Cancer survivors in the US experienced shifting risks of developing subsequent primary cancers over the past four decades, with overall declines in incidence but increasing risk in several clinically important subgroups, according to a study published in PLOS Medicine.

Researchers analyzed Surveillance, Epidemiology, and End Results (SEER) 8 registry data for 3.36 million cancer survivors diagnosed between 1975 and 2019 and followed through 2022. Using age–period–cohort modeling to separate the effects of aging, calendar time, and birth cohort, they evaluated patterns in subsequent primary cancer (SPC) incidence across patient groups.

During 29.5 million person-years of follow-up, 510,340 SPCs were observed.

SPC incidence increased with age at index cancer diagnosis. Among survivors diagnosed at ages 35 to 39 years, incidence was 915 per 100,000 person-years in women and 1,228 per 100,000 in men, rising to 1,980 and 2,945, respectively, among those diagnosed at ages 75 to 79 years.

Diverging trends by age and cancer type

Overall SPC incidence declined modestly over time, by 0.63% annually in women and 0.95% in men. However, these declines were driven largely by survivors diagnosed at younger ages, while those diagnosed at age 65 years or older experienced stable or increasing risk.

This distinction may be clinically important, as adults aged 65 years and older account for about 64% of US cancer survivors, a proportion projected to rise to 73% by 2040.

Trends also varied by index cancer type. Among female survivors of lung and bronchus cancer, SPC incidence increased by 1.12% annually and was 60% higher in 2015 to 2019 compared with 1975 to 1979. Modest increases were also observed among women with skin melanoma and men with urinary bladder cancer, while incidence declined among survivors of breast, colorectal, and prostate cancers overall.

Among melanoma survivors, SPC incidence rose by approximately 15% to 20% across the most recent calendar periods following 2005. Survivors diagnosed with melanoma after age 70 years had the highest SPC risk among all groups in both sexes.

Notable subgroup patterns

Most cancer types showed increasing SPC risk with older age at diagnosis, but female breast cancer survivors were a notable exception. In this group, SPC incidence remained relatively stable across age groups, with only modest variation. The researchers suggested this pattern may reflect differences in genetic risk, hormonal factors, and treatment exposures across age groups.

Among prostate cancer survivors, SPC incidence declined overall but showed a modest uptick in more recent birth cohorts, a pattern the researchers identified as potentially concerning.

Cohort effects and possible drivers

SPC risk varied across birth cohorts, peaking among patients born in the early to mid-20th century — around 1925 to 1930 in men and 1935 in women — and declining in more recent cohorts.

The researchers noted that these patterns may reflect generational exposures. For example, individuals born between 1935 and 1945 were adolescents or young adults during the peak of tobacco use in the 1960s, a period associated with increased long-term cancer risk.

However, cohort-specific increases persisted among female lung cancer survivors and male bladder cancer survivors, suggesting ongoing or emerging risk factors in these groups.

Clinical implications

Although the study was descriptive and did not assess causality, the findings highlight subgroups of survivors who may benefit from targeted surveillance.

These include:

• female survivors of lung cancer, for whom smoking cessation and lung cancer screening may be particularly relevant;

• melanoma survivors, who may benefit from dermatologic surveillance and ultraviolet exposure mitigation; and

• older cancer survivors, who appear to face persistent or increasing SPC risk across multiple cancer types.

Limitations

The researchers emphasized that the analysis could not identify specific causes of SPC trends. SEER data do not include information on treatment exposures, genetic factors, or lifestyle behaviors such as smoking and diet.

In addition, the study did not adjust for multiple comparisons across subgroup analyses, and registry data may include misclassification or incomplete capture of subsequent cancers, particularly if patients move outside registry regions or die before diagnosis.

"SPC risk is shaped by complex, site- and sex-specific temporal patterns," wrote lead study author Hui G. Cheng, of Virginia Commonwealth University, and colleagues.

Disclosures

The study was funded by the US National Cancer Institute. Oxana Palesh reported consulting relationships with Brigham Young University, University of Rochester, Shook, Hardy and Bacon, Elsevier, Merck, the National Institutes of Health, Sage Publisher, and Monash University. The other researchers reported no relevant conflicts of interest.

Source: PLOS Medicine