Retatrutide in Adults with Obesity and Type 2 Diabetes (TRIUMPH-2)
Peer-reviewed phase 3 trial; Lancet 2026, 2026 · paper · Clinical trials
A phase 3 randomized trial (TRIUMPH-2) of retatrutide versus placebo in 1152 adults with obesity and type 2 diabetes. At 80 weeks, body weight fell 11.9–18.8% with retatrutide versus 5.1% with placebo; gastrointestinal side effects were more common.
6 findings
Substance Diarrhoea in the retatrutide 9 mg group is printed as "95 (34%)" / "95 [34%] of 284" in Table 3, the abstract and the Results; 95/284 is 33·5%, which rounds to 33%.
- p. 14 Table 3, Diarrhoea row
95 (34%)
- p. 1 Abstract Findings
95 [34%] of 284 in the 9 mg group
- p. 15 Results text
95 [34%] of 284 in the
- p. 1
95 [34%] of 284
Judge's note: Checked the Table 3 crop and the abstract and page 15 text crops. The 9 mg safety denominator is 284 (column header and text). 95/284 = 33·45%, not a rounding tie. The neighbouring 96/286 = 33·6% correctly rounds to 34%, so the two groups printing the same 34% is not supported by the counts.
Second check: another model agreed, code check
Report a problem with this finding
Substance Vomiting in the retatrutide 4 mg group is printed as "16 (6%)" in Table 3 and "16 [6%]" in the Results; 16/292 is 5·5%, which rounds to 5%.
- p. 14 Table 3, Vomiting row
16 (6%)
- p. 16 Results text
in 16 [6%], 29 [10%], and 45 [16%] vs 12 [4%]
Judge's note: Checked the Table 3 crop and the page 16 text crop. The 4 mg denominator is 292. 16/292 = 5·48%, not a tie. The other vomiting cells round correctly: 29/284 = 10·2%, 45/286 = 15·7%, 12/287 = 4·2%.
Second check: another model agreed
Report a problem with this finding
Substance In Table 3, 13 events in the 4 mg group (n=292) are printed as "13 (5%)" for abdominal distension, fatigue and eructation but as "13 (4%)" for dysesthesia; 13/292 is 4·5%, which rounds to 4%.
- p. 14 Table 3, Abdominal distension, Fatigue and Eructation rows, 4 mg column
13 (5%)
- p. 15 Table 3, Dysesthesia row, 4 mg column
13 (4%)
Judge's note: Checked both Table 3 crops. The denominator is 292 throughout the 4 mg column, so one count cannot round to two values. 13/292 = 4·45%, not a tie. Placebo 13/287 = 4·53% correctly appears as 5%, which may explain the slip.
Second check: another model agreed
Report a problem with this finding
Reference or label Table 1 footnotes define 'Weight gain†' as "use of an SGLT2 inhibitor alone or in combination..." and 'Weight loss‡' as "Use of a sulfonylurea alone or thiazolidinedione alone...". The category labels appear swapped.
- p. 8 Table 1 footnotes † and ‡
†Included use of an SGLT2 inhibitor alone or in combination with any other glucose-lowering agents
Only the authors can say what is right here.
Judge's note: Rests on outside knowledge: SGLT2 inhibitors are associated with weight loss, and sulfonylureas and thiazolidinediones with weight gain. The paper stratifies by each medication's potential effect on bodyweight (page 3), so the footnotes appear attached to the wrong rows.
Second check: another model agreed
Report a problem with this finding
Reference or label REDEFINE-2 HbA1c result ("mean baseline HbA1c of 8·0%.22") is cited to ref 22, the AACE consensus statement (Nadolsky et al), rather than ref 20 (Davies MJ et al, Cagrilintide–semaglutide in adults with overweight or obesity and type 2 diabetes).
- p. 17 Discussion
mean baseline HbA1c of 8·0%.22
- p. 18 Reference 22
Nadolsky K, Garvey WT, Agarwal M, et al. American Association of Clinical Endocrinology Consensus Statement
- p. 18 Reference list, ref 20 and ref 22
Cagrilintide–semaglutide in adults with overweight or obesity and type 2 diabetes
- p. 18 Reference list, entries 20 and 22
Davies MJ, Bajaj HS, Broholm C, et al. Cagrilintide–semaglutide in adults with overweight or obesity and type 2 diabetes.
Judge's note: Viewed the p17 sentence and the p18 reference list: ref 22 is the AACE consensus algorithm, and ref 20 is the cagrilintide–semaglutide T2D trial. The paper elsewhere cites REDEFINE trials as refs 9 and 20.
Second check: another model agreed
Report a problem with this finding
Reference or label SURMOUNT-2 HbA1c result "tirzepatide reduced HbA1c by up to 2·1%.16" is cited to ref 16 (Wharton et al, semaglutide 7·2 mg STEP UP) instead of ref 17 (Garvey et al, SURMOUNT-2).
- p. 17 Discussion
tirzepatide reduced HbA1c by up to 2·1%.16
- p. 18 Reference 16
Once-weekly semaglutide 7·2 mg in adults with obesity (STEP UP)
- p. 18 Reference list, entries 16 and 17
Garvey WT, Frias JP, Jastreboff AM, et al. Tirzepatide once weekly for the treatment of obesity in people with type 2 diabetes (SURMOUNT-2)
- p. 17 Discussion, HbA1c comparison paragraph, last sentence
tirzepatide reduced HbA1c by up to 2·1%.
Judge's note: Viewed the p17 sentence and the p18 reference list: ref 16 is STEP UP and ref 17 is SURMOUNT-2. The Discussion elsewhere cites SURMOUNT-2 as ref 17 ('...by bariatric procedures.17').
Second check: another model agreed
Report a problem with this finding
16 minor (typos and the like; most checked by the judge only)
Wording Abstract Findings describe "672 [58%] of White ethnicity", but Table 1 reports White under Race; ethnicity is reported separately as Hispanic or Latino versus not.
- p. 1 Abstract Findings
672 [58%] of White ethnicity
Judge's note: Checked the abstract crop and Table 1 on page 7. White appears under 'Race', and 'Ethnicity' is a separate block. The abstract wording mislabels a race category as an ethnicity.
Second check: another model agreed
Report a problem with this finding
Wording Figure 1 placebo arm: "287 received study drug", "62 discontinued study drug" and "227 completed study treatment" do not reconcile, since 287 − 62 = 225. The figure 227 equals 289 assigned − 62.
- p. 6 Figure 1, placebo column
62 discontinued study drug
- p. 6 Figure 1, placebo column
227 completed study
Only the authors can say what is right here.
Judge's note: The listed discontinuation reasons sum to 62 and include no 'did not receive' category. The 4 mg and 9 mg arms reconcile from their 'received' counts. The text total of 965 completers uses 227. Only the authors can say which count is right.
Second check: another model agreed
Report a problem with this finding
Wording Figure 1 retatrutide 12 mg arm: "286 received study drug", "40 discontinued study drug" and "247 completed study treatment" do not reconcile, since 286 − 40 = 246.
- p. 6 Figure 1, retatrutide 12 mg column
40 discontinued study drug
- p. 6 Figure 1, retatrutide 12 mg column
247 completed study
Only the authors can say what is right here.
Judge's note: The listed reasons sum to 40. The 4 mg arm (292 − 30 = 262) and 9 mg arm (284 − 55 = 229) reconcile correctly. The figure 247 equals 287 assigned − 40. No explanation is given.
Second check: another model agreed
Report a problem with this finding
Typo The running footer reads "Published online September 29, 2026, 2026", with the year duplicated.
- p. 1 Running header/footer
Published online September 29, 2026, 2026
Judge's note: The crop of the page 1 footer shows the year printed twice. The same footer recurs on every page.
Second check: none (judge only)
Report a problem with this finding
Typo MedDRA abbreviation misspelled at definition: "Medical Dictionary for Regulatory Activities [MedRA], version 29.0" (p4), vs "MedDRA preferred terms" (p5).
- p. 4 Outcomes, safety
[MedRA], version 29.0
- p. 5 Statistical analysis
MedDRA preferred terms
- p. 14 Results, triglycerides
retarutide 9 mg
- p. 16 Results, discontinuation
than in those who recieved retatrutide 4 mg
Judge's note: Page 4 crop shows '[MedRA], version 29.0'; the Statistical analysis section on p5 spells it 'MedDRA'. The other typos listed under this cluster are judged under their own clusters: C009 for '(52%]', C019 for 'retarutide', C024 for 'recieved' and C027 for 'Eli Lily'.
Second check: none (judge only)
Report a problem with this finding
Typo Methods list the combined preferred terms as "hypotension, orthostatic hypotension, or blood pressure", omitting 'decreased'. Table 3 and the Results give the term as blood pressure decreased/decrease.
- p. 5 Statistical analysis
hypotension, orthostatic hypotension, or blood pressure
Judge's note: The crop of page 5 confirms the truncated term. Table 3 on page 15 and the Results on page 16 include 'decreased'/'decrease'.
Second check: none (judge only)
Report a problem with this finding
Typo Mismatched bracket: "554 (48%) were female and 598 (52%] were male" (p6).
- p. 6 Results, baseline paragraph
598 (52%] were male
- p. 6
female and 598 (52%] were
Judge's note: Page 6 crop clearly shows '(52%]' with a square closing bracket in the image itself, so this is not an extraction artefact.
Second check: code check
Report a problem with this finding
Typo Table 1 LDL cholesterol IQRs use full stops as decimal marks, "94·5 (68.6–114.7)" and "94·0 (68.1–124·1)", unlike the mid-dot used everywhere else.
- p. 8 Table 1, LDL cholesterol
94·5 (68.6–114.7)
- p. 8 Table 1, LDL cholesterol row
94·0 (68.1–124·1)
Judge's note: The text layer gives U+002E full stops in these cells and U+00B7 mid-dots in all neighbouring cells and in the same cell ('124·1'). Viewed the crop and the full p8 image; at this resolution the glyph height cannot be made out clearly, so the call rests on the text layer's distinct characters. This is cosmetic.
Second check: code check
Report a problem with this finding
Typo "fewer participants in the retatrutide group remained in class 1, 2, or 3 obesity categories" uses the singular 'group', although the trial has three retatrutide dose groups.
- p. 8 Results, para 1
fewer participants in the retatrutide group remained in class 1, 2, or 3 obesity categories
Judge's note: No pooled retatrutide analysis is described for this endpoint, so the plural 'groups' is intended.
Second check: none (judge only)
Report a problem with this finding
Typo Table 2, triglycerides, 9 mg vs placebo: "ETD –32·3% (–37·3 –27·0)" is missing 'to' between the bounds.
- p. 10 Table 2, triglycerides, 9 mg vs placebo
(–37·3 –27·0)
Judge's note: The crop of page 10 confirms the missing separator. Every other interval in the table uses 'to'.
Second check: none (judge only)
Report a problem with this finding
Typo Table 2: "RD 33·6 (25.6 to 41·5)" uses a full-stop decimal in '25.6'.
- p. 10 Table 2, ≥10% weight reduction, 4 mg RD
RD 33·6 (25.6 to 41·5)
- p. 10
RD 33·6 (25.6 to 41·5),
Judge's note: The crop of page 10 shows a full stop beside mid-dots within the same cell.
Second check: code check
Report a problem with this finding
Typo Table 2 'Percentage change in urinary albumin-to-creatinine ratio' row: the 12 mg cell "–15·7 (4·9)" lacks the % sign carried by the other arms (–2·8%, –24·7%, –17·6%), and its ETDs also lack %.
- p. 11 Table 2, UACR row, 12 mg column
–15·7 (4·9)
Judge's note: Page 11 crop confirms it. Other percentage rows (e.g. fasting insulin) carry % on all cells including the ETDs. The row label gives the unit, so this is cosmetic.
Second check: none (judge only)
Report a problem with this finding
Typo Drug name misspelled: "–33·5% (1·9) with retarutide 9 mg" (p14).
- p. 14 Results
with retarutide 9 mg
- p. 16 Results
recieved
- p. 17 Declaration of interests
Eli Lily, Gila Pharmaceuticals
- p. 14 Results, lipids paragraph
retarutide 9 mg
Judge's note: Page 14 crop shows 'retarutide' in the image. 'recieved' (p16) and 'Eli Lily' (p17) are also real and are recorded under C024 and C027.
Second check: none (judge only)
Report a problem with this finding
Typo Misspelling: "than in those who recieved retatrutide 4 mg" (p16).
- p. 16 Safety results, discontinuation sentence
those who recieved retatrutide 4 mg
- p. 16 Results, safety paragraph
recieved retatrutide 4 mg
- p. 16
than in those who recieved retatrutide 4 mg
Judge's note: Page 16 crop shows 'recieved' in the image.
Second check: none (judge only)
Report a problem with this finding
Typo Company name misspelled in declaration of interests: "Boehringer Ingelheim, Eli Lily, Gila Pharmaceuticals" (p17).
- p. 17 Declaration of interests (CWlR)
Eli Lily, Gila Pharmaceuticals
Judge's note: Page 17 crop shows 'Eli Lily' in the CWlR declaration; 'Eli Lilly' is spelled correctly elsewhere in the same section.
Second check: none (judge only)
Report a problem with this finding
Typo Subject–verb disagreement: "Treatment discontinuation due to gastrointestinal adverse events were numerically low" should read "was".
- p. 17 Discussion, safety paragraph
Treatment discontinuation due to gastrointestinal adverse events were numerically low
Judge's note: Seen in the p 17 crop; the singular subject takes a plural verb.
Second check: none (judge only)