admiralonco: Oncology Extension Package for ADaM in 'R' Asset Library

Description

Programming oncology specific Clinical Data Interchange Standards Consortium (CDISC) compliant Analysis Data Model (ADaM) datasets in 'R'. ADaM datasets are a mandatory part of any New Drug or Biologics License Application submitted to the United States Food and Drug Administration (FDA). Analysis derivations are implemented in accordance with the "Analysis Data Model Implementation Guide" (CDISC Analysis Data Model Team (2021), https://www.cdisc.org/standards/foundational/adam). The package is an extension package of the 'admiral' package.

Author(s)

Maintainer: Stefan Bundfuss stefan.bundfuss@roche.com

Authors:

• Amit Jain

• Vinh Nguyen

• Olga Starostecka

• Kiran Peddamudium

• Tomoyuki Namai

• Ross Farrugia

Other contributors:

• Yirong Cao [contributor]

• Ashwini Weber [contributor]

• F. Hoffmann-La Roche AG [copyright holder, funder]

• GlaxoSmithKline LLC [copyright holder, funder]

• Amgen Inc. [copyright holder, funder]

• Bristol Myers Squibb [copyright holder, funder]

See Also

Useful links:

• https://pharmaverse.github.io/admiralonco/

• https://github.com/pharmaverse/admiralonco

• Report bugs at https://github.com/pharmaverse/admiralonco/issues

Response Analysis Dataset

Description

An example response analysis dataset

Usage

admiral_adrs

Format

An object of class tbl_df (inherits from tbl, data.frame) with 3694 rows and 75 columns.

Source

Derived from the adsl, rs and tu datasets using {admiralonco} (https://github.com/pharmaverse/admiralonco/blob/main/inst/templates/ad_adrs.R)

Map Character Response Values to Numeric Values

Description

Map character response values like "PR" or "SD" to numeric values.

Usage

aval_resp(arg)

Arguments

Value

• 1 if arg equals "CR",

• 2 if arg equals "PR",

• 3 if arg equals "SD",

• 4 if arg equals "NON-CR/NON-PD",

• 5 if arg equals "PD",

• 6 if arg equals "NE",

• 7 if arg equals "MISSING",

• NA_real_ otherwise

Author(s)

Stefan Bundfuss

Examples

aval_resp(c("CR", "PR", "SD", "NON-CR/NON-PD", "PD", "NE", "MISSING", "ND", NA_character_))

Pre-Defined Time-to-Event Source Objects

Description

These pre-defined tte_source objects can be used as input to admiral::derive_param_tte().

Usage

death_event

lastalive_censor

pd_event

lasta_censor

rand_censor

trts_censor

Details

To see the definition of the various objects simply print the object in the R console, e.g. print(death_event). For details of how to use these objects please refer to admiral::derive_param_tte().

Printing an object will display input dataset_name, filter (if applicable), date variable, and appropriate values for EVNTDESC, CNSDTDSC, SRCDOM, SRCVAR, and SRCSEQ.

See Also

admiral::derive_param_tte(), admiral::tte_source(), admiral::event_source(), admiral::censor_source()

Examples

# This shows the definition of all pre-defined `tte_source` objects that ship
# with {admiralonco}
for (obj in admiral::list_tte_source_objects(package = "admiralonco")$object) {
  cat(obj, "\n")
  print(get(obj, envir = getNamespace("admiralonco")))
  cat("\n")
}

Adds a Parameter for Best Overall Response (without confirmation)

Description

The derive_param_bor() function has been superseded in favor of derive_extreme_event().

Adds a parameter for best overall response, without confirmation, optionally up to first progressive disease

Usage

derive_param_bor(
  dataset,
  dataset_adsl,
  filter_source,
  source_pd = NULL,
  source_datasets = NULL,
  reference_date,
  ref_start_window,
  missing_as_ne = FALSE,
  aval_fun,
  set_values_to,
  subject_keys = get_admiral_option("subject_keys")
)

Arguments

pd_date <- date_source(
  dataset_name = "adrs",
  date = ADT,
  filter = PARAMCD == PD
)

Details

Calculates the best overall response (BOR) parameter, as detailed below.

Records after PD can be removed using the source_pd and source_datasets arguments.

Note:

1. All CR, PR and PD response records are considered for Best Overall Response.

2. All SD or NON-CR/NON-PD records where ADT >= reference_date + ref_start_window are also considered for Best Overall Response.

3. Subjects with ONLY an SD or NON-CR/NON-PD records where ADT < reference_date + ref_start_window are assigned a Best Overall Response of NE.

4. The Best Response, from the records in steps 1 to 3, is then selected in the following order of preference: CR, PR, SD, NON-CR/NON-PD, PD, NE, MISSING

5. The AVAL column is added and set using the aval_fun(AVALC) function

6. The columns specified by the set_values_to parameter and records are added to the dataframe passed into the dataset argument

Note: Any responses of SD or NON-CR/NON-PD that occur before reference_date + ref_start_window are ignored in the calculation of BOR. All other responses are included in the calculation of BOR, irrespective of the number of days from the reference date.

Also Note: All columns from the input dataset are kept. For subjects with no records in the input dataset (after the filter is applied) all columns are kept from ADSL which are also in the input dataset. Columns which are not to be populated for the new parameter or populated differently (e.g. RSSTRESC, VISIT, PARCATy, ANLzzFL, ...) should be overwritten using the set_values_to parameter.

Value

The dataframe passed in the dataset argument with additional columns and/or rows as set in the set_values_to argument.

Author(s)

Stephen Gormley

See Also

Other superseded: derive_param_clinbenefit(), derive_param_confirmed_bor(), derive_param_confirmed_resp(), derive_param_response(), filter_pd()

Examples

library(magrittr)
library(dplyr)
library(tibble)
library(lubridate)
library(admiral)

# Create ADSL dataset
adsl <- tribble(
  ~USUBJID, ~TRTSDTC,
  "1",      "2020-01-01",
  "2",      "2019-12-12",
  "3",      "2019-11-11",
  "4",      "2019-12-30",
  "5",      "2020-01-01",
  "6",      "2020-02-02",
  "7",      "2020-02-02",
  "8",      "2020-04-01"
) %>%
  mutate(
    TRTSDT = ymd(TRTSDTC),
    STUDYID = "XX1234"
  )

# Create ADRS dataset
ovr_obs <- tribble(
  ~USUBJID, ~ADTC, ~AVALC, ~ANL01FL,
  "1", "2020-01-01", "PR", "Y",
  "1", "2020-02-01", "CR", "Y",
  "1", "2020-02-16", "NE", "Y",
  "1", "2020-03-01", "CR", "Y",
  "1", "2020-04-01", "SD", "Y",
  "2", "2020-01-01", "SD", "Y",
  "2", "2020-02-01", "PR", "Y",
  "2", "2020-03-01", "SD", "Y",
  "2", "2020-03-13", "CR", "Y",
  "3", "2019-11-12", "CR", "Y",
  "3", "2019-12-02", "CR", "Y",
  "3", "2020-01-01", "SD", "Y",
  "4", "2020-01-01", "PR", "Y",
  "4", "2020-03-01", "SD", "N",
  "4", "2020-04-01", "SD", "Y",
  "4", "2020-05-01", "PR", "Y",
  "4", "2020-05-15", "NON-CR/NON-PD", "Y",
  "5", "2020-01-01", "PR", "Y",
  "5", "2020-01-10", "SD", "Y",
  "5", "2020-01-20", "PR", "Y",
  "5", "2020-05-15", "NON-CR/NON-PD", "Y",
  "6", "2020-02-06", "PR", "Y",
  "6", "2020-02-16", "CR", "Y",
  "6", "2020-03-30", "PR", "Y",
  "6", "2020-04-12", "PD", "Y",
  "6", "2020-05-01", "CR", "Y",
  "6", "2020-06-01", "CR", "Y",
  "7", "2020-02-06", "PR", "Y",
  "7", "2020-02-16", "CR", "Y",
  "7", "2020-04-01", "NE", "N"
) %>%
  mutate(PARAMCD = "OVR")

pd_obs <-
  bind_rows(tribble(
    ~USUBJID, ~ADTC,        ~AVALC,
    "2",      "2020-03-01", "Y",
    "4",      "2020-02-01", "Y"
  ) %>%
    mutate(PARAMCD = "PD"))

adrs <- bind_rows(ovr_obs, pd_obs) %>%
  mutate(
    ADT = ymd(ADTC),
    STUDYID = "XX1234"
  ) %>%
  select(-ADTC) %>%
  derive_vars_merged(
    dataset_add = adsl,
    by_vars     = exprs(STUDYID, USUBJID),
    new_vars    = exprs(TRTSDT)
  )

pd_date <- date_source(
  dataset_name = "adrs",
  date         = ADT,
  filter       = PARAMCD == "PD"
)

aval_fun_pass <- function(arg) {
  case_when(
    arg == "CR" ~ 11,
    arg == "PR" ~ 22,
    arg == "SD" ~ 33,
    arg == "NON-CR/NON-PD" ~ 44,
    arg == "PD" ~ 55,
    arg == "NE" ~ 66,
    arg == "MISSING" ~ 77,
    TRUE ~ NA_real_
  )
}

# Derive best overall response parameter
derive_param_bor(
  adrs,
  dataset_adsl = adsl,
  filter_source = PARAMCD == "OVR" & ANL01FL == "Y",
  source_pd = pd_date,
  source_datasets = list(adrs = adrs),
  aval_fun = aval_fun_pass,
  reference_date = TRTSDT,
  ref_start_window = 28,
  set_values_to = exprs(
    PARAMCD = "BOR",
    PARAM = "Best Overall Response"
  )
) %>%
  filter(PARAMCD == "BOR")

Adds a Parameter for Clinical Benefit

Description

The derive_param_clinbenefit() function has been superseded in favor of derive_extreme_event().

Adds a parameter for clinical benefit/disease control

Usage

derive_param_clinbenefit(
  dataset,
  dataset_adsl,
  filter_source,
  source_resp,
  source_pd = NULL,
  source_datasets,
  reference_date,
  ref_start_window,
  aval_fun,
  clinben_vals = c("CR", "PR", "SD", "NON-CR/NON-PD"),
  set_values_to,
  subject_keys = get_admiral_option("subject_keys")
)

Arguments

Details

Clinical benefit/disease control is first identified by looking for subjects having response status, and then derived for subjects that have at least one evaluable non-PD response assessment prior to first PD (Progressive Disease) (i.e., responses inclusive of CR, PR, SD, and NON-CR/NON-PD) and after a specified amount of time from a reference date (ref_start_window).

Note: The user input values they wish to include when determining clinical benefit using the argument clinben_vals. The default values for this are CR, PR, SD, and NON-CR/NON-PD, as listed above. In the below example, eligible values be limited to CR and PR.

Example: clinben_vals <- c("CR", "PR")

1. The input dataset (dataset) is restricted to the observations matching filter_source and to observations before or at the date specified by source_pd.

2. This dataset is further restricted to include user-generated response assessments from clinben_vals or include response assessments of CR, PR, SD, and NON-CR/NON-PD, exclude missing response assessments, and exclude those less than ref_start_window after reference_date. The earliest assessment by ADT is then selected.

3. The dataset identified by dataset in source_resp is restricted according to its filter argument. The variable corresponding to the date parameter of source_resp is considered together with ADT from the previous step.

4. For the observations being added to dataset, ADT is set to the earlier of the first assessment date representing an evaluable non-PD assessment prior to first PD, or the date representing the start of response.

5. For the observations being added to dataset, AVALC is set to

   • Y for those subjects in the dataset meeting the criteria for clinical benefit above

   • N for subjects not meeting the clinical benefit criteria in dataset or the dataset identified in source_resp

   • N for subjects present in dataset_adsl but not present in dataset or the dataset identified in source_resp.

6. AVAL is derived using AVALC as input to the function specified in aval_fun.

7. The variables specified by set_values_to are added to the new observations with values equal to the values specified in the same.

8. The new observations are added to dataset. Variables held in common between dataset and dataset_adsl are kept for the new observations, and are populated with their values from dataset_adsl.

Value

The input dataset with a new parameter for clinical benefit

Author(s)

Andrew Smith

See Also

Other superseded: derive_param_bor(), derive_param_confirmed_bor(), derive_param_confirmed_resp(), derive_param_response(), filter_pd()

Examples

library(lubridate)
library(dplyr)
library(admiral)

adsl <- tibble::tribble(
  ~USUBJID, ~TRTSDT,
  "01",     ymd("2020-01-14"),
  "02",     ymd("2021-02-16"),
  "03",     ymd("2021-03-09"),
  "04",     ymd("2021-04-21")
) %>%
  mutate(STUDYID = "AB42")

adrs <- tibble::tribble(
  ~USUBJID, ~PARAMCD, ~AVALC, ~ADT,
  "01",     "RSP",    "Y",    ymd("2021-03-14"),
  "02",     "RSP",    "N",    ymd("2021-05-07"),
  "03",     "RSP",    "N",    NA,
  "04",     "RSP",    "N",    NA,
  "01",     "PD",     "N",    NA,
  "02",     "PD",     "Y",    ymd("2021-05-07"),
  "03",     "PD",     "N",    NA,
  "04",     "PD",     "N",    NA,
  "01",     "OVR",    "SD",   ymd("2020-03-14"),
  "01",     "OVR",    "PR",   ymd("2021-04-13"),
  "02",     "OVR",    "PR",   ymd("2021-04-08"),
  "02",     "OVR",    "PD",   ymd("2021-05-07"),
  "02",     "OVR",    "CR",   ymd("2021-06-20"),
  "03",     "OVR",    "SD",   ymd("2021-03-30"),
  "04",     "OVR",    "NE",   ymd("2021-05-21"),
  "04",     "OVR",    "NA",   ymd("2021-06-30"),
  "04",     "OVR",    "NE",   ymd("2021-07-24"),
  "04",     "OVR",    "ND",   ymd("2021-09-04"),
) %>%
  mutate(STUDYID = "AB42", ANL01FL = "Y") %>%
  derive_vars_merged(
    dataset_add = adsl,
    by_vars = exprs(STUDYID, USUBJID),
    new_vars = exprs(TRTSDT)
  )

pd <- date_source(
  dataset_name = "adrs",
  date = ADT,
  filter = PARAMCD == "PD" & AVALC == "Y" & ANL01FL == "Y"
)

resp <- date_source(
  dataset_name = "adrs",
  date = ADT,
  filter = PARAMCD == "RSP" & AVALC == "Y" & ANL01FL == "Y"
)

derive_param_clinbenefit(
  dataset = adrs,
  dataset_adsl = adsl,
  filter_source = PARAMCD == "OVR" & ANL01FL == "Y",
  source_resp = resp,
  source_pd = pd,
  source_datasets = list(adrs = adrs),
  reference_date = TRTSDT,
  ref_start_window = 28,
  set_values_to = exprs(
    PARAMCD = "CBR"
  )
) %>%
  filter(PARAMCD == "CBR")

Adds a Parameter for Confirmed Best Overall Response

Description

The derive_param_confirmed_bor() function has been superseded in favor of derive_extreme_event().

Adds a parameter for confirmed best overall response (BOR)

Usage

derive_param_confirmed_bor(
  dataset,
  dataset_adsl,
  filter_source,
  source_pd = NULL,
  source_datasets = NULL,
  reference_date,
  ref_start_window,
  ref_confirm,
  max_nr_ne = 1,
  accept_sd = FALSE,
  missing_as_ne = FALSE,
  aval_fun,
  set_values_to,
  subject_keys = get_admiral_option("subject_keys")
)

Arguments

pd_date <- date_source(
  dataset_name = "adrs",
  date = ADT,
  filter = PARAMCD == PD
)

Details

1. The input dataset (dataset) is restricted to the observations matching filter_source and to observations before or at the date specified by source_pd.

2. The following potential confirmed responses are selected from the restricted input dataset:

   • "CR": An assessment is considered as complete response (CR) if

     • AVALC == "CR",

     • there is a confirmatory assessment with AVALC == "CR" at least ref_confirm days after the assessment,

     • all assessments between the assessment and the confirmatory assessment are "CR" or "NE", and

     • there are at most max_nr_ne "NE" assessments between the assessment and the confirmatory assessment.

   • "PR": An assessment is considered as partial response (PR) if

     • AVALC == "PR",

     • there is a confirmatory assessment with AVALC %in% c("CR", "PR") at least ref_confirm days after the assessment,

     • all assessments between the assessment and the confirmatory assessment are "CR", "PR", "SD", or "NE",

     • there is no "PR" assessment after a "CR" assessment in the confirmation period,

     • there are at most max_nr_ne "NE" assessments between the assessment and the confirmatory assessment, and

     • if the accept_sd argument is set to TRUE, one "SD" assessment in the confirmation period is accepted. Otherwise, no "SD" assessment must occur within the confirmation period.

   • "SD": An assessment is considered as stable disease (SD) if

     • AVALC %in% c("CR", "PR", "SD") and

     • the assessment is at least ref_start_window days after reference_date.

   • "NON-CR/NON-PD": An assessment is considered as NON-CR/NON-PD if

     • AVALC = "NON-CR/NON-PD" and

     • the assessment is at least ref_start_window days after reference_date.

   • "PD": An assessment is considered as progressive disease (PD) if AVALC == "PD".

   • "NE": An assessment is considered as not estimable (NE) if

     • AVALC == "NE" or

     • AVALC %in% c("CR", "PR", "SD", "NON-CR/NON-PD") and the assessment is less than ref_start_window days after reference_date.

   • "ND": An assessment is considered as not done (ND) if AVALC == "ND".

   • "MISSING": An assessment is considered as missing (MISSING) if a subject has no observation in the input dataset.

     If the missing_as_ne argument is set to TRUE, AVALC is set to "NE" for these subjects.

3. For each subject the best response as derived in the previous step is selected, where "CR" is best and "MISSING" is worst in the order above. If the best response is not unique, the first one (with respect to ADT) is selected. If the selected record is from the input dataset, all variables are kept. If the selected record is from dataset_adsl, all variables which are in both dataset and dataset_adsl are kept.

4. The AVAL variable is added and set to aval_fun(AVALC).

5. The variables specified by the set_values_to parameter are added to the new observations.

6. The new observations are added to input dataset.

Value

The input dataset with a new parameter for confirmed best overall response

Author(s)

Stefan Bundfuss

See Also

Other superseded: derive_param_bor(), derive_param_clinbenefit(), derive_param_confirmed_resp(), derive_param_response(), filter_pd()

Examples

library(dplyr)
library(lubridate)
library(admiral)

# Create ADSL dataset
adsl <- tibble::tribble(
  ~USUBJID, ~TRTSDTC,
  "1",      "2020-01-01",
  "2",      "2019-12-12",
  "3",      "2019-11-11",
  "4",      "2019-12-30",
  "5",      "2020-01-01",
  "6",      "2020-02-02",
  "7",      "2020-02-02",
  "8",      "2020-04-01",
  "9",      "2020-03-01"
) %>%
  mutate(
    TRTSDT = ymd(TRTSDTC),
    STUDYID = "XX1234"
  )

# Create ADRS dataset
ovr_obs <- tibble::tribble(
  ~USUBJID, ~ADTC,        ~AVALC,
  "1",      "2020-01-01", "PR",
  "1",      "2020-02-01", "CR",
  "1",      "2020-02-16", "NE",
  "1",      "2020-03-01", "CR",
  "1",      "2020-04-01", "SD",
  "2",      "2020-01-01", "SD",
  "2",      "2020-02-01", "PR",
  "2",      "2020-03-01", "SD",
  "2",      "2020-03-13", "CR",
  "3",      "2019-11-12", "CR",
  "3",      "2019-12-02", "CR",
  "3",      "2020-01-01", "SD",
  "4",      "2020-01-01", "PR",
  "4",      "2020-03-01", "SD",
  "4",      "2020-04-01", "SD",
  "4",      "2020-05-01", "PR",
  "4",      "2020-05-15", "NON-CR/NON-PD",
  "5",      "2020-01-01", "PR",
  "5",      "2020-01-10", "SD",
  "5",      "2020-01-20", "PR",
  "5",      "2020-05-15", "NON-CR/NON-PD",
  "6",      "2020-02-06", "PR",
  "6",      "2020-02-16", "CR",
  "6",      "2020-03-30", "PR",
  "6",      "2020-04-12", "PD",
  "6",      "2020-05-01", "CR",
  "6",      "2020-06-01", "CR",
  "7",      "2020-02-06", "PR",
  "7",      "2020-02-16", "CR",
  "7",      "2020-04-01", "NE",
  "9",      "2020-03-16", "CR",
  "9",      "2020-04-01", "NE",
  "9",      "2020-04-16", "NE",
  "9",      "2020-05-01", "CR"
) %>%
  mutate(PARAMCD = "OVR", ANL01FL = "Y")

pd_obs <-
  bind_rows(tibble::tribble(
    ~USUBJID, ~ADTC,        ~AVALC,
    "6",      "2020-04-12", "Y"
  ) %>%
    mutate(PARAMCD = "PD", ANL01FL = "Y"))

adrs <- bind_rows(ovr_obs, pd_obs) %>%
  mutate(
    ADT = ymd(ADTC),
    STUDYID = "XX1234"
  ) %>%
  select(-ADTC) %>%
  derive_vars_merged(
    dataset_add = adsl,
    by_vars = exprs(STUDYID, USUBJID),
    new_vars = exprs(TRTSDT)
  )

pd_date <- date_source(
  dataset_name = "adrs",
  date = ADT,
  filter = PARAMCD == "PD" & ANL01FL == "Y"
)

# Derive confirmed best overall response parameter
derive_param_confirmed_bor(
  adrs,
  dataset_adsl = adsl,
  filter_source = PARAMCD == "OVR" & ANL01FL == "Y",
  source_pd = pd_date,
  source_datasets = list(adrs = adrs),
  reference_date = TRTSDT,
  ref_start_window = 28,
  ref_confirm = 28,
  set_values_to = exprs(
    PARAMCD = "CBOR",
    PARAM = "Best Confirmed Overall Response by Investigator"
  )
) %>%
  filter(PARAMCD == "CBOR")

# Derive confirmed best overall response parameter (accepting SD for PR,
# accept two NEs, and considering missings as NE)
derive_param_confirmed_bor(
  adrs,
  dataset_adsl = adsl,
  filter_source = PARAMCD == "OVR" & ANL01FL == "Y",
  source_pd = pd_date,
  source_datasets = list(adrs = adrs),
  reference_date = TRTSDT,
  ref_start_window = 28,
  ref_confirm = 28,
  max_nr_ne = 2,
  accept_sd = TRUE,
  missing_as_ne = TRUE,
  set_values_to = exprs(
    PARAMCD = "CBOR",
    PARAM = "Best Confirmed Overall Response by Investigator"
  )
) %>%
  filter(PARAMCD == "CBOR")

Adds a Parameter for Confirmed Response

Description

The derive_param_confirmed_resp() function has been superseded in favor of derive_extreme_event().

Adds a parameter for confirmed response

Usage

derive_param_confirmed_resp(
  dataset,
  dataset_adsl,
  filter_source,
  source_pd = NULL,
  source_datasets = NULL,
  ref_confirm,
  max_nr_ne = 1,
  accept_sd = FALSE,
  aval_fun,
  set_values_to,
  subject_keys = get_admiral_option("subject_keys")
)

Arguments

pd_date <- date_source(
  dataset_name = "adrs",
  date = ADT,
  filter = PARAMCD == PD
)

Details

1. The input dataset (dataset) is restricted to the observations matching filter_source and to observations before or at the date specified by source_pd.

2. A subject is considered as responder if there is at least one observation in the restricted dataset with

   • AVALC == "CR",

   • there is a confirmatory assessment with AVALC == "CR" at least ref_confirm days after the assessment,

   • all assessments between the assessment and the confirmatory assessment are "CR" or "NE", and

   • there are at most max_nr_ne "NE" assessments between the assessment and the confirmatory assessment.

   or at least one observation with

   • AVALC == "PR",

   • there is a confirmatory assessment with AVALC %in% c("CR", "PR") at least ref_confirm days after the assessment,

   • all assessments between the assessment and the confirmatory assessment are "CR", "PR", "SD", or "NE",

   • there is no "PR" assessment after a "CR" assessment in the confirmation period,

   • there are at most max_nr_ne "NE" assessments between the assessment and the confirmatory assessment,

   • if the accept_sd argument is set to TRUE, one "SD" assessment in the confirmation period is accepted. Otherwise, no "SD" assessment must occur within the confirmation period.

3. For responders AVALC is set to "Y" and ADT to the first date where the response criteria are fulfilled. For all other subjects in dataset_adsl AVALC is set to "N" and ADT to NA.

4. The AVAL variable is added and set to aval_fun(AVALC).

5. The variables specified by the set_values_to parameter are added to the new observations.

6. The new observations are added to input dataset.

Value

The input dataset with a new parameter for confirmed response

Author(s)

Stefan Bundfuss

See Also

Other superseded: derive_param_bor(), derive_param_clinbenefit(), derive_param_confirmed_bor(), derive_param_response(), filter_pd()

Examples

library(dplyr)
library(admiral)

# Create ADSL dataset
adsl <- tibble::tribble(
  ~USUBJID, ~TRTSDTC,
  "1",      "2020-01-01",
  "2",      "2019-12-12",
  "3",      "2019-11-11",
  "4",      "2019-12-30",
  "5",      "2020-01-01",
  "6",      "2020-02-02",
  "7",      "2020-02-02",
  "8",      "2020-04-01",
  "9",      "2020-03-01"
) %>%
  mutate(
    STUDYID = "XX1234"
  )

# Create ADRS dataset
ovr_obs <- tibble::tribble(
  ~USUBJID, ~ADTC,        ~AVALC,
  "1",      "2020-01-01", "PR",
  "1",      "2020-02-01", "CR",
  "1",      "2020-02-16", "NE",
  "1",      "2020-03-01", "CR",
  "1",      "2020-04-01", "SD",
  "2",      "2020-01-01", "SD",
  "2",      "2020-02-01", "PR",
  "2",      "2020-03-01", "SD",
  "2",      "2020-03-13", "CR",
  "3",      "2019-11-12", "CR",
  "3",      "2019-12-02", "CR",
  "3",      "2020-01-01", "SD",
  "4",      "2020-01-01", "PR",
  "4",      "2020-03-01", "SD",
  "4",      "2020-04-01", "SD",
  "4",      "2020-05-01", "PR",
  "4",      "2020-05-15", "NON-CR/NON-PD",
  "5",      "2020-01-01", "PR",
  "5",      "2020-01-10", "SD",
  "5",      "2020-01-20", "PR",
  "5",      "2020-05-15", "NON-CR/NON-PD",
  "6",      "2020-02-06", "PR",
  "6",      "2020-02-16", "CR",
  "6",      "2020-03-30", "PR",
  "6",      "2020-04-12", "PD",
  "6",      "2020-05-01", "CR",
  "6",      "2020-06-01", "CR",
  "7",      "2020-02-06", "PR",
  "7",      "2020-02-16", "CR",
  "7",      "2020-04-01", "NE",
  "9",      "2020-03-16", "CR",
  "9",      "2020-04-01", "NE",
  "9",      "2020-04-16", "NE",
  "9",      "2020-05-01", "CR"
) %>%
  mutate(PARAMCD = "OVR", ANL01FL = "Y")

pd_obs <-
  bind_rows(tibble::tribble(
    ~USUBJID, ~ADTC,        ~AVALC,
    "6",      "2020-04-12", "Y"
  ) %>%
    mutate(PARAMCD = "PD", ANL01FL = "Y"))

adrs <- bind_rows(ovr_obs, pd_obs) %>%
  mutate(
    ADT = lubridate::ymd(ADTC),
    STUDYID = "XX1234"
  ) %>%
  select(-ADTC)

pd_date <- date_source(
  dataset_name = "adrs",
  date = ADT,
  filter = PARAMCD == "PD" & ANL01FL == "Y"
)

# Derive confirmed response parameter
derive_param_confirmed_resp(
  adrs,
  dataset_adsl = adsl,
  filter_source = PARAMCD == "OVR" & ANL01FL == "Y",
  source_pd = pd_date,
  source_datasets = list(adrs = adrs),
  ref_confirm = 28,
  set_values_to = exprs(
    PARAMCD = "CRSP",
    PARAM = "Confirmed Response by Investigator"
  )
) %>%
  filter(PARAMCD == "CRSP")

# Derive confirmed response parameter (accepting SD for PR and two NEs)
derive_param_confirmed_resp(
  adrs,
  dataset_adsl = adsl,
  filter_source = PARAMCD == "OVR" & ANL01FL == "Y",
  source_pd = pd_date,
  source_datasets = list(adrs = adrs),
  ref_confirm = 28,
  max_nr_ne = 2,
  accept_sd = TRUE,
  set_values_to = exprs(
    PARAMCD = "CRSP",
    PARAM = "Confirmed Response by Investigator"
  )
) %>%
  filter(PARAMCD == "CRSP")

Adds a Parameter Indicating If a Subject Had a Response before Progressive Disease

Description

The derive_param_response() function has been superseded in favor of derive_extreme_event().

Adds a parameter indicating if a response has been observed. If a response has been observed, AVALC is set to "Y", AVAL to 1 and ADT is set to the first date when a response has been observed. If a response has not been observed, AVALC is set to "N", AVAL to 0 and ADT is set NA.

Usage

derive_param_response(
  dataset,
  dataset_adsl,
  filter_source,
  source_pd = NULL,
  source_datasets = NULL,
  set_values_to,
  aval_fun,
  subject_keys = get_admiral_option("subject_keys")
)

Arguments

Details

1. The Date of the end of the assessment period (e.g. Progressive disease, as defined by pd_source) is added to the response dataset.

2. The response dataset is restricted to observations occurring before or on the date of progressive disease.

3. For each subject (with respect to the variables specified for the subject_keys parameter), the first observation (with respect to ADT) where the response condition (filter_source parameter) is fulfilled is selected.

4. For each observation in dataset_adsl a new observation is created.

   • For subjects with a response AVALC is set to "Y", AVAL to 1, and ADT to the first date (ADT) where the response condition is fulfilled.

   • For all other subjects AVALC is set to "N", AVAL to 0 and ADT to NA.

5. The variables specified by the set_values_to parameter are added to the new observations.

6. The new observations are added to input dataset.

Value

The input dataset with a new parameter indicating if and when a response occurred

Author(s)

Samia Kabi

See Also

Other superseded: derive_param_bor(), derive_param_clinbenefit(), derive_param_confirmed_bor(), derive_param_confirmed_resp(), filter_pd()

Examples

library(dplyr)
library(admiral)
library(lubridate)
library(tibble)

adsl <- tribble(
  ~USUBJID,
  "1",
  "2",
  "3",
  "4"
) %>%
  mutate(STUDYID = "XX1234")

adrs <- tribble(
  ~USUBJID, ~PARAMCD, ~ADTC,         ~AVALC, ~ANL01FL,
  "1",      "OVR",    "2020-01-02",  "PR",   "Y",
  "1",      "OVR",    "2020-02-01",  "CR",   "Y",
  "1",      "OVR",    "2020-03-01",  "CR",   "Y",
  "1",      "OVR",    "2020-04-01",  "SD",   "Y",
  "1",      "PD",     NA_character_, "N",    "Y",
  "2",      "OVR",    "2021-06-15",  "SD",   "Y",
  "2",      "OVR",    "2021-07-16",  "PD",   "Y",
  "2",      "OVR",    "2021-09-14",  "PD",   "Y",
  "2",      "PD",     "2021-09-14",  "Y",    "Y",
  "3",      "OVR",    "2021-09-14",  "SD",   "Y",
  "3",      "OVR",    "2021-10-30",  "PD",   "Y",
  "3",      "OVR",    "2021-12-25",  "CR",   "Y",
  "3",      "PD",     "2021-10-30",  "Y",    "Y"
) %>%
  mutate(
    STUDYID = "XX1234",
    ADT = ymd(ADTC),
    ANL01FL = "Y"
  ) %>%
  select(-ADTC)

# Define the end of the assessment period for responses:
# all responses before or on the first PD will be used.
pd <- date_source(
  dataset_name = "adrs",
  date = ADT,
  filter = PARAMCD == "PD" & AVALC == "Y"
)
# Derive the response parameter
derive_param_response(
  dataset = adrs,
  dataset_adsl = adsl,
  filter_source = PARAMCD == "OVR" & AVALC %in% c("CR", "PR") & ANL01FL == "Y",
  source_pd = pd,
  source_datasets = list(adrs = adrs),
  set_values_to = exprs(
    AVAL = yn_to_numeric(AVALC),
    PARAMCD = "RSP",
    PARAM = "Response by investigator"
  ),
  subject_keys = get_admiral_option("subject_keys")
) %>%
  arrange(USUBJID, PARAMCD, ADT)

Filter up to First PD (Progressive Disease) Date

Description

The filter_pd() function has been superseded in favor of filter_relative().

Filter a dataset to only include the source parameter records up to and including the first PD (progressive disease). These records are passed to downstream derivations regarding responses such as BOR (best overall response).

Usage

filter_pd(
  dataset,
  filter,
  source_pd,
  source_datasets,
  subject_keys = get_admiral_option("subject_keys")
)

Arguments

Details

1. The input dataset (dataset) is restricted by filter.

2. For each subject the first PD date is derived as the first date (source_pd$date) in the source pd dataset (source_datasets[[source_pd$dataset_name]]) restricted by source_pd$filter.

3. The restricted input dataset is restricted to records up to first PD date. Records matching first PD date are included. For subject without any first PD date, all records are included.

Value

A subset of the input dataset

Author(s)

Teckla Akinyi, Stefan Bundfuss

See Also

Other superseded: derive_param_bor(), derive_param_clinbenefit(), derive_param_confirmed_bor(), derive_param_confirmed_resp(), derive_param_response()

Examples

library(dplyr)
library(lubridate)
library(admiral)
library(admiralonco)

# Filter OVR records up to first PD, first PD date provided in separate BDS dataset (adevent)
adrs <- tibble::tribble(
  ~STUDYID,       ~USUBJID,      ~PARAMCD, ~AVALC, ~ADT,         ~ANL01FL,
  "CDISCPILOT01", "01-701-1015", "OVR",    "CR",   "2016-01-25", "Y",
  "CDISCPILOT01", "01-701-1015", "OVR",    "SD",   "2016-02-22", NA_character_,
  "CDISCPILOT01", "01-701-1015", "OVR",    "PD",   "2016-02-22", "Y",
  "CDISCPILOT01", "01-701-1015", "BOR",    "CR",   "2016-01-25", "Y",
  "CDISCPILOT01", "01-701-1034", "OVR",    "SD",   "2015-12-07", "Y",
  "CDISCPILOT01", "01-701-1034", "OVR",    "PD",   "2016-04-25", "Y",
  "CDISCPILOT01", "01-701-1034", "OVR",    "PD",   "2016-06-25", "Y",
  "CDISCPILOT01", "01-701-1034", "BOR",    "SD",   "2015-12-07", "Y",
  "CDISCPILOT01", "01-701-1035", "OVR",    "SD",   "2016-04-25", "Y",
  "CDISCPILOT01", "01-701-1035", "OVR",    "PR",   "2016-06-25", "Y",
  "CDISCPILOT01", "01-701-1035", "BOR",    "PR",   "2016-06-25", "Y"
) %>% mutate(
  ADT = as_date(ADT)
)

adevent <- tibble::tribble(
  ~STUDYID,       ~USUBJID,      ~PARAMCD, ~AVALC, ~ADT,
  "CDISCPILOT01", "01-701-1015", "PD",     "Y",    "2016-02-22",
  "CDISCPILOT01", "01-701-1034", "PD",     "Y",    "2016-04-25"
) %>% mutate(
  ADT = as_date(ADT)
)

pd <- date_source(
  dataset_name = "adevent",
  date = ADT,
  filter = PARAMCD == "PD"
)

filter_pd(
  dataset = adrs,
  filter = PARAMCD == "OVR" & ANL01FL == "Y",
  source_pd = pd,
  source_datasets = list(adevent = adevent)
)

# Filter OVR records up to first PD, first PD date provided in ADSL dataset
adsl <- tibble::tribble(
  ~STUDYID,       ~USUBJID,      ~PDDT,
  "CDISCPILOT01", "01-701-1015", "2016-02-22",
  "CDISCPILOT01", "01-701-1034", "2016-04-25",
  "CDISCPILOT01", "01-701-1035", ""
) %>% mutate(
  PDDT = as_date(PDDT)
)

pd <- date_source(
  dataset_name = "adsl",
  date = PDDT
)

filter_pd(
  dataset = adrs,
  filter = PARAMCD == "OVR" & ANL01FL == "Y",
  source_pd = pd,
  source_datasets = list(adsl = adsl)
)

# Filter OVR records up to first PD, first PD date provided in input dataset (PD parameter)
adrs <- tibble::tribble(
  ~STUDYID,       ~USUBJID,      ~PARAMCD, ~AVALC, ~ADT,         ~ANL01FL,
  "CDISCPILOT01", "01-701-1015", "OVR",    "CR",   "2016-01-25", "Y",
  "CDISCPILOT01", "01-701-1015", "OVR",    "SD",   "2016-02-22", NA_character_,
  "CDISCPILOT01", "01-701-1015", "OVR",    "PD",   "2016-02-22", "Y",
  "CDISCPILOT01", "01-701-1015", "BOR",    "CR",   "2016-01-25", "Y",
  "CDISCPILOT01", "01-701-1034", "OVR",    "SD",   "2015-12-07", "Y",
  "CDISCPILOT01", "01-701-1034", "OVR",    "PD",   "2016-04-25", "Y",
  "CDISCPILOT01", "01-701-1034", "OVR",    "PD",   "2016-06-25", "Y",
  "CDISCPILOT01", "01-701-1034", "BOR",    "SD",   "2015-12-07", "Y",
  "CDISCPILOT01", "01-701-1035", "OVR",    "SD",   "2016-04-25", "Y",
  "CDISCPILOT01", "01-701-1035", "OVR",    "PR",   "2016-06-25", "Y",
  "CDISCPILOT01", "01-701-1035", "BOR",    "PR",   "2016-06-25", "Y",
  "CDISCPILOT01", "01-701-1015", "PD",     "Y",    "2016-02-22", "Y",
  "CDISCPILOT01", "01-701-1034", "PD",     "Y",    "2016-04-25", "Y"
) %>% mutate(
  ADT = as_date(ADT)
)

pd <- date_source(
  dataset_name = "adrs",
  date = ADT,
  filter = PARAMCD == "PD"
)

filter_pd(
  dataset = adrs,
  filter = PARAMCD == "OVR" & ANL01FL == "Y",
  source_pd = pd,
  source_datasets = list(adrs = adrs)
)

# Filter OVR records up to first PD, first PD date derived from OVR records
adrs <- tibble::tribble(
  ~STUDYID,       ~USUBJID,      ~PARAMCD, ~AVALC, ~ADT,         ~ANL01FL,
  "CDISCPILOT01", "01-701-1015", "OVR",    "CR",   "2016-01-25", "Y",
  "CDISCPILOT01", "01-701-1015", "OVR",    "SD",   "2016-02-22", NA_character_,
  "CDISCPILOT01", "01-701-1015", "OVR",    "PD",   "2016-02-22", "Y",
  "CDISCPILOT01", "01-701-1015", "BOR",    "CR",   "2016-01-25", "Y",
  "CDISCPILOT01", "01-701-1034", "OVR",    "SD",   "2015-12-07", "Y",
  "CDISCPILOT01", "01-701-1034", "OVR",    "PD",   "2016-04-25", "Y",
  "CDISCPILOT01", "01-701-1034", "OVR",    "PD",   "2016-06-25", "Y",
  "CDISCPILOT01", "01-701-1034", "BOR",    "SD",   "2015-12-07", "Y",
  "CDISCPILOT01", "01-701-1035", "OVR",    "SD",   "2016-04-25", "Y",
  "CDISCPILOT01", "01-701-1035", "OVR",    "PR",   "2016-06-25", "Y",
  "CDISCPILOT01", "01-701-1035", "BOR",    "PR",   "2016-06-25", "Y"
) %>% mutate(
  ADT = as_date(ADT)
)

pd <- date_source(
  dataset_name = "adrs",
  date = ADT,
  filter = PARAMCD == "OVR" & ANL01FL == "Y" & AVALC == "PD"
)

filter_pd(
  dataset = adrs,
  filter = PARAMCD == "OVR" & ANL01FL == "Y",
  source_pd = pd,
  source_datasets = list(adrs = adrs)
)

Get CR Records Followed by PR That Lead to a Prior Error

Description

Get CR Records Followed by PR That Lead to a Prior Error

Usage

get_crpr_dataset()

Details

Some {admiralonco} functions check that in the source records CR is not followed by PR and throw an error otherwise. The get_crpr_dataset() function allows one to retrieve the duplicate records that lead to an error.

Note that the function always returns the dataset of duplicates from the last error that has been thrown in the current R session. Thus, after restarting the R sessions get_crpr_dataset() will return NULL and after a second error has been thrown, the dataset of the first error can no longer be accessed (unless it has been saved in a variable).

Value

A data.frame or NULL

Author(s)

Stefan Bundfuss

See Also

signal_crpr()

Utilities for Dataset Checking: signal_crpr()

Examples

library(tibble)
library(dplyr)
library(lubridate)
library(admiralonco)
library(rlang)

adrs <- tribble(
  ~USUBJID, ~ADTC,        ~AVALC,
  "1",      "2020-01-01", "PR",
  "1",      "2020-02-01", "CR",
  "1",      "2020-02-16", "NE",
  "1",      "2020-03-01", "CR",
  "2",      "2020-02-06", "PR",
  "2",      "2020-02-16", "CR",
  "2",      "2020-03-30", "PR",
) %>%
  mutate(
    ADT = ymd(ADTC),
    STUDYID = "XX1234"
  )

signal_crpr(adrs, order = exprs(ADT))

get_crpr_dataset()

Pre-Defined Response Event Objects

Description

These pre-defined event() and event_joined() objects can be used as input to admiral::derive_extreme_event().

Usage

rsp_y

no_data_n

cb_y

bor_cr

bor_pr

bor_sd

bor_non_crpd

bor_pd

bor_ne

no_data_missing

crsp_y_cr

crsp_y_pr

cbor_cr

cbor_pr

Details

To see the definition of the various objects simply print the object in the R console, e.g. bor_sd. For details of how to use these objects please refer to admiral::derive_extreme_event().

It is assumed that dataset_name = "ovr" refers to the dataset of the only overall response assessments at each visit which should be considered for the parameter derivations. For example the dataset should include only post-baseline assessments up to first PD and before start of anti-cancer therapy.

See Also

admiral::derive_extreme_event(), admiral::event(), admiral::event_joined()

Examples

# This shows the definition of all pre-defined `event` objects that ship
# with {admiralonco}
exports <- sort(getNamespaceExports("admiralonco"))
for (obj_name in exports) {
  obj <- getExportedValue("admiralonco", obj_name)
  if (inherits(obj, "event_def")) {
    cat("\n", obj_name, ":\n", sep = "")
    print(obj, indent = 2)
  }
}

Signal CR Records Followed by PR

Description

Signal CR Records Followed by PR

Usage

signal_crpr(
  dataset,
  order,
  msg = "Dataset contains CR records followed by PR.",
  subject_keys = get_admiral_option("subject_keys"),
  check_type = "warning"
)

Arguments

Value

No return value, called for side effects

Author(s)

Stefan Bundfuss

See Also

get_crpr_dataset()

Utilities for Dataset Checking: get_crpr_dataset()

Examples

library(tibble)
library(dplyr)
library(lubridate)
library(admiralonco)
library(rlang)

adrs <- tribble(
  ~USUBJID, ~ADTC,        ~AVALC,
  "1",      "2020-01-01", "PR",
  "1",      "2020-02-01", "CR",
  "1",      "2020-02-16", "NE",
  "1",      "2020-03-01", "CR",
  "2",      "2020-02-06", "PR",
  "2",      "2020-02-16", "CR",
  "2",      "2020-03-30", "PR",
) %>%
  mutate(
    ADT = ymd(ADTC),
    STUDYID = "XX1234"
  )

signal_crpr(adrs, order = exprs(ADT))