Effect of Laparoscopic-Assisted Resection vs Open Resection of Stage II or III Rectal Cancer on Pathologic Outcomes: The ACOSOG Z6051 Randomized Clinical Trial
JAMA, 2015 · paper · Clinical trials
A randomized noninferiority trial of laparoscopic versus open surgery for stage II or III rectal cancer in 486 patients, judged on pathology of the removed tissue. Laparoscopic resection did not meet the noninferiority criterion (81.7% vs 86.9% successful resections).
8 findings
Substance Severe complications "(22.5% vs 22.1%; difference, 0.4%; 95% CI, −4.2% to 2.7%)" (Abstract and Results p5) carry exactly the same CI as readmission "(difference, −0.7%; 95% CI, −4.2% to 2.7%)".
- p. 1 Abstract, Results
(22.5% vs 22.1%; difference, 0.4%; 95% CI, −4.2% to 2.7%)
- p. 5 Results, last paragraph
(difference, 0.4%; 95% CI, −4.2% to 2.7%)
- p. 5 Results, last paragraph
difference, 0.4%; 95% CI, −4.2% to 2.7%
- p. 5 Results, readmission
difference, −0.7%; 95% CI, −4.2% to 2.7%
Only the authors can say what is right here.
Judge's note: Table 4 (p8) gives Clavien-Dindo grade 3–5 counts of 46+6+2=54/240 (22.5%) and 42+5+2=49/222 (22.1%), and rehospitalization of 8/240 vs 9/222. The interval −4.2 to 2.7 is centred near −0.75. That fits the readmission difference, and a Wald interval for readmission reproduces it. It does not fit +0.4: a severe-complication difference with these counts gives a half-width of about 7.6 points. The same interval appears in the Abstract and twice on p5. It looks like the readmission CI was copied, but only the authors can supply the correct interval.
Second check: another model agreed
Report a problem with this finding
Substance In Table 2, Distal margin negative "98.3 (96.7 to 99.95) 98.2 (96.5 to 99.95) −0.1 (−2.3 to 2.5)": laparoscopic minus open is +0.1, and the CI is centred on +0.1.
- p. 5 Table 2, Distal margin negative row
−0.1 (−2.3 to 2.5)
- p. 5 Table 2, Distal margin negative
98.3 (96.7 to 99.95) 98.2 (96.5 to 99.95) −0.1 (−2.3 to 2.5)
Judge's note: The other rows of Table 2 are laparoscopic minus open (87.9−92.3=−4.4; 81.7−86.9=−5.3). 98.3−98.2 is +0.1 (236/240 vs 218/222), and the CI midpoint (−2.3+2.5)/2 is +0.1. Reversing the convention would require an interval of −2.5 to 2.3. The minus sign on the point estimate appears to be wrong.
Second check: another model agreed
Report a problem with this finding
Substance "Open resection never required switching from sphincter-sparing to abdominoperineal resection" contradicts Table 3, where 6 (3.4) of the 174 open-arm patients with sphincter preservation planned before surgery had abdominal perineal resection.
- p. 5 Results, paragraph on conversion
Open resection never required switching from sphincter-sparing to abdominoperineal resection.
- p. 6 Table 3, Sphincter preservation planned before surgery → Abdominal perineal resection
6 (3.4)
Only the authors can say what is right here.
Judge's note: In the p6 Table 3 subsection "Sphincter preservation planned before surgery", the open arm shows 174 patients and abdominal perineal resection 6 (3.4%). The text on p5 says there were no such switches. Only the authors can say which is correct.
Second check: another model agreed
Report a problem with this finding
Substance Table 3 (continued), pathologic stage in the open resection arm (n = 222): 43, 68, 45, 5, 11, 37, 17 and 3 sum to 229, and the printed percentages sum to 103.4%.
- p. 7 Table 3 (continued), Stage, Open Resection column
43 (19.4)
Only the authors can say what is right here.
Judge's note: Read on the crop and the page 7 image. Each open-arm percentage matches its count divided by 222, so the denominator is 222. Footnote a mentions only one laparoscopic patient with missing stage, and nothing explains the 7 extra open-arm patients. The laparoscopic column sums to 239 (240 minus 1 missing), as expected. At least one open-arm count appears wrong, and only the authors can say which.
Second check: another model agreed
Report a problem with this finding
Substance The Discussion says "Pathologic complete response occurred ... 23% and 19% in the current study", but Table 3 reports "Complete pathologic response" as 70 (29.2) and 50 (22.5); 23% and 19% match the Stage 0 row, 55 (23.0) and 43 (19.4).
- p. 9 Discussion, COLOR II comparison
8% to 10% of patients in COLOR II and 23% and 19% in the current study
- p. 7 Table 3 (continued), Complete pathologic response
70 (29.2)
- p. 9 Discussion, COLOR II paragraph
Pathologic complete response occurred in 8% to 10% of patients in COLOR II and 23% and 19% in the current study.
- p. 7 Table 3 (continued), Stage 0 row
55 (23.0)
Only the authors can say what is right here.
Judge's note: Checked on the page 9 and page 7 images. The table has its own labelled pCR row with different values. Its counts (70, 50) also exceed the stage 0 counts (55, 43). The paper does not say which definition the Discussion uses, so the authors need to reconcile the two.
Second check: another model agreed
Report a problem with this finding
Reference or label Table 4 'Other' intraoperative complications are 5 (2.1) laparoscopic and 4 (1.8) open, but footnote c lists 4 laparoscopic events (distal stapler, pancreas, prostate, vagina) and 5 open events (urethra, nerves, urethra, GI NOS leak, rectum leak).
- p. 8 Table 4, Other row
5 (2.1)
- p. 8 Table 4, footnote c
distal stapler (n = 1), pancreas (n = 1), prostate (n = 1), and vagina (n = 1)
Only the authors can say what is right here.
Judge's note: Seen on the page 8 image. The footnote and table counts are reversed between the arms. Several events per patient could not explain a footnote list that is shorter than the patient count (laparoscopic 4 < 5). The arms may be swapped, but only the authors can say which is right.
Second check: another model agreed
Report a problem with this finding
Wording "Patients underwent low anterior resection (76.7%) or abdominoperineal resection (23.3%)" matches the planned procedures in Table 1 (369/481, 112/481), not the operations performed in Table 3.
- p. 1 Abstract, Results
Patients underwent low anterior resection (76.7%) or abdominoperineal resection (23.3%).
- p. 5 Results, first paragraph
Patients underwent low anterior resection (76.7%) or abdominoperineal resection (23.3%) of the rectum.
Only the authors can say what is right here.
Judge's note: Table 1 (p4) shows planned APR 55+57 of 481 and planned LAR 187+182 of 481, which give exactly 23.3% and 76.7%. Table 3 (p6) shows the procedures actually performed in the 462 operated patients: APR 58+47=105 (22.7%), LAR with or without coloanal anastomosis 348 (75.3%), plus Hartmann and proctocolectomy cases. The word "underwent", in the Abstract and on p5, describes the planned figures as performed operations.
Second check: another model agreed
Report a problem with this finding
Wording The Discussion says the COLOR II trial "included 1044 patients with stage I to II rectal cancer within 15 cm of the anal verge"; COLOR II enrolled stage I to III disease.
- p. 9 Discussion, COLOR II paragraph
1044 patients with stage I to II rectal
- p. 9 Discussion
1044 patients with stage I to II rectal cancer
Only the authors can say what is right here.
Judge's note: Rests on outside knowledge: that COLOR II (ref 14, van der Pas et al, Lancet Oncol 2013) enrolled stage I to III rectal cancer. The page 9 image shows "stage I to II" as quoted. The reference entry gives no stage range. Nothing else in the files settles the trial's stage range, but the same paragraph compares stage mix and neoadjuvant therapy, which fits the inclusion of stage III patients. Only the authors can confirm the intended wording.
Second check: another model agreed
Report a problem with this finding
20 minor (typos and the like; most checked by the judge only)
Wording The Discussion gives "entire group 93.6% successful total mesorectal excision" (Table 2), while the Abstract and Results give 93.5% for the same complete-or-nearly-complete TME rate in the same 462 patients.
- p. 7 Discussion, 2nd paragraph
entire group 93.6% suc-cessful total mesorectal excision
- p. 5 Results
nearly complete (16.5%) in 93.5% of the cases (Table 2)
- p. 1 Abstract, Results
complete (77%) and nearly complete (16.5%) in 93.5% of the cases
- p. 7 Discussion paragraph 2
entire group 93.6% suc-
Judge's note: Table 2 counts on page 5: complete TME is 175 (laparoscopic) + 181 (open); nearly complete is 46 + 30. The total is 432/462 = 93.51%, which matches the 93.5% printed in the Abstract (p1) and Results (p5). The 93.6% equals 77.1 + 16.5, a sum of rounded parts. It is the same quantity printed two ways.
Second check: another model agreed
Report a problem with this finding
Wording Methods: distal margin adequate "if it was 2 cm below the line of transection for middle rectal lesions" refers to itself (the line of transection cannot be 2 cm below itself).
- p. 2 Methods, Interventions
if it was 2 cm below the line of transection for middle rectal lesions
- p. 2
ifitwas2cmbelowthelineoftransectionformiddle
Only the authors can say what is right here.
Judge's note: Read on the p2 page image. The parallel rule for upper rectal lesions says "5 cm below the tumor", so "tumor" was probably intended here. Only the authors can confirm.
Second check: another model agreed
Report a problem with this finding
Wording Methods define a negative distal margin as "(>1 mm between the closest tumor to the cut edge", while the Table 2 legend says "distal margin negative (≥1 mm clear)" and "positive (<1 mm clear)".
- p. 3 Methods, Outcomes
distal margin (>1 mm
- p. 5 Table 2, abbreviations
distal margin negative (≥1 mm clear)
- p. 3 Outcomes
distal margin (>1 mm between the closest tumor
- p. 5 Table 2 abbreviations
distal margin negative (ⱖ1 mm clear)
Judge's note: Both pages checked on the images (p3 Outcomes and the p5 Table 2 abbreviations). A margin of exactly 1 mm counts as positive under the Methods definition and negative under the table definition. The effect is likely negligible.
Second check: another model agreed
Report a problem with this finding
Wording The Discussion states "Another secondary outcome of the trial was to determine which instrumentation correlated with failure or success", but this outcome is not among the secondary outcomes listed in the Introduction or Methods, and it is not reported.
- p. 8 Discussion, paragraph 3
Another secondary outcome of the trial was to deter-mine which instrumentation correlated with failure or success
- p. 3 Methods, Outcomes
Secondaryoutcomesincludeddisease-freesurvivalandrate of local recurrence, as well as quality of life and patient-related
Judge's note: Methods (p3) and Introduction (p2) list disease-free survival, local recurrence, quality of life and patient-related benefit. The Methods wording 'included' is not exhaustive, and the funding text (line 1352) mentions an evaluation of instrumentation, so this may be a protocol-level objective. This is a minor consistency point.
Second check: another model agreed
Report a problem with this finding
Wording The design uses a "1-sided z score with α = .10 for falsely declaring noninferiority", but noninferiority is reported with a "1-sided 95% CI" (Table 2 footnote c, Results, Abstract), and the paper does not say how the two relate.
- p. 4 Statistical Analysis
z score with α = .10 for falsely declaring noninferiority
- p. 5 Table 2, footnote c
1-sided 95% CI
- p. 4 Statistical Analysis
α = .10 for falsely declaring noninferiority
Judge's note: The p4 Statistical Analysis and the p5 Table 2 footnote c were checked. A 1-sided 95% interval matches α=.05. Using a conventional 95% interval for reporting could be legitimate, but the paper does not say so; it also gives a final cutoff of z=−1.145. Raised as a reporting question only.
Second check: another model agreed
Report a problem with this finding
Wording The text's 2.7% switch rate "(77.3% planned and 74.6% accomplished)" takes 77.3% from Table 1 (187/242) but 74.6% from Table 3 (179/240), and Table 3 reports 191 (79.6) laparoscopic patients with sphincter preservation planned before surgery.
- p. 4 Table 1, Planned operative procedure, Low anterior resection
187 (77.3)
- p. 6 Table 3, Sphincter preservation planned before surgery
191 (79.6)
- p. 5 Results
(77.3% planned and 74.6% accom-plished)
Only the authors can say what is right here.
Judge's note: Table 1 (p4) shows planned LAR in 187/242 laparoscopic patients. Table 3 (p6) shows 191/240 with sphincter preservation planned before surgery and LAR performed in 69+110=179/240, and 11 (5.8%) of the 191 had APR. The 2.7% is therefore a net difference across two denominators, not the proportion of plans that were changed. The two 'plans' may be recorded at different times (randomization vs pre-surgery), but the paper does not explain this.
Second check: another model agreed
Report a problem with this finding
Wording The Discussion says "Only 29% of patients in COLOR II had tumors in the low rectum (vs 51% in the current study)". Table 1 gives Low as 124 (51.2) laparoscopic and 116 (48.5) open, which is 240/481 (49.9%) for the whole study.
- p. 9 Discussion, COLOR II paragraph
(vs 51% in the current study)
- p. 4 Table 1, Location Low
124 (51.2)
- p. 9 Discussion, COLOR II paragraph
Only 29% of patients in COLOR II had tumors in the low rectum (vs 51% in the current study)
Only the authors can say what is right here.
Judge's note: I checked Table 1 on page 4 against the image: Low is 124 (51.2) in the laparoscopic arm and 116 (48.5) in the open arm. Across the whole study this is 240/481 = 49.9%, about 50%. The 51% in the text matches the laparoscopic arm only, although it is attributed to "the current study". The paper reports no tumour-location figures for another population that would give 51%. The difference is small, so I rate it a wording ambiguity for the authors to clarify.
Second check: another model agreed
Report a problem with this finding
Wording "Overall surgical success, measured by a negative distal and circumferential radial margin result and complete total mesorectal excision" describes the composite as requiring complete TME, but the reported 81.7%/86.9% also count nearly complete TME.
- p. 5 Results, 2nd paragraph
tive distal and circumferential radial margin result and com-plete total mesorectal excision, was higher in the open resec-tion arm (86.9%)
Judge's note: Table 2 (p5) gives success as 156+40=196/240 laparoscopic and 164+29=193/222 open, i.e. complete plus nearly complete TME, as the Methods (p3) define it. Requiring complete TME only would give 65.0%/73.9%. This is a wording error in the Results sentence.
Second check: another model agreed
Report a problem with this finding
Wording Table 2 footnote c states "H0: P1 − P2 ≤ margin; Ha: P1 − P2 > margin (margin = .06)", with P1 the laparoscopic arm. With a positive margin these are superiority-by-6% hypotheses, not noninferiority.
- p. 5 Table 2, footnote c
H0: P1 − P2 ≤ margin; Ha: P1 − P2 > margin (margin = .06)
Judge's note: Read on the p5 image. Reworking: diff = 196/240−193/222 = −0.0527 and SE ≈ 0.0337. This SE reproduces the printed 1-sided lower bound of −10.8, and z = (−0.0527+0.06)/0.0337 ≈ 0.22 gives a 1-sided P ≈ .41, matching the printed P. So the analysis uses margin −0.06. The footnote's sign is wrong, but the reported result stands.
Second check: another model agreed
Report a problem with this finding
Wording The Discussion claims "one of the highest rates of complete total mesorectal excision (entire group 93.6% successful total mesorectal excision)", but 93.5–93.6% is the rate of complete or nearly complete TME; complete TME alone was 77.1%.
- p. 7 Discussion, paragraph 2
one of the highest rates of complete total mesorectal excision (entire group 93.6% suc-cessful total mesorectal excision) (Table 2)
- p. 5 Results, paragraph 2
complete (77.1%)
Only the authors can say what is right here.
Judge's note: The Methods define complete and nearly complete TME separately. The Results (p5) report complete 77.1% and nearly complete 16.5%. The figure quoted for 'complete' TME is the combined rate, so the wording misdescribes the data.
Second check: another model agreed
Report a problem with this finding
Wording Table 4 "30-Day mortality" counts 2 laparoscopic deaths, but footnote f lists "cardiac ischemia/infarction (possibly related; occurred on day 44 after surgery)".
- p. 8 Table 4, footnote f
cardiac ischemia/infarction (possibly related; occurred on day 44 after surgery)
Only the authors can say what is right here.
Judge's note: On the page 8 image, footnote f is attached to the 30-Day mortality row. One of the deaths it lists occurred on day 44 after surgery, outside a 30-day window from surgery. The paper gives no other reference date, so either the row label or the count needs clarification.
Second check: another model agreed
Report a problem with this finding
Wording Table 4, under Days to reoperation, "Missing data 7 8": the open arm has 8 missing values but only "5 (2.3)" reoperations, and a median (range) of 7.0 (3-20) is still reported.
- p. 8 Table 4, Reoperation / Days to reoperation
Missing data
- p. 8 Table 4, Reoperation / Days to reoperation
Missing data 7 8
Only the authors can say what is right here.
Judge's note: Seen on the page 8 image. The number missing cannot exceed the number of reoperations if the row refers to days to reoperation. The table does not say what the Missing data row counts.
Second check: another model agreed
Report a problem with this finding
Typo "venous thrombolic event prophylaxis" should read "venous thromboembolic event prophylaxis".
- p. 2 Methods, Interventions
venous thrombolic event prophylaxis
Judge's note: The misspelling is visible on the p2 page image (Interventions).
Second check: none (judge only)
Report a problem with this finding
Typo Table 1 footnote e misspells oxaliplatin twice: "received oxiliplatin plus radiotherapy" and "received oxiliplatin in addition".
- p. 4 Table 1, footnote e
received oxiliplatin plus radiotherapy
- p. 4 Table 1 footnote e
oxiliplatin plus radiotherapy
Judge's note: Both instances are visible in the p4 image.
Second check: none (judge only)
Report a problem with this finding
Typo Table 2 footnote e reads "213 in the laparascopic" (laparoscopic), and Table 3 uses "Median (IQ)" where the table's abbreviation is IQR.
- p. 5 Table 2, footnote e
213 in the laparascopic
- p. 5 Table 2 footnote e
laparascopic
- p. 6 Table 3
Median (IQ)
Judge's note: The misspelling is visible on p5. The "Median (IQ)" label is visible on p6 under the intensive care unit stay rows, and the Table 3 abbreviation list (p7) defines IQR. Both are cosmetic.
Second check: none (judge only)
Report a problem with this finding
Typo Table 4, Small bowel NOS, open arm "1 (0.4)"; 1/222 = 0.45%, which Table 2 prints as "1 (0.5)".
- p. 8 Table 4, Small bowel NOS, Open
1 (0.4)
- p. 5 Table 2, TME nearly complete CRM ≤1 mm DM(+), Open
1 (0.5)
Judge's note: Seen on the page 8 and page 5 images. The same fraction is rounded differently in the two tables; other 1/222 or 3/222 entries are rounded conventionally.
Second check: none (judge only)
Report a problem with this finding
Typo Table 3, Intensive care unit stay row is labelled "Median (IQ)"; elsewhere in the table it is "Median (IQR)", and IQR is the defined abbreviation.
- p. 6 Table 3, Intensive care unit stay
Median (IQ)
Judge's note: Seen on the page 6 image: the ICU stay subrow reads 'Median (IQ)', while the blood loss, oral analgesics and positive lymph node subrows read 'Median (IQR)'. The footnote defines IQR. This is a typo.
Second check: none (judge only)
Report a problem with this finding
Typo Table 4, Clavien-Dindo grade 4, open arm "5 (2.2)"; 5/222 = 2.25%, which the same table prints as 2.3% for anastomotic leak and reoperation.
- p. 8 Table 4, Clavien-Dindo grade 4, Open
5 (2.2)
Judge's note: Page 8 image: the open arm shows 5 (2.2) for grade 4, but 5 (2.3) for anastomotic leak and for reoperation. This is an inconsistent rounding of the same fraction.
Second check: none (judge only)
Report a problem with this finding
Typo Conflict of Interest Disclosures reads "Dr Change reports serving on the scientific advisory board for Johnson & Johnson"; the author is George Chang.
- p. 9 Conflict of Interest Disclosures
Dr Change reports serving
Judge's note: The crop of page 9 shows "Dr Change" in the disclosures. The byline ("George Chang, MD"), the affiliations ("Chang, Pisters") and the author contributions all spell the name "Chang", and no author is named Change. This is a misspelling.
Second check: none (judge only)
Report a problem with this finding
Typo Disclosures read "Dr Branda reports receiving research support from Covidien", but the byline gives "Megan Branda, MS" and the Author Contributions say "Dr Sargent and Ms Branda".
- p. 9 Conflict of Interest Disclosures
Dr Branda reports receiving research support
- p. 9 Author Contributions
Dr Sargent and Ms Branda
Judge's note: The crops of page 9 show "Dr Branda" in the disclosures and "Ms Branda" in the author contributions. The byline (text line 6) and the group listing (line 1435) both give "Megan Branda, MS". The honorific is inconsistent within the article.
Second check: none (judge only)