The standard of care for cranial bone flap fixation is the use of titanium plates and screws (TPS).1 Hardware fixation may be associated with flap migration due to hardware loosening,2 resorption due to incomplete osseous union with surrounding bone, necrosis,3,4 cosmetic imperfections due to hardware protrusion or flap elevation/depression,5,6 and cerebrospinal fluid (CSF) leakage.7 This study evaluated the efficacy of an osteoconductive, bioresorbable, mineral-organic bone adhesive, Tetranite (RevBio, Inc.), for cranial bone flap fixation. It is composed of tetracalcium phosphate and phosphoserine (TTCP-PS). This adhesive has been demonstrated to be safe and effective for cranial bone flap fixation in an ovine model,8 with adhesive-fixated flaps showing statistically significantly greater strength and osseointegration compared with TPS. The unmet need for an adhesive for bone stabilization is widely recognized.9 The chemical composition and benchtop strength of TTCP-PS adhesive have been studied,10–12 and its biocompatibility and osteoconductive activity in canine, rabbit, and ovine models are well documented.8,12–14
Rigid fixation, resistance to migration, the anatomical profile, cosmesis, protection against CSF egress, and ultimate resorption with replacement by living bone are desired characteristics of an ideal cranial bone flap fixation product. The mechanical stiffness and strength of fixation of cranial bone flaps secured using hardware alone and adhesive alone were measured under impact and quasistatic loading conditions. The resistance of adhesive-fixated cranial flaps to fluid leakage was also evaluated. Furthermore, computed tomography (CT) images were analyzed to quantify the elevation/depression of adhesive-fixated cranial flaps compared with their precraniotomy locations. Penetration depth and kerf area covered by adhesive were measured, and the influence of these factors on mechanical stiffness and strength was investigated.
Methods
A distribution of bone flap size and frequency of different craniotomies was obtained from a survey of 23 neurosurgeons (Table 1). Craniotomy size varied with indication. Based on the survey information, craniotomy widths of 40, 70, and 100 mm were chosen for this study. Bone segments that were circumferentially larger than the intended craniotomy were harvested from 24 fresh frozen adult human cadaver heads.
Craniotomy size and frequency information from the survey of neurosurgeons, as well as location and size of standardized test specimens created for this study
| Craniotomy Survey of Neurosurgeons (n = 23) | |||
|---|---|---|---|
| Size | Indication | Width (mm) | Patient Case Load (%) |
| Small | Dura repair, ICP management, deep brain stimulation implantation, & CSF otorrhea | 62 ± 50; 40 | 24 ± 5; 25 |
| Medium | Tumor diagnosis or removal, aneurysm, & vascular malformation | 76 ± 35; 70 | 53 ± 12; 50 |
| Large | Trauma, stroke, & hemicraniotomy | 113 ± 48; 100 | 36 ± 11; 25 |
| Craniotomy Test Specimens Created for This Study | |||
|---|---|---|---|
| Flap size | Width (mm) | No. of Burr Holes | Location on Skull* |
| 1 | 40 | 1 | Frontal bone |
| 2 | 70 | 2 | Frontal & parietal bone |
| 3 | 70 | 3 | Parietal & occipital bone |
| 4 | 100 | 3 | Parietal bone |
ICP = intracranial pressure.
Values are shown as mean ± SD; nominal unless indicated otherwise.
Images of typical specimens and exploded views are available in Supplemental Table 1.
After the scalp was degloved, 4 segments were marked on each skull using a 3D-printed circular template. One small, 2 medium, and 1 large segment were marked on each skull. These segments were removed from the skull using a neurosurgical drill (Medtronic Midas Rex). The dura mater was not removed. The bone specimens were wrapped in gauze and soaked with saline for hydration. A total of 80 skull segments were removed from the 24 heads. Sixteen segments were not removed because of preexisting cracks in the skull.
The skull segments were embedded in customized fixtures for subsequent testing. The embedding polymethylmethacrylate (PMMA) was shaped like a cylindrical ring with a fixed outer diameter (200 mm). Inner diameter annuli of 55, 85, or 115 mm were used for the small, medium, and large segments, respectively. The space within the inner diameter allowed access to the craniotomy flap for the application of compressive forces, hydrostatic pressure testing, and measurement of displacement during impact testing. In addition to the skull segments, an acrylic ring with Delrin hardware was also embedded in each fixture to enable mold extraction, watertight sealing, and transportation. Also, 4 steel balls (Φ 2.00 mm) were embedded within the PMMA and randomly spaced around the circumference. Three of these served as fiducials for radiographic analysis. The fourth ball was included for redundancy. An exploded illustration of an embedded specimen and photographs are shown in Supplemental Table 1.
The pressure testing setup is shown in Fig. 1. The specimen was clamped on the base with a 1/8-inch rubber gasket, establishing a watertight compartment. A water reservoir was connected to the watertight compartment, and hydrostatic pressure was applied to the inner surface of the specimen by raising the reservoir to the desired height. The apparatus was equipped with a water heater to maintain the test specimen and water at physiological temperatures (37°C). The water was dyed red for easy detection of leaks.
CSF leak test setup. A: Cross-sectional view. B: Exploded view. C: Side view of pressure test setup. D: Image showing leakage from a native specimen. Figure is available in color online only.
Flap sizes 1 and 2 were leak tested to establish baseline data for a pool of specimens in their native (precraniotomy) condition. A total of 36 specimens were leak tested prior to craniotomy. A pressure of 30 mm Hg was applied for 30 seconds. The specimens were frozen until subsequent craniotomy and/or testing.
Craniotomies were completed on a total of 64 specimens after thawing. The remaining 16 specimens were reserved for testing of their mechanical properties in the native (noncraniotomy) condition. To limit variations in cranial bone flap geometry, a single neurosurgeon with extensive experience created all bone flaps. One to 3 burr holes (Table 1) were created in the specimens using a Φ 14-mm perforator. The cutting tool (Medtronic Midas Rex Legend with F2-B1 footed attachment; Φ 2.3 mm) produced a circumferential gap (kerf) between the bone flap and surrounding skull equivalent to the diameter of the cutting drill bit. The specimens were kept frozen until the time of subsequent fixation of the bone flaps to the surrounding skull. They were then transported to the 3 different test sites, where a total of 16 different neurosurgeons fixated the flaps after they were thawed and heated to physiological temperature (37°C) for a minimum of 1 hour prior to the simulated procedure. Each neurosurgeon fixated a total of 4 flaps (flap sizes 1 through 4). They fixated flap sizes 1 and 2 using adhesive alone. Flap sizes 3 and 4 were first fixated using metallic hardware (TPS) alone and subsequently fixated using adhesive alone after testing of the TPS-fixated flaps and removal of hardware. The specimens were divided into 5 test groups with different mechanical testing approaches for each.
The adhesive was used by all surgeons as per the instructions for use. A combination of 3 to 4 intraoperative fixation aids provided in the kit was used to provisionally hold the cranial flap in its anatomical position for adhesive application. Next, the first dose of the adhesive (4 cm3) provided in the kit was mixed and applied to the kerf spaces via a syringe with the provided cannula tip (outer diameter 4.1 mm) to mitigate adhesive over-penetration through the kerf space. At 4.5 minutes, the fixation aids were removed. For flap sizes 1 and 2, the adhesive from the first mix was also used to prepare burr hole plugs that were fabricated using the provided mold from the kit. For these same flaps, the second dose of adhesive (4 cm3) provided in the kit was applied to fill the residual gaps around the burr hole plugs and kerf spaces to provide a watertight seal and to demonstrate optimal cosmesis.
CT was used to measure several geometrical parameters related to the shape of the specimen, the penetration of adhesive in the kerf space, and any elevation or depression of the fixated flap relative to the surrounding skull. The entire process was automatic and coded in Python (version 3.9.2). All specimens were scanned using cone-beam CT (Imaging Sciences International i-CAT 17–19, 0.25-mm resolution) at 3 time points each. They were scanned before craniotomy in their native condition (pretreatment), immediately after craniotomy in the cut condition with the flap removed (resected), and when possible, after adhesive fixation (posttreatment). Specimens that were impacted to failure immediately after fixation were not scanned after fixation with adhesive.
The 3-dimensional coordinates of each fiducial steel ball in each CT data set were automatically obtained using Python. The interfiducial distance was used to register the 3 fiducials between the 3 data sets. This information was then used to establish coordinate systems in each data set, and rotation matrices were calculated to transform coordinates from 1 data set to another. Radial slice images from identical locations on each specimen were then generated in each of the 3 data sets, at 1° increments around the entire circumference of the kerf, utilizing the rotation matrices. The distribution of density within the cut flap was obtained by subtracting the density distribution of the resected data set from the pretreatment data set. The fifth percentile density of the cut flap, inclusive of bone and pore contents, was used as the threshold density for analyzing the slice images. The measurements listed in Table 2 were obtained automatically for each specimen using Python.
Description of CT measurements
| Measurements on Slice i* | Description |
|---|---|
| Kerf location, | 3-dimensional coordinates of kerf |
| Circumference, | Total length of kerf around circumference |
| Kerf cut angle | Angle of kerf cut relative to specimen axis; averaged around circumference |
| Kerf depth, tbi | Distance along depth of kerf occupied by native bone |
| Kerf interface area, | Total area of kerf |
| Adhesive thickness, tai | Distance along depth of kerf occupied by adhesive; averaged around circumference |
| Adhesive interface area, | Total area of kerf occupied by adhesive |
| Outer overfill | Distance along depth of kerf occupied by adhesive in excess of native bone on the outer side of specimen; max around circumference |
| Inner overfill | Distance along depth of kerf occupied by adhesive in excess of native bone on the inner side of specimen; max around circumference |
| Bone curvature angle | Angle perpendicular to bone outer surface relative to specimen axis; averaged around circumference |
| Flap thickness | Thickness of native bone perpendicular to bone outer surface; averaged around circumference |
| Flap elevation | Distance perpendicular to bone outer surface that the flap was elevated/depressed; max absolute value around circumference |
i = 0–359 for each slice image. Indentation indicates the derivation hierarchy.
A typical slice image with the analyzed parameters is shown in Supplemental Table 2. The kerf midline axis for each radial slice was established 1 mm away from the peripheral bone margin. The inner and outer surfaces of bone and adhesive were identified by traversing the kerf midline axis in the pretreatment and posttreatment slice images, respectively. The resulting kerf depth and adhesive thickness were also used to obtain kerf interface and adhesive interface areas, respectively. Adhesive overfill/underfill, when it occurred, was quantified on both inner and outer surfaces.
The bone normal angle was representative of the curvature of the skull segment. The thickness of the bone along this axis was measured. The elevation of the flap relative to native bone was evaluated along the bone normal direction, 10 mm from the kerf, as shown in Supplemental Table 2.
Flap sizes 1 and 2 (n = 32) were used for leak testing and quasistatic compression testing. They were fixated by different neurosurgeons at each test site, who applied adhesive around the entire kerf circumference. Half these flaps (n = 16, consisting of 9 size 1 flaps and 7 pterional size 2 flaps) were tested for leak resistance 10 minutes after the last batch of adhesive used to fixate the flap was mixed. The specimens were maintained hydrated and at physiological temperature (37°C) during this interval. The previously described pressure testing setup was used to leak test the fixated flaps. Hydrostatic pressure was increased from 5 to 40 mm Hg, at steps of 5 mm Hg, with pressure held for 5 seconds at each step. The water was maintained at physiological temperature (37°C) throughout the test. The pressure at which water leaked from the specimen was recorded. Because there was a visible gap around the circumference on the TPS-fixated specimens, pressurized CSF leak testing was not attempted on them.
Afterward, these specimens were returned to the biomechanical testing facility, in a refrigerated state, for quasistatic compression until failure of the adhesive fixation within 10 days of the simulated procedure. Testing was done at room temperature (20°C). After failure, the adhesive was removed from the flap and peripheral bone, and the flaps were fixated using TPS as per each surgeon’s recommendation. The same specimens were then tested again under quasistatic compression.
The quasistatic compression test setup, including images of specimens before and after testing, is shown in Fig. 2. The specimens were compressed at a rate of 1 mm/min using a servohydraulic load frame (Instron 8501). A 3/4-inch cylindrical actuator was used on flap size 1 and a 3-pronged actuator was used for flap size 2.
A–B: Quasistatic compression test setup. C: Typical cranial flaps before and after testing. Figure is available in color online only.
The test endpoint was reached when 4 mm of total displacement occurred or a force of 10% less than the peak value was reached, whichever occurred earlier. Peak force, force at 1 mm of displacement, and stiffness were analyzed. Stiffness—or slope of the force versus displacement data—was determined by using unconstrained linear regression of the data between 25% and 75% of peak force.
Flap sizes 3 and 4 (n = 32) were used for impact testing. They were fixated by the different neurosurgeons at each test site using titanium hardware (Stryker Universal Neuro III 1.5 mm cranial fixation system) placed according to the surgeons’ clinical preference. The mechanical compliance of these specimens was tested at a subthreshold impact energy of 6 J. The metallic hardware was then removed, and the same flaps were fixated by the same neurosurgeons using adhesive alone that was applied partially around the kerf circumference. The newly fixated flaps were again tested at subthreshold impact energies of 6 J and 12 J, 20 minutes after the last batch of adhesive used to fixate the flap had been mixed. The specimen was maintained hydrated and at physiological temperature (37°C) throughout this interval.
Afterward, the specimens were either immediately tested to failure under impact at the test site, or they were returned to the biomechanical testing facility in refrigerated condition for impact to failure testing within 10 days. This testing was done at room temperature (20°C).
The impact test setup is shown in Fig. 3. An impactor with adjustable weight was dropped from different heights to deliver impact energies up to 60 J. The impactor had a 1-inch diameter polyvinyl chloride (PVC) tip and was equipped with a load cell (Loadstar RSB6) to measure the impact force. A linear variable differential transformer (LORD S-LVDT-4) probe was affixed to the inner surface of each flap with cyano-acrylate to measure the displacement of the flap. Impact energy was increased from 6 J, at steps of 6 J, until the specimen failed. At each impact energy level, peak force, maximum deflection, and plastic deflection were recorded. A specimen was considered failed if deflection at the center was greater than 3.9 mm or if it was fractured at either the bone or adhesive.
A–D: Impact test setup. E: Typical cranial flaps before and after testing. PVC = polyvinyl chloride. Figure is available in color online only.
Some specimens were reserved in the intact condition to test the mechanical response of the native skull. No craniotomy or fixation was performed. A total of 8 specimens (n = 8) with flap sizes 1 and 2 were tested under the same quasistatic compression protocol as the resected and fixated specimens. A total of 6 specimens (n = 6) with flap sizes 3 and 4 were tested using the impact protocol.
All statistical analysis was done using R (version 4.1.1). Statistical significance was tested at α = 0.05.
Results
Of the 36 intact (native) skull specimens tested, 18 (50%) were found to leak at a pressure of 30 mm Hg held for 5 seconds. The native leak test results did not correlate with donor anthropometry or a pathological condition.
VIDEO 1. Typical mixing and application of bone adhesive (Tetranite, RevBio Inc.) for cranial flap fixation. © RevBio Inc, published with permission. Click here to view.
The mean ± SD time from mixing the adhesive until the flaps were ready for closure without sealing was 265 ± 24 seconds (n = 66), and the time until full sealing and contouring was 544 ± 85 seconds (n = 30). In comparison, the time taken for fixation with hardware was 293 ± 69 seconds (n = 27).
Table 3 shows the CT measurements grouped by flap size. All measurements varied significantly with flap size with some exceptions. Average flap thickness was not significantly different between flap sizes 1 and 3. Peak flap elevation was not significantly different between flap sizes 1, 3, and 4 but was significantly higher for flap size 2. Average kerf angle and bone normal angle (curvature) were not significantly different between flap sizes 1 and 2. Adhesive interface area was not significantly different between flap sizes 1, 3, and 4 but was significantly higher for flap size 2. Adhesive percent fill over kerf circumference, as well as over total interface area, was significantly lower for flap sizes 3 and 4. Inner and outer overfills have been reported as the peak values around the circumference of each flap. Peak outer overfill did not vary significantly with flap size. On the outer surface, overfill occurred in all specimens of flap sizes 1 and 2, with peak protrusion between 1 and 2 mm. It occurred in 69% of specimens with flap sizes 3 and 4. Peak inner overfill was significantly higher for flap size 2 compared with the other flap sizes. On the inner surface, overfill occurred in nearly all specimens with flap size 2 (pterional flaps) with an average peak protrusion of 3 mm. Among the rest (flap sizes 1, 3, and 4), inner overfill occurred in 56% of specimens with peak protrusion of 1 to 2 mm. An average of 33% to 61% of the kerf depth was filled with adhesive. There was no significant correlation between flap elevation and outer or inner overfill with flap thickness, kerf angle, or bone angle, except peak inner overfill was correlated with flap thickness (p = 0.001, linear model).
CT measurements of the different flap sizes
| Characteristic | Flap Size | |||
|---|---|---|---|---|
| 1 | 2 | 3 | 4 | |
| Circumference, mm | 157 ± 14 (14)* | 247 ± 23 (15)* | 233 ± 17 (15)* | 300 ± 22 (14)* |
| Flap thickness, mm | 7.02 ± 1.86 (14)* | 5.91 ± 0.64 (15) | 6.39 ± 1.02 (15)* | 5.81 ± 0.97 (14)* |
| Angle relative to axis over circumference, ° | ||||
| Kerf cut | 27 ± 7 (14)* | 30 ± 3 (15) | 32 ± 3 (15)* | 37 ± 4 (14)* |
| Bone normal | 21 ± 4 (14)* | 23 ± 2 (15) | 26 ± 3 (15)* | 33 ± 3 (14)* |
| Adhesive fill, % | ||||
| Over circumference | 99 ± 2 (14)* | 99 ± 2 (15)* | 71 ± 9 (10) | 72 ± 9 (10)* |
| Over interface area | 70 ± 18 (14)* | 81 ± 18 (15)* | 45 ± 17 (10) | 53 ± 18 (10) |
| Over kerf depth | 71 ± 18 (14)* | 82 ± 19 (15) | 54 ± 15 (10)* | 58 ± 17 (10) |
| Thickness along depth of kerf, mm | ||||
| Full kerf | 7.25 ± 1.47 (14)* | 5.68 ± 1.02 (15)* | 6.83 ± 0.74 (10) | 5.96 ± 0.99 (10)* |
| Adhesive | 5.03 ± 1.15 (14)* | 4.61 ± 1.11 (15) | 3.64 ± 1.01 (15)* | 3.45 ± 1.00 (14)* |
| Total interface area, mm2 | ||||
| Full kerf | 1141 ± 269 (14)* | 1405 ± 286 (15)* | 1598 ± 176 (10) | 1793 ± 315 (10) |
| Adhesive | 777 ± 180 (14)* | 1122 ± 284 (15)* | 697 ± 218 (10)* | 928 ± 250 (10)* |
| Alignment | ||||
| Flap elevation, mm | 1.17 ± 0.51 (14)* | 2.00 ± 0.93 (15) | 1.02 ± 0.29 (10)* | 1.00 ± 0.53 (10)* |
| Peak overfill, mm | ||||
| Outer | 1.26 ± 0.93 (14)* | 1.86 ± 0.69 (15) | 1.34 ± 0.89 (10) | 1.04 ± 0.44 (10) |
| Inner | 1.12 ± 0.98 (9) | 2.99 ± 1.39 (14)* | 1.91 ± 1.00 (6) | 2.23 ± 1.40 (9) |
Values are shown as number or mean ± SD (sample size) unless indicated otherwise.
Measurement was significantly different from the flap size (a = 0.05, linear model).
Of the 16 adhesive-fixated craniotomy specimens tested (9 flaps were size 1 and 7 flaps were size 2), 14 (88%) did not leak during testing up to 40 mm Hg. Both specimens that leaked were flap size 2. One specimen leaked at 25 mm Hg and the other leaked at 40 mm Hg.
The distribution of the different flap sizes into the different mechanical test groups and the corresponding mechanical test results are shown in Table 4.
Mechanical test groups and test results by flap size
| Characteristic | Flap Size | |||
|---|---|---|---|---|
| 1 | 2 | 3 | 4 | |
| Sample size | ||||
| Total | 21 | 20 | 18 | 19 |
| Mechanical testing group | ||||
| Quasistatic testing | ||||
| ≤10 days after fixation | 16 | 17 | ||
| Native specimens | 5 | 3 | ||
| Impact testing | ||||
| ≤10 days after fixation | 10 | 9 | ||
| 20 mins after fixation | 6 | 6 | ||
| Native specimens | 2 | 4 | ||
| Quasistatic test results | ||||
| Peak force, N | ||||
| Adhesive | 5125 ± 2248 (16) | 5694 ± 1528 (17) | ||
| TPS | 113 ± 104 (15) | 116 ± 78 (16) | ||
| Native | 7171 ± 2303 (5) | 5383 ± 1775 (3) | ||
| Stiffness, N/mm | ||||
| Adhesive | 4771 ± 1912 (16) | 3679 ± 761 (17) | ||
| TPS | 58 ± 44 (15) | 147 ± 86 (16) | ||
| Native | 4719 ± 2419 (5) | 2528 ± 982 (3) | ||
| Force at 1 mm, N | ||||
| Adhesive | 3314 ± 1861 (16) | 3074 ± 850 (17) | ||
| TPS | 50 ± 29 (15) | 90 ± 35 (16) | ||
| Native | 3285 ± 2267 (5) | 2094 ± 668 (3) | ||
| 6-J impact test results | ||||
| Total deflection, mm | ||||
| Adhesive | 248 ± 58 (16) | 234 ± 74 (15) | ||
| TPS | 3114 ± 823 (16) | 2762 ± 790 (15) | ||
| Native | 298 ± 53 (2) | 364 ± 131 (4) | ||
| Plastic deflection, mm | ||||
| Adhesive | 64 ± 46 (16) | 61 ± 39 (15) | ||
| TPS | 1731 ± 1061 (15) | 1648 ± 819 (15) | ||
| Native | 143 ± 29 (2) | 184 ± 75 (4) | ||
| Peak force, N | ||||
| Adhesive | 481 ± 379 (16) | 331 ± 329 (15) | ||
| TPS | 746 ± 916 (16) | 377 ± 226 (15) | ||
| Native | 389 ± 172 (2) | 683 ± 438 (4) | ||
| 12-J impact test results | ||||
| Total deflection, mm | ||||
| Adhesive | 438 ± 301 (16) | 425 ± 272 (15) | ||
| Native | 187 ± 21 (2) | 329 ± 155 (4) | ||
| Plastic deflection, mm | ||||
| Adhesive | 188 ± 207 (16) | 150 ± 90 (15) | ||
| Native | 181 ± 12 (2) | 363 ± 155 (4) | ||
| Peak force, N | ||||
| Adhesive | 771 ± 523 (16) | 666 ± 628 (15) | ||
| Native | 535 ± 52 (2) | 903 ± 433 (4) | ||
| Staircase impact to failure results | ||||
| Prefailure deflection, mm | ||||
| Adhesive | 703 ± 412 (5) | 1552 ± 968 (15) | ||
| Native | 1976 ± 788 (2) | 1742 ± 425 (4) | ||
| Failure force, N | ||||
| Adhesive | 1053 ± 351 (16) | 1109 ± 464 (15) | ||
| Native | 1574 ± 421 (2) | 1790 ± 147 (4) | ||
| Failure energy, J | ||||
| Adhesive | 24 ± 11 (16) | 32 ± 14 (14) | ||
| Native | 42 ± 15 (4) | |||
| Sustained energy, J | ||||
| Adhesive | 19 ± 10 (15) | 28 ± 16 (15) | ||
| Native | 48 ± 17 (2) | 36 ± 15 (4) | ||
Values are shown as number or mean ± SD (sample size) unless indicated otherwise.
VIDEO 2. Quasistatic compression response of the adhesive-fixated (left) and TPS-fixated (right) bone flaps. Force versus displacement data of the 2 flaps were plotted. © RevBio Inc, published with permission. Click here to view.
Correlation of quasistatic (A), subfailure impact (B), and staircase to failure impact (C) mechanical parameters with fixation type. Asterisks indicate a statistically significant difference (α = 0.05, t-test; p values indicating significant differences are shown). The labels show the average value and sample size. Figure is available in color online only
None of the parameters were significantly different between adhesive fixation and native skull. Thus, adhesive fixation of the bone flaps was mechanically superior to TPS fixation and statistically equivalent to native skull.
Linear regression indicated the correlation of all the quasistatic response parameters of the adhesive-fixated specimens with average flap thickness. Peak force in adhesive-fixated specimens also correlated with adhesive fill area and kerf interface area. Only the quasistatic force response at 1 mm of deflection of the TPS-fixated specimens correlated with the bone angle (curvature of the bone flap).
The distribution of the subthreshold impact response values by fixation type are shown in Fig. 4B. The results for native skull are also shown. The Student t-test was used to compare the 2 fixation types and native skull. Total and plastic deflection under 6-J impact were significantly less for adhesive-fixated flaps compared with TPS-fixated specimens. They were also significantly less for the native specimens compared with TPS (Video 3).
VIDEO 3. Impact response of TPS-fixated (left) and adhesive-fixated (center and right) bone flaps at impact energies of 6, 6, and 12 J, respectively. Peak total deflection, residual plastic deformation, and impact force were charted for the 3 flaps. The video has been paused at time points corresponding to peak deflection and end of elastic recovery. © RevBio Inc, published with permission. Click here to view.
There was no significant difference in any subthreshold impact metric between adhesive-fixated and native specimens at both impact energy levels, with 1 exception. Plastic deflection at 6-J impact was significantly less in adhesive-fixated flaps than native specimens. Peak force was not significantly different between adhesive-fixated, TPS-fixated, and native specimens at both energy levels.
The distributions of staircase impact to failure measurements for adhesive-fixated flaps and native specimens are shown in Fig. 4C. The Student t-test was used to compare adhesive-fixated flaps and native skull. All metrics except failure energy were significantly different.
Use of the t-test showed that the results of the staircase impact to failure tests on adhesive-fixated flaps were not significantly different between those tested 20 minutes after fixation and those tested 10 days later (p > 0.05).
Linear regression revealed the correlation of total deflection of TPS-fixated flaps under 6-J impact with bone normal angle (flap curvature). Plastic deflection of TPS-fixated flaps under 6-J impact correlated with total kerf interface area, while total deflection of adhesive-fixated flaps under 12-J impact correlated with bone normal angle (flap curvature). None of the staircase impact to failure measurements correlated significantly with CT measurements.
One each of the adhesive-fixated and native specimens did not fail at an impact energy up to 60 J. In the majority of adhesive-fixated flaps, both adhesive and bone broke. Only bone broke in 1 case and only adhesive broke in 7 cases.
Discussion
This article presents data from craniotomy test specimens that were fixated in a simulated surgical environment by 16 neurosurgeons. Although identical instructions for use of the adhesive were provided to each surgeon, they had no prior experience with the material. Therefore, the spectrum of outcomes represents a worst-case scenario.
Leakage of CSF reportedly occurs in 4.6% of craniotomies in general15 and ranges as high as 13% in posterior fossa surgery.16 Many studies also reported that the use of dural sealants did not change the CSF leakage rate.7,15,17 Using a novel test apparatus, seepage of fluid through the natural vasculature of native skull bone was discovered, with 50% of intact skull specimens leaking at a pressure of 30 mm Hg held for 5 seconds. In total, 88% of the adhesive-fixated craniotomy specimens resisted leakage under 40 mm Hg pressure, indicating that the adhesive can serve as a secondary barrier to CSF leakage. This could potentially reduce the likelihood of infection.
Wang et al.18 previously tested cranial flaps under quasistatic compression. They reported peak forces of 83 ± 9, 333 ± 53, and 385 ± 63 N for flaps fixated using suture, wire, and titanium clamps, respectively. In this study, adhesive fixation was much stronger (5418 ± 1902 N) than hardware fixation (115 ± 90 N); in fact, it produced bone flap strength comparable to intact skull (6501 ± 2188 N). The superior strength of adhesive fixation relative to hardware fixation under impact-type loading and its comparability to intact skull were also demonstrated, despite the kerf in these specimens being only partially filled (72% ± 9%) around the circumference. Reoperations necessitated by bone flap depression have been reported in 3% of patients who underwent craniotomy.19 Greater mechanical fixation strength could minimize or even eliminate bone flap depression. In fact, in this study, all the quasistatic strength parameters and impact failure energies were statistically indistinguishable when adhesive-fixated cranial bone flaps were compared with native skull. This may allow patients with adhesive-fixated bone flaps to return sooner to vigorous activity (e.g., sports).
Spetzler6 and Di Lorenzo et al.5 have written that even low-profile miniplates can disfigure the contour of the skin in patients with thin scalp. Patients frequently palpate the skull hardware and cause skin irritation. In this study, we used CT to accurately measure flap elevation and adhesive overfill. Median flap elevation and outer overfill were 1.24 mm and 1.33 mm, respectively. Although this overfill is comparable to the thickness of miniplates (1.5 mm),1 the adhesive is spread over a greater distance with a shallow ramp leading up to the peak and thus has a smooth contour without the step-off that can be palpated at the edge of a plate. For this reason, the adhesive should be inconspicuous to the patient.
Measurements from CT also demonstrated the efficacy of the adhesive and delivery system design (cannula size and viscosity) to mitigate adhesive overfill and extravasation beneath the skull. Peak inner overfill (maximum protrusion of adhesive around the entire circumference of the kerf on the inner surface) was 2.20 mm. In the flaps where the entire circumference of the bone flaps had been filled, the adhesive penetrated only 71% to 82% of the kerf depth on average but penetrated only 54% to 58% of the kerf depth in the others.
The adhesive can be drilled/cut using standard neurosurgical tools, even at advanced time points. In cases of reoperation and need for emergency decompression, the adhered flap can be removed by drilling the original kerf.
Although we attempted to simulate full in vivo conditions, including temperature, wet bone, and CSF pressurization after fixation, a limitation of this study was the absence of active bleeding at the surgery site. The lack of the dura, brain, and scalp in our specimens was another deviation from in vivo conditions and may be another limitation. This was considered a worst-case scenario for testing resistance to CSF leakage, resistance to fracture under impact loading, and adhesive overfill on the inner surface of the skull. Because we tested cadaver tissue, adhesive material resorption and bone remodeling were absent. Therefore, the measurement of mechanical strength and stiffness of our specimens, particularly at 10 days after flap fixation, may differ from in vivo strength. The fact that the same adhesive was tested in vivo in a sheep craniotomy model and showed significantly greater bone flap mechanical strength at 1 year postoperatively compared with adhesive-fixated flaps at 12 weeks after surgery is a manifestation of the osteoconductive properties of the material in a living recipient.8
Conclusions
In this laboratory study that compared the use of standard TPS with an osteoconductive, bioresorbable, mineral-organic bone adhesive for human cranial bone flap fixation, the bone adhesive was biomechanically superior. In fact, adhesive-fixated flaps were comparable in strength to the intact skull itself. Additionally, the adhesive sealed the kerfs around the bone flaps and mitigated simulated fluid leakage. This novel material has the potential to improve cranial bone flap fixation biomechanically, cosmetically, and hydrostatically.
Acknowledgments
Research reported in this publication was supported by the National Institute of Neurological Disorders and Stroke of the National Institutes of Health (award no. R44NS115386). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. The following surgeons contributed to this study by conducting fixation on the cranial bone flaps: Mohammad A. Aziz-Sultan, Michael Anthony Mooney, Nirav Patel, and Alaa Montaser at Brigham and Women’s Hospital Study Group, Boston, MA; Sanjay Konakondla, Michel Lacroix, Edward Monaco, Clemens M. Schirmer, and Cameron J. Brimley at Geisinger Medical Center Study Group, Danville, PA; and Frederick Boop, Madison Michael, Kenan Arnautovic, Frank Farokhi, Vincent Nguyen, and Daniel Hoit at Semmes-Murphey Clinic Study Group, Memphis, TN.
Disclosures
Dr. Smith is a consultant for RevBio, Inc. Dr. Foley is a consultant for Medtronic; owns stock in Accelus, DiscGenics, DuraStat, Medtronic, NuVasive, RevBio, Spine Wave, Tissue Differentiation Intelligence, True Digital Surgery, and Vori Health; holds patents with DiscGenics, Medtronic, and NuVasive; receives royalties from Medtronic; and serves on the board of directors of DiscGenics, DuraStat, RevBio, Tissue Differentiation Intelligence, and True Digital Surgery. Dr. Boruah is an employee of RevBio, Inc. Dr. Slotkin owns stock in RevBio. Dr. Woodard is a consultant for RevBio and Vallum, Inc.; and owns stock in Medtronic. Dr. Lazor owns stock in Clozex and received study-related clinical or research support from RevBio. Mr. Brown owns stock in RevBio; is an employee of RevBio; and holds patents with RevBio. Ms. McDonough is an employee of RevBio. Mr. Hess owns stock in RevBio; is an employee of RevBio; and holds patents with RevBio. Dr. Citters received study-related clinical or research support from RevBio; and receives non–study-related clinical or research support from DePuy Synthes Joint Reconstruction, Medacta, and ConforMIS.
Supplemental Information
Videos
Video 1. https://vimeo.com/765639554.
Video 2. https://vimeo.com/765641722.
Video 3. https://vimeo.com/765642278.
Online-Only Content
Supplemental material is available with the online version of the article.
Supplemental Tables 1 and 2. https://thejns.org/doi/suppl/10.3171/2022.10.JNS221657.
Previous Presentations
Podium presentation (Smith T, Foley K, Slotkin J, et al. Novel osteoconductive, bioresorbable bone adhesive provides superior cranial flap fixation strength and reduced CSF leak potential in a cadaver model) at the Congress of Neurosurgeons 2021 Annual Meeting, Austin, TX, October 20, 2021.
Author Contributions
Conception and design: Boruah, Smith, Foley, Cavaleri, Brown, McDonough, Hess, Van Citters. Acquisition of data: Boruah, Smith, Foley, Slotkin, Woodard, Lazor, Cavaleri, McDonough, Van Citters. Analysis and interpretation of data: Boruah, Cavaleri, Van Citters. Drafting the article: Boruah, Foley. Critically revising the article: Boruah, Smith, Foley, Slotkin, Lazor, Cavaleri, Brown, McDonough, Hess, Van Citters. Reviewed submitted version of manuscript: Boruah, Smith, Foley, Slotkin, Lazor, McDonough, Van Citters. Approved the final version of the manuscript on behalf of all authors: Boruah. Statistical analysis: Boruah. Administrative/technical/material support: Smith, Foley, Slotkin, Woodard, Lazor, Cavaleri, Brown, McDonough, Hess, Van Citters. Study supervision: Smith, Brown, Hess, Van Citters.
References
- 1↑
Marciano FF, Vishteh AG. Fixation techniques for cranial flap replacement. Oper Techn Neurosurg. 1998;1(1):50–56.
- 2↑
Rashidi A, Adolf D, Karagiannis D, Melhem OB, Luchtmann M. Incidence and risk factors for skull implant displacement after cranial surgery. World Neurosurg. 2019;126:e814–e818.
- 3↑
Schwarz F, Dünisch P, Walter J, Sakr Y, Kalff R, Ewald C. Cranioplasty after decompressive craniectomy: is there a rationale for an initial artificial bone-substitute implant? A single-center experience after 631 procedures. J Neurosurg. 2016;124(3):710–715.
- 4↑
Kim JH, Kim JH, Kwon TH, Chong K, Hwang SY, Yoon WK. Aseptic bone flap resorption after cranioplasty with autologous bone: incidence, risk factors, and clinical implications. World Neurosurg. 2018;115:e111–e118.
- 5↑
Di Lorenzo N, Mouchaty H, Shamsaldin M, Gallina P, Maleci A. Cranial bone flap fixation with microplates and screws: a new application technique. Technical note. J Neurosurg Sci. 2004;48(1):55–56.
- 6↑
Spetzler RF. Bone flap fixation: a new technique. Technical note. J Neurosurg. 1997;87(3):475–476.
- 7↑
Shi ZH, Xu M, Wang YZ, et al. Post-craniotomy intracranial infection in patients with brain tumors: a retrospective analysis of 5723 consecutive patients. Br J Neurosurg. 2017;31(1):5–9.
- 8↑
Foley KT, Woodard EJ, Slotkin JR, et al. Cranial flap fixation in sheep using a resorbable bone adhesive. J Neurosurg. 2021;134(2):621–629.
- 9↑
Norton MR, Kay GW, Brown MC, et al. Bone glue - the final frontier for fracture repair and implantable device stabilization. Int J Adhes Adhes. 2020;102:102647.
- 10↑
Kirillova A, Kelly C, von Windheim N, Gall K. Bioinspired mineral-organic bioresorbable bone adhesive. Adv Healthc Mater. 2018;7(17):e1800467.
- 11
Kesseli FP, Lauer CS, Baker I, Mirica KA, Van Citters DW. Identification of a calcium phosphoserine coordination network in an adhesive organo-apatitic bone cement system. Acta Biomater. 2020;105:280–289.
- 12↑
Kirillova A, Nillissen O, Liu S, Kelly C, Gall K. Reinforcement and fatigue of a bioinspired mineral-organic bioresorbable bone adhesive. Adv Healthc Mater. 2021;10(2):e2001058.
- 13
Sugita R, Jones AA, Kotsakis GA, Cochran DL. Radiographic evaluation of a novel bone adhesive for maintenance of crestal bone around implants in canine oversized osteotomies. J Periodontol. 2022;93(6):924–932.
- 14↑
Cochran DL, Jones A, Sugita R, et al. Immediate dental implant stabilization in a canine model using a novel mineral-organic adhesive: 4-month results. Int J Oral Maxillofac Implants. 2020;35(1):39–51.
- 15↑
Kehler U, Hirdes C, Weber C, et al. CSF leaks after cranial surgery — a prospective multicenter analysis. Innov Neurosurg. 2013;1(1):49–53.
- 16↑
Dubey A, Sung WS, Shaya M, et al. Complications of posterior cranial fossa surgery—an institutional experience of 500 patients. Surg Neurol. 2009;72(4):369–375.
- 17↑
Kinaci A, Algra A, Heuts S, O’Donnell D, van der Zwan A, van Doormaal T. Effectiveness of dural sealants in prevention of cerebrospinal fluid leakage after craniotomy: a systematic review. World Neurosurg. 2018;118:368–376.e1.
- 18↑
Wang YR, Su ZP, Yang SX, Guo BY, Zeng YJ. Biomechanical evaluation of cranial flap fixation techniques: comparative experimental study of suture, stainless steel wire, and rivetlike titanium clamp. Ann Plast Surg. 2007;58(4):388–391.
- 19↑
Tsang ACO, Hui VKH, Lui WM, Leung GK. Complications of post-craniectomy cranioplasty: risk factor analysis and implications for treatment planning. J Clin Neurosci. 2015;22(5):834–837.
