EMBRACE-II Variable Description • embraceR

Version Date: March 21, 2025
Author: Stefan Ecker, PhD

Quick Reference: All Added Variables EMBRACE-II

This document describes the custom variables added to the EMBRACE-II dataset during data processing. Variables are organized by functional categories.

1. Patient and Study Information

Variable	Details
`age`	Patient age at diagnosis Calculation: `histology_assessment_date` `year_of_birth` Implementation: `add_age()` Unit: Years

`included_in_study`	Indicates whether a patient is included in the final study cohort Calculation: Based on the study’s inclusion/exclusion criteria documented in the official study flowchart Implementation: `load_embrace_ii()` Values: TRUE = included in study, FALSE = excluded from study

`withdrew_consent`	Indicates if a patient withdrew consent from the study Calculation: Based on `followup_m` columns Implementation:* `add_lost_to_fu()` Values: TRUE = consent withdrawn, FALSE = consent maintained

`is_lost_to_fu`	Indicates if a patient was lost to follow-up during the study Calculation: Based on `followup_m` columns Implementation:* `add_lost_to_fu()` Values: TRUE = patient lost to followup, FALSE = patient not lost to followup

2. Diagnostic Measurements

Variable	Details
`max_tumor_dimension_` `sta_d`	Maximum tumor dimension at diagnosis Calculation: Overall maximum value across both MRI and gynecological exam measurements (`mri_tumor__sta_d` and `gyn_tumor__sta_d`) Implementation: `add_max_tumor_dimension()` Unit: Millimeters

`gyn_max_parametrium_` `sta_d`	Maximum parametrial involvement from gynecological examination at diagnosis Calculation: Maximum value between left and right parametrial involvement (`gyn_left_parametrium_sta_d` and `gyn_right_parametrium_sta_d`) Implementation: `add_parametrial_involvement()` Values: 0 = Not involved, 1 = Proximal, 2 = Distal, 3 = To pelvic wall

`gyn_max_tumor_` `dimension_sta_d`	Maximum tumor dimension from gynecological examination Calculation: Maximum value across all gynecological tumor measurements (width, height, thickness) Implementation: `add_max_tumor_dimension()` Unit: Millimeters

`mri_max_parametrium_` `sta_d`	Maximum parametrial involvement from MRI at diagnosis Calculation: Maximum value between left and right parametrial involvement (`mri_left_parametrium_sta_d` and `mri_right_parametrium_sta_d`) Implementation: `add_parametrial_involvement()` Values: 0 = Not involved, 1 = Proximal, 2 = Distal, 3 = To pelvic wall

`mri_max_tumor_` `dimension_sta_d`	Maximum tumor dimension from MRI Calculation: Maximum value across all MRI tumor measurements (width, height, thickness) Implementation: `add_max_tumor_dimension()` Unit: Millimeters

3. Brachytherapy Treatment and Technique

Variable	Details
`icis`	Intracavitary/Interstitial classification Calculation: Based on `fraction__technique_tdvh` columns Implementation:* `add_icis()` Values: Binary (1 = ICIS, 0 = IC) Note: Derived from technique values across fractions

`icis_parallel_oblique`	Detailed ICIS needle classification Calculation: Based on `fraction__technique_tdvh` columns Implementation:* `add_parallel_oblique_needles()` Values: “no needles”, “parallel needles”, “oblique needles”, “other”

`average_nr_active_` `needles`	Average number of active needles Calculation: Mean of `fraction__active_needles` columns Implementation:* `add_average_active_needles()` Unit: Number of needles

`time_to_bt`	Time from EBRT start to first BT fraction Calculation: `fraction01date_tdvh` - `ebrt_start_date_tdvh` Implementation: `add_time_to_bt()` Unit: Days

`time_to_bt_percent`	Relative timing of brachytherapy Calculation: `time_to_bt` / `ott_ebrt` Implementation: `add_time_to_bt()` Unit: Proportion

`ott_ebrt`	Overall treatment time of EBRT Calculation: `ebrt_end_date_tdvh` - `ebrt_start_date_tdvh` Implementation: `add_time_to_bt()` Unit: Days

`ott`	Overall Treatment Time from start of EBRT to last treatment Calculation: Time interval between `ebrt_start_date_tdvh` and `last_treatment_date` (which is the latest of EBRT end date and all fraction dates) Implementation: `emii_add_ott()` Unit: Days

`last_treatment_date`	Date of the last treatment received Calculation: Maximum date among EBRT end date (`ebrt_end_date_tdvh`) and all brachytherapy fraction dates (`fractiondate_tdvh`) Implementation:* `emii_add_ott()` Unit: Date

`fraction01hrctv_` `volume_bins`	CTV-HR volume categories Calculation: Based on `fraction01hrctv_volume_tdvh` Implementation: `add_hrctv_volume_bins()` Values: 0: < 30 cm³, 1: 30-45 cm³, 2: > 45 cm³, NA: missing

`trak_total_sum`	Total Reference Air Kerma (TRAK) summed across all brachytherapy fractions Calculation: Sum of all `fractiontrak_tdvh` values Implementation:* `emii_add_trak_absolute()` Unit: cGy at 1m

`trak_needles_sum`	Total TRAK contribution from needles across all brachytherapy fractions Calculation: Sum of `fraction_needles_abs` values, which are derived from percentage contributions (`trak trak_needles_pct / 100`) Implementation: `emii_add_trak_absolute()` Unit: cGy at 1m

`trak_vaginal_` `applicator_sum`	Total TRAK contribution from vaginal applicator across all fractions Calculation: Sum of `fraction_vaginal_applicator_abs` values, which are derived from percentage contributions (`trak trak_vaginal_applicator_pct / 100`) Implementation: `emii_add_trak_absolute()` Unit: cGy at 1m

`trak_tandem_` `applicator_sum`	Total TRAK contribution from tandem applicator across all fractions Calculation: Sum of `fraction_tandem_applicator_abs` values, which are derived from percentage contributions (`trak trak_tandem_applicator_pct / 100`) Implementation: `emii_add_trak_absolute()` Unit: cGy at 1m

`imp1gyn_max_` `parametrium_sta_b`	Maximum parametrial involvement from gynecological examination at first implant Calculation: Maximum value between left and right parametrial involvement (`imp1gyn_left_parametrium_sta_b` and `imp1gyn_right_parametrium_sta_b`) Implementation: `add_parametrial_involvement()` Values: 0 = Not involved, 1 = Proximal, 2 = Distal, 3 = To pelvic wall

`imp1image_max_` `parametrium_sta_b`	Maximum parametrial involvement from imaging at first implant Calculation: Maximum value between left and right parametrial involvement (`imp1image_left_parametrium_sta_b` and `imp1imaging_right_parametrium_sta_b`) Implementation: `add_parametrial_involvement()` Values: 0 = Not involved, 1 = Proximal, 2 = Distal, 3 = To pelvic wall

4. External Beam Radiotherapy (EBRT)

Variable	Details
`ebrt_elective_target_` `algorithm`	Algorithm-recommended treatment target volume for elective nodal irradiation Calculation: Determined based on EMBRACE-II protocol Implementation: `emii_add_elective_targets()` Values: ‘Small Pelvis’, ‘Large Pelvis’, ‘Large Pelvis + Para-aortic’, ‘Large Pelvis + Inguinal’, ‘Large Pelvis + Para-aortic + Inguinal’

`elective_high_risk`	Indicates if high-risk criteria for nodal treatment are met Calculation: TRUE if pathological nodes present AND >= 3 nodes found, OR if any common iliac or para-aortic nodes present Implementation: `emii_add_elective_targets()` Values: TRUE/FALSE

`elective_inguinal`	Indicates if inguinal node irradiation criteria are met Calculation: TRUE if maximum vaginal involvement = 3 (distal third) OR any inguinal nodes present in right or left groin Implementation: `emii_add_elective_targets()` Values: TRUE/FALSE

`elective_low_risk`	Indicates if low-risk criteria for nodal treatment are met Calculation: TRUE if NO pathological nodes, tumor <= 40mm, TNM stage 1/2/3/5, squamous cell histology, and no corpus uteri involvement on MRI Implementation: `emii_add_elective_targets()` Values: TRUE/FALSE

`any_node_ci_pa`	Indicates presence of any common iliac or para-aortic lymph nodes at diagnosis Calculation: TRUE if any node has anatomical position of “R com iliac”, “L com iliac”, or “Para Aortic” Implementation: `emii_add_elective_targets()` Values: TRUE/FALSE

5. Lymph Nodes and Metastases

Variable	Details
`nodal_classification`	Classification of nodal involvement status Calculation: Based on pathological nodes present (`pathological_nodes_present`) and Para-Aortic lymph node involvement at diagnosis Implementation: `emii_add_nodal_classification()` Values: ‘N0’ (no pathological nodes), ‘N1PAN’ (pathological nodes with Para-Aortic involvement), ‘N1pelvic’ (pathological nodes without Para-Aortic involvement)

`number_common_iliac_ln_` `stat_d`	Count of common iliac lymph nodes at diag. Calculation: Sum of left and right common iliac lymph nodes from staging data Implementation: `emii_add_number_common_iliac_ln_stat_d()` Unit: Count

`number_paraaortic_ln_` `stat_d`	Count of para-aortic lymph nodes at diag. Calculation: Count of lymph nodes with “Para Aortic” anatomical position from staging data Implementation: `add_number_paraaortic_ln_stat_d()` Unit: Count

`has_*_followup`	Indicates lymph node involvement during follow-up by specific anatomical location Calculation: Checks if any lymph node documented during follow-up has this specific anatomical position Implementation: `emii_add_recurrent_nodes()` Example variables: `has_L.ext.iliac_followup`, `has_Para.Aortic_followup`, etc. Values: TRUE/FALSE

`has_*_diagnosis`	Indicates lymph node involvement at diagnosis by specific anatomical location Calculation: Checks if any lymph node documented at diagnosis has this specific anatomical position Implementation: `emii_add_diagnostic_nodes()` Example variables: `has_L.ext.iliac_diagnosis`, `has_Para.Aortic_diagnosis`, etc. Values: TRUE/FALSE

`has_*_metastases`	Indicates presence of specific metastatic disease sites Calculation: Checks if any follow-up visit reports metastasis at this site Implementation: `add_metastases()` Example variables: `has_liver_metastases`, `has_bone_metastases`, `has_brain_metastases`, `has_lung_metastases`, `has_other_metastases` Values: TRUE/FALSE

`has_*_nodes`	Indicates presence of distant nodal metastases Calculation: Checks if any follow-up visit reports nodes at these sites Implementation: `add_metastases()` Example variables: `has_paraaortic_nodes_above_l2`, `has_supraclavicular_nodes`, `has_mediastinal_nodes` Values: TRUE/FALSE

`has_abdominal_` `carcinomatosis`	Indicates presence of abdominal carcinomatosis Calculation: Checks if any follow-up visit reports abdominal carcinomatosis Implementation: `add_metastases()` Values: TRUE/FALSE

6. Event Endpoints

Variable	Endpoint	Details
`event_localfailure`	Local Control	Calculation: Triggered when any followup `disease_local_status_` column has a value of 2, indicating recurrence/progression Implementation:* `add_local_failure_event()` Values: 1 = failure (event occurred), 0 = no failure (no event)

`event_nodalfailure`	Nodal Control (lower PAO)	Calculation: Triggered when any followup `disease_nodal_status_` column has a value of 2, indicating nodal recurrence or progression Implementation:* `add_nodal_failure_event()` Values: 1 = failure (event occurred), 0 = no failure (no event)

`event_nodalcontrol_` `incl_pao`	Nodal control (incl. all PAO)	Calculation: Triggered when either pelvic nodal failure (`event_pelvic_nodal`) OR para-aortic nodal failure (`event_paraaortic_nodal`) occurs Implementation: `emii_add_nodalcontrol_incl_pao()` Values: 1 = failure (event occurred), 0 = no failure (no event)

`event_systemicfailure`	Systemic Control (incl. upper PAO)	Calculation: Triggered when any followup `disease_systemic_status_` column has a value of 2, indicating systemic recurrence or progression Implementation:* `add_systemic_failure_event()` Values: 1 = failure (event occurred), 0 = no failure (no event)

`event_systemic_excl_pao`	Systemic control (excl. PAO)	Calculation: Triggered when systemic failure (`event_systemicfailure`) occurs AND at least one of: supraclavicular nodes, mediastinal nodes, liver, bone, brain, lung metastases, abdominal carcinomatosis, or other metastases is present Implementation: `emii_add_systemic_excl_pao()` Values: 1 = failure (event occurred), 0 = no failure (no event)

Variable	Endpoint	Details
`event_vitalstatus`	Overall Survival	Calculation: Uses the value from `vital_status` when available; otherwise sets to 0 (alive) if a disease assessment date exists (`latest_assessment_date_ disease`) Implementation: `add_vitalstatus_event()` Values: 1 = death (event occurred), 0 = alive (no event)

`event_pelvic_nodal`	Pelvic Nodal control	Calculation: Triggered when nodal failure (`event_nodalfailure`) occurs AND at least one node is positive in any pelvic location: external iliac, internal iliac, common iliac, or parametrial/paracervical regions (left or right) Implementation: `emii_add_pelvic_nodal_event()` Values: 1 = failure (event occurred), 0 = no failure (no event)

`event_pelvic`	Pelvic control (local + nodal)	Calculation: Triggered when either local failure (`event_localfailure`) OR pelvic nodal failure (`event_pelvic_nodal`) occurs Implementation: `emii_add_pelvic_event()` Values: 1 = failure (event occurred), 0 = no failure (no event)

`event_paraaortic_nodal`	Para-aortic nodal control (below L2)	Calculation: Triggered when either para-aortic nodes above L2 (`has_paraaortic_nodes_above_l2`) OR para-aortic nodes below L2 (`has_Para.Aortic_followup`) is detected Implementation: `emii_add_paraaortic_nodal()` Values: 1 = failure (event occurred), 0 = no failure (no event)

`event_locoregional`	Locoregional (Pelvic/PAO)	Calculation: Triggered when either local failure (`event_localfailure`) OR nodal control including para-aortic (`event_nodalcontrol_incl_pao`) occurs Implementation: `add_locoregional_event()` Values: 1 = failure (event occurred), 0 = no failure (no event)

Variable	Endpoint	Details
`event_locoregional_` `alone`	Locoregional alone (no metastases)	Calculation: Triggered when a locoregional failure occurs (`event_locoregional=1`) WITHOUT any concurrent distant metastases (`event_systemic_excl_pao=0`) Implementation: `add_locoregional_event()` Values: 1 = isolated locoregional recurrence, 0 = no recurrence or recurrence with distant metastases

`event_cancer_specific`	Cancer-Specific Survival	Calculation: Triggered when death occurs (`event_vitalstatus=1`) AND the main cause of death is primary cancer (codes 1 or 3 in `vital_status_cause_of_death_vital_` `status`) Implementation: `emii_add_cancer_specific()` Values: 1 = cancer death, 0 = alive or non-cancer death

`event_disease_control`	Disease Control	Calculation: Triggered when any failure occurs at any site: local failure OR nodal failure OR systemic failure Implementation: `emii_add_disease_control()` Values: 1 = any disease failure, 0 = disease controlled at all sites

`event_progression_free`	Progression-Free Survival	Calculation: Triggered when any disease event (`event_disease_control=1`) OR death occurs (`event_vitalstatus=1`) Implementation: `emii_add_progression_free_survival()` Values: 1 = progression or death, 0 = alive without progression

`event_distant_alone`	Distant Metastases Alone	Calculation: Triggered when systemic failure occurs (`event_systemicfailure=1`) WITHOUT para-aortic nodes above L2 AND WITHOUT concurrent local failure or pelvic nodal failure Implementation: `emii_add_distant_alone()` Values: 1 = isolated distant metastases, 0 = no distant metastases or concurrent locoregional disease

Variable	Endpoint	Details
`timetoevent_disease`	Time to Disease Event	Calculation: Time interval in months from histology assessment date to either the first disease event date or the latest assessment date if no event occurred Implementation: `add_time_to_diseaseevent()` Values: Number of months

`timetoevent_vitalstatus`	Time to Last Vital Status	Calculation: Time interval in months from histology assessment date to the latest vital status date Implementation: `add_time_to_last_vitalstatus()` Values: Number of months

`latest_vital_status_` `date`	Latest Vital Status Date	Description: The date of the latest available information on vital status, calculated from assessment dates, last info date, and date of death when available Implementation: `add_time_to_last_vitalstatus()` Values: Date format

7. Late Toxicity and Morbidity (only followup assessments)

Variable	Details
`max_value_*`	Maximum grade of a specific toxicity or morbidity across all follow-up time points Calculation: Maximum CTCAE grade value observed for the specific toxicity across all follow-up assessments Implementation: `get_max_side_effects()` Values: 0-5 (CTCAE grade scale) Example variables: `max_value_bladder_cystitis` `max_value_vagina_stenosis` `max_value_fistula` `max_value_gastro_proctitis`

`max_timepoint_*`	Follow-up timepoint at which the maximum grade of toxicity was observed Calculation: Time point (e.g., “3m” for 3 months) at which maximum CTCAE grade was observed for each specific toxicity Implementation: `get_max_side_effects()` Values: Time point indicators (e.g., “3m”, “6m”, “12m”, etc.) Example variables: `max_timepoint_bladder_cystitis` `max_timepoint_vagina_stenosis` `max_timepoint_fistula` `max_timepoint_gastro_proctitis`

`overall_max_morbidity_` `grade`	Overall maximum morbidity grade across all toxicity endpoints and all time points Calculation: The maximum value across all `max_value_` variables for each patient Implementation:* `emii_add_max_morbidity()` Values: 0-5 (CTCAE grade scale) Purpose: Represents the highest toxicity experienced by a patient regardless of specific type or timepoint

The toxicity and morbidity data include various organ systems and toxicity types:

Bladder toxicity: Cystitis, frequency, incontinence, urgency, stenosis/stricture, bleeding
Gastrointestinal toxicity: Abdominal pain/cramping, constipation, diarrhea, incontinence, proctitis, bleeding (at different sites), stenosis/stricture
Vaginal toxicity: Bleeding, discharge, dryness, mucositis, stenosis
Fistula: Grading and localization (bladder, rectum, vagina, etc.)
Lymphatic complications: Edema (limb, trunk/genital), lymphocele, thromboembolic events
Muscle/skeletal toxicity: Fibrosis, fracture, back/pelvic pain
Other toxicities: Fatigue, hot flashes, insomnia, and other patient-specific issues