Anavex Life Sciences Corp (AVXL)

[XenaLives]

2014 paper on diagnostic accuracy in Alzheimer's in Australia:

My read is that 20% is a better figure for possible miss diagnosis and the age of participants reduces that likelihood further. Of those misdiagnosed, they would most likely have another form of dementia. The input of Dr. MacFarlane would also be likely to reduce the odds of misdiagnosis further because he got stung by that bee in an earlier trial of another drug.....

The second key finding is that >80% of individuals meeting the criteria for prodromal AD, that is, clear objective EM impairment with a positive biomarker for AD,6 will develop dementia and meet clinical criteria for probable AD within 3 years if the AD biomarker is Aß imaging. In this study, although there was overlap of CIs on bivariate analysis, Aß imaging with PiB PET appeared to perform substantially better than MRI measures of HA in identifying individuals with MCI who would progress to Alzheimer dementia (OR = 14.5 vs 3.8). This was supported by the results of multivariate analysis that showed PiB PET to be the strongest predictor of progression to AD. Nevertheless, there was considerable gain in positive predictive power from the presence of both imaging biomarkers (OR = 44, PPV = 83%). However, even greater predictive accuracy was obtained by combining clear objective evidence of EM impairment (ie, amnestic MCI) with a positive PiB scan (PPV = 86%, NPV = 100%), consistent with past reports that EM is the cognitive domain most affected in early AD7, 22, 23 and has good predictive accuracy for AD in MCI cohorts.24-26

The predictive value of Aß imaging for progression from MCI to AD has been reported previously. Pooling those studies that, like the present study, used neuropsychological criteria for MCI, reveals progression to AD occurred in 60% if PiB+ compared to 7% if PiB- over an average of 2 years of follow-up.27-30 We observed a higher rate of progression to a clinical diagnosis of AD in the PiB- MCI patients of 19% (5 of 27). Given the well-documented deficiencies in the clinical diagnosis of probable AD compared to postmortem neuropathological diagnosis,31 particularly in distinguishing AD from other forms of dementia, the observation that 9% (5 of 58) of the MCI cohort that progressed to probable AD were PiB- is consistent with the reported rate of misdiagnosis. All 5 did not carry the apolipoprotein E e4 allele that is present in the majority of patients with AD, further suggesting that these individuals have been incorrectly diagnosed.

https://onlinelibrary.wiley.com/doi/full/10.1002/ana.24040