Anna L V Johansson1, Irma Fredriksson2, Lene Mellemkjaer3, Hanne Stensheim4, Päivi Lähteenmäki5, Jeanette F Winther6, Gustav J Ullenhag7, Frida E Lundberg8. 1. Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, PO Box 281, Stockholm, SE-17177, Sweden; Cancer Registry of Norway, Oslo, Norway. Electronic address: anna.johansson@ki.se. 2. Department of Breast, Endocrine Tumors and Sarcoma, D2:02, Karolinska University Hospital, Stockholm, SE-16757, Sweden; Department of Molecular Medicine and Surgery, Karolinska Institutet, Stockholm, Sweden. Electronic address: irma.fredriksson@sll.se. 3. Breast Cancer Epidemiology Group, Danish Cancer Society Research Center, Strandboulevarden 49, Copenhagen, DK-2100, Denmark. Electronic address: lene@cancer.dk. 4. Cancer Registry of Norway, Department of Research, Box 5313 Majorstuen, Oslo, N-0304, Norway. Electronic address: hanne.stensheim@kreftregisteret.no. 5. Swedish Childhood Cancer Registry, Karolinska Institutet, Tomtebodavägen 18A, Solna, SE-17165, Sweden; Turku University Hospital, Turku, Finland. Electronic address: paivi.lahteenmaki@ki.se. 6. Childhood Cancer Research Group, Danish Cancer Society Research Center, Strandboulevarden 49, Copenhagen, DK-2100, Denmark; Department of Clinical Medicine, Faculty of Health, Aarhus University and University Hospital, Aarhus, Denmark. Electronic address: jeanette@cancer.dk. 7. Department of Oncology, Entrance 101, Uppsala University Hospital, Uppsala, SE-751 85, Sweden; Department of Immunology, Genetics and Pathology, Uppsala University, Uppsala, Sweden. Electronic address: gustav.ullenhag@igp.uu.se. 8. Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, PO Box 281, Stockholm, SE-17177, Sweden; Department of Oncology-Pathology, Karolinska Institutet, Stockholm, Sweden. Electronic address: frida.lundberg@ki.se.
Abstract
BACKGROUND: Pregnancy-associated cancer (PAC) is increasing over time in many countries. We provide a comprehensive, population-based overview of cancer survival in women with PAC across five decades. METHODS: We performed a nationwide cohort study of 121,382 women diagnosed with cancer at age 15-49 between 1970 and 2018 using birth and cancer registers in Sweden. Pregnancy-associated cancer was defined as diagnosed during pregnancy and within one year of delivery, while non-PAC was outside this window. Cox regression estimated adjusted hazard ratios (HRs) with 95% confidence intervals (CIs) comparing cancer mortality for PAC versus non-PAC. RESULTS: In total, 5079 women had a diagnosis of PAC. Cutaneous malignant melanoma, breast, cervical, thyroid and central nervous system (CNS) were the most common sites of PAC. A higher cancer mortality was observed in PAC versus non-PAC for breast (HR = 1.72, 95% CI 1.54-1.93) and uterine cancer (myometrium/unspecified) (8.62, 2.80-26.53), in which all PAC deaths were uterine sarcomas. Increased mortality was also observed in upper digestive tract cancer diagnosed during pregnancy and colon cancer diagnosed during first year after delivery. Contrary, the HR for CNS tumours was significantly decreased (0.71, 0.55-0.91). Survival after PAC improved for most sites over time, with survival after breast cancer during pregnancy in recent years being similar to that of non-pregnancy associated breast cancer. CONCLUSION: For the majority of sites, PAC was not associated with poorer prognosis compared to non-PAC, a finding which was stable over time. The main exceptions were breast cancer and rarer cancers, such as uterine sarcoma.
BACKGROUND: Pregnancy-associated cancer (PAC) is increasing over time in many countries. We provide a comprehensive, population-based overview of cancer survival in women with PAC across five decades. METHODS: We performed a nationwide cohort study of 121,382 women diagnosed with cancer at age 15-49 between 1970 and 2018 using birth and cancer registers in Sweden. Pregnancy-associated cancer was defined as diagnosed during pregnancy and within one year of delivery, while non-PAC was outside this window. Cox regression estimated adjusted hazard ratios (HRs) with 95% confidence intervals (CIs) comparing cancer mortality for PAC versus non-PAC. RESULTS: In total, 5079 women had a diagnosis of PAC. Cutaneous malignant melanoma, breast, cervical, thyroid and central nervous system (CNS) were the most common sites of PAC. A higher cancer mortality was observed in PAC versus non-PAC for breast (HR = 1.72, 95% CI 1.54-1.93) and uterine cancer (myometrium/unspecified) (8.62, 2.80-26.53), in which all PAC deaths were uterine sarcomas. Increased mortality was also observed in upper digestive tract cancer diagnosed during pregnancy and colon cancer diagnosed during first year after delivery. Contrary, the HR for CNS tumours was significantly decreased (0.71, 0.55-0.91). Survival after PAC improved for most sites over time, with survival after breast cancer during pregnancy in recent years being similar to that of non-pregnancy associated breast cancer. CONCLUSION: For the majority of sites, PAC was not associated with poorer prognosis compared to non-PAC, a finding which was stable over time. The main exceptions were breast cancer and rarer cancers, such as uterine sarcoma.