Customized and Cost-Effective 3D Printed Mold for Cranioplasty: India's First Single Center Experience
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.319221
Source of Support: None, Conflict of Interest: None
Keywords: 3D cranioplasty, 3D molds, PEEK, PMMAKey Message: Customized 3D printing and PMMA cranioplasty is a cost effective and patient satisfying procedure.
The term cranioplasty refers to surgical reconstruction of a skull vault defect following previous surgical intervention for any intracranial pathology. The choice of material used for cranioplasty can be either biological or nonbiological. Factors like ideal timing of cranioplasty, surgical technique, and the material used has continuously evolved over time. However, the patient's primary expectation from undergoing the procedure is repair of the cosmetic defect. Along with cosmetics, cranioplasty also provides a good protective covering to the intracranial contents and helps in re-establishing the normal CSF and blood flow dynamics, which would have been modified by craniectomy (1). A recent detailed review by Craniosafe group evaluating the safety of cranioplasty materials reported that autologous bone is the most commonly used material owing to its cost effectiveness and easy availability. However, the chances of bone resorption and infection preclude its routine use (2, 3). Other studies also conclude that the autologous bone carries a higher reoperation rate mainly due to resorption. Previous bone flap infection and tumor erosion preclude its use and institutional bone banks cause a heavy financial burden. The other available materials like titanium, PolyEtherEtherKetone (PEEK), poly methyl methacrylate (PMMA) have their own benefits and shortcomings. The available evidence is inconclusive about the superiority of one material over the other. The most cost effective and widely used option is PMMA, a polymer with thermoplastic and radiolucent properties comparable to bone strength. It is easier to design handmade implants using PMMA, to fit different shapes of skull defects. However, when the defects are larger as in decompressive craniectomy, the cosmetic outcome is usually less than ideal. It is difficult to get an ideal fit, desired thickness, contour, and a smooth surface, each of which are important factors in the outcome of the surgery. Certain techniques like the use of native bone as a template, neuro navigation, and computer-assisted design (CAD) can help in overcoming these problems., Three-dimensional (3D) printing has gained widespread medical attention in planning, execution, and analysis for complex surgical scenarios. Three-dimensional printed titanium flaps are expensive and take several weeks to plan and is not universally accepted due to financial reasons. Hence, by combining CAD and 3D printing technology, we adapted a cost-effective method and provided a cosmetically accurate implant using PMMA in patients having large craniectomy defects without native bone flaps. The current paper is first of its kind in India from a single tertiary care center, where the first flap was created in 2015, briefing our experience and outcomes in this technology.
The study period extended for 3 years with prospective inclusion of patients from July 2015 to April 2018. A total of 114 cranioplasty were done during this time period and 25/114 underwent cranioplasty for large skull defects for nonavailability of autologous bone flap. They were included in the current study and underwent cranioplasty using 3D printed template molds designed by an indigenously developed software that utilizes Computer Assisted Designing (CAD). These molds were used intraoperatively to create PMMA implants. The clinical demographics, primary pathology for undergoing decompressive craniectomy, timing of cranioplasty after initial surgery, reason for nonavailability of the bone flap, duration of the surgery, postoperative complications, duration and outcomes during follow-up, and patient satisfaction index were analyzed. The visual analog score for cosmesis (VASC) is scored by the patients with scale ranging from 1 to 10. Odom's score is a four-point rating scale used to assess surgical outcomes usually after cervical spine procedures based on overall patient satisfaction outcomes. The same rating scale was modified to fit for post cranioplasty outcomes as – excellent, good, fair, and poor. The criteria used for this scoring include postoperative cosmetic outcome based on the overall fit and symmetry. The study could not get approval by local institutional ethics committee, as the study antedated the committee and necessary consent was obtained from patients included for the study.
The workflow included two stages:
Preoperative design and manufacturing of the implant mold
The basic approach involved in modeling cranioplasty implants is reconstruction of the patient skull along with the defect window and creation of an intact skull using mirrored imaging technique from the unaffected side. The final step is carving out a mold having contours of the outer, inner, and contact surfaces. For reconstruction, the primary step is data acquisition of the skull by thin slice [<0.6 mm] CT scans acquired in DICOM format. CT scan is the preferred imaging technique because of its inherent image contrast between bone and soft tissue. This provides ease in post processing techniques and creation of 3D rendering of the skull. The 3D solid model of the skull obtained in the stereo lithographic format is processed using indigenously developed software created by Osteo3D Inc., Bangalore, India. It is further processed using a solid surface modeling method called nonuniform rational B-Spline-based method (NURB method).
In this NURB method, the defective skull in the solid 3D model [Figure 2] is used as a starting design data. By applying mathematical modeling, the patient's intact skull is generated by mirroring the unaffected side of the skull [Figure 3]. From the newly reconstructed skull [Figure 4] generated using this method, the area covering the defect is used to contour the outer, inner, contact surface, and thickness of the model. Finally, with these contours, a mold is carved out from a rectangular block [Figure 5]. The digital mold model created is loaded to a commercial polylactic acid 3D printer that prints based on Fused Deposition Modelling. The final 3D printed mold has a convex and a concave cavity pertaining to the shape of the inner and outer surfaces of the designed bone defect [Figure 6]. It also has a 5–6 mm cavity for the PMMA implant to be molded.
Intraoperative implant molding [Figure 1]
The 3D printed mold is labeled with patient identification details. The mold is sterilized with ethylene oxide [ETO] gas sterilization and stored in sterile packing. The scalp region is thoroughly cleaned and prepared using chlorhexidine scrub and iodine solution. Hair is clipped wide around the incision site. Skin flap is raised both towards and away from the defect preserving the dural covering and to achieve a lax closure. Meticulous haemostasis of the craniectomy cavity is secured. After the craniectomy defect is exposed, the process of cement preparation begins in a sterile bowl where the antibiotic premixed PMMA polymer and monomer are mixed in appropriate ratio. The mixture is continuously kneaded till it polymerizes to form a thick paste. Continuous kneading is the key for uniform mixing of the powder into a paste without clumping. Meanwhile, both surfaces of the mold are first layered with sterile paraffin oil to avoid sticking of the PMMA. Once the PMMA forms into a thick paste, it is poured into the concave surface of the mold cavity and then evenly layered. The convex side of the mold is then placed over the concave cavity as it fits the contour. The template mold is held tightly in place with the holders. Any spill over from the sides of the mold is removed flush from the edges. The template mold is left untouched for about 5–7 min until the PMMA sets in and becomes firmer in consistency, thus obtaining the desired variable thickness and generated by the CAD. Once the mixture is set, it can be delivered out like a coconut kernel out of its shell taking the shape of the desired implant. The exothermic reaction is allowed to happen on the table. Hanging thin slivers of PMMA can be cut away with scissors or a knife or can be nibbled away as deemed necessary.
The exposed craniectomy site is irrigated with saline and cleaned. Haemostasis is maintained throughout, and the PMMA implant prepared using the 3D printed mold is fitted over the defect. The corresponding author (SG) prefers separating the temporalis muscle off the dura for better anchoring of the implant and cosmesis. The PMMA 3D cranioplasty implant is fixed using miniplates and screws, and the wound is closed with a wound drain in place. A couple of 5 mm perforations are made in the created 3D flap to prevent collection beneath.
The mean age of the patients is 38.43 ± 14.6 years of age (Range, 9–66 years). The male to female ratio is 2.57: 1. The most common primary pathology for undergoing craniectomy is stroke (n = 13; 52%). Among the stroke cases, 7/12 had ischemic stroke and 6/12 had a haemorrhage. Ten (40%) patients underwent craniectomy following traumatic brain injury and two cases (8%) had tumor. The secondary pathologies leading to nonavailability of bone flap were infection (n = 14; 56%), flap resorption (n = 4; 16%), and other causes like trauma or tumor in seven patients (28%). [Table 1] gives the demographic details, primary cranial pathology, and causes for a 3D designed bone flap.
The mean dimensions of the bone flap are 12.43 ± 1.61 mm × 14.97 ± 2.14 mm with an average thickness of 4.9 ± 0.7 mm. The mean time required for manufacturing the 3D printed template was 13.24 ± 2.18 h. The average delay from bone flap infection to cranioplasty is nine months.
All the patients were reviewed on follow up with a mean duration of 11.8 ± 3.7 months. On follow up, the median Odom's score is excellent in 52% of cases [n-13], good in 40% of cases [n-10], and fair in 8% of cases [n-2]. The mean VASC score was 8.19 ± 1.36. Three patients developed complications post cranioplasty that includes 3D-bone flap infection, flap exposure due to skin necrosis, and extradural collection post-surgery, which required immediate re-exploration, evacuation of the clot, and replacement of the customized flap. The other two patients required redo construction of the bone flap using their previous template. The redo surgery was performed after adequate treatment of the infection with good outcomes.
Cranioplasty is an ancient neurosurgical procedure, with historical reports of gold and silver plates being used for covering trephined skull defects as evidenced in 2000 year old Peruvian skulls. The first reported case was by Van Meekren in 1670 where he used a xenograft from a dog to repair a human skull. Interestingly due to religious opposition, the successful bone graft was forced to be removed as it was described as “marring Gods image” by using beast bones on humans. Since then a range of options from goose skull to rib grafts and boiled grafts to dead bone grafts have been reported in the technique evolution for cranioplasty. The physiological and psychological impact of having a large skull defect has been emphasized enough in the available literature. An ideal result through cranioplasty is to get a good cosmetic correction and restoration of the physiological fluid dynamics of a closed skull. Although autologous bone gives an ideal cosmetic result, recent studies have proved them inferior to alloplastic materials. Various materials have been tried till date as basic ingredient of the implant. Titanium is widely considered as a viable alternative owing to its good biocompatibility, infection resistance, and biomechanical stability. The real limitation with titanium is its cost, poor mold-ability intraoperatively according to a surgeon's need, radio-opacity, and conduction of heat and cold., PMMA, among all other materials, is easily available and has been proven to have good mechanical strength from long-term clinical experience in prosthodontics. It is radiolucent and can be hand made during the operative procedure and is relatively inexpensive. PMMA is also conducive to CT and MRI imaging and has minimum artefactual tendency.
We reviewed the major studies that had used PMMA as either a “prefabricated” – the implant is manufactured prior to the surgery using CAD – and directly used during surgery or “template molded” – where surgeon prepares the implant intraoperatively using 3D printed molds designed using CAD. [Table 2] shows the various studies on CAD-designed PMMA implants using both the techniques. There is a clear superiority of alloplastic implants made of any material over autologous bone grafts. Although clear cut superiority of one alloplastic implant over other is not clearly established, PMMA has a long clinical usage and easily available and molded. However, the natural contours present in the skull makes it difficult to accurately shape using manual handmade techniques especially in larger craniectomy defects. The other major issue with PMMA implants is the risk of infection, which is partially attributable to its irregular rough surface when it is handmade. The study by Lee et al. has concurred with this observation and noted that infection rate of prefabricated PMMA prosthesis are lower than hand molded ones. Implant failure due to infection is the major complication of cranioplasty. The added sophistication of using smooth surface for molding the implant using a template reduces the risk of infection and secondary implant failure.
In our experience over the past four years of treating craniectomy defects, we found loss of autologous bone as the major reason for the use of 3D-designed templates. Infection was the primary reason for nonavailability of the autologous bone flap in our study. Patients with stroke undergoing craniectomy for decompression have their bone flap preserved in the anterior abdominal wall as the preferred method, which is supported by many studies.[18–20] The other method used is cryopreservation, which has higher chances of nonviability of the bone flap and infection. The high cost is another major factor for it being not widely available. Hence, when the bone preserved using any of these techniques gets either infected or resorbed and in rare instances like bone erosion by tumor or trauma induced damage leads to unavailability, it leaves us with limited options. Even in our series, we found that stroke and trauma being the major cause of need for decompressive craniectomy. The diffuse nature of these pathologies and the prolonged rehabilitation needed for these patients render the bone flap stored in the anterior abdominal wall prone to resorption and infection. In 1994, Mankovich et al. first described the utility of 3D printing in surgical planning and for modeling implants. Direct 3D printing of the implants has been described using titanium and PEEK materials as shown in [Table 2]. Directly printing the required implant is definitely a viable option. However, certain factors like the high direct cost of the material and the expenditure involved in setting up the manufacturing unit and strict quality control regulations for using prefabricated implants precludes its routine availability. Also, the flexibility we get during intraoperative template-based preparation to further remodel for a more accurate fit is not available with prefabricated implants. The other practical advantage noted is availability of the template in case of implant infection or breakage and a new fresh implant can be made without adding financial burden to the patient.
The notable shortcoming from this method is the asymmetry in the temporal region observed in few patients due to atrophied temporalis muscle. Though the postoperative outcomes were excellent, in terms of cosmesis almost 40% of cases had “Good” Odom's score rather than Excellent due to this temporal asymmetry. The problem becomes obvious when the hairline is receding. However, this issue is present in other types of implants also. A viable solution we suggest is to provide variable thickness in the temporal region.
In a recent WHO survey, it has been noted that “out-of-pocket” payments for health care expenditure in India is common as only 15% of the population is covered by health insurance., Given these limitations, our technique is cost effective, easier to use, and has consistently reproducible results in all the 25 cases. We have used 3D template molds created using a local start-up company, Osteo3D Inc, Bengaluru, India using locally available materials. The whole production cost is one-fifth of the total amount used in developed countries with an average estimate of about INR 20,000 (<$250 including all without any subsidy). The cost is less than a generic non molded titanium plate, and the production time is less than 13 h as the CT data can be uploaded online through a cloud-based software. This method using 3D printed molds is quick, straightforward, cosmetically accurate, and biomechanically stable. It also avoids direct contact of the PMMA implant to tissues caused by exothermic reaction and also helps to create a smoother surface thereby reducing the risk of infection.,,,,,,,,,
An effective technology should provide an ideal patient outcome and economic feasibility to be used by all. The technique described in this study using CAD and 3D printing technology gives ideal cosmetic effect and also cost effectiveness. The major advantage of this system is reusability without adding any additional production costs. This method however needs detailed evaluation in a larger study population to ascertain their long-term functional and aesthetic outcomes and to look at various viable ways of improvement and refinement.
Declaration of patient consent
The authors certify that they have obtained all appropriate patient consent forms. In the form, the patient(s) has/have given his/her/their consent for his/her/their images and other clinical information to be reported in the journal. The patients understand that their names and initials will not be published and due efforts will be made to conceal their identity, but anonymity cannot be guaranteed.
We thank Osteo3d® for assistance in printing 3D template molds using Computer Assisted Designing. https://www.osteo3d.com
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5], [Figure 6]
[Table 1], [Table 2]