How can I create a cross-table by multiple variables in R?

I have a dummy dataset df which contains data on the treatment, disease grade and survival outcome of 50 people.

library(Hmisc)# Generate initial cases.set.seed(123)n <- 50df <- data.frame(  treatment = sample(0:1, n, replace = TRUE),  grade = sample(0:1, n, replace = TRUE),  death = sample(0:1, n, replace = TRUE))# Add labels to columns.label(df$treatment) <- "Drug Intervention"label(df$grade) <- "Cancer Grade"label(df$death) <- "Death Occurrence"# Factor the values.df$treatment <- factor(df$treatment, levels = c(0, 1), labels = c("Drug A", "Drug B"))df$grade <- factor(df$grade, levels = c(0, 1), labels = c("Grade 1", "Grade 2"))df$death <- factor(df$death, levels = c(0, 1), labels = c("Survived", "Died"))

I want to generate a 2 x 3 cross table which shows the number of people who survived by grade and treatment. I am using tbl_strata() and tbl_summary() from gtsummary to do this.

This code is getting close to the desired outcome:

library(tidyverse)library(gtsummary)df %>%   tbl_strata(    strata = grade,    ~.x %>%      tbl_summary(        by = death,        percent = "row"      ))

Which produces a plot that looks like this:

However the desired output is:

How can I use gtsummary to filter out/collapse the 'Survived` columns to simplify the table, and is it possible to add a Fisher's exact p-value for the relationship between death and grade (for the two separate drugs)?