Response Letter to Dr. Cummings

To the Editor: We thank Cummings et al. for their comments. Our original search strategy identified the study by Hoffman et al.,1 but it was excluded from the final review. We agree with the limitations of this study, which Silverman et al. highlighted in an accompanying editorial:2 small sample size, inadequate clinical detail on patients (e.g., Mini Mental State Examination scores not provided), and difficulty in assessing the applicability of the results to routine practice (i.e., subjects had difficult-to-diagnose dementia and were recruited nonconsecutively from a memory disorder clinic). The study by Silverman et al.3 was published after our systematic review was completed. This study was described in our Discussion section. Although this study is a major improvement over the previous literature, we still have several concerns relating to its methodology. The method by which subjects were chosen for inclusion and the indications for performance of positron emission tomography (PET) are unclear. The results presented do not allow the incremental benefit of PET over standard clinical evaluation to be determined. Neither of these two studies changes our conclusion that the value of adding PET to the routine evaluation of patients with possible Alzheimer's disease (AD) has not been convincingly established. This conclusion is in agreement with the American Academy of Neurology's practice parameter on the diagnosis of dementia.4 We could not compare the sensitivity, specificity, or accuracy of PET with conventional diagnostic approaches to dementia because of the deficiencies and heterogeneity of the literature. Our findings highlight the need for better-quality studies of diagnostic tests such as PET. We hope that the Standards for Reporting of Diagnostic Accuracy initiative will provide direction to future studies of PET use in dementia.5 Cummings et al. suggest that the studies of PET's diagnostic accuracy conducted before the availability of cholinesterase inhibitors should be given less emphasis. Although we agree that the presence of drug therapy for dementia makes accurate diagnosis even more clinically important than before, we do not understand why this should affect the diagnostic accuracy of PET. Our goal was not to present an economic evaluation of PET for the diagnosis of AD. The cost-effectiveness analysis cited by Cummings et al.6 concluded that PET is associated with a cost savings, although the sensitivity analysis reveals this conclusion to be dependent upon the assumed sensitivity of PET. Furthermore, this hypothetical model makes important assumptions relating to the effectiveness and costs of cholinesterase inhibitor treatment for early AD, which we do not believe real world studies have substantiated. Cummings et al. correctly note that “probable AD” is not synonymous with “advanced AD,” but as dementia progresses, it may be easier to obtain a diagnosis of probable AD as defined by the National Institute of Neurological and Communicative Diseases and Stroke/Alzheimer's Disease and Related Disorders Association criteria. Furthermore, if a patient has probable AD by these criteria, another means of diagnosis may not be needed, although additional confirmatory tests such as PET might have value in selected patients with difficult-to-diagnose dementia. We agree that dementia is often underrecognized and that better methods of identifying and characterizing dementia are needed. None of the diagnostic studies we reviewed were conducted in primary care. Accordingly, we are unaware of evidence to support the use of PET in this setting. Although the possibility that PET might improve diagnosis in primary care is exciting, there are many reasons for caution: •PET might become a substitute for careful clinical evaluation. •The sensitivity and specificity of PET in the early stages of dementia may be lower than in the studies done to date. •Many patients without AD may be falsely labeled as having AD (the relatively low specificity of PET is potentially a major problem in populations with a low prevalence of AD). •The cost of PET, and subsequent tests and therapies, in this large patient population could be high. We look forward to future, methodologically sound studies that address the use of PET in the beginning of AD.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.