Sage Journals: Discover world-class research

Abstract

Purpose:

How to select operative methods for delayed paralysis after osteoporotic vertebral fracture remains a matter of debate. It is unclear which procedure provides better neurological and functional recovery. This study aimed to compare the various operative methods and investigate postoperative outcomes.

Methods:

The study included 42 delayed paralysis patients (mean age 73.7 years). The operative methods were posterior spinal shortening osteotomy with direct decompression and short or long fusion and posterior laminectomy and short fusion with vertebroplasty. All patients in the vertebroplasty group were fitted with hooks to prevent correction loss. We investigated postoperative complications, surgery time, intraoperative blood loss, and changes in walking ability and local kyphotic angle (LKA). Patients were divided into three groups: shortening with a peripheral fusion range of one intervertebral level (the SS group, n = 18), the same procedure with a peripheral fusion range of ≥2 intervertebral levels (the SL group, n = 12), and posterior laminectomy and short fusion with vertebroplasty (the VP group, n = 12).

Results:

Postoperative complications were observed in five patients, all of whom were in the SS and SL group. There were no significant differences between the SS and SL groups. The VP group experienced significantly shorter mean surgery times, significantly lower mean intraoperative blood loss, and an improvement in paralysis, walking ability, LKA, and correction loss. Overall, the VP group had significantly better outcomes.

Conclusion:

Posterior laminectomy and short fusion with vertebroplasty with the addition of hooks is useful in cases of delayed paralysis in elderly patients.

Keywords

delayed spinal cord paralysis operative methods osteoporotic vertebral fracture

Introduction

There has been an increasing incidence of osteoporotic vertebral fractures associated with the growing elderly population in Japan. This type of fracture results in issues such as pain and negatively affects activities of daily living and quality of life. In some cases of osteoporotic vertebral fracture, the vertebrae heal in place, albeit with a slight localized kyphosis, and clinical issues arising from this type of fracture are relatively rare.^1,2 However, recent increases in the numbers of postvertebral fracture kyphotic deformity cases have resulted in increased cases of chronic lower back pain and various internal organ dysfunctions such as reflux esophagitis and respiratory disorders.^3

–6 Thus, vertebral fractures lead to increased mid- to long-term mortality risk.⁷

In 1958, Kempinsky et al.⁸ reported on patients presenting with neurological disorders because of advanced postvertebral fracture compression. Since then, we have learned that it is not rare for delayed bone fusion to result in pseudarthrosis and persistent lingering pain, or for advanced vertebral compression to cause delayed paralysis.^9
–11

Operative techniques for pain caused by postvertebral fracture, delayed paralysis, and spinal deformity include spinal fusion using implants and spinal shortening. These procedures can be used in cases where nonsurgical management is ineffective, but they have not been tested against conservative treatment in randomized control trials. Therefore, the appropriate operative method can only be selected according to the condition of each individual patient.

There are various operative methods currently used in these types of cases, but there is no consensus on the appropriate use and limitations of each method. We had been using posterior spinal shortening for cases of delayed paralysis accompanying local kyphosis. However, reports have indicated that the use of this method in elderly patients should be restricted because of its highly invasive nature; therefore, in 2013, we switched to posterior indirect decompression and short fusion with vertebroplasty and the additional use of pedicle screws and hooks for fixation. In this study, we investigated the postoperative outcomes of this method and compared it with the posterior spinal shortening.

Materials and methods

Patient population

A total of 42 patients who were diagnosed with delayed paralysis after an osteoporotic vertebral fracture at this medical department from January 2005 to February 2015 who were available for at least a 1-year observation period were selected. The mean age at the time of surgery was 73.7 years (range, 39–89 years). There were 12 males and 30 females. All patients had had an osteoporotic vertebral fracture, and because they had no obvious paralysis at the time of the fracture, they had been treated with conservative therapy. However, delayed paralysis subsequently occurred. Patients that presented with paralysis associated with obvious burst fractures at the onset of the condition, and those with paralysis associated with metastatic spinal neoplasm or infection were excluded. Spinal shortening was performed in 30 subjects and posterior fusion with vertebroplasty was performed in 12 subjects.

The study was approved by the Institutional Review Board of Yamaguchi University Hospital, and it adhered to the tenets of the Declaration of Helsinki. Written informed consent was obtained from all patients.

Study groups

There are no clear selection criteria for the various surgical procedures, and so the type of procedure used mainly depended on when the surgery was carried out. We performed posterior spinal shortening between 2005 and 2012. During the first half of this period (from 2005 to 2008), we mainly performed the fusion procedure by fusing two vertebrae above and below the shortened vertebra (posterior spinal shortening osteotomy with direct anterior decompression and long fusion; the SL group, n = 12). During the second half of this period (from 2009 to 2012), we mainly performed the fusion procedure by fusing one distal vertebra along with the simultaneous use of a screw and hook (posterior spinal shortening osteotomy with direct anterior decompression and short fusion; the SS group, n = 18). In all patients treated from 2013 to 2015, we performed posterior decompression and fusion with vertebroplasty (posterior laminectomy and short fusion with vertebroplasty; the VP group, n = 12; Figure 1).

Figure 1.

Radiographs of three typical surgical procedures. (a), (b) Postoperative plain radiographs at 1 year after posterior spinal shortening osteotomy with direct decompression and long fusion (SL group). (c), (d) Postoperative plain radiographs at 1 year after posterior spinal shortening osteotomy with direct decompression and short fusion (SS group). (e), (f) Postoperative plain radiographs at 1 year after posterior indirect decompression and short fusion with vertebroplasty (VP group). (g)An enlarged view of the vertebral body with vertebroplasty. Hydroxyapatite blocks can be seen. SL: posterior spinal shortening osteotomy with direct neural decompression and long fusion; SS: posterior spinal shortening osteotomy with direct decompression and short fusion; VP: posterior indirect decompression and short fusion with vertebroplasty.

Clinical outcomes

We retrospectively reviewed clinical findings such as duration of surgery, intraoperative blood loss, neurological status, walking function, radiographic results, and postoperative complications. We evaluated clinical results 1 year after surgery.

Preoperative and postoperative neurological impairment was assessed using the Frankel Grade classification (Table 1).¹³ Walking ability was assessed using a 7-step scale ranging from needing a wheelchair (astasis) to independent walking. We assessed differences between preoperative walking ability and walking ability 1 year following surgery (Table 2). We subtracted the preoperative score from the final score to determine the score change and compared the score changes among the three groups.

Table 1.

Evaluation of neurological impairment using the Frankel Grade classification.

Grade	Neurological status	Score (points)
A	Complete motor and sensory loss	0
B	Sensory only	1
C	Motor useless	2
D	Motor useful	3
E	Recovery	4

Table 2.

Walking function scores.

Walking ability	Score (points)
Astasia	0
Stand while holding on to things	1
Walking while holding on to things	2
Walker walking	3
Assistant walking	4
Crutch walking	5
Independent walking	6

Radiological examination

Plain radiographs were reviewed before surgery, immediately after surgery, and 1 year after surgery. Lateral plain radiographs of patients in the standing or sitting position were used for measurement.

We measured the local kyphotic angle (LKA) between the fused vertebrae using lateral radiographs. Plain radiographs before surgery were compared with those 1 year after surgery, and the presence or absence of adjacent vertebral fractures was assessed.

Surgical procedures

Using the posterior approach for the spinal shortening groups, we temporarily fixed the central and peripheral vertebrae on either side of the affected vertebra by inserting pedicle screws. We also resected the posterior elements of the affected vertebra, resected the bone fragments extending into the spinal canal anterior to the dura mater, and performed posterior spinal shortening between the adjacent vertebrae of the affected vertebrae using the pedicle screw-and-rod system.^14,15

For the VP group, we inserted hydroxyapatite blocks (HOYA Technosurgical Co., Tokyo, Japan) using the transpedicular approach. We used a hydroxyapatite block because it is associated with a smaller risk of leakage into the spinal canal and tissue damage caused by heat than bone cement, and it has bone affinity. Posterior indirect decompression was performed using laminectomy at the affected vertebral level. The main reason for laminectomy in the VP group was to decompress the neural tissue; it also enabled fixing of the claw hook and avoided paralysis caused by restenosis, for example, correction loss. None of the patients in the VP group received treatments for retropulsed posterior cortex such as dissection of bony fragments or impaction of bony fragments into the vertebral body. Posterior spinal fixation from two levels above to one level below the affected vertebra was performed using the pedicle screw-and-rod system. We used a sublaminar hook on the distal end of the vertebra and attached a claw hook to the proximal side. It has been reported that large correction loss occurs during pedicle screw fixation with vertebroplasty,¹² so we additionally used a claw hook at the proximal side and a sublaminar hook at the distal end to prevent correction loss. The number of fused vertebrae, duration of surgery, estimated intraoperative blood loss, and intraoperative complications were obtained from medical records.

Statistical analyses

We used Microsoft Excel Statistics for statistical analysis. The Kruskal–Wallis test was used to compare the three groups, and the Wilcoxon test was used to compare pairs of groups. The significance level was set at 5%.

Results

Patient demographics

Table 3 shows the demographics for all 42 patients. The power and sample size calculation confirmed that the sample size for each group was acceptable. Patients in the VP group were significantly older than those in the SL group and those in the SS group (SL group, 69.8 ± 7.2 years; SS group, 72.6 ± 11.8 years; VP group, 79.2 ± 6.6 years; p = 0.007). There were no significant differences among the three groups in terms of sex. One patient in SL group had injured 2 vertebral bodies (L1 and L2), but the other 41 patients had injured single vertebral bodies. Preoperative neurological score and walking ability were significantly higher in the SS group than in the SL and VP groups.

Table 3.

Patients’ demographics by group.^a

	SL group	SS group	VP group	p Value
Number of patients	12	18	12	NS
Age (years)	69.8 ± 7.2^b	72.6 ± 11.8	79.2 ± 6.6^b
Sex male/female (n)	4/8	4/14	3/9	NS
Number of collapsed vertebrae
Above Th10	0	0	1
Th11	0	0	1
Th12	9	8	5
L1	3	4	5
L2	1	4	0
Below L3	0	2	0
Preoperative neurological score	2.3 ± 0.5^c	2.7 ± 0.5^c,d	2.2 ± 0.4^d
Preoperative walking ability score	1.4 ± 1.4^e	2.6 ± 1.3^e,f	1.3 ± 1.0^f

NS: not significant; SL: posterior spinal shortening osteotomy with direct decompression and long fusion; SS: posterior spinal shortening osteotomy with direct decompression and short fusion; VP: posterior indirect decompression and short fusion with vertebroplasty.

^aValues are shown as means ± standard deviation.

^b p = 0.007.

^c p = 0.03.

^d p = 0.009.

^e p = 0.025.

^f p = 0.008.

Clinical outcomes

Tables 4 and 5 show surgical and functional outcomes. The mean surgery time was 325 min and mean intraoperative blood loss was 642 g. Postoperative complications included epidural hematoma (n = 1), deep vein thrombosis (n = 1), postoperative wound infection (n = 2), and aspiration pneumonitis (n = 1). All subjects with postoperative complications had undergone shortening. One subject in the VP group suffered delayed surgical wound healing, but there were no serious complications (Table 4). Mean neurological scores improved from 2.4 ± 0.5 points before surgery to 3.3 ± 0.6 points at 1 year after surgery (an increase of 0.9 ± 0.6). Mean walking function score improved from 1.9 ± 1.4 points before surgery to 4.4 ± 1.5 points at 1 year after surgery (an increase of 2.5 ± 1.6). Both paralysis and walking ability generally improved following surgery, but the walking ability of one patient who underwent a tracheotomy because of aspiration pneumonitis worsened.

Table 4.

Surgical outcomes.^a

	SL group	SS group	VP group
Duration of surgery (min)	332 ± 77	359 ± 61^b	267 ± 53^b
Estimated intraoperative blood loss (g)	916 ± 602^c	718 ± 515^d	231 ± 185^c,d
Postoperative complications	Epidural hematoma (n = 1)	Aspiration pneumonitis (n = 1); postoperative wound infection (n = 2); deep vein thrombosis (n = 1)	Treatment delay of operative wound (n = 1)

SL: posterior spinal shortening osteotomy with direct decompression and long fusion; SS: posterior spinal shortening osteotomy with direct decompression and short fusion; VP: posterior indirect decompression and short fusion with vertebroplasty.

^aValues are shown as means ± standard deviation.

^b p = 0.0004.

^c p = 0.0015.

^d p = 0.0081.

Table 5.

Functional outcomes.^a

	SL group	SS group	VP group
Preoperative neurological score	2.3 ± 0.5^b	2.7 ± 0.5^b,c	2.2 ± 0.4^c
Postoperative neurological score	2.8 ± 0.6^d,e	3.4 ± 0.6^d	3.5 ± 0.5^e
Score change	0.6 ± 0.5^f	0.8 ± 0.5^g	1.3 ± 0.5^f,g
Preoperative walking ability score	1.4 ± 1.4^h	2.6 ± 1.3^h,i	1.3 ± 1.0ⁱ
Postoperative walking ability score	3.3 ± 1.7^j	4.4 ± 1.5	5.3 ± 0.6^j
Score change	1.9 ± 1.6^k	1.9 ± 1.2^l	4.0 ± 0.7^k,l

^aValues are shown as means ± standard deviation. Score change = postoperative score minus preoperative score.

^b p = 0.03.

^c p = 0.009.

^d p = 0.014.

^e p = 0.012.

^f p = 0.004.

^g p = 0.007.

^h p = 0.025.

ⁱ p = 0.008.

^j p = 0.004.

^k p = 0.003.

^l p < 0.0001.

Comparison of the surgical procedures indicated that patients in the VP group had shorter surgery times and lesser blood loss than those in the SS and SL groups. Neurological score change and walking score change were significantly greater in the VP group (Table 5).

Radiological outcomes

Table 6 shows radiological data. The mean preoperative LKA was 22°, whereas the mean postoperative kyphotic angle was 5.2°. The mean angle at the final examination was 9.2°, the mean corrected angle was 17°, and the mean correction loss was 4°. Patients in the VP group had significantly larger preoperative LKAs, but there were no differences between the groups in terms of postoperative corrected angle. Correction loss 1 year after surgery was significantly smaller in the VP group compared with the SL group.

Table 6.

Radiographic outcomes.^a

	SL group	SS group	VP group	p Value
LKA within the fused vertebra
Preoperative LKA (°)	17.3 ± 12.6^b	18.4 ± 10.8^c	32.5 ± 12.6^b,c
Postoperative LKA1 (°)	0.1 ± 5.0^d	2.0 ± 7.9^e	15.1 ± 7.6^d,e
Correction angle (°)	17.2 ± 11.7	16.4 ± 7.9	17.5 ± 9.9	NS
Postoperative LKA2 (°)	5.7 ± 5.3^f	6.0 ± 7.8^g	17.5 ± 8.8^f,g
Correction loss (°)	5.6 ± 4.6^h	4.0 ± 3.8	2.4 ± 4.9^h
Presence of subsequent vertebral fracture 1 year after surgery (n)	3	7	3	NS

LKA: local kyphotic angle; NS: not significant; SL: posterior spinal shortening osteotomy with direct decompression and long fusion; SS: posterior spinal shortening osteotomy with direct decompression and short fusion; VP: posterior indirect decompression and short fusion with vertebroplasty; LKA1: LKA immediately after surgery; LKA2: LKA at 1-year follow-up.

^aValues are shown as mean ± standard deviation.

^b p = 0.009.

^c p = 0.01.

^d p < 0.0001.

^e p = 0.0003.

^f p = 0.002.

^g p = 0.004.

^h p = 0.02.

Postoperative adjacent vertebral fractures 1 year after surgery occurred in 13 of the 42 patients (31%). These fractures occurred in 10 of the 30 patients in the shortening groups (33%) and in 3 of the 12 patients in the VP group (25%). No patients experienced new neurological deficits associated with adjacent vertebral fractures.

Discussion

Operative methods for delayed paralysis after osteoporotic vertebral fracture include vertebroplasty, anterior decompression and fusion, posterior decompression and fusion, and posterior shortening. However, each of these methods has both advantages and disadvantages, and no clear criteria exist for deciding which method to use. Posterior spinal shortening was first reported by Saita et al.,¹⁴ and although several studies of the procedure have been published, few of these reported any definitive conclusion.^15

–18

Recent studies on spinal shortening have reported a large number of complications, although the patients were young, indicating that extreme care is required during these procedures.^16,17 In the present study, we found that spinal shortening maintained sufficient correction; even in patients with a kyphotic angle of ≥30°, all of the patients who did not have postoperative complications experienced improvement. We believe this indicates that good surgical outcomes can be expected if the operative method is carefully selected. Posterior decompression with vertebroplasty would be the preferred option for elderly patients who have a high risk of complications because of its relatively low invasiveness. However, Kashii et al.¹² reported that correction loss occurred in 88% of cases in which this operative method was utilized, indicating that preventing correction loss remains an issue with this surgical technique. In 2013 at our medical facility, we switched to less invasive posterior fusion with vertebroplasty. Since then, we have prevented correction loss caused by posterior fusion using screws and hooks. This strategy demonstrated a mean correction loss at 1 year after surgery of 2.4° (a correction loss rate of 14%). Further, vertebroplasty was not inferior to shortening, with a correction loss of 4.6° (a correction loss rate of 27.5%), and satisfactory corrections were maintained. Therefore, we believe that the use of laminar and claw hooks, which attach to the relatively sturdy cortical bone, is an effective postvertebroplasty correction loss countermeasure, even when used on osteoporotic bones.

Our investigation of the number of fused intervertebral foramina indicated that there were no significant differences between the SS and SL groups. In particular, on the distal side, it was desirable to use sublaminar hooks to fuse a small number of intervertebral foramina to promptly preserve the mobile segment. Proximal junctional kyphosis is an issue when performing fusion on the proximal side, so the number of vertebrae to fuse must be carefully considered in each case. When using a claw hook on the proximal side, it is necessary to fenestrate the adjacent vertebra on the upper side. This results in the risk of increased instability of the upper fused vertebra. Some patients in this study had a vertebral fracture in the cranial portion of the uppermost fused vertebra. Therefore, we believe that it is desirable to avoid attaching the claw hook to the vertebral arch of the uppermost fused vertebra whenever possible and instead attach it to the vertebral arch of the vertebra immediately below, and preserve the interspinous ligament between the uppermost fused vertebrae.

Our investigation of improvements in postoperative paralysis and walking ability indicated that subjects in the VP group had significantly higher neurological and walking scores. We believe that this was because the VP group had significantly shorter surgery times, significantly less intraoperative blood loss, and no serious postoperative complications, compared with the two spinal osteotomy groups. Therefore, we concluded that the less invasive nature of the vertebroplasty procedure resulted in satisfactory functional improvements.

Kashii et al.¹² reported that postoperative adjacent vertebral fractures occurred 2 years after surgery in 71% of patients that underwent spinal shortening and in 38% of patients that underwent posterior decompression and fusion with vertebroplasty. In our study, 13 of the 42 patients (31%) suffered adjacent vertebral fractures 1 year after surgery. Previous studies reported that teriparatide is effective in postvertebroplasty adjacent vertebral fractures.^19,20 However, in our investigation of patients who were administered postoperative teriparatide (n = 19) and those who were not (n = 23) indicated that within 1 year of surgery, 16% of the teriparatide group (3 patients out of 19) compared with 43% of the non-teriparatide group (10 patients out of 23) experienced adjacent vertebral fractures. This difference was significant. Teriparatide may be effective in the prevention of postoperative adjacent vertebral fractures in cases of delayed paralysis after osteoporotic vertebral fractures, but this issue requires further investigation.

This study has several limitations. It was a retrospective, single-center study, and the number of patients was small. The surgeries were performed by specialist spine surgeons with more than 5 years of experience with spine surgery, but there were several surgeons because of the retrospective long-term nature of the study. Since most patients were transferred to another hospital 1 or 2 months postoperatively, the rehabilitation programs may have differed between patients. In more recent cases, measures to improve the problem of osteoporotic bone were taken, such as use of claw hooks and administration of teriparatide; these factors could have influenced the results of this study.

Conclusions

We performed a comparative investigation of operative methods used on 42 patients who experienced delayed paralysis after an osteoporotic vertebral fracture. Although spinal shortening resulted in satisfactory correction, it is a highly invasive procedure and its use on elderly patients must be restricted. Vertebroplasty with hooks is less invasive than spinal shortening osteotomy, with fewer complications and comparable functional (walking) and radiographic improvement. Therefore, we believe that it should be preferred in the treatment of delayed paralysis in elderly patients.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Itoi

Sakurai

Mizunashi

. Long-term observations of vertebral fractures in spinal osteoporosis. Calcif Tissue Int 1990; 47: 202–208.

Lee

Yip

KMH

. The osteoporotic spine. Clin Orthop 1996; 323: 91–97.

Yamaguchi

Sugimoto

Yamada

. The presence and severity of vertebral fractures is associated with the presence of esophageal hiatal hernia in postmenopausal women. Osteoporos Int 2002; 13: 331–336.

Yamaguchi

Sugimoto

Yamauchi

. Multiple vertebral fractures are associated with refractory reflux esophagitis in postmenopausal women. J Bone Miner Metab 2005; 23: 36–40.

Harrison

Siminoski

Vethanayagam

. Osteoporosis-related kyphosis and impairments in pulmonary function: a systematic review. J Bone Miner Res 2007; 22: 447–457.

Schlaich

Minne

Bruckner

. Reduced pulmonary function in patients with spinal osteoporotic fractures. Osteoporos Int 1998; 8: 261–267.

van Staa

Dennison

Leufkens

. Epidemiology of fractures in England and Wales. Bone 2001; 29: 517–522.

Kempinsky

Morgan

PP,

Boniface

. Osteoporotic kyphosis with paraplegia. Neurology 1958; 8: 181–186.

Kim

Lee

Jang

. Intervertebral vacuum phenomenon in osteoporotic compression fracture: report of 67cases with quantitative evaluation of intravertebral instability. J Neurosurg 2004; 100: 24–31.

10.

Lee

. Classification of symptomatic osteoporotic compression fractures of the thoracic and lumbar spine. J Clin Neurosci 2006; 13: 31–38.

11.

Tsujio

Nakamura

Terai

. Characteristic radiographic of magnetic resonance images of fresh osteoporotic vertebral fractures predicting potential risk for nonunion: a prospective multicenter study. Spine 2011; 36: 1229–1235.

12.

Kashii

Yamazaki

Yamashita

. Surgical treatment for osteoporotic vertebral collapse with neurological deficits: retrospective comparative study of three procedures—anterior surgery versus posterior spinal shortening osteotomy versus posterior spinal fusion using vertebroplasty. Eur Spine J 2013; 22: 1633–1642.

13.

Frankel

Hancock

Hyslop

. The value of postural reduction in the initial management of closed injuries of the spine with paraplesia and tetraplesia. Paraplesia 1969; 7: 179–192.

14.

Saita

Hoshino

Kikkawa

. Posterior spinal shortening for paraplegia after vertebral collapse caused by osteoporosis. Spine 2000; 25: 2832–2835.

15.

Saita

Hoshino

Higashi

. Posterior spinal shortening for paraplegia following vertebral collapse due to osteoporosis. Spinal Cord 2008; 46: 16–20.

16.

Buchowski

Bridwell

Lenke

. Neurologic complications of lumbar pedicle subtraction osteotomy: a 10-year assessment. Spine 2007; 32: 2245–2252.

17.

Kanchiku

Imajo

Suzuki

. Surgical application and limitation of spinal shortening for delayed paralysis after osteoporotic vertebral fracture. J Spine Res 2015; 6: 899–904.

18.

Suk

Kim

Lee

. Anterior-posterior surgery versus posterior closing wedge osteotomy in posttraumatic kyphosis with neurological compromised osteoporotic fracture. Spine 2003; 28: 2170–2175.

19.

Tseng

Lui

. Prospective comparison of the therapeutic effect of teriparatide with that of combined vertebroplasty with antiresorptive agents for the treatment of new-onset adjacent vertebral compression fracture after percutaneous vertebroplasty. Osteoporos Int 2012; 23: 1613–1622.

20.

Yang

. Comparison of the therapeutic effect of teriparatide with that of combined vertebroplasty with antiresorptive agents for the treatment of new-onset adjacent vertebral compression fracture after percutaneous vertebroplasty. J Spinal Disord Tech 2013; 26: 200–206.