Effects of Early Human Immunodeficiency Virus Infection on Cerebral White Matter Blood Flow Autoregulation

SUMMARY

Blood flow autoregulation in cerebral white matter was measured before and after acute nicardipine-induced changes in mean arterial pressure of 10–21% in 21 treatment naïve HIV-positive adults and 32 controls. The Autoregulatory Index (-%CBF change/% MAP change) was not different at baseline (p= 0.71) or after one year of treatment (n=11, p=0.17). We found no autoregulatory defect to explain the increased stroke risk or the development of cerebral white damage in people with HIV.

INTRODUCTION

An increased risk of ischemic stroke independent of traditional vascular risk factors has been reported in people with HIV (PWHIV). ^[1–3] White matter (WM) hyperintensities on neuroimaging indicating tissue damage are more common in PWHIV than in age matched controls, although not in all series. ^{[1, 4–6]} The mechanism of these have not been fully elucidated. ^{[1, 7]}

Normally, the physiological mechanism of cerebral autoregulation maintains cerebral blood flow (CBF) ) within a narrow range by compenatory arteriolar vasodilation during 20–30% reductions in mean arterial pressure (MAP), thus precventing ischemic damage. ^[8–10] In PWHIV, impaired cerebral vasodilatory responses to carbon dioxide and acetazolamide have been reported. ^{[11, 12]} While these studies do not provide specific data on WM vascular function nor on the autoregulatory vasodilatory responses to changes in MAP, they suggest that impaired white matter autoregulation resulting in ischecmic damage may be important in the cerebral pathology of HIV.

MATERIALS AND METHODS

HIV-positive adults without evidence of AIDS-defining illness who had not yet begun ART were recruited through the University of North Carolina Infectious Diseases Clinic. HIV-negative controls matched by gender and age (±2 years) were recruited through IRB approved local, newspaper, and email listserv advertisements and flyers. Exclusion criteria are described in a previous publication ^[13].

CBF was measured using pseudo-continuous arterial spin labeling (pCASL) at ambient MAP and after acute MAP reduction. ^{[14, 15] [16]} MAP was lowered with intravenous nicardipine infusion with a target of ≥10% but no more than 15% reduction to a maximum dose of 15 mg/hr. CBF measurements were repeated in the PWHIV after 12 months of ART. Image analysis was performed as described previously. ^[17]

Autoregulation was measured by the Autoregulatory Index (AI) separately for WM and GM. AI= −(%Change in CBF)/(%Change in MAP). ^{[18, 19]}

The primary analysis was the comparison of WM AI in control participants to PWHIV at baseline and after 12 months of ART with the threshold for statistical significance set at p= 0.05 using the single-step multiple contrast test procedure (mctp “Dunnett”). ^[20] A secondary analysis of GM AI was also performed. The p-values for the secondary analyses are presented without correction for multiple comparisons and should be interpreted with this limitation. Statistical calculations were performed using R 4.0.2. (www.r-project.org).

This study was approved by the University of North Carolina Chapel Hill Institutional Review Board. All participants provided written informed consent before participating.

RESULTS

Of 46 control participants enrolled, 32 achieved target MAP reduction. Of 41 PWHIV enrolled, 21 achieved target MAP reduction at baseline. Final nicardipine doses were 9.1±4.4 mg/hr in the 32 controls and 8.7±4.2 mg/hr in the 21 PWHIV (p=0.76). At 1 year, 30 PWHIV returned. 12 achieve target MAP reduction. One was eliminated as an outlying value, leaving 11 for analysis. Four PWHIV were in both baseline and 1-year groups.

Ages were 29.6±7.7 years in controls, 27.5±5.6 years in baseline PWHIV, and 32.2±10.2 years in 1-year PWHIV groups. Since 19/21 of the PWHIV participants were men, the control group is mostly (30/32) men as well. At 1 year, 11/11 of the PWHIV were men. Groups were well-matched except there were more smokers in the PWHIV groups at baseline (52%) and one year (82%) as compared to controls (22%). Hypertension was present in 1 (3%) of controls, 3 (14%) of baseline PWHIV and 2 (19%) of 1-year PWHIV. Diabetes was present in 0 controls, 1 (5%) of baseline PWHIV and 1 (9%) of 1-year PWHIV. Hyperlipidemia was present in 1(3%) of controls, 1(5%) of baseline PWHIV and 1(9%) of 1-year PWHIV. In baseline PWHIV participants, median CD4 counts were 473 [interquartile range (IQR) 362-606] cells/mm³ and median viral copies were 38445 (IQR 6584-114695)/mL. At 1 year all 11 of the PWHIV participants were prescribed multi-drug highly active anti-retroviral therapy. Median CD4 counts were 798 (IQR 501-851) cells/mm³ and median viral copies were 39 (IQR 39-44)/mL.

In the primary analysis, WM AI in the PWHIV group was not significantly different from controls at baseline (mean difference 0.29, 95% CI −0.64 to 1.23, p= 0.71) or at one year (mean difference 0.87, 95% CI −0.29 to 2.03, p=0.17) (Table). Effect sizes (Cohen’s d) associated with the lower limits of the 95% CI of the differences for WM AI for the two pairwise comparisons are −0.49 for the former and −0.19 for the latter. These exclude with 95% confidence effect sizes for defective (more negative) AI in the PWHIV that are than medium or more (< −0.50) and small or more (< −0.20) respectively. ^[21]

Table:

White Matter Autoregulatory Index (WM AI), mean arterial pressure (MAP, mm Hg) and white matter cerebral blood flow (WM CBF, mL 100g⁻¹ min⁻¹) in controls and people with the human immunodeficiency virus (PWHIV) at baseline and at one year. Mean ± standard deviation

	Controls (N=32)	PWHIV Baseline (N=21)	p-value vs Controlsa	PWHIV 1 year (N=11)	p-value vs Controlsa
WM AI	−0.65±1.26	−0.36±1.40	0.71 a	0.22±2.05	0.17 a
MAP
Initial	85.0±9.5	86.2±9.0	0.84	88.6±9.0	0.26
Reduced	73.2±7.9	74.6±7.5	0.83	76.7±5.7	0.09
Change	−13.8±2.6%	−13.4±2.9%	0.78	−13.2±2.6%	0.63
Range	−10 to −20%	−10 to −21%		−10 to −19%
WM CBF
Initial MAP	30.4±7.6	31.0±8.6	0.93a	30.8±9.3	0.94a
Reduced MAP	28.2±9.0	28.9±8.5	0.77a	31.3±8.6	0.24a
p-value initial vs. reduced	0.02 b	0.09c		0.76c

^amctp “Dunnett”

^bPaired Wilcoxon signed rank test

^cPaired t-test

GM AI in the PWHIV group at baseline or at one year was not different from controls (p= 0.67 and 0.60. Data not shown).

DISCUSSION

Strengths of our study include quantitative regional measurements of CBF specifically in WM, which is commonly affected in PWHIV. ^[22] We assessed the stability of CBF to changes in MAP, which is directly relevant to the potential causation of naturally occurring cerebral ischemic damage, rather to changes carbon dioxide and acetazolamide administration, which have unclear clinical significance. We studied both PWHIV when treatment-naive and after 12months on ART to assess any effect of ART.

The major limitation of our study is that almost all of the participants were men. We studied PWHIV early in the course of disease with minimal or no MRI WM damage, so these results do not apply to PWHIV with longer duration of disease, more severe WM damage or longer treatment with ART. We chose a target reduction in MAP of 10–15% of baseline to remain within the normal autoregulatory range. ^{[23, 24]} We cannot determine if the limits of autoregulation below these levels of MAP are different in PWHIV.

In conclusion, the stability of CBF in response to these MAP changes was similar between PWHIV cases and controls and remained so after 12 months of ART. We found no evidence of a defect in autoregulation to explain the high risk of ischemic stroke documented in other studies or the development of WMH in PWHIV.