[R] Adjustment for multiple-comparison for log-rank test

[Marco Barbàra]

DeaR experts,

I was asked for a log-rank pairwise survival comparison. I've a straightforward way
 to do this using the SAS system:

http://support.sas.com/documentation/cdl/en/statug/63033/HTML/default/viewer.htm#/documentation/cdl/en/statug/63033/HTML/default/statug_lifetest_sect019.htm

What I've found in R is shown below, but it's not a logrank test, 
I suppose. (The documentation talks about "Tukey pairwise-comparisons").

Is it possible to carry out a "pairwise" logrank test? 
Am I totally misguided?

Thank you very much for help.
 

################################### R code #################################################
> data(pbc)
> pbc$stage <- factor(pbc$stage)
> (fit <- coxph(Surv(time,status==2)~stage,data=pbc))
Call:
coxph(formula = Surv(time, status == 2) ~ stage, data = pbc)


       coef exp(coef) se(coef)    z       p
stage2 1.10      3.01    0.737 1.50 0.13000
stage3 1.53      4.63    0.722 2.12 0.03400
stage4 2.53     12.57    0.717 3.53 0.00041

Likelihood ratio test=65.1  on 3 df, p=4.84e-14  n=412 (6 observations deleted due to missingness)

> summary(glht(fit,linfct=mcp(stage="Tukey"),alternative="g"))

	 Simultaneous Tests for General Linear Hypotheses

Multiple Comparisons of Means: Tukey Contrasts


Fit: coxph(formula = Surv(time, status == 2) ~ stage, data = pbc)

Linear Hypotheses:
           Estimate Std. Error z value Pr(>z)    
2 - 1 <= 0   1.1027     0.7374   1.495  0.237    
3 - 1 <= 0   1.5318     0.7224   2.120  0.068 .  
4 - 1 <= 0   2.5311     0.7168   3.531 <0.001 ***
3 - 2 <= 0   0.4291     0.2544   1.686  0.169    
4 - 2 <= 0   1.4284     0.2375   6.013 <0.001 ***
4 - 3 <= 0   0.9994     0.1816   5.502 <0.001 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1 
(Adjusted p values reported -- single-step method)