Introduction - Protocol
Four groups were studied:
Group A: HIV negativei) Clinically, no history of dementia or neurocognitive disability
ii) Pathologically, normal neuropathologically
Group B: HIV positive, not clinically impaired.
i) Clinically, diagnosed normal using NNTC neuropsychology test battery
ii) Pathologically, normal neuropathologically
Group C: HIV positive, neurocognitively impaired with NNTC clinical rating of over 7, and brain specimen does not contain HIVE
i) Clinically, diagnosed with neurocognitive impairment using NNTC test criteria.
ii) Pathologically, did not have HIVE and did not have another pathological change that would produce impairment
Group D: HIV positive, neurocognitively impaired with NNTC clinical rating of over 7, and brain specimen does contain HIVE
i) Clinically, diagnosed with neurocognitive impairment using NNTC test criteria.
ii) Pathologically, met the Budka & Wiley criteria for HIVE and did not have another pathological change that would have produced dementia.
GROUP COMPARISONS
Comparison #1: Group A compared to Group B
This comparison will determine what brain genes are abnormally expressed in neurocognitively normal people who have HIV infection, as compared with normal people who were not infected with HIV. The comparison of two classes of nondemented people is biologically relevant. These data will identify changes in people with HIV infection that have little to do with being demented. It also is possible that this comparison would identify brain gene abnormalities that lead to future impairment, but are not linked temporally to the presence of brain dysfunction. Such genes might be irrelevant to the pathophysiology of HIV dementia because they occur in people who were not impaired. It is, nevertheless, possible that they represent precursors or “harbingers” of future neurocognitive impairment (i.e., a transitory step towards future dysfunction akin to immediate early gene responses).
Comparison #2: Group B compared to Group C
This comparison will determine what brain genes are abnormally expressed in HIV infected people who had impairment, but did not have HIV encephalitis (HIVE). It is proposed here to use HIV infected people who were not impaired as the comparison group. One could argue that Groups A and B could be combined because they both do not have impairment and are not neuropathologically abnormal. If the genes that are abnormal in comparison 1 are excluded from the analysis, one might have a legitimate way to squeeze out some more power by combining these two control Groups A and B. There are some problems with this as well, and future consultations with the statisticians will be undertaken concerning lumping HIV infected and noninfected “normal” people together.
Comparison #3: Group B compared to Group D
This comparison will determine what brain genes are abnormally expressed in HIV infected people who are impaired (or demented), and also had HIVE at autopsy (again using infected people who are not demented as controls).
i) The genes that are abnormal in comparison 3 AND comparison 2 both are highly likely to be dementia-related genes that do not depend on the presence or absence of HIVE neuropathologically.
ii) The genes that are different in comparison 2 but not in comparison 3 would be dementia-related genes that do not depend on HIVE neuropathologically.
iii) The genes that are different in comparison 3, but not comparison 2, would be dementia genes that depend on having HIVE neuropathologically.
Comparison #4: Groups A and B combined compared to groups C and D combined
This comparison is between all people who were not impaired or demented versus all people who were impaired or demented, regardless of HIV infection or brain pathology due to HIVE. This pooling of groups would select for dementia-related genes regardless of the neuropathological findings or HIV infection status (yes/no). This is the most robust and powerful comparison between demented people and nondemented people that one could assemble. It would, however, reveal little about the effects of HIV infection and HIVE per se, and would tend to “dilute” some of the changes that occur. For example, if only the HIVE subjects have a gene change, one might not see a significant effect overall, because the nonHIVE group was not affected. Thus, pooling of neurocognitive functional groups would dilute away changes that depend on a particular neuropathology. The same would hold when if a gene were altered only in people without HIVE. It is also possible that pooling will produce “cancellation.” Here the two pathological situations (HIVE versus no HIVE) could produce multidirectional changes in the demented people (i.e., one pathology increases the gene, the other decreases it). The pooling of pathological groups could statistically neutralize the two opposing changes. Comparisons 2 and 3 above are designed to avoid these “dilution” and “cancellation” effects.
These comparisons would produce statistically validated lists of gene transcripts that match highly important and searched-for categories in the neuroAIDS field:
1) A list of brain gene changes that segregate with HIV infection per se with no clear linkage to having impairment or dementia (comparison 1).
2) A report of brain gene expression selective for impaired or demented people without HIVE (comparison 2).
3) A report of brain gene expression selective for impaired or demented people that do have HIVE (comparison 3)
4) A report of brain gene expression common to most people with impairment or dementia regardless of whether HIVE was present or not (comparison 4).
