Individual patient data network meta-analysis using either restricted mean survival time difference or hazard ratios: is there a difference? A case study on locoregionally advanced nasopharyngeal carcinomas.
Hazard ratio
Nasopharyngeal carcinoma
Network meta-analysis
Restricted mean survival time difference
Survival analysis
Journal
Systematic reviews
ISSN: 2046-4053
Titre abrégé: Syst Rev
Pays: England
ID NLM: 101580575
Informations de publication
Date de publication:
15 04 2019
15 04 2019
Historique:
received:
19
12
2017
accepted:
11
03
2019
entrez:
17
4
2019
pubmed:
17
4
2019
medline:
6
5
2020
Statut:
epublish
Résumé
This study aimed at applying the restricted mean survival time difference (rmstD) as an absolute outcome measure in a network meta-analysis and comparing the results with those obtained using hazard ratios (HR) from the individual patient data (IPD) network meta-analysis (NMA) on the role of chemotherapy for nasopharyngeal carcinoma (NPC) recently published by the MAC-NPC collaborative group (Meta-Analysis of Chemotherapy [CT] in NPC). Twenty trials (5144 patients) comparing radiotherapy (RT) with or without CT in non-metastatic NPC were included. Treatments were grouped in seven categories: RT alone (RT), induction CT followed by RT (IC-RT), RT followed by adjuvant CT (RT-AC), IC followed by RT followed by AC (IC-RT-AC), concomitant chemoradiotherapy (CRT), IC followed by CRT (IC-CRT), and CRT followed by AC (CRT-AC). The primary endpoint was overall survival (OS); secondary endpoints were progression-free survival and locoregional control. The rmstD was estimated at t* = 10 years in each trial. Random-effect frequentist NMA models were applied. P score was used to rank treatments. Heterogeneity and inconsistency were evaluated. The three treatments that had the highest effect on OS with rmstD were CRT-AC, IC-CRT, and CRT (respective P scores of 92%, 72%, and 64%) compared to CRT-AC, CRT, and IC-CRT when using HR (respective P scores of 96%, 71%, and 63%). Of the 32 HR and rmstD analyzed, 5 had a different interpretation, 3 with a direction change (different direction of treatment effect) and 2 with a change in significance (same direction but a change in statistical significance). Results for secondary endpoints were overall in agreement. The use of either HR or rmstD impacts the results of NMA. Given the sensitivity of HR to non-proportional hazards, this finding could have implications in terms of meta-analysis methodology.
Sections du résumé
BACKGROUND
This study aimed at applying the restricted mean survival time difference (rmstD) as an absolute outcome measure in a network meta-analysis and comparing the results with those obtained using hazard ratios (HR) from the individual patient data (IPD) network meta-analysis (NMA) on the role of chemotherapy for nasopharyngeal carcinoma (NPC) recently published by the MAC-NPC collaborative group (Meta-Analysis of Chemotherapy [CT] in NPC).
PATIENTS AND METHODS
Twenty trials (5144 patients) comparing radiotherapy (RT) with or without CT in non-metastatic NPC were included. Treatments were grouped in seven categories: RT alone (RT), induction CT followed by RT (IC-RT), RT followed by adjuvant CT (RT-AC), IC followed by RT followed by AC (IC-RT-AC), concomitant chemoradiotherapy (CRT), IC followed by CRT (IC-CRT), and CRT followed by AC (CRT-AC). The primary endpoint was overall survival (OS); secondary endpoints were progression-free survival and locoregional control. The rmstD was estimated at t* = 10 years in each trial. Random-effect frequentist NMA models were applied. P score was used to rank treatments. Heterogeneity and inconsistency were evaluated.
RESULTS
The three treatments that had the highest effect on OS with rmstD were CRT-AC, IC-CRT, and CRT (respective P scores of 92%, 72%, and 64%) compared to CRT-AC, CRT, and IC-CRT when using HR (respective P scores of 96%, 71%, and 63%). Of the 32 HR and rmstD analyzed, 5 had a different interpretation, 3 with a direction change (different direction of treatment effect) and 2 with a change in significance (same direction but a change in statistical significance). Results for secondary endpoints were overall in agreement.
CONCLUSION
The use of either HR or rmstD impacts the results of NMA. Given the sensitivity of HR to non-proportional hazards, this finding could have implications in terms of meta-analysis methodology.
Identifiants
pubmed: 30987679
doi: 10.1186/s13643-019-0984-x
pii: 10.1186/s13643-019-0984-x
pmc: PMC6463649
doi:
Types de publication
Journal Article
Research Support, Non-U.S. Gov't
Langues
eng
Sous-ensembles de citation
IM
Pagination
96Références
PLoS One. 2016 Mar 09;11(3):e0150032
pubmed: 26960150
Cancer. 1998 Dec 1;83(11):2270-83
pubmed: 9840526
BMC Med Res Methodol. 2016 Mar 29;16:37
pubmed: 27025706
Stat Med. 2011 Aug 30;30(19):2409-21
pubmed: 21611958
J Clin Epidemiol. 2011 Sep;64(9):985-92
pubmed: 21330105
N Engl J Med. 2017 Jul 27;377(4):338-351
pubmed: 28578639
J Clin Oncol. 2014 Aug 1;32(22):2380-5
pubmed: 24982461
Res Synth Methods. 2012 Dec;3(4):312-24
pubmed: 26053424
Cancer. 2013 Jun 15;119(12):2230-8
pubmed: 23576020
J Natl Cancer Inst. 2010 Aug 4;102(15):1188-98
pubmed: 20634482
Stat Med. 2002 Jun 15;21(11):1539-58
pubmed: 12111919
Int J Radiat Oncol Biol Phys. 1996 Jun 1;35(3):463-9
pubmed: 8655368
Ann Intern Med. 2015 Jul 21;163(2):127-34
pubmed: 26054047
Res Synth Methods. 2010 Jul;1(3-4):258-71
pubmed: 26061470
Stat Med. 2015 Sep 20;34(21):2881-98
pubmed: 26099573
Lancet Oncol. 2017 Sep;18(9):1221-1237
pubmed: 28757375
BMJ. 2017 May 25;357:j2250
pubmed: 28546261
Ann Oncol. 2013 Aug;24(8):2131-6
pubmed: 23661293
J Clin Oncol. 2004 Jul 1;22(13):2643-53
pubmed: 15226332
J Natl Cancer Inst. 2016 Dec 7;109(4):
pubmed: 27927756
J Clin Oncol. 2017 Feb 10;35(5):498-505
pubmed: 27918720
Ann Oncol. 2016 Oct;27(10):1818-28
pubmed: 27436850
Oral Oncol. 2012 Oct;48(10):1038-1044
pubmed: 22591726
BMC Med Res Methodol. 2013 Dec 07;13:152
pubmed: 24314264
J Natl Cancer Inst. 2005 Apr 6;97(7):536-9
pubmed: 15812080
J Clin Oncol. 2016 May 20;34(15):1813-9
pubmed: 26884584
Ann Intern Med. 2015 Jun 2;162(11):777-84
pubmed: 26030634
Oral Oncol. 2014 Feb;50(2):71-6
pubmed: 24315404
Cancer. 2002 Apr 15;94(8):2217-23
pubmed: 12001120
BMC Med Res Methodol. 2018 Mar 9;18(1):27
pubmed: 29523081
J Clin Oncol. 1998 Apr;16(4):1310-7
pubmed: 9552031
Stat Med. 2002 Aug 15;21(15):2175-97
pubmed: 12210632
Int J Radiat Oncol Biol Phys. 2002 Apr 1;52(5):1238-44
pubmed: 11955734
Int J Radiat Oncol Biol Phys. 1995 Oct 15;33(3):569-77
pubmed: 7558945
JAMA. 2012 Sep 26;308(12):1246-53
pubmed: 23011714
J Clin Oncol. 2009 Jan 10;27(2):242-9
pubmed: 19064973
Lancet Oncol. 2012 Feb;13(2):163-71
pubmed: 22154591
Lancet Oncol. 2015 Jun;16(6):645-55
pubmed: 25957714
Res Synth Methods. 2017 Dec;8(4):451-464
pubmed: 28742955
BMC Med Res Methodol. 2015 Jul 31;15:58
pubmed: 26227148
Radiother Oncol. 2011 Jan;98(1):15-22
pubmed: 20971520
Ann Oncol. 2012 Feb;23(2):427-35
pubmed: 21525406
J Clin Oncol. 2016 Oct 1;34(28):3474-6
pubmed: 27507871
Pharm Stat. 2015 Sep-Oct;14(5):409-17
pubmed: 26218830
J Natl Cancer Inst. 2011 Dec 7;103(23):1761-70
pubmed: 22056739
J Clin Oncol. 2005 Sep 20;23(27):6730-8
pubmed: 16170180