Skip to main content
letter
. 2017 Mar 28;116(9):1223–1228. doi: 10.1038/bjc.2017.68

Table 2. The association between regular pre-diagnosis use of common analgesic medications and overall survival following a diagnosis of invasive ovarian cancer, by histologic subtype.

  Aspirin (Regulara vs no regular use) (12 studies)
 Non-aspirin NSAIDs (Regulara vs no regular use) (10 studies)
 Any NSAIDs (Regulara vs no regular use) (12 studies)
 Acetaminophen (Regulara vs no regular use) (11 studies)
Histologic Subtype nb pHRc 95% CI I2 nb pHRc 95% CI I2 nb pHRc 95% CI I2 nb pHRc 95% CI I2
Serousd 4386 1.04 0.92–1.17 24.0 4034 1.01 0.90–1.14 28.9 4394 1.01 0.93–1.09 2.7 4255 1.02 0.93–1.13 1.3
 High-grade 4065 1.02 0.91–1.13 11.0 3728 1.00 0.88–1.13 30.6 4067 0.99 0.91–1.09 9.5 3933 1.01 0.91–1.12 1.8
 Low-grade 265 1.71 0.76–3.87 31.2 252 1.28 0.63–2.59 46.9 277 0.99 0.56–1.75 25.6 246 1.61 0.87–2.99 34.6
Mucinous 248 0.92 0.42–2.01 0.0 287 1.50 0.80–2.80 0.0 353 0.86 0.49–1.48 0.0 209 2.17 0.82–5.74 34.2
Endometrioid 1126 0.87 0.63–1.21 0.0 1000 0.93 0.69–1.27 0.0 1118 0.94 0.66–1.33 37.5 982 0.96 0.60–1.53 29.2
Clear cell 474 1.22 0.74–2.02 14.2 452 0.87 0.50–1.53 36.2 548 1.01 0.61–1.67 43.6 392 0.93 0.37–2.33 61.6e

Abbreviations: CI=confidence interval; NSAID=nonsteroidal anti-inflammatory drug; pHR=pooled hazard ratio.

a

‘Regular' use defined as at least once per week (depending on the question used by each study to collect information on medication use).

b

Participants in model (regular users plus participants with no regular use).

c

Adjusted for age (in years), ethnicity (if <95% of participants were of the same ethnicity) (White/Hispanic/Black/Asian/Other), and education (Less than high-school/Completed high-school including some college/College graduate/Education status unknown).

d

The number of cases included in high and low-grade serous analyses do not add to the total number of cases included in serous analyses, because some participants (particularly with low-grade serous cancers) could not be included in analyses if there were insufficient cases from their study to estimate a site-specific hazard ratio. A number of serous cancers of unknown grade (8.5% of high-grade serous cancers) were assumed to be high-grade for these analyses.

e

Statistically significant heterogeneity in the pHR across studies.