Table e-1: Methodological quality of included studies (12).

Study / Study participation
The study sample represents the population of interest on key characteristics, sufficient to limit potential bias on the results. / Study attrition
Loss to follow-up (from sample to study population) is not associated with key characteristics (i.e., the study data adequately represent the sample), sufficient to limit potential bias. / Prognostic factor measurement
The prognostic factor of interest is adequately measured in study participants to sufficiently limit potential bias. / Outcome measurement
The outcome of interest is adequately measured in study participants to sufficiently limit potential bias. / Confounding measurement and account
Important potential confounders are appropriately accounted for, limiting potential bias with respect to the prognostic factor of interest.
(6) / Yes / Yes / Yes / Yes / Yes
(5) / Yes / Yes / Yes / Yes / Yes
(9) / Yes / Yes / Yes / Yes / Yes
(23) / Yes / Yes / Yes / Yes / Yes
(3) / Yes / Yes / Yes / Yes / Yes
(24) / Yes / Partly / Yes / Yes / No
(7) / Yes / Partly / Yes / Yes / Yes
(25) / Yes / Partly / Yes / Yes / Partly
(8) / Yes / Yes / Yes* / Yes / Yes

*Only patients with a CO2R test (hypercapnic challenge) were considered (an acetazolamide test has been performed in the remainder).

1

Table e-2: Influence of baseline characteristics on CO2 reactivity (pCi)

in the MCA ipsilateral to carotid stenosis or occlusion.

Parameter pCi / n=754 patients / n=647 patients
(excluding one study (3))
Estimate / SE / P value** / Estimate / SE / P value**
(Intercept)* / 32.25 / 10.22 / < 0.0001 / 32.57 / 9.22 / < 0.0001
Method of Hypercapnia
Breath-holding (reference) / 0.00 / - / 0.00 / - / -
Inhalation of CO2 5% / 4.81 / 14.37 / 0.36 / 2.44 / 13.06 / 0.56
Inhalation of CO2 6% / -6.10 / 17.56 / -5.41 / 15.95
Inhalation of CO2 7% / 11.03 / 14.28 / 9.48 / 12.97
Inhalation of CO2 8% / 18.30 / 17.49 / Not included in model
Age (in years) / 0.22 / 0.08 / 0.0059 / 0.13 / 0.08 / 0.093
Female sex (vs. male) / -0.31 / 1.68 / 0.89 / -0.55 / 1.64 / 0.76
Occlusion vs. Stenosis
Stenosis 70-89% (reference) / 0.00 / - / - / 0.00 / - / -
Stenosis 90-99% / -5.95 / 2.21 / < 0.0001 / -6.60 / 2.14 / 0.0004
Occlusion / -14.83 / 2.43 / -10.06 / 2.63
Contralateral stenosis ≥70% / -4.40 / 1.97 / 0.021 / -3.25 / 1.93 / 0.083
Symptoms within previous
3 months / -10.68 / 2.06 / < 0.0001 / -10.64 / 1.81 / < 0.0001
Hypertension / Not included in model / -0.28 / 1.54 / 0.95
Diabetes / Not included in model / -3.05 / 1.62 / 0.057
Current smoking / Not included in model / -4.13 / 1.52 / 0.0078

Linear mixed effects model with study as random effect. On the right side the model includes vascular risk factors which were available for 647 patients (excluding one study).

* Intercept corresponds to a patient with the following baseline characteristics: breath-holding method, age: 67 years, sex: male, stenosis 70-89%, contralateral stenosis <70%, asymptomatic (for model considering vascular risk factors: no hypertension, no diabetes, no current smoking).

** P-value from likelihood ratio test.

pCi, percentage increase of cerebral blood flow velocity.

1

Figure e-1:

A Ipsilateral Stroke

B IpsilateralStroke or TIA

Forest plot showing the prognostic value of impaired CO2 reactivity (pCi <20%) on ipsilateral ischemic events for individual studies. Each solid square represents an unadjusted hazard ratio (HR) for individual studies, and the size of the square represents the weight of the individual study in the meta-analysis. For the endpoint stroke results of the studies of Marshall (24) and Markus (3) could not be analysed separately since no stroke was observed in either the group of patients with pCi ≥20% (24) or pCi<20% (3). In the ACES study (8), no ischemic event was observed in any group. Horizontal lines indicate 95% CIs. The width of the diamond shows the 95% CI for the pooled HRs. FE=fixed-effect. W(fixed)=weight based on fixed-effect model.

1