Lindner A, Rass V, Ianosi BA, et al. The Importance of PbtO2 Probe Location for Data Interpretation in Patients with Intracerebral Hemorrhage. Neurocritical Care. 2020 Sep 11:1-0.
Reviewed by Wazim Mohamed, MD
Read the Article
Background: Secondary brain injury after spontaneous intracerebral hemorrhage (ICH) continues to have a significant impact on long-term morbidity and mortality. Brain tissue oxygen tension monitoring (PbtO2) is an additional neuromonitoring modality that may provide insight into the extent of secondary brain injury after ICH, helping to identify targets to improve brain tissue hypoxia (BTH). One of the traditional arguments against PbtO2 monitoring has been the localized nature of its monitoring capability, confined to the peri-probe brain tissue. In patients with traumatic brain injury or subarachnoid hemorrhage, there have been consensus statements for the location of probe placement, in relation to site of injury but no similar data exists for ICH. The current study hypothesized that probe location impacts low PbtO2 level (<20 mmHg) detection and is associated with outcomes after ICH.
Methods: This study is a retrospective analysis of prospectively collected data on ICH patients undergoing hematoma evacuation and PbtO2 monitoring over 7 years at a tertiary care center. The decision to evacuate the hematoma was made by a multidisciplinary team consisting of neurointensivists and neurosurgeons. Whenever possible, the monitoring probes were placed in the perihematomal brain tissue after hematoma evacuation. The cohort was divided in to 2 groups based on probe location (perilesional vs. normal brain tissue). The probe location was classified based on a CT scan obtained within 24 hours of placement and considered perilesional if the tip was within 1 cm of focal hemorrhage or focal hypodensity (edema or ischemia). Intralesional probe placement was excluded from analysis. The Glasgow coma scale (GCS), APACHE-II score and ICH scores were used to evaluate disease severity. Demographics, clinical characteristics, neuromonitoring modalities (ICP, CPP, PbtO2), complications, treatment interventions and outcomes were prospectively recorded. A 3-month blinded functional outcome was assessed by a dichotomized modified Ranking scale (mRS) in which an mRS ≤ 4 was compared to an mRS 5 or 6. Brain tissue hypoxia (BTH) were events with PbtO2 < 20 mm Hg and were treated based on the institutional protocol. The time period for data collection was considered acute phase at ≤ 72 hours from probe placement or subacute at 73 – 168 hours. Multivariable models were corrected for covariate confounders (age, GCS, ICH volume) and adjusted odds ratios (aOR) were reported.
Results: Data were collected from 51 patients of which 11 had either dysfunctional probes (7) or recording time was less than 12 hours (4). The final cohort consisted of 40 patients of which 32 received craniotomy and evacuation and 8 patients who underwent hemicraniectomy. The baseline characteristics were similar between both groups. The patients were predominantly male (70%) with a median age of 60 years. Median time to surgery was 8.5 hours and PbtO2 monitoring was done for a median of 6 days (IQR 2-11. Perilesional probe placement occurred in 55% vs normal brain tissue in 45%. During the monitoring period, the CPP did not differ significantly between the 2 groups. In addition, episodes of elevated ICP or mean administration of hyperosmolar therapy was not significantly different between the groups, however there was a trend towards more episodes of elevated ICP in the perilesional group (14% vs. 5%; p=0.06). Although the mean PbtO2 levels recorded were similar between perilesional and normal brain tissue locations (25±12 mmHg vs. 29±10 mmHg; p=0.14), episodes of BTH were slightly more common in the perilesional location (31% vs. 18%; p=0.07) during the entire monitoring period. Overall, the mean PbtO2 during the acute phase was significantly lower compared to the subacute phase (21±12 mmHg vs. 29±10mmHg; p=0.01) among all patients. Patients with perilesional probes had higher incidence of BTH compared to probes in normal tissue after adjusting for the previously stated covariates (46% vs. 19%; aOR 4.0[95% CI 1.54 - 10.58]; p=0.005). The episodes of BTH significantly reduced over time in the perilesional group (p<0.001) and remained unchanged in the normal brain location group (p=0.49). In the subacute phase, there was no association between probe location and BTH.
Although there was a higher incidence of BTH and lower overall PbtO2 readings in patients with mRS ≥ 5 at 3 months, these did not reach statistical significance. In perilesional group, there was a trend towards lower PbtO2 in patients with mRS ≥ 5 (23 [23-29] mmHg) when compared to patients with better outcomes (27 [17-36] mmHg); Adjusted OR was 0.92 with 95% CI 0.83-1.01 (p=0.09). In patients with perilesional probe placement, there was a significant association between BTH and mRS ≥ 5 (aOR 6.6 [95% CI 1.3-33.8]; p=0.02). There were no significant associations between BTH (or PbtO2 levels) and outcomes when the probe was placed in normal brain tissue.
Commentary: The authors have several conclusions based on the results. 1) BTH was more common in perilesional areas in patients ICH. 2) BTH was more common in the early phase of the disease process. 3) Episodes of BTH in the perilesional group decreased over time. 4) There was a significant association between BTH and outcomes when the probe was placed in a perilesional area. PbtO2 levels in the brain can be influenced by a variety of factors but can ultimately be divided in to two groups – factors that affect oxygen delivery and factors that affect oxygen consumption. In patients with traumatic brain injury, recommendations are to insert the PbtO2 probe in the most affected hemisphere targeting the areas vulnerable to secondary brain injury. In patients with subarachnoid hemorrhage, the probe should be placed in the watershed zone of the aneurysm bearing vessel. There are no recommendations regarding probe location in patients with ICH. This is the first study that evaluated the role of probe location in patients with ICH and evaluated ongoing PbtO2 levels. The results of the study support including probe location as a factor that affects PbtO2 levels when interpreting data.
In patients with ICH, the perihematomal tissue is most susceptible to brain injury due to ongoing inflammatory changes, edema and secondary injury cascades. This is highlighted by the current study which showed BTH in these areas. It is encouraging to note that BTH decreased over time suggesting that there is potential for tissue recovery in these areas. In the current study, it is unclear whether this was due to the natural disease progression or therapeutic effects to prevent BTH. In patients with ICH, the association between BTH and outcomes has not been adequately studied and is an area open for future research. The current study showed an association between probe location and outcomes in patients with ICH but the degree to which treatment alters BTH and how this affects outcome after ICH requires more investigation.
There are several limitations to the study. The data, although prospectively collected, was analyzed retrospectively. In such a scenario, study conclusions are limited to associations. The patients in the study represent the cohort of ‘sick’ ICH patients that require surgery and may not truly represent the entire ICH populations. Moreover, the dichotomized outcomes with an mRS cut off at 4 is not standard and has not been validated. The probe location was classified as perilesional based on CT scans and may have missed subtle hemorrhages or areas of ischemia that may influence PbtO2 levels. The technique for probe placement was not image guided and even with the intent of placing the probe in the perilesional area, only 45% of patients achieved this target. Häni et al. demonstrated that CT guided probe insertion is a reliable and easy technique to target the area of interest in a recently published study. This may alleviate the disadvantages associated with ‘blind’ insertion. Overall, the authors of the current study make a valid argument regarding incorporating probe locations in data analysis of PbtO2 levels, especially in ICH patients. Their data also demonstrates multiple areas for future research and areas to target.