Cerebral microdialysis and PtiO2 to decide unilateral decompressive craniectomy after brain gunshot.

Decompressive craniectomy (DC) following brain injury can induce complications (hemorrhage, infection, and hygroma). It is then considered as a last-tier therapy, and can be deleteriously delayed. Focal neuromonitoring (microdialysis and PtiO2) can help clinicians to decide bedside to perform DC in case of intracranial pressure (ICP) around 20 to 25 mmHg despite maximal medical treatment. This was the case of a hunter, brain injured by gunshot. DC was performed at day 6, because of unstable ICP, ischemic trend of PtiO2, and decreased cerebral glucose but normal lactate/pyruvate ratio. His evolution was good despite left hemiplegia due to initial injury.

INTRODUCTION

This case report focuses on the help of focal neuromonitoring (microdialysis [MD] and PtiO2) after brain gunshot, to decide to perform a craniectomy before deleterious effects of brain ischemia. To our knowledge, this is the first one after brain injury.

CASE REPORT

We report the case of a hunter, brain injured by an accidental head gunshot. Prehospital examination showed a Glasgow Coma Scale score of 3 with a right dilated fixed pupil. The patient was sedated, intubated, and mechanically ventilated. Single-dose hypertonic saline was infused. At the arrival in the emergency department, right dilated fixed pupil was still present. Computed tomography (CT) scan 2 hours after accident showed more than 30 intraparenchymal small shots localized in the right hemisphere with brain edema and signs of intracranial hypertension (ICHT) [Figure 1]. There was intraventricular hemorrhage. Basal cisterns were visible.

Figure 1

Head CT scan showing small shots in the right Rolando area and PtiO2 and MD catheters (black arrow)

In intensive care unit (ICU), intracranial pressure (ICP) was monitored (Codman, Johnson and Johnson, USA) with initial value of 35 mmHg, decreasing to 10 mmHg after hypertonic saline infusion. 18 hours later, neuromonitoring was upgraded with cerebral MD (CMA 70, CMA 106, CMA, Sweden) and PtiO2 (Licox, Integra NeuroSciences, USA) placed in the right frontal lobe, through a triple-lumen access device. Initial values were normal (lactate/pyruvate ratio [LPR] between 25 and 30, cerebral glucose >1 mmol/l), except for glutamate (>10 μmol/l).

From days 2 to 5, there was no ICHT, despite two cerebral CT scans (days 2 and 5) showing brain edema, minor hydrocephalous, and visible interpeduncular cistern. At day 5, transcranial Doppler was showing decreased diastolic velocities (22 cm/s) and ICP was >20 mm Hg under conventional sedation. Third-tier treatment was introduced (propofol sedation, moderate hypothermia [33°C]). Six days after injury, ICP was still above 20 mm Hg with cerebral perfusion pressure (CPP) dropping below 60 mm Hg. Discussion was conducted whether to perform decompressive craniectomy (DC) or not. Despite unstable ICP and CT scan, neurosurgeon first decided to continue maximal medical treatment. At that time, cerebral MD showed subnormal values, with normal LPR, but glucose was regularly below 0.8 mmol/l. PtiO2 was showing brief ischemic episodes [Figures 2 and 3], due to CPP decreasing and respiratory failure (Pa02 <80 mm Hg). A bronchofibroscopy led to severe ICHT (50 mm Hg) with clinical signs of transtentorial herniation (bilateral fixed dilated pupils) and was discontinued. ICP decreased to 16 mm Hg and pupils returned to normal diameter. Right unilateral DC was then decided on the following arguments: unstable ICP under maximal medical treatment, subischemic pattern with combined PtiO2 drops, and decreased cerebral glucose values below 0.8 mmol/l and normal LPR, meaning no mitochondrial dysfunction.

Figure 2

ICP, CPP, and PtiO2 trends. Note episodes of high ICP around 50mm Hg, with ischemia (PtiO2 <15mm Hg), prior to craniectomy (EVD: External Ventricular Drainage)

Figure 3

Lactate/pyruvate ratio and cerebral glucose trends. Note normal LPR around 25 during the hours preceding craniectomy. Prior to craniectomy, glucose values were decreased below 0.8mmol/l, which indicates ongoing ischemia (correlation with PtiO2)

After surgery, PtiO2 returned to normal values. Cerebral glucose was still decreased, which can indicate hyperglycolysis. Awakening began after tracheostomy. At day 27, patient was conscious with left hemiplegia. He left the hospital on day 48 for reeducational center, severely disabled (left hemiplegia, Glasgow Outcome Scale score of 3).

DISCUSSION

This case report highlights the potential interest of focal neuromonitoring in neuroinjury. Level I studies are lacking to prove its prognosis value, but some clinical studies are showing trends toward improved neurological outcome after traumatic brain injury (TBI) when using oxygen supply-targeted therapy.[1-3]

PtiO2 is considered to be the final result of balance between oxygen demand and supply. For demand, treatments like sedation are lowering it, on the contrary of epilepsy and awakening. For supply, CPP is one of the main determinants of cerebral blood flow (CBF), with PaCO2 as the second one. Other determinants of oxygen supply are SaO2, PaO2, and hemoglobin. CPP alone is insufficient to tell clinicians whether the chosen threshold is high enough or not for brain oxygen supply. In our case, despite CPP above the classical threshold of 60 to 70 mm Hg, PtiO2 decreased sometimes below 15 mm Hg, which we considered as the ischemic threshold. These episodes had at least the following three mechanisms: primary ICHT with lowering of CPP, hypoxia, in a context of ventilator assisted pneumonia (VAP), and hypocapnia, introduced for high ICP and aggravated by hypothermia (data not shown).

Cerebral MD give additional informations on deleterious effect or not of ischemia on mitochondrial metabolism.[4] In our patient, cerebral MD showed almost unaffected local brain metabolism, except for cerebral glucose. There is no consensus on the MD values defining focal ischemia. In our unit, it is based on the association of cerebral glucose <0.8 mmol/l (with blood glucose >6 mmol/l or sub-cutaneous (SC) glucose >4.5 mmol/l) and LPR >40.[5] In our case, cerebral glucose drops were probably related to decreased CBF, as it is most closely related to level of perfusion,[6] except 3 hours before craniectomy, where it was also due to low blood glucose (data not shown). There was no deleterious effect on mitochondrial metabolism (normal LPR), possibly due to neuroprotection induced by hypothermia. Furthermore, normal LPR was in favor of brain edema due to prolonged blood-brain barrier disruption and not by mitochondrial dysfunction. For us, it was an argument to avoid CPP values above 80 mm Hg. In fact, high CPP induces increased hydrostatic pressure and can aggravate edema.[7] Of course, our choice can be considered as irrelevant because of a higher autoregulation threshold in the patient. Moreover, the ischemic threshold for PtiO2 (in depth and duration) is still a matter of debate. Then, one can consider that there was no ischemia and that low cerebral glucose value associated to normal or subnormal LPR prior to craniectomy was already due to hyperglycolysis. Again, depending on PtiO2 threshold for ischemia, one would have considered deepening of hypocapnia to control ICHT. However, we decided DC in the absence of mitochondrial metabolic distress, before subsequent degradation of focal metabolism in a context of intractable ICHT. We think that it allowed the good evolution.

CONCLUSION

We previously described the potential help of focal monitoring to decide craniectomy after TBI.[8] Craniectomy is now a well-recognized last-tier therapy for post-traumatic intractable ICHT, and retrospective studies are showing improved neurological outcome.[910] A now-recruiting study (the DECRA Trial, see clinicaltrials.com website) will prospectively study this gain. But, craniectomy implies some complications (infection, hygroma, hemorrhage, ...) and cannot be proposed as a routine. For this reason, multimodal neuromonitoring, even if focal, can help the clinician to decide whether it is time or not to perform DC.