|Home | About | Journals | Submit | Contact Us | Français|
Sacral insufficiency fractures can cause severe, debilitating pain to patients concerned. The incidence of this fracture type correlates with the appearance of osteoporosis in the elderly population. A polymethylmethacrylate (PMMA) cement injection procedure called sacroplasty has been recently described as an optional method for the treatment of this fracture type. However, the correct cement placement in the complex anatomical structure of the sacrum is a surgical challenge. The aim of the study is to compare the precision, safety, and radiation exposure of standard multiplanar fluoroscopy and computed tomography (CT) guidance for PMMA application to the sacrum using both balloon-assisted sacroplasty and conventional sacroplasty. A controlled experimental investigation in a human cadaver trial has been performed. Two imaging and two application modalities to monitor percutaneous PMMA injection to the sacrum were examined. The application forms were randomized from side to side of the pelvis. We found less cement extravasation in the CT-guided groups, but also a significant higher radiation exposure (P < 0.05) by using CT guidance. The conventional fluoroscopy-guided sacroplasty revealed the shortest procedure time (incision to closure time) of all treatment groups (P < 0.01). These findings show no difference regarding cement extravasation between ballon-assisted and conventional sacroplasty. Further, in comparison to fluoroscopy-assisted technique, the CT-guided cement injection seems to decrease the risk of cement extravasation, irrespective of the use of an additional balloon assistance. However, we have to consider a greater radiation exposure using CT guidance. Further investigations will proof the suitability in the normal course of clinical life.
Sacral insufficiency fractures (SIFs) were originally described by Lourie in 1982  and become increasingly an important clinical entity of both social and economic significance as the western population ages continuously .
Sacral insufficiency fractures primarily occur in elderly women, the prevalence within general population in women, aged over 55 years, who presented to hospital with low-back pain is estimated to be around 2% [16, 32, 42]. About 2% of those fractures are accountable for lumbosacral nerve compression syndromes, particularly causing urinary sphincter dysfunction or lower limb paresthesia .
Risk factors are osteoporosis, osteopenia, rheumatoid arthritis, prolonged steroid administration, and other causes of secondary osteoporosis, radiation therapy, primary bone tumors, and metastatic disease, whereas postmenopausal osteoporosis represents the main cause among them [29, 40]. SIFs are often overlooked due to lack of clinical suspicious and non-specific nature of symptoms . Frequently, standard X-rays do not visualize the fracture and further specific imaging like magnetic resonance imaging (MRI), bone scintigraphy, and computed tomography (CT) are necessary to identify a suspected fracture [24, 43].
Treatments for this fracture type are commonly conservative including bed rest, moderate weight-bearing exercise, analgesia, osteoporosis medication, which often lead to complete healing within 9 months [23, 29]. However, physicians have to be aware of complications from immobility including DVT, pulmonary embolism, pneumonia, progressive osteopenia, lingering illness, and death [2, 9, 18].
Inspired by the positive experiences of vertebroplasty in the treatment of painful vertebral compression fractures, the augmentation of SIFs with bone cement presents an alternative treatment by using a percutaneous procedure called sacroplasty [4, 20, 31]. CT and fluoroscopic imaging modalities are described for the proper cannula placement prior to cement injection [5, 6, 11, 19, 26, 35, 44]. The most evident complications related to sacroplasty are cannula malperforation and cement leakage . In the vertebral column, the balloon kyphoplasty technique potentially provides a secure and effective method to reduce extravasation of the cement. The advantage of a preformed cavity with compaction of the cancelleous bone allows the controlled placement of bone cement in the vertebral body . Furthermore, the created void allows the use of an exact defined amount and highly viscous bone cement with a lower risk for extravasation. The balloon kyphoplasty has been previously described for the treatment of SIFs with good results .
Currently, there are still controversies as to which procedure, either balloon-assisted or without balloons, provides better conditions regarding minimizing procedure duration, optimizing procedure precision, and minimizing radiation exposure as well. In order to address these parameters, a controlled experimental investigation was conducted.
The aim of this study was to compare the precision, safety, and radiation exposure of standard multiplanar fluoroscopy and CT guidance for polymethylmethacrylate (PMMA) application to the sacrum using both the balloon sacroplasty and conventional sacroplasty.
The study was conducted on 10 human cadaver specimens previously fixed and conserved in Jores’s solution [50 g Karlsbad salt (mixture of 22 g NaSO4, 1 g K2SO4, 9 g NaCl, 18 g NaHCO3), 50 ml formaldehyde, 50 ml chloralhydrated water, 1,000 ml H2O] . Sacral vertebral augmentation with PMMA (High Viscosity, Radiopaque Bone Cement; KyphX® HV-R™, Kyphon Inc., Sunnyvale, USA) was performed bilaterally. The cement application was supported by kyphoplasty balloons (Kyphon, Brussels, Belgium) on the right-hand side and without balloons on the left-hand side. In five specimens (mean age 79 ± 6.3 years; range 63–88; f/m ratio: 3m:2f), the procedures were performed fluoroscopically using a Ziehm Exposcop 8000 (Ziehm Imaging GmbH, Nürnberg, Germany) with a total filtration of 4.0 mm aluminum. After positioning the specimen prone on a radiolucent table, standard radiographic images (anteroposterior, inlet, outlet and lateral views) were obtained and the appropriate settings marked. Small cannulas (20G × 23/4 hypodermic-needle) were inserted from dorsal directed from medial to lateral, perpendicular to the dorsal boundary of the S1 vertebral body (Fig. 1a). After identifying the S1 neuroforamina, the cannula were inserted about 2 cm caudad to the superior endplate of S1 and about 2 cm medial to the sacro-iliac joint (Fig. 1b). Jamshidi needles (11 Gauge) were placed instead of the small cannulas after stab incisions and pushed forward until the tip perforated the dorsal cortical surface by around 5 mm, subsequently the guide wires were introduced and large bore Kyphoplasty trocars were positioned. In the next step, a pusher was inserted through the cannula to form a channel (Fig. 1c). In the lateral view, the “iliac cortical density” (ICD) was respected. On the true lateral projection, the ICD adjacent to the sacroiliac joint parallels the sacral alar slope; therefore if the tip of the inserted device exceeds the ICD in the lateral view a breach of the anterior sacrum with subsequent cement extravasation is most likely (Fig. 1d). On the right-hand side, a 15 mm balloon was then inserted through the trocar and inflated with 3 ml contrast agent to create an intramedullary void (Fig. 1d, e). The balloon pressure was hold below 200 psi throughout the procedure. After deflation of the balloons, 3 ml of a high viscous PMMA (“toothpaste texture,” as recommended by the producer) was installed under fluoroscopic control until the void were filled. On the left-hand side, the same amount of PMMA was simultaneously installed without prior balloon preparation. After completion of sacroplasty, CT scanning was performed to evaluate the cement distribution in the sacrum (Fig. 1f).
In another five specimens (mean age 79.8 ± 7.5 years; range 69–88; f/m ratio: 4m:1f), CT interventions were performed on a 64-slice computed tomograph (Brilliance 64, Philips Medical Systems, Cleveland, USA). The cadaver was placed in head-first and prone position. A K-wire was placed on the skin as landmark for better orientation regarding the entry-point, penetration depth, and angle of divergence during the positioning of the punction needle (Fig. 2a). After generating the surview image, a spiral scan for the planning of the procedure was performed with a collimation of 64 × 0.625 mm, a pitch of 0.98 (39.32 mm table movement per rotation), a rotation time of 1 s, 120 kV, and 100 eff. mAs. The scan range for the planning procedure was 10 cm. The images were reconstructed with a slice width and an increment of 3 mm using a bone kernel (kernel D); additionally images with a slice width of 1 mm and an increment of 0.5 mm were performed for multiplanar reformations (MPR). These dose adjustments were chosen, since most of the patients who could be treated with the use of CT are old patients with osteoporotic fractures of the sacrum. In order to identify the corticalis, especially of the neuroforamen higher tube settings need to be chosen. Each step of the sacroplasty procedure was controlled with a single sequential scan that could be exposed with a foot pedal from inside the scanner room and displayed on a monitor inside the scanner room (Fig. 2). The sequential scan was performed with a collimation of 32 × 1.25 mm, a rotation time of 0.5 s, 120 kV, and 100 eff. mAs. From this single scan, the CT reconstructed four slices with a slice width and increment of 10 mm, resulting in a total scan range of 4 cm. After finishing the interventional procedure, a spiral control scan with the same scan parameters as the planning scan was acquired.
The cement was applied with high viscosity after penetration of the dorsal cortical surface of the massa lateralis of both sides. After creation of a void via balloon sacroplasty, on the right-hand side, and without balloon preparation, on the left-hand side, respectively, we used the same amount of PMMA.
The effective doses from the fluoroscopy procedure were calculated from the dose area product (DAP) according to the conversion coefficients by Gosch et al. . Since the conversion factors differ according to the direction of the applied radiation (anterior–posterior (a.p.) vs. lateral projections), we presumed that approximately two-third of the procedure to be anterior–posterior projection and one-third in lateral projection.
Since the cadavers used in the study were preserved and not very slim, a near round shape in the scan region was found, therefore the DAP could be divided into one-third (a.p.) and two-third (lateral) for multiplication with the according conversion factors depending on the projection. Unfortunately, the fluoroscopy doses could not be calculated sex specifically, since the conversion factors are sex-averaged (the mathematical phantom for calculations of the conversion factors was a hermaphrodite ).
The effective doses of the CT scans were calculated with the PC program CT Expo version 1.6 (G. Stamm, Hannover und H.D. Nagel, Hamburg, Deutschland; demo version: http://www.mh-hannover.de/fileadmin/kliniken/diagnostische_radiologie/download/ct-expo-d.zip) . Dose calculations relating to standard patients were made for the central part of the sacrum (z = 13–23 (women) and z = 14–24 (men) for both the planning and control scan; z = 16–20 (women) and z = 17–21 (men) for the sequential scans during intervention, where z is the position on the patient’s longitudinal axis for calculation with CT Expo). CT Expo calculates the effective doses on the basis of the used CT-settings and scanner specific influences. Furthermore, the calculations can be done for female and male separately, since there are differences in the anatomical distribution of the radiation sensitive organs (i.e., gonads, female breast).
Evaluation of cement distribution was performed by two experienced reviewers in consensus. Cement penetration observed outside of the sacral alae (e.g., neuroforamina, spinal canal, and iliosacral joint) was defined as cement leakage.
All data were given as mean value ± SD. Differences between the trial groups were evaluated by a standard t-test. Statistical analysis was performed with SigmaStat (Version 2.03) by SPSS.
The procedure time (incision to closure time) of the different procedure modalities revealed a significant prolongation by using both CT guidance and balloon assistance. Furthermore, the CT-guided balloon-assisted sacroplasty was significant more time consuming than the fluoroscopy-guided balloon-assisted sacroplasty (Table 1).
Presuming a scan range of 10 cm for the planning scan and the control scan, each scan calculates to an effective dose of 2.1 mSv for female and 1.1 mSv for male patients. Furthermore, each sequential scan calculates to an effective dose of 0.3 mSv for women and 0.2 mSv for men. The calculation results are summarized in Table 2. The doses per patient, which were calculated from the doses for the planning and the control scan, and the amount of sequential scans needed for the procedure are listed in Table 3.
The dose calculations for the fluoroscopy procedures are summarized in Table 4. We calculated a mean effective dose of 5.6 mSv (SD ± 0.55) in total for fluoroscopy only. The effective dose using CT guidance was significant higher compared to the fluoroscopy guidance, albeit this significant difference is only obvious for females (P < 0.001). However, considering a postoperative control CT scan to examine the correct cement placement in the fluoroscopy-guided groups, we found still a significant higher effective dose for females by using CT guidance (P < 0.05) (Table 1).
Cement malpenetration was found twice in the fluoroscopy-guided procedure (Table 1). In one case, cement malpenetration into the S1-neuroforamen was found after fluoroscopy-guided sacroplasty without using a balloon (Fig. 3a). Furthermore, an anterior cement malpenetration was found in the fluoroscopy-guided balloon-assisted sacroplasty (Fig. 3b). In that case, we presume that the cannula tip or the pusher breached of the anterior sacrum and therefore constituted a way for cement leakage. This underlines how important it is to avoid exceeding the ICD in the lateral view. In the CT-guided groups, PMMA cement was confined to the lateral sacrum, without leakage into the foramen or outside the cortical bone.
Sacroplasty describes the percutaneous vertroplasty of the sacrum. We find the first description of a fluoroscopic-guided sacroplasty for the treatment of a painful, osteoporosis associated SIF in the year 2002 . In 2005, the first CT-guided procedure was presented in a patient with sacral metastases from hepatocellular carcinoma . Most of the studies in the literature are case reports and predominantly provide technical support for the special features of sacroplasty. The exact deposit of the cement in the sacrum is the main problem in sacroplasty , particularly to avoid neural injuries by exothermic reaction during hardening of extraosseous cement [21, 28]. It is well known that the advantage of the kyphoplasty in the vertebral column over the vertebroplasty is the reduction of cement leakage . In our study, we could show that the balloon-assisted sacroplasty is a feasible technique for correct cement placement in the sacrum after creating a void, albeit no difference to the leakage-rate of the nonballoon-assisted sacroplasty was obvious. Bohner et al.  could demonstrate that cement extravasation can be drastically reduced by increasing the cement viscosity. By performing balloon assistance, it might be possible to use cement of higher viscosity due to the originated cavity. Therefore, the extravasation rate might be reduced in comparison to the technique without balloon assistance. In the present study, we used the same cement with similar viscosity for both application techniques to avoid further bias. However, we have to keep in mind that the chosen cadaveric trial does not realize a fracture model. The risk of cement leakage in the presence of a fracture gap is inevitable higher than in our cadaveric model. Furthermore, the CT-guided sacroplasty presents a safe technique with calculable radiation exposure. The CT-guided technique in cooperation between a surgeon and a radiologist provides a reliable procedure for exact positioning of the cannulas and the subsequent cement placement.
The calculation of the effective doses for the fluoroscopy procedure was done by dose estimation using conversion factors on the applied DAP. According to a previous study on dose application for kyphoplasty, we also divided the procedure in parts of anterior–posterior fluoroscopy (approx. two-third of the DAP) and lateral fluoroscopy (approx. one-third of the DAP) . The conversion factors used for this study were taken from an actual publication from Gosch et al. . However, Boszczyk et al.  report a mean effective dose of 4.28 mSv for their 16 estimated patients for kyphoplastie, which is quite similar to our estimated 5.6 mSv. A reason for our slightly higher effective dose can be the presence of the ovaries and uterus by applying radiation to the pelvis/sacrum and using a sex-averaged conversion factor.
The general disadvantage of CT guidance is the application of a higher effective radiation dose. In our study, this is even more relevant for female examinations due to the localization of the uterus and ovaries near and/or within the scan range. Compared to the mean sex-averaged effective dose for the fluoroscopy procedure, the mean effective dose for the CT was 1.3 mSv higher for male and 5.7 mSv higher for female calculations. Since the fluoroscopy data were sex averaged, the effective dose difference for male might be a little higher than stated above and for female a little lower. Nevertheless, assuming a maximum dose difference from ~5.7 mSv, this would correspond to little more than the double the annual natural effective radiation dose, which is reported to be ~2.4 mSv . The benefits of more accuracy might outweigh the disadvantage of the increased radiation exposure by using CT guidance. Additionally, it has to be discussed that the dose of the CT guidance depends very much on the amount of sequential scans needed to guide the procedure (see Table 2). With an increasing experience in this procedure, there is a further possibility to safe dose during CT-guided intervention. Another aspect is the age of the patients with mostly old patients, where the possible advantage of an early mobilization and pain relief might outweigh the risk of developing a malignant tumor due to radiation exposure. Einstein et al.  could show that this risk decreases with increasing age, they estimated in a dose calculation for cardiac CT scans, that a 20-year-old female has an approx. ninefold higher relative risk of attributable cancer incidence associated with a single CT coronary angiography scan than a 80-year-old female.
Currently, there are no reliable data in the literature that sacroplasty provides clinical and functional improvement, especially regarding the alleviation of pain.
Data from biomechanical studies suggest that the cement augmentation does not apply for a higher bony stiffness, thus the pain relief may contribute by the local effect of the PMMA [1, 38]. The injection of PMMA into the fracture cleft may contribute to temporary pain relief, but the cement also inhibits fracture healing and may cause incident fractures, especially in patients with osteoporotic insufficiency fractures [14, 39]. It remains to be seen to what extent alternative materials (e.g., bisphenol-a-glycidyl dimethacrylate) with altered features may modify these influences .
Up to date, there are no guidelines concerning preoperative diagnosis and indications for sacroplasty. Furthermore, it is not known what sort of cement placement would be ideal for what type of fracture, what happens if the balloon expands within the fracture line. In contrast to kyphoplasty, there is no potential for reduction of a fracture. There are still a lot of general questions which need to be looked at before general recommendations may be further defined. In our opinion, only SIFs without concomitant instability of the pelvic ring are appropriate indications for sacroplasty. Finally, appropriate training and experience in pelvic surgery is necessary prior performing sacroplasty.
Our findings indicate no difference regarding cement extravasation between ballon-assisted and conventional sacroplasty. Further, in comparison to fluoroscopy-assisted technique, the CT-guided cement injection seems to decrease the risk of cement extravasation, irrespective of the use of an additional balloon assistance. However, the CT guidance is accompanied with a significant increased radiation exposure. Further investigations have to proof the suitability in the normal course of clinical life.
The authors would like to thank Kyphon® Europe for excellent technical support during the study.
Conflict of interest statement The authors ensure that there were no conflicts of interest related to companies or products named in the manuscript.
L. Grossterlinden and P. G. C. Begemann have contributed equally.