0.99386387
## Waiting for profiling to be done...
##          Coef           OR      2.5 %        97.5 %    p.value
## 1:  N.Status0 2.000000e-01 0.03073790  7.587157e-01 0.03773005
## 2:  N.Status1 2.857143e-01 0.04257115  1.182077e+00 0.11817732
## 3: N.Status1a 5.000000e-01 0.02323546  5.220162e+00 0.57142620
## 4: N.Status1b 3.333333e-01 0.04883842  1.446755e+00 0.17845744
## 5:  N.Status2 2.350463e-08         NA           Inf 0.99645728
## 6: N.Status2c 5.000000e-01 0.02323546  5.220162e+00 0.57142620
## 7:  N.Status3 9.375000e-01 0.45887033  1.905041e+00 0.85748691
## 8:  N.Statusx 2.350463e-08         NA 6.529915e+121 0.99386387

looking at M1 stage status, regression and fisher’s exact

## 
##  Fisher's Exact Test for Count Data
## 
## data:  lea.m
## p-value = 4.341e-05
## alternative hypothesis: two.sided
##                                      term    estimate    std.error
## 1 factor(M.Stadium, levels = c(2, 1, 3))2  -2.0149030    0.7527727
## 2 factor(M.Stadium, levels = c(2, 1, 3))1 -18.5660685 1743.2484934
## 3 factor(M.Stadium, levels = c(2, 1, 3))3   0.1541507    0.3212080
##     statistic     p.value
## 1 -2.67664216 0.007436401
## 2 -0.01065027 0.991502474
## 3  0.47990916 0.631291986
## Waiting for profiling to be done...
##                                       Coef           OR      2.5 %
## 1: factor(M.Stadium, levels = c(2, 1, 3))2 1.333333e-01 0.02101077
## 2: factor(M.Stadium, levels = c(2, 1, 3))1 8.646869e-09         NA
## 3: factor(M.Stadium, levels = c(2, 1, 3))3 1.166667e+00 0.62156694
##          97.5 %     p.value
## 1: 4.724765e-01 0.007436401
## 2: 5.528461e+42 0.991502474
## 3: 2.212424e+00 0.631291986

looking at number of mets, regression and fisher’s exact

## 
##  Fisher's Exact Test for Count Data
## 
## data:  lea.num
## p-value = 4.621e-12
## alternative hypothesis: two.sided
##          term   estimate    std.error    statistic     p.value
## 1 nummets6-15  -1.386294    1.1180340 -1.239939371 0.214997820
## 2  nummets1-5 -20.566069 2955.0616519 -0.006959607 0.994447082
## 3  nummets>15   1.145132    0.4339489  2.638864118 0.008318432
## Waiting for profiling to be done...
##           Coef           OR     2.5 %       97.5 %     p.value
## 1: nummets6-15 2.500000e-01 0.0127798 1.689944e+00 0.214997820
## 2:  nummets1-5 1.170226e-09 0.0000000 6.348999e+50 0.994447082
## 3:  nummets>15 3.142857e+00 1.4106570 7.950633e+00 0.008318432

Multivariate logistic regression based on significant factors

Age, LDH, S100, M1 stage, Number of mets

##                                       term    estimate    std.error
## 1                              (Intercept) -16.9552451    12.783444
## 2                   Age.at.first.diagnosis   0.3730655     0.264492
## 3                    factor(LDH.elevated)2  -6.0606707     4.662929
## 4                    factor(LDH.elevated)3 -35.7779734 19267.576102
## 5                    factor(S100elevated)1   0.5806059    10.880053
## 6                    factor(S100elevated)3  15.3263384 22879.638228
## 7  factor(M.Stadium, levels = c(2, 1, 3))1 -18.8446301 14978.470385
## 8  factor(M.Stadium, levels = c(2, 1, 3))3   8.0910664     6.187591
## 9                               nummets1-5 -28.9521574  8715.441872
## 10                              nummets>15  -3.2994021    11.130990
##       statistic   p.value
## 1  -1.326344087 0.1847257
## 2   1.410498551 0.1583925
## 3  -1.299756230 0.1936845
## 4  -0.001856901 0.9985184
## 5   0.053364256 0.9574417
## 6   0.000669868 0.9994655
## 7  -0.001258114 0.9989962
## 8   1.307627843 0.1909996
## 9  -0.003321938 0.9973495
## 10 -0.296415863 0.7669125
## Waiting for profiling to be done...
##                                        Coef           OR        2.5 %
##  1:                             (Intercept) 4.329429e-08 6.206882e-25
##  2:                  Age.at.first.diagnosis 1.452180e+00 1.093713e+00
##  3:                   factor(LDH.elevated)2 2.332836e-03 1.168112e-09
##  4:                   factor(LDH.elevated)3 2.896167e-16           NA
##  5:                   factor(S100elevated)1 1.787121e+00 7.507362e-07
##  6:                   factor(S100elevated)3 4.530480e+06 0.000000e+00
##  7: factor(M.Stadium, levels = c(2, 1, 3))1 6.544570e-09           NA
##  8: factor(M.Stadium, levels = c(2, 1, 3))3 3.265168e+03 1.297734e+00
##  9:                              nummets1-5 2.668319e-13           NA
## 10:                              nummets>15 3.690523e-02 7.306527e-12
##           97.5 %   p.value
##  1: 3.416256e-01 0.1847257
##  2: 3.297522e+00 0.1583925
##  3: 1.558884e+00 0.1936845
##  4:          Inf 0.9985184
##  5: 5.332256e+07 0.9574417
##  6:          Inf 0.9994655
##  7:          Inf 0.9989962
##  8: 4.446614e+11 0.1909996
##  9:          Inf 0.9973495
## 10: 1.807537e+03 0.7669125

making forestplot with all factors

## Loading required package: grid
## Loading required package: magrittr
## Loading required package: checkmate