CHOP Chemotherapy plus Rituximab Compared with CHOP Alone in Elderly Patients with Diffuse Large-B-Cell Lymphoma
2002 · paper · Clinical trials
A randomized trial in 399 patients aged 60 to 80 with diffuse large-B-cell lymphoma, comparing CHOP chemotherapy with and without rituximab. Adding rituximab raised complete response rates and prolonged event-free and overall survival without clinically significant added toxicity.
5 findings
Substance The abstract and Results (p4) report a CR/CRu rate of "76 percent" with CHOP plus rituximab. Table 3 gives 106 (52) + 46 (23) of 202 = 152/202 = 75.2%.
- p. 4 Results, Efficacy paragraph
achieved in 76 percent of the patients treat-
- p. 5 Table 3, complete and unconfirmed complete response rows
106 (52)
- p. 1 Abstract, Results
(76 percent vs. 63 percent, P=0.005)
- p. 4 Results, Efficacy paragraph
was achieved in 76 percent of the patients treat-
Only the authors can say what is right here.
Judge's note: I checked Table 3 on the p5 image. The R-CHOP column sums to 202 and the CHOP column to 197. CHOP's 124/197 = 62.9% matches the stated 63%. The only untreated patient was in the CHOP arm, so no smaller R-CHOP denominator is stated anywhere. The 76% cannot be reached from the table counts, and only the authors can say which figure is right.
Second check: another model agreed
Report a problem with this finding
Substance Grade 3/4 cardiac toxicity is given as "the same in both groups (8 percent)" (text and Table 4), but grade 3/4 "cardiac failure or left ventricular dysfunction occurred in 18 patients in the CHOP-plus-rituximab group and 19 patients in the CHOP group" (8.9% and 9.6%). Each subset is larger than its whole category.
- p. 5 Results, Adverse Effects
failure or left ventricular dysfunction occurred in 18
- p. 6 Table 4, Cardiac toxicity, grade 3 or 4
Cardiac toxicity
- p. 5 Results, Adverse Effects
failure or left ventricular dysfunction occurred in 18 patients in the CHOP-plus-rituximab group and 19 patients in the CHOP group.
- p. 5 Results, Adverse Effects
incidence of grade 3 or 4 cardiac toxicity was the same in both groups (8 percent)
Only the authors can say what is right here.
Judge's note: Table 4 on p6 shows 8 and 8 in the grade 3 or 4 cardiac row, with the unit 'percentage of patients with an event in at least 1 cycle', which is the same unit (patients) as the text. 8% can hold at most 17 of 202 or 16 of 197 patients. The text presents cardiac failure inside the grade 3/4 cardiac toxicity paragraph. Only the authors can say which figure is right.
Second check: another model agreed
Report a problem with this finding
Reference or label Table 1 IPI blocks are misaligned. The values "20 (10) 21 (11)" and "29 (14) 23 (12)" sit on the 'Index score§' header line, so every later value sits one row too high and the rows '3' (age-adjusted) and '4–5' (standard) are blank.
- p. 3 Table 1, Age-adjusted and Standard International Prognostic Index score rows
Index score§
Judge's note: The page 3 image confirms the layout. In each block, four value pairs belong to four categories, and each column sums to its arm N. The data are therefore intact, but read row by row the printed labels carry the wrong values (for example, score 0 appears as 61 (30)/56 (28)).
Second check: another model agreed
Report a problem with this finding
Reference or label In Table 1, 'No. of extranodal sites' uses the categories 0, 1 and ">2". This leaves out patients with exactly 2 sites, yet the rows sum to the full arms, so the last category must be ≥2.
- p. 3 Table 1, No. of extranodal sites
>2
Judge's note: The page 3 image shows '>2', while the same table prints '≥75 yr' correctly. The rows sum to 202 and 197. The Results text (p5) refers to 'more than one extranodal site'.
Second check: another model agreed
Report a problem with this finding
Reference or label Table 2's lower event rows are shifted up one line. "29 (14) 49 (25)" sits on the wrapped 'response' line, Relapse shows 21 (10)/13 (7), Death without progression shows 12 (6)/11 (6), During treatment shows 9 (4)/2 (1), and After treatment is blank.
- p. 4 Table 2, rows 'Progression after partial response' to 'After treatment'
29 (14)
Judge's note: The page 4 image confirms the layout. The intended alignment (relapse 29/49, death 21/13, during 12/11, after 9/2) is consistent with the event totals and with the p6 text (34 deaths, 21 R-CHOP and 13 CHOP; 23 during treatment, 11 after, of which 2 CHOP and 9 R-CHOP).
Second check: another model agreed
Report a problem with this finding
10 minor (typos and the like; most checked by the judge only)
Wording The Methods say treatment-effect estimates "after adjustment for these prognostic factors, are expressed as risk ratios". The abstract and Table 2, however, give the unadjusted risk ratios (0.58 [0.44 to 0.77], 0.64 [0.45 to 0.89]) without calling them unadjusted. The adjusted values (0.55, 0.53) appear only in the Results text.
- p. 2 Methods, Statistical Analysis
after adjustment for these prognostic factors, are expressed as risk ratios
- p. 1 Abstract, Results
risk ratios, 0.58 [95 percent confi-
- p. 5 Results, Cox paragraph
the unadjusted values of 0.58 (0.44 to 0.77)
Judge's note: I checked the p2 Methods sentence, the p1 abstract and Table 2 on p4. The p5 Results text explicitly calls 0.58 and 0.64 'the unadjusted values', so the numbers are internally consistent. The only problem is that the abstract and table carry no label, while the Methods description suggests adjusted estimates. This is an ambiguity, not a wrong number.
Second check: another model agreed
Report a problem with this finding
Wording In the CHOP arm, Table 2 has "Progression during treatment 44 (22)" but Table 3 has "Progressive disease 43 (22)". The R-CHOP arm agrees (19 in both), and the difference is not explained.
- p. 4 Table 2, Progression during treatment, CHOP column
44 (22)
- p. 5 Table 3, Progressive disease, CHOP column
43 (22)
Judge's note: Each table sums correctly with its own value (Table 2 events 120; Table 3 total 197), so the two rows may count slightly different things (an event category versus a response category at assessment). The paper does not say so, and the text gives only 22%. This is a minor unexplained difference.
Second check: another model agreed
Report a problem with this finding
Wording In the CHOP arm at 3 years, Figure 1 gives an event-free-survival number at risk of "17", which is more than the overall-survival number at risk of "16" in Figure 2.
- p. 4 Figure 1 number at risk
197 144 101 72 42 17
- p. 5 Figure 2 number at risk
197 171 136 96 58 16
Only the authors can say what is right here.
Judge's note: The images of p4 and p5 show 17 and 16. Patients still at risk for event-free survival should be a subset of those at risk for overall survival, and Table 2 footnote censoring (most recent assessment or cutoff) does not allow longer follow-up for event-free survival. All other time points satisfy EFS ≤ OS. The authors should check this.
Second check: another model agreed
Report a problem with this finding
Wording Herpes zoster comparison 'a few more cases of herpes zoster after treatment ... (nine vs. two, P=0.40)': the printed P value and counts do not correspond under a standard test of 9/202 vs 2/197; which test and denominators were used?
- p. 6 Results, Adverse Effects, herpes zoster sentence
vs. two, P=0.40)
- p. 6 Results, last paragraph before Discussion
(nine vs. two, P=0.40)
- p. 6 Results, Adverse effects
vs. two, P=0.40
Only the authors can say what is right here.
Judge's note: Checked the p6 image: 'nine vs. two, P=0.40' is as quoted. Statistical Analysis (p4) names only log-rank and Cox tests for survival, says all P values are two-tailed, and does not name the test for adverse-event proportions. With arms of 202 and 197, an uncorrected chi-square on 9 vs 2 gives about 4.4 (P about 0.04), and Fisher's exact test gives about 0.06. Neither reaches 0.40. The denominators for 'after treatment' are not stated, so only the authors can say which test and denominators gave this value.
Second check: another model agreed
Report a problem with this finding
Typo The affiliations give "Hôpital Henri-Mondor, Paris (E.L., P.G., F.R.)" but the Appendix lists "Hôpital Henri Mondor, Creteil — F. Reyes and C. Haioun".
- p. 1 Affiliations footnote
Hôpital Henri-Mondor, Paris (E.L., P.G., F.R.)
- p. 7 Appendix
Creteil — F. Reyes and C. Haioun
Judge's note: Rests on outside knowledge: Hôpital Henri-Mondor is located in Créteil. Both entries appear on the page images (p1 footnote, p7 Appendix), so the internal inconsistency is real. Giving 'Paris' for the Paris area is a common convention, so this is a minor consistency point rather than an error.
Second check: none (judge only)
Report a problem with this finding
Typo Unbalanced parentheses in the Cox covariate list: "(specifically, age, ... B symptoms [weight loss, fever, and night sweats]), and age-adjusted International Prognostic Index scores) on event-free". There is one closing parenthesis too many.
- p. 2 Methods, Statistical Analysis
night sweats]),
- p. 2 Methods, Statistical Analysis
night sweats]), and age-adjusted International Prognostic Index scores) on event-
Judge's note: The page 2 image shows ']),' after 'night sweats' and a second ')' after 'scores'. Only one '(' was opened, at '(specifically'. The likely fix is to delete the ')' after ']', which keeps the IPI score inside the list.
Second check: none (judge only)
Report a problem with this finding
Typo ECOG is expanded as "Eastern Clinical Oncology Group" in the Methods (p2) and in the Table 1 footnote † (p3).
- p. 2 Methods, Patients
of the Eastern Clinical Oncology Group
- p. 3 Table 1 footnote †
Clinical Oncology Group
- p. 2 Patients
Eastern Clinical Oncology Group
- p. 2 Methods, Patients
criteria of the Eastern Clinical Oncology Group
Only the authors can say what is right here.
Judge's note: Rests on outside knowledge: the performance-status scale comes from the Eastern Cooperative Oncology Group. The page images show 'Clinical' in both places (p2 Patients paragraph, p3 Table 1 footnote †).
Second check: none (judge only)
Report a problem with this finding
Typo The regimen is spelled out as 'doxorubicin, cyclophosphamide, vindesine, bleomycin, and prednisone (ACVB)', but the cited reference 23 is titled 'A randomized comparison of ACVBP and CHOP'.
- p. 6 Discussion, final paragraph
nisone (ACVB) with CHOP in elderly patients
- p. 8 Reference 23
ACVBP and CHOP
- p. 6 Discussion, third paragraph
cyclophosphamide, vindesine, bleomycin, and pred- nisone (ACVB)
- p. 8 Reference 23
A randomized comparison of ACVBP and CHOP
Judge's note: Checked the p6 Discussion and reference 23 on p8. The abbreviation ACVB leaves out the P for prednisone, which appears in the paper's own spelled-out list of drugs. It also differs from 'ACVBP' in the title of the cited study. Either form might be a house abbreviation, so this is a minor consistency point.
Second check: none (judge only)
Report a problem with this finding
Typo Misspelling in the Appendix: 'GELA clincial research' should read 'GELA clinical research'.
- p. 7 Appendix
GELA clincial research
Judge's note: The misspelling 'clincial' is visible on the p7 page image (Appendix).
Second check: none (judge only)
Report a problem with this finding
Typo Reference 29 title reads 'results of randomized study of the European Organization...' with the article 'a' missing.
- p. 8 Reference 29
results of randomized study
- p. 8 Reference 29
results of randomized study of the
Judge's note: The p8 image of reference 29 shows a visible double-width gap between 'of' and 'randomized', consistent with a dropped 'a'. The phrase is ungrammatical as printed.
Second check: none (judge only)