implications for survivorship may include amputation, eye removal, disfigurement, or growth abnormalities. Exposure to therapeutic agents during the rapid and dramatic physiologic and psychologic changes occurring from infancy to early adulthood can result in specific tissue or organ damage, or alteration of normal patterns of growth and development. Long-term sequelae of chemotherapy and radiation are common, may be mild or severe, and may be asymptomatic for extended periods. As many as two-thirds of survivors will experience a late effect of chemotherapy or radiation, defined as any chronic or late occurring outcome—physical or psychosocial—that persists or develops beyond five years from the diagnosis of the cancer (Garre et al., 1994; Oeffinger et al., 2000; Stevens et al., 1998; Vonderweid et al., 1996). These late effects include cognitive impairment, fertility problems, alterations in growth and development, organ system damage, chronic hepatitis, and second malignant neoplasms (DeLaat and Lampkin, 1992; Donaldson, 1993; Dreyer et al., 2002; Friedman and Meadows, 2002; Marina, 1997; Meister and Meadows, 1993; Neglia and Nesbit, 1993; Schwartz, 1995). Survivors frequently have more than one late effect, with perhaps as many as a quarter of survivors experiencing one that is severe or life-threatening (Garre et al., 1994; Oeffinger et al., 2000; Stevens et al., 1998).
The seriousness of the consequences of late effects is evident in studies of premature death following cancer treatment. In one study of late mortality among 20,227 5-year survivors of childhood cancer diagnosed with cancer from 1970 to 1986, there was a 10.8-fold excess in overall mortality (Mertens et al., 2001). Ten percent of these individuals had died by 1996. Figure 4.1 shows all-cause mortality in this cohort as compared to age-adjusted expected survival rates for the U.S. population. Survival is shown by original cancer diagnosis in Figures 4.2a and 4.2b.
Among those for whom cause of death was ascertained, relapse of the primary cancer accounted for 67.4 percent of deaths and treatment-related consequences accounted for 21.3 percent of deaths. The remaining 11.3 percent of deaths were caused by non-treatment external causes (e.g., motor vehicle accidents) or medical conditions (e.g., HIV, pneumonia) (Table 4.1). The three most common treatment-related causes of death observed were (1) the development of a secondary or subsequent cancer, (2) cardiac toxicity, and (3) pulmonary complications.
Among the Childhood Cancer Survivor Study (CCSS) cohort, the overall all-cause absolute excess risk was 8.8 deaths per 1,000 person-years. Within treatment-related cause-specific categories (i.e., excluding recurrences and non-treatment-related deaths), the absolute excess risk was 1.26, 0.27, and 0.015 deaths per 1,000 person-years for secondary and subsequent cancers, cardiac causes, and pulmonary causes, respectively. These treatment-related deaths account for 18 percent of the excess risk of death