Cancer in the Elderly—Reply

Abstract

We wish to thank Harding and colleagues for taking their time reading and commenting on our analysis of cancer incidence in advanced age (1), and would want to take this opportunity to rectify some misinterpretations.Harding and colleagues claim that prevalent autopsy findings overestimate the true cancer occurrence in advanced age and that the overestimation increases gradually with age due to the accumulation of indolent, i.e., nonclinically significant, tumors in the elderly. Harding and colleagues also quote us stating “that Sweden's medical system fails to diagnose or treat most cancers in the very old”. We argue that the age-incidence pattern is indeed driven by underreporting and underdiagnosing of cancer but refrain from interpreting this as being a “system failure”. Underreporting in the elderly is certainly an issue since the proportion of pathology-confirmed cancers declines over age (Fig. 1; ref. 1), and the completeness of the Swedish Cancer Register is ensured by double-reporting from both clinical and pathologic departments.The objection regarding the potential accumulation of nonrelevant tumors in the very old may have some merits (2). The analyses incorporating unexpected autopsy findings (Fig. 3; ref. 1), is however not the only result supporting the notion that the registered cancer decline in the very old is factitious. The previously mentioned decreasing proportion of pathology-confirmed cases over age (Fig. 1; ref. 1), particularly noticeable in cancers where tumor biopsy implicate a risk of potentially fatal complications reinforce this conclusion. So also does the finding that the decrease in old age is most pronounced in tumors requiring invasive diagnostics while absent in malignant melanoma of the skin (Fig. 1; ref. 1). We are unsure whether Harding and colleagues argue that the latter is an artefact arising from overdiagnosis of skin cancer in the elderly.More important is the lack of evidence supporting Harding and colleagues’ key argument that autopsy-discovered cancers are indolent and that slow-growing, nonclinically significant tumors accumulate with age (3). Our joint clinical experience is that pancreatic cancer, one of the fastest declining cancers in the oldest, is a highly aggressive malignancy across all ages and overdiagnosis and overtreatment of indolent tumors is a larger issue in younger age groups included in screening. This is supported by a more advanced stage distribution and poorer relative survival in elderly cancer patients together with a less impaired age-incidence pattern where cancer registers are complemented with data from cause of death certificates, unlikely to comprise indolent, nonfatal tumors (4). Given the general consistency of the data provided in our manuscript, we maintain that the decrease of cancer incidence in old age is largely an artifact and not the effect of an actual, biological cancer decline in the elderly.See the original Letter to the Editor, p. 1505No disclosures were reported.