High ICP

Sequelae of high ICP
In TBI patients, high ICP (>40 mmHg for longer than 1 hour) is associated with a drop in brain oxygenation measured via PBtO2, as well as impaired autoregulation (as measured by pressure reactivity index (PRx) values).

Optic nerve sheath diameter
Data suggests that the highest measured axial measurement in either eye is the most predictive of ICP. A cutoff of 6.2 mm had 100% sensitivity for identifying patients at risk for decompensation. A ratio of the optic nerve sheath diameter (ONSD) to the eyeball transverse diameter (ETD) (ONSD/ETD) greater than 0.25 was found to be more sensitive (90% vs 80%), more specific (82% vs 79%) and with greater area under the curve (0.92 vs. 0.87) for detecting ICP >20 mmHg compared with ONSD alone. Moreover, the ONSD/ETD ratio was more reliably measured (kappa = 0.710) compared with ONSD alone (kappa 0.602).

Retinal venous pulsations
Absence of retinal venous pulsations correlates with elevated ICP, but its utility for this remains uncertain.

Treatment with positioning
Remove cervical collars or loosen other tight bands around the neck (e.g. endotracheal tube or tracheostomy collars) if clinically appropriate

Cervical collars
A meta-analysis of 5 studies including 86 patients showed that ICPs are elevated in patients with cervical collar by an average of 4.4 mmHg, and that removal of the cervical collar lowers ICP by 3 mmHg on average.

Sodium-based solutions (saline, Ringer's, sodium sulphate, sodium bicarbonate)
In 1919, Weed and McKibben at the Army Neuro-surgical Laboratory at Johns Hopkins University described some experiments in cats measuring suboccipital CSF pressure while administering various intravenous solutions and noted that with ringers (slightly hypotonic), the pressure transiently rose from ~120 to ~180 over approximately 25 minutes. In those given sterile water the ICP rose more significantly, by approximately 50 to 150 mm, and continued to rise for up to 50 minutes after injection. However, when administering 30% NaCl the cats had an initial rise in pressure over 10 minutes, and then an abrupt and significant fall in pressure to over the next 10 minutes, which was maintained for over an hour. Similar patterns were observed with administration of concentrated sodium bicarbonate. Of note, they did notice some cardiorespiratory instability when giving 30% NaCl, but did not qualify this with more detail. Importantly, this acute instability (likely hypotension) may have contributed to the initial rise in ICP that was seen, which was pointed out years later by Milles and Hurwitz in 1932. Weed and McKibben subsequently administered 30% NaCl intravenously to cats in which the skull had been opened, and watched the brain fall away from the inner table of the skull after injection. With hypotonic intravenous injection of free water, the brain swelled several millimeters through trephined openings. As it occurred in the trephined skull, they hypothesized that these changes were not due to osmotic shifts involving the CSF, but instead the brain itself. In 1920 Cushing and Foley published an abstract in which they administered 20-30 mL concentrated sodium chloride to animals and found similar findings. In 1920 Foley and Putnam confirmed the prior findings of Weed and McKibben and also administered doses of 50 mL 30% NaCl rectally or into the duodenum of animals and noted similar findings. However, there was a 30 minute delay when hypertonic saline was administered enterally.

In 1921, Sachs and Malone reported injecting 30% NaCl IV into dogs with trephinations and found a decrease in pressure occurring within 10 minutes. They also noted that the rate of administration mattered -- if administered at a faster rate than 1 mL/minute the dogs would often develop hypotension and shallow respirations, but at slower rates of administration this did not occur. In administering 30% glucose they found no changes, but noted that this only had an osmotic pressure equivalent to 4.5% NaCl, so it would have to be much more concentrated. Also in 1921, Weed and Hughson administered 30% NaCl intravenously to cats at 2 mL/min over 6 minutes and noted no disturbances of respiration but confirmed a decrease in blood pressure that coincided to a transient rise in ICP, prior to ICP falling.

The first human report of hypertonic saline was in 1920, when Cushing and Foley reported in an abstract that they had used concentrated sodium chloride on patients and noted improvement in decreasing tension from brain mass effect. Sachs and Belcher published a case later in 1920 of using "saturated salt solution" in a patient with a brain tumor and noted significant improvement in his clinical status, from tumor to being wide awake. When he again deteriorated 12 hours later, he was given another injection and yet again improved. This was done a total of three times and then he gradually improved thereafter. They subsequently used it on several other cases and reported positive results without providing details. Also in 1920, Ebaugh and Stevenson published a case of an epileptic patient who agreed to undergo intracranial pressure monitoring via prior craniotomy sites for epilepsy surgery. They observed falls in ICP of approximately 10-15 mmH2O when given 19% hypertonic Ringer's solution orally, occurring over approximately one hour. With intravenous 30% glucose, they observed a fall in ICP with a delay of approximately 1 hour before falling at all. With administering oral free water, they noted an immediate gradual rise in ICP of approximately 20-25 mmH2O to a peak value 1-2 hours later. When given hypertonic sodium sulfate IV they noted an immediate drop in ICP of approximately 50 mm over one hour. In 1921, Foley published an additional series of cases with improvement in edema or symptoms after hypertonic saline administration, most of whom had brain mass lesions.

Probably the first case of hypertonic saline use in closed head injury TBI was published by Dowman in 1922, describing a man with a head injury who was unconscious for 36 hours. He was given 50 mL of 30% NaCl IV with improvement, and was given this every 6 hours for the next 36 hours, at the end of which he was conscious. He did very well and was reportedly discharged neurologically intact. Of note, he was also administered oral magnesium sulphate for subsequent maintenance. In 1923, Fay dedscribed 15 patients given 15-35% (30-140 mL) NaCl IV with a fall in ICP in most. He reported that larger volumes of 15% NaCl were more effective than smaller amounts of higher concentrations, so they began to use 50-120 mL of 15% NaCl as standard practice, but no data to this effect was presented. These reports were followed in subsequent years by several other cases, using various solutions at various concentrations.

There were, however, some concerns. Ebaugh and Stevenson's epileptic patient was actually noted to have a slight rise in ICP after the initial fall, occurring at 7-8 hours after administration. Fay briefly mentioned a similar phenomenon and hypothesized that this was due to a subsequent diffusion of NaCl into the brain, causing a delayed osmotic shift of water into the brain, and noted that this did not appear to occur with glucose. Milles and Hurwitz in 1932 noted rebound in ICP for both hypertonic saline and 50% glucose solutions in dogs. As such, the use of hypertonic saline was curbed for a number of years. However, some research continued. In 1951, Wilson et al. published a series of dogs given equiosmolar sodium succinate (18%), sodium lactate (11.2%), or sodium chloride (5.8%) and noted that they were well-tolerated and all had significant drops in ICP for 2.5-4 hours and never rose above the levels initially seen.

Magnesium sulphate
In 1920, Foley and Putnam published a combined rectal injection of hypertonic saline and magnesium sulphate. In 1922, Dowman published a case of TBI in which he was given 30% IV NaCl with improvement, but then oral magnesium sulphate for maintenance. He described using magnesium sulphate enterally regularly, due to its effect in avoiding "accumulation of fluid" via causing diarrhea. In 1923, Fay described 16 patients with various brain tumors given magnesium sulfate (90g in 175 mL water) rectally with improvement in symptoms of signs and symptoms of elevated ICP and strongly advocated for its use. This seemed to avoid the secondary rise in pressure, but also caused vomiting and diarrhea, and the ensuing dehydration could be dangerous to patients with possible hemorrhagic shock. Many years later in 1958, with an intravenous form, Guillaume et al. reported complications of sweating, facial redness, and nausea which sometimes forced them to stop the infusion. Moreover, in their studies of 7 patients with elevated ICP (mostly from tumors), 4 patients improved, 2 had no change, and 1 had worsened results. Several patients had transient ICP elevations which were attributed, along with its side effects, to a general vasodilatory effect.

Glucose
Weed and McKibben also published reports of 30% IV glucose being effective in lowering ICP in cats in 1919. The same year, Haden administered 25% glucose solution to a patient with meningitis who improved, and this was attributed to a possible effect on intracranial pressure, although pressures were not measured and the improvement may instead have been due to the correction of undiagnosed hypoglycemia. In 1920, Ebaugh and Stevenson published their case of an epileptic patient undergoing ICP monitoring (see above), and observed a fall in ICP with 30% IV glucose with a delay of approximately one hour, without a terminal rise in ICP that had been seen with hypertonic saline. In 1925, Peet advocated strongly for the utility of hypertonic glucose solutions, given other reports of their benefit in shock and published a case of the utility of 50% IV glucose in a 12 year-old boy with hydrocephalus. Use of hypertonic glucose solutions continued to be used during the 1920s, and glucose was in many cases favored over hypertonic saline due to its appearing to be safer given less cardiorespiratory side effects through the 1930s.

However, in 1930 Browder published a cautionary series of TBI patients given 75-100 mL of 50% glucose and found that it worked for only about half an hour to treat ICP in some patients, but that in others they had a transient rise in ICP, and in one patient there was a severe rise in ICP and death. Milles and Hurwitz in 1932 noted rebound in ICP for both hypertonic saline and 50% glucose solutions in dogs. In 1933, Jackson et al. noted similar results to Browder in patients with TBI, that half of their patients had a fast rise in ICP with 50% IV glucose. In the first large series of 69 healthy volunteers given hypertonic glucose of various concentrations, Masserman reported a rise in ICP of up to 30 mmH2O seen initially over the first 30 minutes, followed by a fall down to 70 mmH2O below baseline with nadir at 2 hours. In the subsequent hours, however, the ICP rose back up and by 3 hours was rising above baseline. This was hypothesized to be due to either 1) the loss of glucose in the urine causing the blood to become relatively hypotonic, and/or 2) the deposition of glucose in the nervous tissues due to a damaged blood-brain barrier. Bullock et al. hypothesized that this was probably due to glucose freely entering into the CSF. In 1958 Guilllaume et al. administered IV glucose to several patients and had no improvement, despite it having a higher osmolarity than patients given sucrose. They hypothesized that the glucose is immediately metabolized by the body and has insufficient time at a high concentration to cause osmotic shifts.

Sodium arabinate
As glucose was also considered to possibly have temporary effects since it will diffuse away and be metabolized, Hughes and Laplace in 1930 investigated sodium arabinate, a pentose sugar made from gum acacia. Sure enough, glucose solutions given to dogs showed a rapid fall in ICP that lasted at most for 30 minutes and then returned back to normal. With 25% sodium arabinate, the ICP fell and stayed down for hours. However, this was complicated to prepare, which probably limited its clinical use.

Sucrose
Given the issues of rebound ICP noted with hypertonic saline and glucose solutions and the difficulty of preparation of sodium arabinate, in 1935 Bullock et al. hypothesized that a sucrose solution may be better since it would not be metabolized and would not penetrate the blood-brain barrier. In a group of dogs, they injected 50% sucrose solution, 50% glucose solution, of 30% NaCl. They showed that the 50% glucose group had ICP lowered for only 2 hours, the 30% NaCl had ICP lowered for only 4 hours, but 50% sucrose lowered ICP for 7 hours. Moreover, there was a rebound increase in ICP for both 50% glucose and 30% NaCl, but no such rebound for 50% sucrose. The sucrose was excreted via diuresis in the subsequent hours Masserman administered 50% sucrose IV to normal humans and showed a reduction in ICP of 37% for up to 5 hours, but with diuresis of up to four times that administered. Then, in 1937 Jackson et al. gave 50% sucrose IV to eight patients with TBI, and found a reduction of ICP of 50% in 2/3rd of cases, but in two cases there was a secondary rise of ICP. However there were concerns about renal injury with hypertonic sucrose based on prior animal studies, and several years later in 1939-1940 reports emerged of renal injury in humans given hypertonic sucrose solutions. Although the severity and clinical significant of these problems was questioned, it made researchers begin looking for alternative agents. Despite this, however, it was used again in a French study in 1958, who administered it to 3 patients and noted significant drops in ICP without systemic hypotension.

Mannitol
Mannitol is a sugar alcohol that was first characterized by French chemist Joseph-Louis Proust in 1806 from the "manna ash" tree, Fraxinus ornus and later that year described in more detail by French chemist Louis Jacques Thenard and French anatomist and surgeon Guillaume Dupuytren as part of a study on sugars in the urine of diabetics. It was originally called mannite. In a thesis on the topic written in 1915, Braham suggested that the term mannite was first used in 1831. However, in 1815 Louis Jacques Thénard reported that he designated the term mannite to refer to the solid, white, odorless, and sweet material that formed needle-like crystals obtained from manna, the sap of Fraxinus ornus, a claim that is corroborated separately by Laugier and Boillon-la-Grange. By 10 years later the term was well known and used by many, including in Botanique Medicale by Achille Richard. In this botanical he also mentions its use, that the plant itself is a frequently used drug, being a sweet purgative that can be administered easily.

Manna as a purgative***

Beginning in 1937, mannitol was able to be manufactured by chemical reactions with glucose or invert sugar.

In 1960 Scharfetter aministered mannitol to 5 patients and reported that it worked well, but that it did not work as well as urea and that it required a much larger volume of infusion than urea as it has a larger molecular weight and only moderate solubility in water. In 1961, Wise and Chater administered 1.5-2 g/kg of 25% mannitol to 24 patients and reported strong efficacy for decreasing brain bulk in surgery and decreasing intracranial pressure.

A patient level meta-analysis and general meta-analysis done in 2020 clearly showed efficacy of mannitol for lowering ICP.

Urea
In 1914, Hertel administered IV urea to animals and noted that it lowered intraocular pressure. Then, in 1927 Fremont-Smith and Forbes administered 50% urea intraperitoneally to cats and documented a concurrent decrease in intraocular pressure and ICP. One year later in 1928, Wolff and Forbes published a similar decrease in ICP with 50% IV urea. Apparently in 1936 IV urea was administered to 10 patients and was noted to have a significant effect on ICP, but this report was not published until 1960 by Irvine after urea had already begun being used again! Instead, from 1928 to 1950 there were no further reports on the use of IV urea for ICP management, and then in 1950 Smythe et al. published a small series of monkeys given osmotic agents, and noted that 50% urea (5 mL/kg) was far more effective at lowering ICP than the same doses of 50% glucose or 50% sucrose. In 1955, Javid and Settlage administered 30% urea solutions IV (in D5W) to 21 patients with intracranial mass lesions or hydrocephalus and noted that it was well-tolerated and caused significant decreases in ICP and further reported on its utility in 1957. In 1958 Guillaume et al. published its effects on 3 patients with elevated ICP and determined that it seemed to be more effective than IV sucrose, IV magnesium sulfate, or IV glucose. Then in 1960, Katz published a case in the New England Journal of Medicine of its utility in lead encephalopathy, and Stubbs and Pennybacker published on 129 patients given 30% urea IV to neurosurgical patients and noted decreased brain bulk during surgery, decreased lumbar CSF pressure, and no major toxicity. They did note a diuresis beginning shortly after administration, and were concerned about intraoperative bleeding seeming slightly increased. Their series included several patients with TBI, as well as tuberculous meningitis and intracranial hemorrhage. Also in 1960, Scharfetter reported on 100 patients given 30% urea for elevated ICP who had clinical improvement and fast drops in ICP, reaching a minimum ICP about 20-30 minutes after the start of the infusion with effects lasting at least for several hours, and seemed to work better than mannitol. In 1961, Javid published a report of 700 patients given IV urea, at a dose of 1g/kg in a 30% solution with 10% invert sugar and reported its utility, although not many details were provided. Importantly, in patients who herniate osmotherapy with urea was not necessarily sufficient to reverse the herniation in one study. Also in 1961, there were several cases reported of a secondary rise in ICP after urea administration (4 of 8 patients studied), similar to those observed with other agents.

The mechanism of action of urea appears to be an osmotic shift from the brain, and does not depend on diuresis. In monkeys who underwent bilateral nephrectomy, the effect on ICP persisted and was better maintained.

Side effects: Urea was generally fairly well-tolerated, with reports of headaches and sporadic vomiting, occasionally local thrombosis at the infusion site. With extravasation it caused skin blebs with small amounts, and sloughing of the skin in large amounts. Urea does not accumulate significantly in patients with normal renal function, and 12-24 hours later there was no detectable increase in several reports. Javid suggested that it is contraindicated in patients with severe renal damage (as it would accumulate), active intracranial bleeding (reason not clear, but probably because of some reports that it could worsen bleeding), or severe volume depletion (as it would cause a diuresis). Additional fluid needs to be given in some cases due to the diuresis.

One problem with urea is its stability. In the solution of 30% urea with 10% invert sugar, used by Scharfetter et al. and Javid, it forms ammonium carbamate and carbonate at room temperature, changing its color and becoming alkaline, so it had to be kept at 4°C. If not, the ammonia released may contribute to hemolysis. This is in comparison with mannitol that must be kept at 20-30°C (room temperature) to avoid crystal formation

Rate of administration
In the initial experiments by Weed and McKibben in 1919, they noted a transient rise in ICP with hypertonic saline, and also reported that the cats developed cardiorespiratory instability. Most likely these cats had significant hypotension when the solution was injected quickly. Sachs and Malone in 1921 noted no problems in injecting a 30% NaCl solution IV unless it was given at faster than 1 mL/minute, in which case the dogs had hypotension and shallow respirations.

Correction of hyponatremia
In 1933, Fremont-Smith and Merritt evaluated oral water intake and its effect on ICP. When an average of 1.7L of free water was administered orally to normal subjects, ICP did not change. However, if also given DDAVP to inhibit free wate rdiuresis, then the ICP rose from ~130 mmH2O up to over 200 mmH2O on average. D5W administered to dogs raised ICP.