Cancer linked to Alzheimer disease but not vascular dementia
William J.Burke, Saint Louis University Medical School, 1438 S. Grand, St Louis, MO 63104[email protected]
NONE
Submitted June 22, 2010
Regarding the occurrence of cancer in AD [1], we examined all histopathology reports of complete autopsies of patients 50-100 years old at Saint Louis University Group Hospitals from 1983-1988 for diagnoses of AD and cancer. [9]
Control patients had no histopathological evidence of AD. Of the 900 patient reports reviewed, 19.7% of AD and 29.3% of non-AD controls had a pre-mortem diagnosis of cancer. These cases were eliminated from the complete analysis to control for clinical selection bias that results when previously diagnosed cases are included [10], leaving 71 AD and 575 control cases for further analysis.
Sixteen organs were examined for evidence of cancer. We found no statistical difference in total, lung, or prostate cancer between AD and control patients. There was a 6.7-fold increased prevalence of pancreatic cancer in AD patients compared to controls. Controlling for multiple comparisons using Bonferroni’s correction, the odds ratio for pancreatic cancer was significantly (p<0.001) different between AD patients and controls.
We conclude that some of the apparent pre-mortem reduced risk of cancer in AD [1] is due to the difficulty in making a diagnosis of cancer in AD patients versus controls. Specific cancer subtypes may be increased in AD.
References
9. Burke WJ, McLaughlin JR, Chung HD, et al. Occurrence of cancer in Alzheimer and elderly control patients: an epidemiologic necropsy study. Alzheimer Disease and Related Disorders 1994; 8:22-28.
Regarding the occurrence of cancer in AD [1], we examined all histopathology reports of complete autopsies of patients 50-100 years old at Saint Louis University Group Hospitals from 1983-1988 for diagnoses of AD and cancer. [9]
Control patients had no histopathological evidence of AD. Of the 900 patient reports reviewed, 19.7% of AD and 29.3% of non-AD controls had a pre-mortem diagnosis of cancer. These cases were eliminated from the complete analysis to control for clinical selection bias that results when previously diagnosed cases are included [10], leaving 71 AD and 575 control cases for further analysis.
Sixteen organs were examined for evidence of cancer. We found no statistical difference in total, lung, or prostate cancer between AD and control patients. There was a 6.7-fold increased prevalence of pancreatic cancer in AD patients compared to controls. Controlling for multiple comparisons using Bonferroni’s correction, the odds ratio for pancreatic cancer was significantly (p<0.001) different between AD patients and controls.
We conclude that some of the apparent pre-mortem reduced risk of cancer in AD [1] is due to the difficulty in making a diagnosis of cancer in AD patients versus controls. Specific cancer subtypes may be increased in AD.
References
9. Burke WJ, McLaughlin JR, Chung HD, et al. Occurrence of cancer in Alzheimer and elderly control patients: an epidemiologic necropsy study. Alzheimer Disease and Related Disorders 1994; 8:22-28.
10. McFarlane MJ, Feinstein AR, Wells CK, Chan CK. The epidemiologic necropsy. JAMA 1987; 258:331-338.
Disclosure: The author reports no disclosures.