Clinical Application of a 3D-Printed Positioning Module and Navigation Template for Percutaneous Vertebroplasty

Introduction

Osteoporotic vertebral compression fracture (OVCF) is the most common type of osteoporotic fracture in elderly individuals. ¹ At present, percutaneous vertebroplasty (PVP) is the most widespread technique used for the treatment of OVCFs. The procedure is minimally invasive and provides rapid pain relief; however, the operation is mainly performed under indirect vision, and failure to properly locate the site of intraoperative puncture may lead to spinal dura mater injury, pedicle rupture, and even neurological damage. Therefore, accurate puncture site localization is key to the success of PVP. ² In addition, nearly 50% of patients have local complications from traditional PVP, 95% of which are due to leakage of bone cement into the surrounding tissues or paraspinal venous embolism. Repeated correction of the position with a positioning needle under X-ray fluoroscopy may also increase radiation exposure to the patient and surgeon and prolong the operative time. Alongside the advent of precision surgical technology, computer-assisted 3D-printing navigation technology has demonstrated obvious advantages in guiding vertebroplasty and improving surgical accuracy and safety. ³ At present, various image navigation systems are used for pedicle puncture site localization and pedicle screw placement. ⁴ However, according to the literature, the error rates range between 1.2% and 15.7%, and reoperation rates between 0 and 7.42% continue to be reported. ⁵ In addition, expensive image navigation equipment is not feasible for developing countries. Therefore, it is very important that a simple, inexpensive, accurate, and effective individualized assisted puncture technique be developed. In our study, a PVP preoperative navigation module based on preoperative CT scan data was designed to assist pedicle puncture positioning and was effectively applied in clinical practice. ⁶

Materials and Methods

General Information

The study was reviewed and approved by the institutional ethics board of the hospital. All patients and their family members agreed to the operation and signed the surgical consent forms. A total of 22 patients with vertebral compression fracture admitted to our hospital from January 2016 to January 2017 were included. The experimental group consisted of 11 patients, 2 males and 9 females, with a mean age of 67.27 ± 11.86 years (range: 48 to 80 years), and the control group consisted of 11 patients, 3 males and 8 females, with a mean age of 74.27 ± 7.24 years (range: 63 to 89 years). The difference in age between the experimental group and the control group was not statistically significant (t = −1.954, P = .079) (Tables 1 and 2).

OPEN IN VIEWER

Table 1.

General Information of the Experimental Group.

Case	Age (years)	Sex	Affected Vertebral Segments	Side	Duration of Puncture Site Localization (min)	Number of Fluoroscopy Sessions	Postoperative RNP	Preoperative VAS Score	Postoperative VAS Score
1	58	Male	T10	Unilateral	4	4	NA	8	1
2	74	Female	T12	Bilateral	2	8	NA	9	1
3	48	Female	T12	Unilateral	3	9	NA	9	0
4	77	Female	T12, L1	Unilateral	5	5	NA	9	2
5	78	Female	L1	Bilateral	4	4	NA	8	3
6	50	Female	L1	Bilateral	5	7	NA	6	3
7	77	Female	L1	Bilateral	5	5	NA	6	3
8	57	Female	L1	Bilateral	4	8	NA	6	2
9	80	Male	L2	Unilateral	5	6	NA	8	1
10	68	Female	L2	Bilateral	4	6	NA	7	3
11	73	Female	L3, L5	Bilateral	4	6	NA	9	4
Mean	67.27	-	-	-	4.09	6.18	-	7.72	2.09

OPEN IN VIEWER

Table 2.

General Information of the Control Group.

Case	Age (years)	Sex	Affected Vertebral Segments	Side	Duration of Puncture Site Localization (min)	Number of Fluoroscopy Sessions	Postoperative RNP	Preoperative VAS Score	Postoperative VAS Score
1	72	Female	T12	Bilateral	16	14	NA	9	2
2	64	Female	T12	Bilateral	15	15	NA	9	0
3	76	Male	T12	Bilateral	16	16	NA	8	2
4	79	Male	T12, L1	Unilateral	14	14	NA	7	1
5	76	Female	Ll	Bilateral	17	17	NA	8	3
6	63	Female	L1	Bilateral	18	15	NA	9	1
7	89	Male	L1	Bilateral	16	15	NA	9	1
8	77	Female	L1	Bilateral	18	18	NA	8	2
9	70	Female	L1	Bilateral	17	14	NA	8	3
10	73	Female	L1	Bilateral	18	15	NA	7	2
11	78	Female	L2	Bilateral	17	17	NA	8	3
Mean	74.27	-	-	-	16.55	15.45	-	8.18	1.82

Methods

Design and manufacture of the 3D-printing positioning module and surgical guide plate

To obtain the target vertebral body CT scan data, the patient was first placed in the prone position on a spinal cord stereotaxic frame. Then, according to the patient’s preoperative X-ray, the positioning module was placed on the corresponding location on the skin over the diseased vertebra, and the location of the positioning module on the patient’s skin was marked. Next, scans were obtained using 64-slice dual-source CT device (Siemens, Germany) from our hospital (layer thickness: .6 mm, tube voltage: 140 kV, and tube current: 800 mA). ⁷ The scope of the CT scan included one vertebra above and one vertebra below the affected vertebra, and the data were saved in DICOM format (Figure 1).

OPEN IN VIEWER

Figure 1.

Preoperative CT examination and data extraction. (a and b) CT scan for locating the module and target vertebral body; (c and d) patient was placed in the prone position on a spinal cord stereotaxic frame; and (e) the compressed vertebral body, skin, and positioning module 3D-reconstructed with DICOM CT scan data (red arrows: the location of the module; blue arrows: spinal cord stereotactic frame).

Image Processing

The CT data of the vertebral body, skin, and positioning module were imported into MIMICS software, where a 3D model of the structures was obtained through image filtering, threshold segmentation, editing, 3D reconstruction, and other processes. The 3D model was then imported into Geomagic Studio 12.0 software (Raindrop, USA) in STL format for physical image encapsulation, and the results were stored in IGS format (Figure 2).

OPEN IN VIEWER

Figure 2.

The 3D model was imported into Geomagic Studio 12.0 software in STL format for physical image encapsulation and then stored in IGS format. (a) Model file processed by MIMICS. (b) Model files processed by Geomagic Studio.

Design and Manufacture of the Individualized Surgical Guide Plate

According to the puncture site established above, the area of the skin containing the positioning module was selected and combined with the skin patch for drawing. A Boolean operation was used to obtain a guide plate that could be fitted to the skin such that the positioning hole on the guide plate coincided with the positioning module in the preoperative CT scan. The length of the guide sleeve on the sides of the guide plate was measured to ensure that the puncture needle could not be advanced beyond its depth limit. Finally, the designed guide plate data model was exported to a Creatbot 3D printer in STL format, and the percutaneous vertebral body-shaping guide plate was printed in 3D (Figure 3).

OPEN IN VIEWER

Figure 3.

Design the navigation template model and 3D-printed. (a and b) Design the navigation template model according to the location of the module and skin surface geometry. (c) Template designed to ensure proper pedicle puncture angle and depth. (d and e) Designed navigation template model input into the 3D-printing software for editing. (f) 3D-printed navigation template preoperative simulated puncture.

Surgical Method

All operations were performed by the same team of surgeons. The patient was placed in the prone position, the surgical area was disinfected, and a surgical sheet was draped over the surgical area. After sterilization with low-temperature plasma, the hole in the personalized guide plate was matched with the positioning module area marked on the patient’s back. Then, after local infiltration anesthesia, based on the orientation of the guide plate, a bilateral or unilateral incision measuring approximately .5 cm in length was made, and the puncture needle was inserted through the guide plate hole to the pedicle surface. A C-arm X-ray device was used to observe the site and angle of needle insertion. Holding the guide plate in one hand, the operator then drove the puncture needle to its proper depth with a bone hammer on one side or both sides simultaneously. The other procedures were the same as those of traditional PVP or percutaneous kyphoplasty (PKP) (Figure 4).

OPEN IN VIEWER

Figure 4.

Intraoperative positioning, X-ray fluoroscopy, and bone cement injection. (a) The position of the preoperative navigation module is marked on the skin. (b) The positioning window of the navigation template overlaps with the marked module position. (c-f) Needles are inserted into the vertebral body, and bone cement is injected.

Results

The intraoperative puncture operation was successful for all patients, and intraoperative and postoperative X-ray examinations confirmed that the distribution of the injected bone cement injection was satisfactory, the spinal canal and pedicle structures were complete, and no nerve root or spinal cord injuries or other complications were reported. On average, the patients’ postoperative pain was significantly relieved, and the patients were able to perform therapeutic, functional exercises on the second day after surgery while wearing braces.

The mean duration of puncture localization for the experimental group (4.09 ± .94 min) was significantly shorter than that for the control group (16.55 ± 1.29 min; P < .05). Similarly, fluoroscopy was performed significantly fewer times in the experimental group (4.09 ± .94) than in the control group (15.45 ± 1.67; P < .05). The average preoperative VAS score of the experimental group was 7.73 ± 1.27, and the average VAS score 2 days after surgery was 2.09 ± 1.22. The average preoperative VAS score in the control group was 8.18 ± .75, and the average VAS score 2 days after surgery was 1.82 ± .98. Neither the preoperative nor the postoperative VAS scores were significantly different between the experimental group and the control group (P > .05 for both) (Table 3).

OPEN IN VIEWER

Table 3.

Comparison of Related Indicators.

Group	Number	Age (years)		Duration of Puncture Localization (min)	Number of Fluoroscopy Sessions	Preoperative VAS Score	Postoperative VAS Score
Experimental group	11		67.27 ± 11.86	4.09 ± .94	6.18 ± 1.66	7.73 ± 1.27	2.09 ± 1.22
Control group	11	74.27 ± 7.24		16.55 ± 1.29	15.45 ± 1.67	8.18 ± .75	1.82 ± .98
t value		−1.954		−28.691	−17.143	−.922	.637
P value		.079		.000	.000	.378	.539

Discussion

The accuracy of preoperative vertebral body-surface localization and intraoperative pedicle puncture site localization are key to the success of PVP. Most patients with vertebral compression fractures are elderly patients with complications such as cardiopulmonary insufficiency ⁸ ; PVP is performed in the prone position, which is not conducive to the respiratory and circulatory system health of such patients. Therefore, it is very important to minimize the operation time as much as possible. In the traditional procedure, the patient typically needs four to six fluoroscopies to confirm the puncture site and mark the skin. During PVP, fluoroscopy should be used to select the side opening distance, side opening angle, and head tilt angle of the needle entry point and judge the entry depth. An individualized 3D-printed positioning module and surgical guide plate were used to identify the diseased vertebra and the puncture insertion point(s) and needle angle(s) according to the position marked on the patient’s skin before surgery. ⁹ Furthermore, the length of the guide plate injection tube was designed to limit the depth to which the puncture needle could be driven to ensure that it would not penetrate the anterior edge of the vertebral body and cause iatrogenic injury. All of these factors contributed to reducing the time of surgical risk exposure and enhancing the safety of the surgery. ¹⁰

The design of the 3D-printed individualized surgical guidance plate was obtained by using the positioning module during the patient’s CT scan to guide 3D reconstruction of the two-dimensional skin and bone imaging data using related software. ¹¹ Preoperative planning, computer simulation of the puncture process, and determination of the operation plan were also involved in the ultimate design of the surgical guide plate. The adjustment accuracy of the computer software used in the preoperative planning was .01 mm, and the angle accuracy was .01°. However, the method of puncture into the needle point for the control group was more dependent on the operator’s experience and ability to palpate the spinous processes, and its poor repeatability is not conducive to rapid use by younger doctors. ¹²

To overcome the shortcomings of traditional X-ray fluoroscopy-assisted PVP, we initially designed a 3D-printed navigation template that matched the shape of the skin and did not involve the use of the CT scan positioning module described earlier. Instead, the template was designed to allow metal wires to be placed onto a guide plate for easy perspective positioning (Figure 5). However, this navigation template was unable to localize the diseased vertebra with one fluoroscopic scan before puncture; indeed, the metal positioning line had to be placed above the puncture point of the diseased vertebra and scanned with fluoroscopy 4–6 times, increasing the number of times fluoroscopy had to be used. To address this problem, we researched and designed the positioning module described in this study. Before surgery, the positioning module was placed onto the area above the skin of the affected vertebra as determined by the patient’s X-ray. During the preoperative CT scan and surgery, the patient was placed in the prone position on a spinal cord stereotactic frame. Thus, the patient was fixed in position during the intraoperative and preoperative scan, reducing the positioning error. (Figures 1C-1D)

OPEN IN VIEWER

Figure 5.

Wire positioning navigation template. (a) Wire is placed into the navigation plate and used to locate the vertebral body for puncture. (b) Intraoperative fluoroscopic positioning using the navigation plate.

The module was included in the CT scan, and the patient’s back skin was marked. Then, based on the geometric relationship among the diseased vertebrae, the skin, and the module, the 3D-printed navigation template was designed. The location of the module on the navigation template was marked, and a hole was made; then, prior to puncture, the positioning hole on the navigation template was directly matched with the preoperative position marked on the patient’s back with the positioning module (Figure 6). Thus, we could directly determine the relationship between the navigation template and the surgical vertebral body without repeated fluoroscopy ¹³ .

OPEN IN VIEWER

Figure 6.

Navigation module positioning. (a) The positioning window of the navigation template overlaps with location on the patient’s back previously marked using the positioning. (b) Preoperative CT scan positioning module.

We found that to reduce the influence of the sterile operation sheet on the guide plate, the length of the latter should not exceed the distance between the two posterior axillary lines. Similarly, the width of the guide plate should be greater than the height of one and a half vertebrae; this can not only increase the stability of the guide plate but also shorten the time required to print it in 3D. ¹⁴

The guide plate is made of polylactic acid, which requires low-temperature plasma disinfection (45°C60°C) prior to use. The disinfection process may lead to slight expansion of the puncture hole in the guide plate. ¹⁵ In this study, the diameter of the puncture needle sleeve for PVP was 4.3 mm, and the designed diameter of the puncture hole in the guide plate varied between 4.8 and 5.0 mm throughout the experiment. This maximized the stability of the needle and did not result in an increase in needle resistance. During the operation, the position of the patient was the same as that in the preoperative CT scan, and movement of the back skin was minimal and thus did not affect the puncture positioning process with the navigation template. ¹⁶

Traditional percutaneous vertebroplasty is performed with repeated fluoroscopy. The resulting X-ray radiation delivered during surgery can eventually affect the health of the patients and operators as it accumulates in the tissues, which in serious cases can lead to cancer. Our study showed that compared with the control group, the experimental group, treated using 3D-printed personalized surgical navigation templates, could significantly reduce the number of fluoroscopy procedures performed. ¹⁷ The main purpose of the template is to locate the affected vertebra and determine the puncture site and angle. These results suggest that the use of 3D-printing to create personalized surgical navigation templates can reduce the number of intraoperative fluoroscopy procedures.

Our experimental group included a small number of patients; therefore, the conclusions should be verified by multicenter and large-sample studies. Some patients have not been able to perform long-term follow-up, and therefore the long-term efficacy of our 3D-printed navigation template-based surgeries cannot be evaluated. ¹⁸

The preliminary clinical application of PVP assisted by a positioning module and a 3D-printed personalized surgical navigation template achieved effective clinical effects, shortening the operation time, reducing the number of fluoroscopies, and increasing the safety of the operation. This work is meaningful because it can effectively improve the surgical skill of young orthopedic surgeons from primary hospitals and shorten the learning curve of the orthopedic surgeons.

With the emergence of precision medicine in recent years, individualized patient diagnosis and treatment, decreased trauma, and increased precision have gradually occupied a dominant position in the development of medicine.