ggsurvplot not displaying number at risk for time 0

I want to generate a Kaplan-Meier plot from a dataset (data2), based on a original dataset from which two data points were removed (data1) to generate data2.

When I generate the KM plot for the original dataset (data1) the risk table displays properly (left plot), when using data2, no numbers at risk are shown at time 0 (right plot).

Km plots for data1 and data2.

library(survival)
library(survminer)

data1 <- data.frame(
  time_OS = c(11.47,6.57,3.57,21.73,10.67,12.40,7.17,16.37,15.43,4.93,13.43,
              11.87,21.27,14.83,3.47,4.40,8.33,13.57,20.33,17.33,5.60,4.23,6.50,
              4.63,13.73,14.03,12.93,1.13, 1.73),
  cnsr_os_recode = c(2,2,2,1,2,2,2,2,1,2,2,1,1,1,2,2,1,1,1,1,1,1,1,2,2,1,1,2,2),
  copies_gr_median = c(FALSE,TRUE,TRUE,FALSE,FALSE,FALSE,TRUE,TRUE,FALSE,TRUE,TRUE,
                       FALSE,FALSE,FALSE,TRUE,TRUE,TRUE,FALSE,FALSE,FALSE,FALSE,
                       TRUE,FALSE,TRUE,FALSE,FALSE,TRUE, TRUE, TRUE)
)

data2 <- data.frame(
  time_OS = c(11.47,6.57,3.57,21.73,10.67,12.40,7.17,16.37,15.43,4.93,13.43,
              11.87,21.27,14.83,3.47,4.40,8.33,13.57,20.33,17.33,5.60,4.23,6.50,
              4.63,13.73,14.03,12.93),
  cnsr_os_recode = c(2,2,2,1,2,2,2,2,1,2,2,1,1,1,2,2,1,1,1,1,1,1,1,2,2,1,1),
  copies_gr_median = c(FALSE,TRUE,TRUE,FALSE,FALSE,FALSE,TRUE,TRUE,FALSE,TRUE,TRUE,
                       FALSE,FALSE,FALSE,TRUE,TRUE,TRUE,FALSE,FALSE,FALSE,FALSE,
                       TRUE,FALSE,TRUE,FALSE,FALSE,TRUE)
)

func1 <- Surv(time_OS, cnsr_os_recode) ~ copies_gr_median
sf1 <- survminer::surv_fit(func1, data = data1)
    
g1 <- survminer::ggsurvplot(
  fit = sf1, 
  risk.table = TRUE, 
  conf.int = TRUE, 
  palette = 'jco',
  ggtheme = theme_bw()
)

func2 <- Surv(time_OS, cnsr_os_recode) ~ copies_gr_median
sf2 <- survminer::surv_fit(func2, data = data2)
    
g2 <- survminer::ggsurvplot(
  fit = sf2, 
  risk.table = TRUE, 
  conf.int = TRUE, 
  palette = 'jco',
  ggtheme = theme_bw()
)
  
splots <- list()
splots[[1]] <- g1
splots[[2]] <- g2

# Arrange multiple ggsurvplots and print the output
arrange_ggsurvplots(splots, print = TRUE,
  ncol = 2, nrow = 1, risk.table.height = 0.3)

I have tried googling and recoding the variables, but have not had any success.

Solution

That looks like a bug to me, which can be circumvented by adding the break.time.by argument to ggsurvplot.

g2 <- survminer::ggsurvplot(
  fit = sf2, 
  risk.table = TRUE, 
  conf.int = TRUE,
  break.time.by=5   # <---- HERE
)

g2