TY - JOUR
T1 - Ovarian cancer risk factors by tumor aggressiveness
T2 - An analysis from the Ovarian Cancer Cohort Consortium
AU - Fortner, Renée T.
AU - Poole, Elizabeth M.
AU - Wentzensen, Nicolas A.
AU - Trabert, Britton
AU - White, Emily
AU - Arslan, Alan A.
AU - Patel, Alpa V.
AU - Setiawan, V. Wendy
AU - Visvanathan, Kala
AU - Weiderpass, Elisabete
AU - Adami, Hans Olov
AU - Black, Amanda
AU - Bernstein, Leslie
AU - Brinton, Louise A.
AU - Buring, Julie
AU - Clendenen, Tess V.
AU - Fournier, Agnès
AU - Fraser, Gary
AU - Gapstur, Susan M.
AU - Gaudet, Mia M.
AU - Giles, Graham G.
AU - Gram, Inger T.
AU - Hartge, Patricia
AU - Hoffman-Bolton, Judith
AU - Idahl, Annika
AU - Kaaks, Rudolf
AU - Kirsh, Victoria A.
AU - Knutsen, Synnove
AU - Koh, Woon Puay
AU - Lacey, James V.
AU - Lee, I. Min
AU - Lundin, Eva
AU - Merritt, Melissa A.
AU - Milne, Roger L.
AU - Onland-Moret, N. Charlotte
AU - Peters, Ulrike
AU - Poynter, Jenny N.
AU - Rinaldi, Sabina
AU - Robien, Kim
AU - Rohan, Thomas
AU - Sánchez, Maria José
AU - Schairer, Catherine
AU - Schouten, Leo J.
AU - Tjonneland, Anne
AU - Townsend, Mary K.
AU - Travis, Ruth C.
AU - Trichopoulou, Antonia
AU - van den Brandt, Piet A.
AU - Vineis, Paolo
AU - Wilkens, Lynne
AU - Wolk, Alicja
AU - Yang, Hannah P.
AU - Zeleniuch-Jacquotte, Anne
AU - Tworoger, Shelley S.
N1 - Publisher Copyright:
© 2018 UICC
PY - 2019/7/1
Y1 - 2019/7/1
N2 - Ovarian cancer risk factors differ by histotype; however, within subtype, there is substantial variability in outcomes. We hypothesized that risk factor profiles may influence tumor aggressiveness, defined by time between diagnosis and death, independent of histology. Among 1.3 million women from 21 prospective cohorts, 4,584 invasive epithelial ovarian cancers were identified and classified as highly aggressive (death in <1 year, n = 864), very aggressive (death in 1 to < 3 years, n = 1,390), moderately aggressive (death in 3 to < 5 years, n = 639), and less aggressive (lived 5+ years, n = 1,691). Using competing risks Cox proportional hazards regression, we assessed heterogeneity of associations by tumor aggressiveness for all cases and among serous and endometrioid/clear cell tumors. Associations between parity (phet = 0.01), family history of ovarian cancer (phet = 0.02), body mass index (BMI; phet ≤ 0.04) and smoking (phet < 0.01) and ovarian cancer risk differed by aggressiveness. A first/single pregnancy, relative to nulliparity, was inversely associated with highly aggressive disease (HR: 0.72; 95% CI [0.58–0.88]), no association was observed for subsequent pregnancies (per pregnancy, 0.97 [0.92–1.02]). In contrast, first and subsequent pregnancies were similarly associated with less aggressive disease (0.87 for both). Family history of ovarian cancer was only associated with risk of less aggressive disease (1.94 [1.47–2.55]). High BMI (≥35 vs. 20 to < 25 kg/m2, 1.93 [1.46–2.56] and current smoking (vs. never, 1.30 [1.07–1.57]) were associated with increased risk of highly aggressive disease. Results were similar within histotypes. Ovarian cancer risk factors may be directly associated with subtypes defined by tumor aggressiveness, rather than through differential effects on histology. Studies to assess biological pathways are warranted.
AB - Ovarian cancer risk factors differ by histotype; however, within subtype, there is substantial variability in outcomes. We hypothesized that risk factor profiles may influence tumor aggressiveness, defined by time between diagnosis and death, independent of histology. Among 1.3 million women from 21 prospective cohorts, 4,584 invasive epithelial ovarian cancers were identified and classified as highly aggressive (death in <1 year, n = 864), very aggressive (death in 1 to < 3 years, n = 1,390), moderately aggressive (death in 3 to < 5 years, n = 639), and less aggressive (lived 5+ years, n = 1,691). Using competing risks Cox proportional hazards regression, we assessed heterogeneity of associations by tumor aggressiveness for all cases and among serous and endometrioid/clear cell tumors. Associations between parity (phet = 0.01), family history of ovarian cancer (phet = 0.02), body mass index (BMI; phet ≤ 0.04) and smoking (phet < 0.01) and ovarian cancer risk differed by aggressiveness. A first/single pregnancy, relative to nulliparity, was inversely associated with highly aggressive disease (HR: 0.72; 95% CI [0.58–0.88]), no association was observed for subsequent pregnancies (per pregnancy, 0.97 [0.92–1.02]). In contrast, first and subsequent pregnancies were similarly associated with less aggressive disease (0.87 for both). Family history of ovarian cancer was only associated with risk of less aggressive disease (1.94 [1.47–2.55]). High BMI (≥35 vs. 20 to < 25 kg/m2, 1.93 [1.46–2.56] and current smoking (vs. never, 1.30 [1.07–1.57]) were associated with increased risk of highly aggressive disease. Results were similar within histotypes. Ovarian cancer risk factors may be directly associated with subtypes defined by tumor aggressiveness, rather than through differential effects on histology. Studies to assess biological pathways are warranted.
KW - aggressiveness
KW - ovarian cancer
KW - prospective cohort
KW - risk factors
KW - subtypes
UR - http://www.scopus.com/inward/record.url?scp=85060039750&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85060039750&partnerID=8YFLogxK
U2 - 10.1002/ijc.32075
DO - 10.1002/ijc.32075
M3 - Article
C2 - 30561796
AN - SCOPUS:85060039750
SN - 0020-7136
VL - 145
SP - 58
EP - 69
JO - International Journal of Cancer
JF - International Journal of Cancer
IS - 1
ER -
