Table 1 Study characteristics
Author and publication year | Period of first BC1 dx2 for cohort | Follow-up period | Study design | Country and centre of data derivation | Definition of cohort | Definition of second primary cancers |
|---|---|---|---|---|---|---|
AIRTUM Working Group [1] | Dx: 1976–2010 (varied by registry) | Start: At BC dx End: At SPC dx, death, date of last known vital status, or end of last year of registration (dates varied by registry) | Retrospective cohort | Italy (Multiple cancer registries covering 48% of the population) | All patients dx with a first cancer, although melanoma skin cancer cases, cases based on death certificate only, cases based on autopsy only, and cases with follow-up time equal to zero were excluded. Cohort was stratified by first cancer site, allowing analysis for first BC | IARC/IACR4 rules |
Andersson [2] | Dx 1977–2001 | Start: 1y after BC dx End: SPC dx, death, emigration, or study end (2002) | Retrospective cohort | Denmark (Danish Breast Cancer Cooperative Group) | Female BC patients with record of BC dx at under age 90 in both the Danish Breast Cancer Cooperative Group and the Danish Cancer Register, who survived at least 1y5 post-BC dx, with no prior cancer history other than non-melanoma skin cancer, treated and followed accorded to a Danish Breast Cancer Cooperative Group protocol | SPC6 coding rules unstated, but the Danish Breast Cancer Cooperative Group is linked to the Danish Cancer Register, which uses IARC/IACR rules |
Brenner [3] | Dx 1968–1987 | Start: 1y after BC dx End: study end given as 1987. No details of other censoring events provided | Retrospective cohort | Germany (Saarland Cancer Registry) | Women dx with a first BC (first 1y post dx excluded from analysis) | SPC coding rules unstated, but German registries use IARC/IACR rules. Secondary malignancies and tumours of unspecified location, the skin, the bone, the brain and nervous system, the lung and the liver were excluded |
Brown [4] | Dx 1943–1999 (Denmark), 1953–2002 (Finland), 1953–2000 (Norway), 1958–2002 (Sweden) | Start: 1y after BC dx End: SPC dx, death, or study end (1999–2002, depending on registry) | Retrospective cohort | Denmark, Finland, Norway, Sweden (all national registries) | Women dx with a first BC, who survived for at least 1y (first 1y post dx excluded from analysis) | SPC coding rules unstated, but all participating registries use IARC/IACR rules. Non-haematological malignancies excluded |
Chen [5] | Dx 1997–2010 | Start: At BC dx End: SPC dx, death, emigration, or study end (2010) | Retrospective cohort | Germany (12 German cancer registries covering 33% of population). Data was also reported for Sweden, but is not included here due to fully overlapping with several larger studies | Patients aged 15y or over at dx of a first primary malignant tumour. Patients with only death certificate/autopsy information were excluded. Cohort was stratified by first cancer site, allowing analysis for first BC | According to IARC/IACR rules, not including non-melanoma skin cancer. All cancers must be discordant. 95% + of were cancers microscopically verified |
Diab [6] | Dx 1973–2012 | Start: At BC dx End: SPC dx, death, or study end (2015) | Retrospective cohort | The USA—9 SEER7 registries (Connecticut, Detroit, Atlanta, San Francisco (Oakland), Hawaii, Iowa, New Mexico, Seattle (Puget Sound), Utah) | Women dx with breast cancer. In situ malignancies, dx made without microscopic confirmation, and dx from death certificates and autopsy reports were not included | SEER rules |
Evans [7] | Two cohorts pooled: Dx 1961–1970 and dx 1971–1995 | Start: At BC dx End: SPC dx, death, loss to follow-up, 85th birthday, or study end (1982 for those dx 1961–70, 1996 for those dx 1971–1995) | Retrospective cohort | England (Thames Cancer Registry) | Women dx with first BC at under age 85 | Second tumours at a separate anatomical site and of a distinct histological type to the first tumour, or stated as a new tumour by the treating clinician. Non-melanoma skin cancers, non-malignant cancers, second cancers occurring within 1y of the initial cancer at the same site with the same laterality and histology, or cancers in patients without residency information available at date of dx or a without given date of dx were all excluded |
Gulhan [18] | 1992–2006 | Start: 1y after BC dx End: study end given as 2006. No details of other censoring events provided | Retrospective cohort | Turkey (Izmir Cancer Registry) | Women aged at least 20 with histologically confirmed invasive BC, with at least 1m8 of follow-up | IARC/IACR rules |
Harvey [8] | 1935–1982 | Start: 2m after BC dx End: At SPC dx, death, date of last known vital status, or study end (1982) | Retrospective cohort | The USA (Connecticut Tumour Registry) | Individuals diagnosed with a first primary invasive BC when they were resident in Connecticut, that survived without a second cancer developing for at least 2 m after the diagnosis, who were observed for at least 2 m after the diagnosis, and whose cancer was not diagnosed only from an autopsy report or death certificate | Most of the data used in this study predates the publication of the SEER SPC coding rules, the most common rules applied in North America. However, SPC coding rules used in this study were very similar. Briefly, study defined SPCs as invasive cancers that developed at least 2 m after the first cancer, excluding in situ cancers or non-melanoma skin cancers. SPCs diagnosed only from autopsy reports or death certificates were included |
Hung 2016 [19] | 1997–2010 | Start: At BC dx End: SPC dx, death, dropout from programme providing study data, or study end (2011) | Retrospective cohort | Taiwan (Registry of Catastrophic Illness) | Patients dx with a first BC | SPC coding rules unstated, but the registry histologically confirms cancer cases. Oncologists are required to give evidence of the diagnosis, including cytology reports, pathology reports, laboratory studies, and imaging studies, for review by commissioned expert panels |
Jégu [9] | Dx 1989–2004 | Start: 2 m (62 days) after BC dx End: At SPC dx, death, date of last known vital status, or study end (2007) | Retrospective cohort | France (10 registries covering the Bas-Rhin, Calvados, Doubs, Hérault, Isère, Manche, Somme and Tarn administrative regions) | Patients dx with a first cancer, who did not develop a SPC within 2 m (62 days) after their first cancer. Cohort was stratified by first cancer site, allowing analysis for first BC | IARC/IACR rules |
Jung [20] | Dx 1989–2014 | Start: At BC dx End: At SPC dx, death, date of last known hospital visit, or study end (2014) | Retrospective cohort | Korea (3 medical centres in Soeul, Bucheon, and Choenan) | Women aged at least 20y dx with BC and with at least 1 visit to the Soeul, Bucheon, or Choenan centres within 2 m from dx and with treatment records, who contributed at least 2 m of follow-up time | SPC coding rules unspecified, but second cancers must be at discordant sites, dx at least 2 m after BC diagnosis, with each case "thoroughly reviewed, and misleading information from breast cancer metastasis excluded” |
Lee [21] | Dx 1979–2003 | Start: At BC dx End: At SPC dx, death, or study end (2003) | Retrospective cohort | Taiwan (National Cancer Registry) | Women dx with first BC, without missing dates of birth, follow-up dates or death statuses, and who survived without a second cancer for at least 1 m post-BC dx | According to IARC/IACR rules. Second cancers reported within 1 m of BC dx excluded |
Levi [10] | Dx 1974–1998 | Start: At BC dx End: At SPC dx, death, emigration, or study end (1998) | Retrospective cohort | Switzerland (Swiss Cancer Registries of Vaud and Neuchâtel) | Women dx with a first BC with at least 1 m of follow-up | SPC rules unstated, but the Vaud and Neuchâtel registries use IARC/IACR rules. Second cancers diagnosed at autopsy, death, by death certification alone, or within 1 m of first BC were excluded. Second cancers must be morphologically different or at different anatomical sites |
Mellemkjaer [33] | Australia, New South Wales: 1972–1997, Canada, British Colombia: 1970–1998, Canada, Manitoba: 1970–1998, Canada, Saskatchewan: 1967–1998, Denmark: 1943–1997, Finland: 1953–1998, Iceland: 1955–2000, Norway: 1953–1999, Singapore: 1968–1992, Slovenia: 1961–1998, Spain, Zaragoza: 1978–1998, Sweden: 1961–1998, UK, Scotland: 1960–1996 | Start: At BC dx End: At SPC dx, death, emigration, or study end (between 1992 and 2000, depending on registry) | Retrospective cohort | 13 large cancer registries. Canada (British Columbia, Manitoba and Saskatchewan), Singapore, Slovenia, Norway, Denmark, Scotland, Australia (New South Wales), Sweden, Finland, Iceland, Spain (Zaragoza) | Women dx with a first BC | IARC/IACR rules. Tumours identified by following the recording practices of the included registries |
Molina-Montes [11] | Dx 1985–2007 | Start: At BC dx End: At SPC dx, death, or study end (2007) | Retrospective cohort | Spain (Granada Cancer Registry) | Women dx with a first BC, aged 15y or over at BC dx | According to IARC/IACR rules. Second cancers only included if they occurred at least 3 m after the BC dx |
Murakami [22] | Dx 1965–1982 | Start: Unstated, but less than 1y after BC dx End: At SPC dx, death, or study end (1983) | Retrospective cohort | Japan (Osaka Cancer Registry) | Women dx with a first BC who survived at least 3 m after the BC dx | SPC rules unspecified but Osaka Cancer Registry follows IARC/IACR rules. Second cancers only included if they occurred at least 3 m after the BC dx |
Odani [23] | 2000–2014 | Start: 3 m after BC dx End: At SPC dx, death, 10y after BC dx, or study end (2015) | Retrospective cohort | Japan (Osaka Cancer Registry) | Dx with first primary invasive cancer, aged 15–79 years and resident in Osaka at dx. Dx with death certificate only were excluded. Cohort was stratified by first cancer site, allowing analysis for first BC | IARC/IACR rules |
Ricceri [12] | Individuals recruited to cohort of generally healthy individuals between 1992 and 1998. The subset of these that developed a first primary BC during 11y of follow-up was taken as the cohort of BC survivors in this study | Start: At BC dx End: at SPC dx, death, or end of study (year of study end unstated) | Prospective cohort | EPIC9 cohort is drawn from generally healthy individuals aged 35–70 from 23 centres from Denmark, France, Germany, Greece, Italy, the Netherlands, Norway, Spain, Sweden and the UK. Follow-up for cancer was based on population cancer registries except in France, Germany and Greece, where a combination of methods including health insurance records, cancer and pathology registries and active follow-up were used | Female subset of EPIC cohort that developed a first BC after recruitment into study, or that developed a BC as their second cancer after a first non-melanoma skin cancer. Cases identified using death certificate only were excluded | IARC/IACR rules. Second cancers dx on same date as initial BC or identified using death certificate only were excluded |
Rubino [13] | 1954–1984 | Start: 1973, for those dx with BC 1954–1971. 1y after BC dx, for those dx with BC 1972–1984 End: At SPC dx, death, loss to follow-up, or study end (1992) | Retrospective cohort | France (Institut Gustave Roussy) | Women dx with first BC, born and living in France, with at least 1y of follow-up since BC dx | SPC rules unspecified. All second malignancies were histologically confirmed. Second bilateral BCs and non-melanoma skin cancers were excluded |
Schaapveld [14] | Groningen and Amsterdam: 1989–2003 Eindhoven: 1989–2002 | Start: At BC dx End: at SPC dx, death, or end of study (Groningen and Amsterdam: 2005. Eindhoven: 2004) | Retrospective cohort | The Netherlands (Comprehensive cancer centres of Groningen, Amsterdam, and Eindhoven) | Women dx with first BC with no prior cancer history, or a first BC following non-melanoma skin cancer | According to IARC/IACR coding rules. All unknown primary adenocarcinomas, meningiomas, myelodysplastic syndromes, polycythemia veras, and non-melanoma skin cancers were excluded as second cancers. A cancer occurring after a non-melanoma skin cancer that followed the BC was classed as the second cancer rather than the non-melanoma skin cancer |
Schottenfeld [15] | Treatment (rather than dx) of breast, endometrial, ovarian, vagina, vulva, or cervix uteri cancers at Memorial Sloan Kettering Cancer Centre between 1949 and 1962 | Start: Unstated End: Unstated. Study ended in 1962 | Retrospective cohort | The USA (Memorial Sloan Kettering Cancer Centre) | Patients with cancer of the breast, endometrium, ovary, vagina, vulva, or cervix uteri treated at the Memorial Sloan Kettering Cancer Centre between 1949 and 1962. Cohort was stratified by first cancer site, allowing analysis for first BC | The study predates the publication of the SEER SPC coding rules, the most common rules applied in North America. However, medical records were reviewed to validate the pathologic findings (where presumably recurrences and metastases were ruled out) whenever SPC incidence "increased significantly" |
Silverman [24] | 1990–2006 | Start: At BC dx. Also provided results for a start of follow-up at 6 m after BC dx End: at SPC dx, death, or end of study (2011) | Retrospective cohort | Israel—Israel National Cancer Registry | Women with first BC, excluding breast lymphomas | SPC coding rules unstated, but Israel National Cancer Registry uses IARC/IACR rules with the following optional rules: 1: Two tumours of different laterality, but of the same morphology, diagnosed in paired organs (e.g. breast) are registered separately unless stated to have originated from a single primary 2: Cancers that occur in any 4th character subcategory of colon (C18) and skin (C44) are registered as multiple primary cancers |
Tabuchi [25] | Dx 1985–2004 | Start: 3 m after BC dx End: At SPC dx, death, 10y after BC dx, 80th birthday, or study end (2005) | Retrospective cohort | Japan (Osaka Cancer Registry) | All individuals aged 0–79 dx with a first primary cancer who survived at least 3 m. Cohort was stratified by first cancer site, allowing analysis for first BC | According to IARC/IACR rules. Only discordant second cancers included |
Trama [16] | Dx at and followed up until various periods starting from 1976, respectively, according to the establishment dates of and the most recent incidence data entry dates of the registries in study | Start: At BC dx End: At SPC dx, death, emigration, or end of last year of data entry into registry records (dates varied by registry) | Retrospective cohort | Italy—34 cancer registries covering 43% of Italian population as of 2019 | Individuals diagnosed with a first primary cancer (invasive or of uncertain behaviour), aged 15–39 at the first cancer diagnosis, who survived at least 5y after the first diagnosis. Cohort was stratified by first cancer site, allowing analysis for first BC | IARC/IACR rules |
Tsukuma [26] | 1966–86, but information on standardized incidence ratios for SPCs following BC only available for those dx 1978–86 | Start: At BC dx End: At SPC dx, death, 80th birthday, or study end (1989) | Retrospective cohort | Japan (Osaka Cancer Registry) | All individuals aged 0–79 dx with a first primary cancer, who survived at least 3 m after the first cancer dx. Cohort was stratified by first cancer site, allowing analysis for first BC | IARC/IACR rules. Second cancers only included if they occurred at least 3 m after the BC dx |
Utada [27] | 1985–2007 | Start: At BC dx End: At SPC dx, death, or study end (2008) | Retrospective cohort | Japan (Nagasaki Cancer Registry) | All individuals dx with a first primary cancer. Cohort was stratified by first cancer site, allowing analysis for first BC | IARC/IACR rules. Only discordant second cancers included |
Zheng [17] | 1958–2015 | Start: At BC dx End: At SPC dx, death, emigration, or study end (2015) | Retrospective cohort | Sweden (FCD10) | The Swedish FCD is composed of two separate cohorts. 1: Swedish people born after 1931 (“offspring generation”), and 2: their parents (“parental generation”). This study examined the subset of the offspring generation dx with BC between 1958 and 2015 | Swedish FCD data is linked to national registry, which uses IARC/IACR rules. All second cancers undergo "rigorous histological diagnostics". A request for separate and consistent tumour notifications from clinicians and pathologists is required |