We read the article by Zhong and colleagues (1) with great interest. We sincerely appreciate the tremendous effort made by the authors for presenting this extremely valuable outcome. However, some worthwhile issues also need to be explored.

First, the authors need to interpret the results with caution when isolating the HRs and CIs for survival due to the different median follow-up times. Although the patients were followed up to 10 years in references 8 and 11, patients in some other studies were followed up to only 5 years, such as references 5, 9, and 10. Interestingly, when the authors isolated the HRs and CIs for survival associated with mutation in BRCA1 or BRCA2 from reference 7, the adjusted HRs for 5-year survival, rather than 10-year survival, were chosen. Anyway, the authors did not provide the methods of standardization or stratification. Therefore, the HRs and CIs may not be meaningful and cannot really be interpreted.

In addition, when the authors calculated the HRs and CIs from the Kaplan–Meier curves in reference 41, an HR of 0.68 (95% CI, 0.43–1.07) means the overall survival was not significantly better in BRCA1 carriers than that for mutation-negative ones. Surprisingly, the original data from reference 41 are that BRCA1-linked cases survived significantly longer than sporadic cases (P = 0.008). We think that the authors should describe why the conclusions are contradictory.

Last but not least, prospective and retrospective studies are two fundamental but different kinds of research endeavors (2). As in this article, 20 studies are retrospective studies, and only 7 are prospective. Thus, it is worth mentioning that if the design types could be added into the subgroup analysis, the data could be presented in a more convincing way. In addition, the authors mistook “0.70 (0.36–1.38)” for “0.70 (0.36–1.36),” “1.16 (0.72–1.87)” for “1.16 (0.72–1.85)” and “0.20 (0.06–0.65)” for “0.20 (0.06–0.67)” in Fig. 2, or these may be typos. Above all, we respect the contribution of the authors and we do not doubt the reliability of the measurements.

See the Response, p. 3807

No potential conflicts of interest were disclosed.

This work is supported by the National Natural Science Foundation of China (grant 81172200).

1.
Zhong
Q
,
Peng
HL
,
Zhao
X
,
Zhang
L
,
Hwang
WT
. 
Effects of BRCA1- and BRCA2-related mutations on ovarian and breast cancer survival: a meta-analysis
.
Clin Cancer Res
2015
;
21
:
211
20
.
2.
van der Steen
JT
,
Kruse
RL
,
Szafara
KL
,
Mehr
DR
,
van der Wal
G
,
Ribbe
MW
, et al
Benefits and pitfalls of pooling datasets from comparable observational studies: combining US and Dutch nursing home studies
.
Palliat Med
2008
;
22
:
750
9
.