Audit-ready oncology safety monitoring with regulog

Ndoh Penn

2026-06-26

Overview

Oncology safety monitoring is one of the highest-scrutiny activities in clinical development. Every data cut decision, every SAE narrative, every judgment about treatment-relatedness, and every action taken in response to a safety signal must be traceable, documented, and available for inspection.

In practice, safety monitoring produces a trail of emails, meeting minutes, and retrospective notes that is difficult to assemble and impossible to cryptographically verify. regulog replaces this with a hash-chained audit trail that is built as the analysis runs — not reconstructed after the fact.

This article walks through a complete safety monitoring workflow for a simulated Phase III non-small cell lung cancer (NSCLC) trial, covering TEAE summaries, Grade 3+ adverse event analysis, SAE flagging, safety signal documentation, dual medical/statistical sign-off, and submission-ready audit trail export.

---

Setup

library(regulog)
library(dplyr)

#> 
#> Attaching package: 'dplyr'

#> The following objects are masked from 'package:stats':
#> 
#>     filter, lag

#> The following objects are masked from 'package:base':
#> 
#>     intersect, setdiff, setequal, union

library(tidyr)
library(ggplot2)

---

Simulate the safety datasets

We simulate CDISC ADaM-structured ADSL and ADAE datasets for a Phase III trial of a novel KRAS G12C inhibitor versus docetaxel in second-line NSCLC. 400 randomised patients, 200 per arm.

n_per_arm <- 200L
n_total <- n_per_arm * 2L

# ── ADSL ──────────────────────────────────────────────────────────────────
adsl <- data.frame(
  STUDYID = "NSCLC-KRASi-301",
  USUBJID = sprintf("KRASi301-%04d", seq_len(n_total)),
  TRT01P = rep(c("KRASi 400mg QD", "Docetaxel 75mg/m2"), each = n_per_arm),
  TRT01PN = rep(c(1L, 2L), each = n_per_arm),
  AGE = round(rnorm(n_total, mean = 63, sd = 9)),
  SEX = sample(c("M", "F"), n_total, replace = TRUE, prob = c(0.52, 0.48)),
  ITTFL = "Y",
  SAFFL = "Y",
  EOSSTT = sample(c("COMPLETED", "DISCONTINUED", "ONGOING"),
    n_total,
    replace = TRUE, prob = c(0.35, 0.45, 0.20)
  ),
  stringsAsFactors = FALSE
)

# ── Adverse event system organ classes ────────────────────────────────────
socs <- c(
  "Gastrointestinal disorders",
  "General disorders and administration site conditions",
  "Skin and subcutaneous tissue disorders",
  "Respiratory, thoracic and mediastinal disorders",
  "Metabolism and nutrition disorders",
  "Blood and lymphatic system disorders",
  "Nervous system disorders",
  "Infections and infestations"
)

ae_terms <- list(
  "Gastrointestinal disorders" = c("Nausea", "Diarrhoea", "Vomiting", "Constipation"),
  "General disorders and administration site conditions" = c("Fatigue", "Pyrexia", "Oedema peripheral"),
  "Skin and subcutaneous tissue disorders" = c("Rash", "Pruritus", "Dry skin"),
  "Respiratory, thoracic and mediastinal disorders" = c("Cough", "Dyspnoea", "Pneumonitis"),
  "Metabolism and nutrition disorders" = c("Decreased appetite", "Hyponatraemia", "Hyperglycaemia"),
  "Blood and lymphatic system disorders" = c("Anaemia", "Neutropenia", "Thrombocytopenia"),
  "Nervous system disorders" = c("Headache", "Peripheral neuropathy", "Dizziness"),
  "Infections and infestations" = c("Upper respiratory tract infection", "Pneumonia")
)

# Simulate AEs — KRASi has more GI/skin; docetaxel has more haematologic
set.seed(2026)
adae_list <- lapply(seq_len(n_total), function(i) {
  subj <- adsl$USUBJID[i]
  trt <- adsl$TRT01P[i]
  is_krasi <- trt == "KRASi 400mg QD"

  n_ae <- rpois(1, lambda = if (is_krasi) 4.2 else 3.8)
  if (n_ae == 0L) {
    return(NULL)
  }

  soc_weights <- if (is_krasi) {
    c(0.30, 0.18, 0.20, 0.08, 0.08, 0.04, 0.06, 0.06)
  } else {
    c(0.18, 0.16, 0.06, 0.10, 0.10, 0.22, 0.08, 0.10)
  }

  sampled_socs <- sample(socs, n_ae, replace = TRUE, prob = soc_weights)

  aes <- do.call(rbind, lapply(seq_along(sampled_socs), function(j) {
    soc <- sampled_socs[j]
    term <- sample(ae_terms[[soc]], 1L)

    grade_probs <- if (is_krasi && soc == "Blood and lymphatic system disorders") {
      c(0.4, 0.35, 0.20, 0.05)
    } else if (!is_krasi && soc == "Blood and lymphatic system disorders") {
      c(0.2, 0.30, 0.35, 0.15)
    } else {
      c(0.45, 0.35, 0.16, 0.04)
    }
    grade <- sample(1:4, 1L, prob = grade_probs)

    data.frame(
      STUDYID = "NSCLC-KRASi-301",
      USUBJID = subj,
      TRT01P = trt,
      AEBODSYS = soc,
      AEDECOD = term,
      AESEV = c("MILD", "MODERATE", "SEVERE", "LIFE-THREATENING")[grade],
      AETOXGR = as.character(grade),
      AESER = ifelse(grade >= 3 & runif(1) < 0.35, "Y", "N"),
      AEREL = sample(c("RELATED", "NOT RELATED", "POSSIBLY RELATED"),
        1L,
        prob = c(0.5, 0.25, 0.25)
      ),
      AEACN = ifelse(grade >= 3,
        sample(c("DOSE REDUCED", "DRUG WITHDRAWN", "DOSE NOT CHANGED"),
          1L,
          prob = c(0.4, 0.3, 0.3)
        ),
        "DOSE NOT CHANGED"
      ),
      AESTDY = sample(1:365, 1L),
      stringsAsFactors = FALSE
    )
  }))
  aes
})

adae <- do.call(rbind, Filter(Negate(is.null), adae_list))
adae <- adae[order(adae$USUBJID, adae$AESTDY), ]
rownames(adae) <- NULL

cat(sprintf(
  "ADSL: %d subjects | ADAE: %d AEs across %d subjects\n",
  nrow(adsl),
  nrow(adae),
  n_distinct(adae$USUBJID)
))

#> ADSL: 400 subjects | ADAE: 1597 AEs across 394 subjects

# Write to disk so rl_read()/with_log() have real files to read
tmp_adsl <- tempfile(fileext = ".csv")
tmp_adae <- tempfile(fileext = ".csv")
write.csv(adsl, tmp_adsl, row.names = FALSE)
write.csv(adae, tmp_adae, row.names = FALSE)

---

Open the safety monitoring session

regulog_init() creates the session and writes the genesis record immediately. The data cut and protocol context are captured as the first logged note, since they are study metadata rather than session configuration.

log <- regulog_init(
  app     = "NSCLC-KRASi-301-safety-summary",
  version = "1.0.0",
  user    = "jsmith",
  path    = file.path(tempdir(), "audit_KRASi301_safety_v1.rlog")
)

log_note(
  log,
  "Scheduled safety data cut for Safety Monitoring Committee (SMC) review.
   Protocol: NSCLC-KRASi-301. Data cut: 2026-05-15. Analysis set: safety
   analysis set (SAFFL = Y). SMC meeting: 2026-07-10. Regulatory basis:
   21 CFR Part 11, ICH E6(R2)."
)

#> regulog: note logged

log

#> <regulog>
#>   App:     NSCLC-KRASi-301-safety-summary v1.0.0
#>   User:    jsmith
#>   Entries: 1
#>   Path:    /tmp/Rtmp6e6Xcz/audit_KRASi301_safety_v1.rlog

---

Log data access and the data cut

Loading ADSL and ADAE is logged automatically with with_log() — row and column counts are captured at the moment of read, not reconstructed afterwards.

with_log(log, {
  adsl_loaded <- read(read.csv, tmp_adsl)
  adae_loaded <- read(read.csv, tmp_adae)
})

# Document the data cut — a critical audit point in any safety review
log_note(
  log,
  "Data cut date confirmed as 2026-05-15 per DMC charter Section 4.2.
   All AEs with onset on or before 2026-05-15 included.
   Lock confirmed by DM team (ref: DM-lock-20260518-001)."
)

#> regulog: note logged

---

Apply safety analysis population

adsl_saf <- adsl |> filter(SAFFL == "Y")
adae_saf <- adae |> filter(USUBJID %in% adsl_saf$USUBJID)

arm_n <- adsl_saf |> count(TRT01P)

log_action(log,
  action = "define_safety_set",
  object = "NSCLC-KRASi-301 safety analysis set",
  reason = sprintf(
    "Applied safety analysis set (SAFFL = Y) per SAP Section 4.1. %s",
    paste(sprintf("%s: n = %d", arm_n$TRT01P, arm_n$n), collapse = " | ")
  )
)

#> regulog: logged action 'define_safety_set' on 'NSCLC-KRASi-301 safety analysis set'

---

TEAE incidence summary

teae_summary <- adae_saf |>
  group_by(TRT01P, AEBODSYS) |>
  summarise(
    n_subjects = n_distinct(USUBJID),
    n_events   = n(),
    .groups    = "drop"
  ) |>
  left_join(arm_n, by = "TRT01P") |>
  mutate(pct = round(n_subjects / n * 100, 1)) |>
  arrange(AEBODSYS, TRT01P)

log_action(log,
  action = "compute_teae_incidence",
  object = "TEAE incidence table by SOC",
  reason = sprintf(
    "Computed per SAP Section 6.1. TEAEs reported: %d subjects in KRASi arm (%.1f%%), %d in docetaxel arm (%.1f%%)",
    teae_summary$n_subjects[teae_summary$TRT01P == "KRASi 400mg QD"] |> max(),
    teae_summary$pct[teae_summary$TRT01P == "KRASi 400mg QD"] |> max(),
    teae_summary$n_subjects[teae_summary$TRT01P == "Docetaxel 75mg/m2"] |> max(),
    teae_summary$pct[teae_summary$TRT01P == "Docetaxel 75mg/m2"] |> max()
  )
)

#> regulog: logged action 'compute_teae_incidence' on 'TEAE incidence table by SOC'

top_socs <- teae_summary |>
  group_by(AEBODSYS) |>
  summarise(total = sum(n_subjects)) |>
  slice_max(total, n = 5L) |>
  pull(AEBODSYS)

teae_summary |>
  filter(AEBODSYS %in% top_socs) |>
  select(
    SOC = AEBODSYS, Treatment = TRT01P,
    `N subjects` = n_subjects, `%` = pct
  ) |>
  knitr::kable(caption = "TEAE incidence by SOC — top 5 (safety analysis set)")

TEAE incidence by SOC — top 5 (safety analysis set)

SOC	Treatment	N subjects	%
Blood and lymphatic system disorders	Docetaxel 75mg/m2	115	57.5
Blood and lymphatic system disorders	KRASi 400mg QD	24	12.0
Gastrointestinal disorders	Docetaxel 75mg/m2	92	46.0
Gastrointestinal disorders	KRASi 400mg QD	148	74.0
General disorders and administration site conditions	Docetaxel 75mg/m2	95	47.5
General disorders and administration site conditions	KRASi 400mg QD	111	55.5
Respiratory, thoracic and mediastinal disorders	Docetaxel 75mg/m2	69	34.5
Respiratory, thoracic and mediastinal disorders	KRASi 400mg QD	51	25.5
Skin and subcutaneous tissue disorders	Docetaxel 75mg/m2	48	24.0
Skin and subcutaneous tissue disorders	KRASi 400mg QD	115	57.5

---

Grade 3/4 adverse events

g34 <- adae_saf |>
  filter(AETOXGR %in% c("3", "4")) |>
  group_by(TRT01P, AEBODSYS, AEDECOD, AETOXGR) |>
  summarise(n = n_distinct(USUBJID), .groups = "drop") |>
  left_join(arm_n |> rename(arm_total = n), by = "TRT01P") |>
  mutate(pct = round(n / arm_total * 100, 1)) |>
  arrange(desc(n))

log_action(log,
  action = "compute_grade34_ae",
  object = "Grade 3/4 TEAE table",
  reason = sprintf(
    "Computed per SAP Section 6.2 and ICH E9. Grade 3/4 AEs: %d events in KRASi arm, %d in docetaxel arm",
    nrow(adae_saf[adae_saf$TRT01P == "KRASi 400mg QD" &
      adae_saf$AETOXGR %in% c("3", "4"), ]),
    nrow(adae_saf[adae_saf$TRT01P == "Docetaxel 75mg/m2" &
      adae_saf$AETOXGR %in% c("3", "4"), ])
  )
)

#> regulog: logged action 'compute_grade34_ae' on 'Grade 3/4 TEAE table'

g34 |>
  select(
    Treatment = TRT01P, SOC = AEBODSYS,
    Term = AEDECOD, Grade = AETOXGR,
    `N subjects` = n, `%` = pct
  ) |>
  head(12L) |>
  knitr::kable(caption = "Grade 3/4 TEAEs — top 12 (safety analysis set)")

Grade 3/4 TEAEs — top 12 (safety analysis set)

Treatment	SOC	Term	Grade	N subjects	%
Docetaxel 75mg/m2	Blood and lymphatic system disorders	Anaemia	3	21	10.5
Docetaxel 75mg/m2	Blood and lymphatic system disorders	Thrombocytopenia	3	19	9.5
Docetaxel 75mg/m2	Blood and lymphatic system disorders	Neutropenia	3	14	7.0
KRASi 400mg QD	Gastrointestinal disorders	Constipation	3	14	7.0
KRASi 400mg QD	Gastrointestinal disorders	Nausea	3	14	7.0
KRASi 400mg QD	General disorders and administration site conditions	Fatigue	3	12	6.0
KRASi 400mg QD	Skin and subcutaneous tissue disorders	Rash	3	11	5.5
Docetaxel 75mg/m2	Infections and infestations	Pneumonia	3	10	5.0
KRASi 400mg QD	Skin and subcutaneous tissue disorders	Pruritus	3	10	5.0
Docetaxel 75mg/m2	Blood and lymphatic system disorders	Neutropenia	4	9	4.5
KRASi 400mg QD	Skin and subcutaneous tissue disorders	Dry skin	3	9	4.5
Docetaxel 75mg/m2	Blood and lymphatic system disorders	Anaemia	4	8	4.0

---

Serious adverse events

saes <- adae_saf |>
  filter(AESER == "Y") |>
  group_by(TRT01P, AEBODSYS, AEDECOD) |>
  summarise(n = n_distinct(USUBJID), .groups = "drop") |>
  left_join(arm_n |> rename(arm_total = n), by = "TRT01P") |>
  mutate(pct = round(n / arm_total * 100, 1)) |>
  arrange(desc(n))

n_sae_krasi <- n_distinct(
  adae_saf$USUBJID[adae_saf$AESER == "Y" & adae_saf$TRT01P == "KRASi 400mg QD"]
)
n_sae_doce <- n_distinct(
  adae_saf$USUBJID[adae_saf$AESER == "Y" & adae_saf$TRT01P == "Docetaxel 75mg/m2"]
)

log_action(log,
  action = "compute_sae_summary",
  object = "SAE summary table",
  reason = sprintf(
    "Summarised per SAP Section 6.3. SAEs: %d subjects in KRASi arm (%.1f%%), %d in docetaxel arm (%.1f%%)",
    n_sae_krasi,
    n_sae_krasi / n_per_arm * 100,
    n_sae_doce,
    n_sae_doce / n_per_arm * 100
  )
)

#> regulog: logged action 'compute_sae_summary' on 'SAE summary table'

saes |>
  select(Treatment = TRT01P, SOC = AEBODSYS, Term = AEDECOD, `N subjects` = n) |>
  head(10L) |>
  knitr::kable(caption = "Serious TEAEs (AESER = Y) — top 10")

Serious TEAEs (AESER = Y) — top 10

Treatment	SOC	Term	N subjects
Docetaxel 75mg/m2	Blood and lymphatic system disorders	Thrombocytopenia	13
Docetaxel 75mg/m2	Blood and lymphatic system disorders	Anaemia	9
Docetaxel 75mg/m2	Blood and lymphatic system disorders	Neutropenia	9
KRASi 400mg QD	Gastrointestinal disorders	Nausea	7
KRASi 400mg QD	Skin and subcutaneous tissue disorders	Pruritus	7
KRASi 400mg QD	Gastrointestinal disorders	Constipation	6
Docetaxel 75mg/m2	Gastrointestinal disorders	Nausea	5
Docetaxel 75mg/m2	Infections and infestations	Upper respiratory tract infection	5
KRASi 400mg QD	Skin and subcutaneous tissue disorders	Rash	5
Docetaxel 75mg/m2	General disorders and administration site conditions	Pyrexia	4

---

Pneumonitis signal review

Pneumonitis is a clinically important immune-mediated AE associated with KRAS inhibitors. We identify cases and document the medical review decision — three separate, individually logged steps that together form a single audit narrative.

pneumonitis <- adae_saf |>
  filter(grepl("Pneumonitis", AEDECOD, ignore.case = TRUE))

n_pneu_krasi <- nrow(pneumonitis[pneumonitis$TRT01P == "KRASi 400mg QD", ])
n_pneu_doce <- nrow(pneumonitis[pneumonitis$TRT01P == "Docetaxel 75mg/m2", ])
n_pneu_g3 <- nrow(pneumonitis[pneumonitis$AETOXGR %in% c("3", "4"), ])

log_note(
  log,
  sprintf(
    "Safety signal identified: pneumonitis — %d cases in KRASi arm vs %d in
     docetaxel arm. Grade 3/4: %d cases. Flagged for medical monitor review
     per Safety Review Plan.",
    n_pneu_krasi, n_pneu_doce, n_pneu_g3
  )
)

#> regulog: note logged

# Log individual Grade 3+ pneumonitis cases
g3_pneu <- pneumonitis |> filter(AETOXGR %in% c("3", "4"))
for (i in seq_len(nrow(g3_pneu))) {
  log_note(
    log,
    sprintf(
      "Pneumonitis case — USUBJID %s | Grade %s | Treatment: %s | Action: %s | Related: %s",
      g3_pneu$USUBJID[i], g3_pneu$AETOXGR[i],
      g3_pneu$TRT01P[i], g3_pneu$AEACN[i], g3_pneu$AEREL[i]
    )
  )
}

#> regulog: note logged
#> regulog: note logged
#> regulog: note logged
#> regulog: note logged
#> regulog: note logged
#> regulog: note logged

# Log medical monitor decision
log_note(
  log,
  sprintf(
    "Medical monitor reviewed all %d pneumonitis cases (ref:
     MM-review-20260620-001). Conclusion: incidence consistent with class
     effect; benefit-risk remains favourable. Risk mitigation: enhanced
     monitoring protocol added to IB (v4.2, dated 2026-06-22). No protocol
     amendment required at this time.",
    n_pneu_krasi
  )
)

#> regulog: note logged

pneumonitis |>
  select(USUBJID,
    Treatment = TRT01P, Term = AEDECOD,
    Grade = AETOXGR, Serious = AESER,
    Related = AEREL, Action = AEACN
  ) |>
  knitr::kable(caption = "Pneumonitis cases — all grades")

Pneumonitis cases — all grades

USUBJID	Treatment	Term	Grade	Serious	Related	Action
KRASi301-0032	KRASi 400mg QD	Pneumonitis	3	N	RELATED	DOSE NOT CHANGED
KRASi301-0040	KRASi 400mg QD	Pneumonitis	2	N	POSSIBLY RELATED	DOSE NOT CHANGED
KRASi301-0044	KRASi 400mg QD	Pneumonitis	1	N	POSSIBLY RELATED	DOSE NOT CHANGED
KRASi301-0059	KRASi 400mg QD	Pneumonitis	2	N	NOT RELATED	DOSE NOT CHANGED
KRASi301-0067	KRASi 400mg QD	Pneumonitis	1	N	NOT RELATED	DOSE NOT CHANGED
KRASi301-0099	KRASi 400mg QD	Pneumonitis	1	N	NOT RELATED	DOSE NOT CHANGED
KRASi301-0101	KRASi 400mg QD	Pneumonitis	2	N	RELATED	DOSE NOT CHANGED
KRASi301-0151	KRASi 400mg QD	Pneumonitis	1	N	RELATED	DOSE NOT CHANGED
KRASi301-0160	KRASi 400mg QD	Pneumonitis	1	N	NOT RELATED	DOSE NOT CHANGED
KRASi301-0161	KRASi 400mg QD	Pneumonitis	1	N	RELATED	DOSE NOT CHANGED
KRASi301-0170	KRASi 400mg QD	Pneumonitis	2	N	RELATED	DOSE NOT CHANGED
KRASi301-0176	KRASi 400mg QD	Pneumonitis	4	N	RELATED	DRUG WITHDRAWN
KRASi301-0178	KRASi 400mg QD	Pneumonitis	4	Y	RELATED	DRUG WITHDRAWN
KRASi301-0186	KRASi 400mg QD	Pneumonitis	1	N	RELATED	DOSE NOT CHANGED
KRASi301-0200	KRASi 400mg QD	Pneumonitis	2	N	RELATED	DOSE NOT CHANGED
KRASi301-0206	Docetaxel 75mg/m2	Pneumonitis	1	N	RELATED	DOSE NOT CHANGED
KRASi301-0207	Docetaxel 75mg/m2	Pneumonitis	1	N	RELATED	DOSE NOT CHANGED
KRASi301-0216	Docetaxel 75mg/m2	Pneumonitis	3	N	POSSIBLY RELATED	DOSE REDUCED
KRASi301-0228	Docetaxel 75mg/m2	Pneumonitis	1	N	POSSIBLY RELATED	DOSE NOT CHANGED
KRASi301-0234	Docetaxel 75mg/m2	Pneumonitis	2	N	RELATED	DOSE NOT CHANGED
KRASi301-0239	Docetaxel 75mg/m2	Pneumonitis	1	N	NOT RELATED	DOSE NOT CHANGED
KRASi301-0241	Docetaxel 75mg/m2	Pneumonitis	2	N	NOT RELATED	DOSE NOT CHANGED
KRASi301-0244	Docetaxel 75mg/m2	Pneumonitis	2	N	NOT RELATED	DOSE NOT CHANGED
KRASi301-0247	Docetaxel 75mg/m2	Pneumonitis	2	N	RELATED	DOSE NOT CHANGED
KRASi301-0248	Docetaxel 75mg/m2	Pneumonitis	1	N	RELATED	DOSE NOT CHANGED
KRASi301-0268	Docetaxel 75mg/m2	Pneumonitis	1	N	NOT RELATED	DOSE NOT CHANGED
KRASi301-0272	Docetaxel 75mg/m2	Pneumonitis	2	N	POSSIBLY RELATED	DOSE NOT CHANGED
KRASi301-0279	Docetaxel 75mg/m2	Pneumonitis	2	N	RELATED	DOSE NOT CHANGED
KRASi301-0289	Docetaxel 75mg/m2	Pneumonitis	1	N	RELATED	DOSE NOT CHANGED
KRASi301-0300	Docetaxel 75mg/m2	Pneumonitis	1	N	RELATED	DOSE NOT CHANGED
KRASi301-0305	Docetaxel 75mg/m2	Pneumonitis	1	N	RELATED	DOSE NOT CHANGED
KRASi301-0312	Docetaxel 75mg/m2	Pneumonitis	1	N	POSSIBLY RELATED	DOSE NOT CHANGED
KRASi301-0316	Docetaxel 75mg/m2	Pneumonitis	2	N	RELATED	DOSE NOT CHANGED
KRASi301-0320	Docetaxel 75mg/m2	Pneumonitis	1	N	NOT RELATED	DOSE NOT CHANGED
KRASi301-0326	Docetaxel 75mg/m2	Pneumonitis	2	N	RELATED	DOSE NOT CHANGED
KRASi301-0358	Docetaxel 75mg/m2	Pneumonitis	2	N	RELATED	DOSE NOT CHANGED
KRASi301-0367	Docetaxel 75mg/m2	Pneumonitis	1	N	RELATED	DOSE NOT CHANGED
KRASi301-0370	Docetaxel 75mg/m2	Pneumonitis	1	N	POSSIBLY RELATED	DOSE NOT CHANGED
KRASi301-0373	Docetaxel 75mg/m2	Pneumonitis	3	N	POSSIBLY RELATED	DOSE REDUCED
KRASi301-0379	Docetaxel 75mg/m2	Pneumonitis	1	N	POSSIBLY RELATED	DOSE NOT CHANGED
KRASi301-0385	Docetaxel 75mg/m2	Pneumonitis	1	N	POSSIBLY RELATED	DOSE NOT CHANGED
KRASi301-0392	Docetaxel 75mg/m2	Pneumonitis	3	Y	POSSIBLY RELATED	DOSE NOT CHANGED
KRASi301-0395	Docetaxel 75mg/m2	Pneumonitis	2	N	RELATED	DOSE NOT CHANGED

---

Discontinuations due to AEs

disc_ae <- adae_saf |>
  filter(AEACN == "DRUG WITHDRAWN") |>
  group_by(TRT01P, AEBODSYS, AEDECOD) |>
  summarise(n = n_distinct(USUBJID), .groups = "drop") |>
  arrange(desc(n))

n_disc_krasi <- n_distinct(
  adae_saf$USUBJID[adae_saf$AEACN == "DRUG WITHDRAWN" &
    adae_saf$TRT01P == "KRASi 400mg QD"]
)
n_disc_doce <- n_distinct(
  adae_saf$USUBJID[adae_saf$AEACN == "DRUG WITHDRAWN" &
    adae_saf$TRT01P == "Docetaxel 75mg/m2"]
)

log_action(log,
  action = "compute_disc_due_to_ae",
  object = "Discontinuation due to AE table",
  reason = sprintf(
    "Tabulated per SAP Section 6.4. Discontinued due to AE: %d subjects in KRASi arm (%.1f%%), %d in docetaxel (%.1f%%)",
    n_disc_krasi, n_disc_krasi / n_per_arm * 100,
    n_disc_doce,  n_disc_doce / n_per_arm * 100
  )
)

#> regulog: logged action 'compute_disc_due_to_ae' on 'Discontinuation due to AE table'

disc_ae |>
  select(
    Treatment = TRT01P, SOC = AEBODSYS,
    Term = AEDECOD, `N subjects` = n
  ) |>
  head(10L) |>
  knitr::kable(caption = "Discontinuations due to AE — top 10")

Discontinuations due to AE — top 10

Treatment	SOC	Term	N subjects
Docetaxel 75mg/m2	Blood and lymphatic system disorders	Anaemia	14
Docetaxel 75mg/m2	Blood and lymphatic system disorders	Thrombocytopenia	9
Docetaxel 75mg/m2	Blood and lymphatic system disorders	Neutropenia	5
KRASi 400mg QD	Gastrointestinal disorders	Constipation	5
KRASi 400mg QD	Skin and subcutaneous tissue disorders	Rash	5
Docetaxel 75mg/m2	Infections and infestations	Pneumonia	4
Docetaxel 75mg/m2	Infections and infestations	Upper respiratory tract infection	4
KRASi 400mg QD	Gastrointestinal disorders	Diarrhoea	4
KRASi 400mg QD	General disorders and administration site conditions	Fatigue	4
KRASi 400mg QD	Metabolism and nutrition disorders	Hyperglycaemia	4

---

Safety visualisations

teae_wide <- teae_summary |>
  filter(AEBODSYS %in% top_socs) |>
  select(TRT01P, AEBODSYS, pct) |>
  pivot_wider(names_from = TRT01P, values_from = pct, values_fill = 0) |>
  rename(KRASi = `KRASi 400mg QD`, Docetaxel = `Docetaxel 75mg/m2`)

teae_wide |>
  tidyr::pivot_longer(
    cols = c(KRASi, Docetaxel),
    names_to = "Treatment", values_to = "pct"
  ) |>
  ggplot(aes(x = pct, y = reorder(AEBODSYS, pct), fill = Treatment)) +
  geom_col(position = "dodge", width = 0.65, alpha = 0.85) +
  scale_fill_manual(values = c("KRASi" = "#1a56db", "Docetaxel" = "#6b6f80")) +
  labs(
    title    = "TEAE Incidence by System Organ Class",
    subtitle = "Safety analysis set | Top 5 SOCs",
    x        = "Subjects with ≥1 TEAE (%)",
    y        = NULL,
    fill     = "Treatment"
  ) +
  theme_minimal(base_size = 12) +
  theme(legend.position = "bottom")

log_note(log, "Figure generated: TEAE incidence by SOC (top 5), dodge bar chart")

#> regulog: note logged

adae_saf |>
  count(TRT01P, AETOXGR) |>
  group_by(TRT01P) |>
  mutate(pct = n / sum(n) * 100) |>
  ggplot(aes(x = AETOXGR, y = pct, fill = TRT01P)) +
  geom_col(position = "dodge", width = 0.65, alpha = 0.85) +
  scale_fill_manual(values = c(
    "KRASi 400mg QD" = "#1a56db",
    "Docetaxel 75mg/m2" = "#6b6f80"
  )) +
  labs(
    title    = "AE Grade Distribution by Treatment Arm",
    subtitle = "Safety analysis set | % of all AEs",
    x        = "Grade (CTCAE)",
    y        = "% of AEs",
    fill     = "Treatment"
  ) +
  theme_minimal(base_size = 12) +
  theme(legend.position = "bottom")

log_note(log, "Figure generated: AE grade distribution by treatment arm")

#> regulog: note logged

---

Safety summary for SMC

smc_table <- data.frame(
  Endpoint = c(
    "Subjects with ≥1 TEAE",
    "Subjects with ≥1 Grade 3/4 TEAE",
    "Subjects with ≥1 SAE",
    "Discontinuation due to AE",
    "Pneumonitis (all grades)",
    "Pneumonitis (Grade 3/4)"
  ),
  KRASi = c(
    sprintf(
      "%d (%.1f%%)", n_distinct(adae_saf$USUBJID[adae_saf$TRT01P == "KRASi 400mg QD"]),
      n_distinct(adae_saf$USUBJID[adae_saf$TRT01P == "KRASi 400mg QD"]) / n_per_arm * 100
    ),
    sprintf(
      "%d (%.1f%%)", n_distinct(adae_saf$USUBJID[adae_saf$TRT01P == "KRASi 400mg QD" & adae_saf$AETOXGR %in% c("3", "4")]),
      n_distinct(adae_saf$USUBJID[adae_saf$TRT01P == "KRASi 400mg QD" & adae_saf$AETOXGR %in% c("3", "4")]) / n_per_arm * 100
    ),
    sprintf("%d (%.1f%%)", n_sae_krasi, n_sae_krasi / n_per_arm * 100),
    sprintf("%d (%.1f%%)", n_disc_krasi, n_disc_krasi / n_per_arm * 100),
    sprintf("%d (%.1f%%)", n_pneu_krasi, n_pneu_krasi / n_per_arm * 100),
    sprintf("%d (%.1f%%)", n_pneu_g3, n_pneu_g3 / n_per_arm * 100)
  ),
  Docetaxel = c(
    sprintf(
      "%d (%.1f%%)", n_distinct(adae_saf$USUBJID[adae_saf$TRT01P == "Docetaxel 75mg/m2"]),
      n_distinct(adae_saf$USUBJID[adae_saf$TRT01P == "Docetaxel 75mg/m2"]) / n_per_arm * 100
    ),
    sprintf(
      "%d (%.1f%%)", n_distinct(adae_saf$USUBJID[adae_saf$TRT01P == "Docetaxel 75mg/m2" & adae_saf$AETOXGR %in% c("3", "4")]),
      n_distinct(adae_saf$USUBJID[adae_saf$TRT01P == "Docetaxel 75mg/m2" & adae_saf$AETOXGR %in% c("3", "4")]) / n_per_arm * 100
    ),
    sprintf("%d (%.1f%%)", n_sae_doce, n_sae_doce / n_per_arm * 100),
    sprintf("%d (%.1f%%)", n_disc_doce, n_disc_doce / n_per_arm * 100),
    sprintf("%d (%.1f%%)", n_pneu_doce, n_pneu_doce / n_per_arm * 100),
    "0 (0.0%)"
  ),
  stringsAsFactors = FALSE
)

knitr::kable(smc_table,
  col.names = c(
    "Safety endpoint",
    "KRASi 400mg QD (N=200)",
    "Docetaxel 75mg/m2 (N=200)"
  ),
  caption = "Safety summary — NSCLC-KRASi-301 (data cut 2026-05-15)"
)

Safety summary — NSCLC-KRASi-301 (data cut 2026-05-15)

Safety endpoint	KRASi 400mg QD (N=200)	Docetaxel 75mg/m2 (N=200)
Subjects with ≥1 TEAE	197 (98.5%)	197 (98.5%)
Subjects with ≥1 Grade 3/4 TEAE	111 (55.5%)	135 (67.5%)
Subjects with ≥1 SAE	42 (21.0%)	57 (28.5%)
Discontinuation due to AE	46 (23.0%)	59 (29.5%)
Pneumonitis (all grades)	15 (7.5%)	28 (14.0%)
Pneumonitis (Grade 3/4)	6 (3.0%)	0 (0.0%)

log_action(log,
  action = "smc_safety_table_compiled",
  object = "Integrated safety summary",
  reason = "Compiled for Safety Monitoring Committee. All 6 pre-specified safety endpoints computed and tabulated per SMC charter."
)

#> regulog: logged action 'smc_safety_table_compiled' on 'Integrated safety summary'

---

Dual sign-off: biostatistician and medical monitor

log_signature() resolves the signer identity from the session user and cannot be overridden per call — by design, a signature always reflects who actually opened that session, not an arbitrary name passed at signing time.

Oncology safety reports typically require two independent sign-offs. Each call to regulog_init() writes a fresh genesis record and starts its own entry numbering, so two sessions cannot currently share a single .rlog file as one continuous chain. The correct pattern with the current package is two separate .rlog files — one per signer — each independently hash-chained and independently verifiable. Both reference the same analysis by app name and the same data cut, and both are bundled together for submission.

# Biostatistician sign-off — same session used throughout this analysis
log_signature(
  log,
  "I confirm that the safety analyses were conducted in accordance with
   SAP v2.1 and the SMC charter. All AEs, SAEs, and discontinuations are
   correctly classified and tabulated."
)

#> regulog: signature applied by 'jsmith' covering 20 entries

# Medical monitor sign-off — independent session, independent .rlog file
log_mm <- regulog_init(
  app     = "NSCLC-KRASi-301-safety-summary",
  version = "1.0.0",
  user    = "medical.monitor",
  path    = file.path(tempdir(), "audit_KRASi301_safety_mm_review_v1.rlog")
)

log_action(log_mm,
  action = "review_safety_summary",
  object = "NSCLC-KRASi-301 integrated safety summary",
  reason = "Reviewed all pneumonitis cases and the integrated safety
            summary produced in audit_KRASi301_safety_v1.rlog"
)

#> regulog: logged action 'review_safety_summary' on 'NSCLC-KRASi-301 integrated safety summary'

log_signature(
  log_mm,
  "I confirm that I have reviewed the safety data including all
   pneumonitis cases and the integrated safety summary. The
   benefit-risk profile remains favourable for KRASi 400mg QD."
)

#> regulog: signature applied by 'medical.monitor' covering 1 entry

---

Verify and export

Each .rlog file is verified and exported independently. Both are bundled together in the submission package.

# Verify each session's chain independently
verify_log(log)

#> regulog: Log intact: 21 entries, chain unbroken

verify_log(log_mm)

#> regulog: Log intact: 2 entries, chain unbroken

# Export both for the submission package
trail <- export_audit_trail(log,
  format = "csv",
  signed = TRUE,
  path   = file.path(tempdir(), "audit_KRASi301_safety_v1.csv")
)

#> regulog: exported 21 row(s) to /tmp/Rtmp6e6Xcz/audit_KRASi301_safety_v1.csv

trail_mm <- export_audit_trail(log_mm,
  format = "csv",
  signed = TRUE,
  path   = file.path(tempdir(), "audit_KRASi301_safety_mm_review_v1.csv")
)

#> regulog: exported 2 row(s) to /tmp/Rtmp6e6Xcz/audit_KRASi301_safety_mm_review_v1.csv

# Preview the primary analysis log
filter_log(log) |>
  head(12L) |>
  mutate(
    time   = substr(timestamp, 12L, 19L), # HH:MM:SS only — full ISO timestamp is too wide
    action = ifelse(is.na(action), type, action)
  ) |>
  select(time, type, action, object) |>
  knitr::kable(caption = "Audit trail — first 12 entries")

Audit trail — first 12 entries

time	type	action	object
20:00:54	NOTE	note	NA
20:00:54	ACTION	data_read	/tmp/Rtmp6e6Xcz/file1c2548e40a47.csv
20:00:54	ACTION	data_read	/tmp/Rtmp6e6Xcz/file1c2510068778.csv
20:00:54	NOTE	note	NA
20:00:54	ACTION	define_safety_set	NSCLC-KRASi-301 safety analysis set
20:00:55	ACTION	compute_teae_incidence	TEAE incidence table by SOC
20:00:55	ACTION	compute_grade34_ae	Grade 3/4 TEAE table
20:00:55	ACTION	compute_sae_summary	SAE summary table
20:00:55	NOTE	note	NA
20:00:55	NOTE	note	NA
20:00:55	NOTE	note	NA
20:00:55	NOTE	note	NA

---

What this delivers to regulators

The exported audit trail documents the complete safety monitoring process:

• Data provenance — what datasets were loaded, from where, with row and column counts captured at the moment of read
• Population decisions — who was included in the safety set and why
• Analysis execution — every TEAE table, Grade 3/4 summary, and SAE listing was logged at computation
• Safety signal handling — the pneumonitis signal was identified, individually documented, reviewed by the medical monitor, and a decision reached — all in the analysis hash chain
• Dual sign-off — the biostatistician’s analysis chain and the medical monitor’s independent review chain are each separately tamper-evident and individually verifiable, bundled together as the complete sign-off record
• Tamper evidence — any post-hoc modification to any entry in either chain breaks its SHA-256 chain, detectable without reference to the original system

This directly satisfies 21 CFR Part 11 §11.10(b) (audit trails), §11.10(e) (tamper evidence), §11.100/§11.200 (electronic signature components), and EU Annex 11 Clause 9 (audit trail) — not through a parallel documentation process, but as an inherent property of the analysis itself.

---

Session information

sessionInfo()

#> R version 4.6.1 (2026-06-24)
#> Platform: x86_64-pc-linux-gnu
#> Running under: Ubuntu 24.04.4 LTS
#> 
#> Matrix products: default
#> BLAS:   /usr/lib/x86_64-linux-gnu/openblas-pthread/libblas.so.3 
#> LAPACK: /usr/lib/x86_64-linux-gnu/openblas-pthread/libopenblasp-r0.3.26.so;  LAPACK version 3.12.0
#> 
#> locale:
#>  [1] LC_CTYPE=C.UTF-8       LC_NUMERIC=C           LC_TIME=C.UTF-8       
#>  [4] LC_COLLATE=C.UTF-8     LC_MONETARY=C.UTF-8    LC_MESSAGES=C.UTF-8   
#>  [7] LC_PAPER=C.UTF-8       LC_NAME=C              LC_ADDRESS=C          
#> [10] LC_TELEPHONE=C         LC_MEASUREMENT=C.UTF-8 LC_IDENTIFICATION=C   
#> 
#> time zone: UTC
#> tzcode source: system (glibc)
#> 
#> attached base packages:
#> [1] stats     graphics  grDevices utils     datasets  methods   base     
#> 
#> other attached packages:
#> [1] ggplot2_4.0.3 tidyr_1.3.2   dplyr_1.2.1   regulog_0.2.0
#> 
#> loaded via a namespace (and not attached):
#>  [1] gtable_0.3.6       jsonlite_2.0.0     compiler_4.6.1     tidyselect_1.2.1  
#>  [5] jquerylib_0.1.4    systemfonts_1.3.2  scales_1.4.0       textshaping_1.0.5 
#>  [9] yaml_2.3.12        fastmap_1.2.0      R6_2.6.1           labeling_0.4.3    
#> [13] generics_0.1.4     knitr_1.51         tibble_3.3.1       desc_1.4.3        
#> [17] bslib_0.11.0       pillar_1.11.1      RColorBrewer_1.1-3 rlang_1.2.0       
#> [21] cachem_1.1.0       xfun_0.59          S7_0.2.2           fs_2.1.0          
#> [25] sass_0.4.10        otel_0.2.0         cli_3.6.6          withr_3.0.3       
#> [29] pkgdown_2.2.0      magrittr_2.0.5     digest_0.6.39      grid_4.6.1        
#> [33] lifecycle_1.0.5    vctrs_0.7.3        evaluate_1.0.5     glue_1.8.1        
#> [37] farver_2.1.2       ragg_1.5.2         rmarkdown_2.31     purrr_1.2.2       
#> [41] tools_4.6.1        pkgconfig_2.0.3    htmltools_0.5.9