|Home | About | Journals | Submit | Contact Us | Français|
Background: The prosthetic disc nucleus (PDN) device offers an adjunct treatment for patients with degenerative disc disease and herniation, who necessitate surgical intervention, avoiding total-disc replacement or fusion. This prospective, clinical study aimed to gauge the long-term effectiveness of microdiscectomy followed by PDN implantation in relieving pain and improving functional status in patients with symptomatic degenerative lumbar disc disease and herniation.
Methods: Ten patients with a) at least 6 months low back pain and/or sciatica resistant to conservative treatment and b) radiologically documented degenerative lumbar disc disease and herniation have been selected. Follow-up at 6 weeks, 3, 12, 48, and 96 months postoperatively included physical examination, radiological investigation (plain and dynamic radiographs and magnetic resonance imaging), and self-completion of outcome scales (visual analogue, Oswestry, and Prolo functional status). Short Form-36 version 2 Health Survey patient profile at 96 months completed the image of health related quality of life.
Results: Patients' mean follow-up was 100.6 months. Significant improvements in Oswestry, Prolo, and VAS scores were documented (p: 0.004 in all scales at 48 months). Generic health status was rated within the average lumbar disease population (46.36.8 for physical component summary and 45.29.6 for mental component summary). Lumbar spine range of motion (20.211.8 at 96 months) was restricted in relation to normal, but maintained considerable mobility. Treated disc height increased postoperatively (p:0.002) and its maintenance could also be documented in all cases. Disc height at the level above did not show any significant modification. All postoperative MRI showed a non-clinically significant high signal of end-plate on T2 sequences. Clinically relevant complications included one case of pulmonary thrombosis and one case of device extrusion, which was subsequently explanted.
Conclusions: After implantation, most patients continue to enjoy significant pain relief, a considerable amount of mobility is conserved and the disease specific functional outcome is excellent and remains for long, although it could not be supported that the generic health related quality of life is that of the general population.
Successful treatment of moderate degenerative disc disease of the lumbar spine has been a longstanding challenge for medical community. When conservative treatment fails to improve patient's symptoms, few surgical options are available. Mechanical stabilization can provide pain relief, but the procedure is extensive and not always successful. Moreover, the possibility of promoting degeneration in adjacent segments still exists, as the latter must compensate for the immobilized section of the spine1,2–4. One alternative is simple microdiscectomy, which typically eliminates leg pain and may also improve low back pain5. This procedure, though, carries the disadvantage of further destabilizing the segment6–11, since nucleus material must be removed to reduce the possibility of reherniation and eliminate the source of physiodiscogenic pain from the anaerobic metabolic by-products of the diseased disc12. The optimal solution for treating symptomatic degenerative disc disease would be to replace all or part of the disc in such a way as to mimic its normal function. In this regard, total-disc replacement devices13–18 offer a more biomechanically-friendly solution by allowing mobility in the affected segment. Nonetheless, the implant procedure is extensive and these devices do not restore the cushioning function of the replaced disc.
For cases of degenerative disc disease with disc herniation where the degree of degeneration is not sufficiently advanced to warrant a total-disc replacement or fusion, but where there is sufficient pain to necessitate surgical intervention, nucleus replacement technology offers a good alternative. The PDN (Raymedica Inc., MN) prosthetic disc nucleus device, introduced by C.D. Ray in 199619,20 is the first nucleus-replacement device made available, designed to replace the functions of the disc that were lost through degeneration. First clinical trials started in 199619,21, with promising results, and the device and technique of implantation has been modified over the years in term to minimize the risk of migration. Indications for PDN implantation have been recently summarized by Ray22. Briefly, candidates for implantation should not present advanced degeneration, disc height should be more than 5 mm, body mass index (BMI: weight in Kg/height in m2) less than 30, and body weight less than 90 kg if the L5-S1 space is treated. Anterior-posterior dimension of the space indicates the use of two implants if more than 37 mm or one if smaller.
The present study was conducted in a prospective manner in term to evaluate the long-term results of PDN implantation after lumbar microdiscectomy in a group of patients suffering from degenerative disc disease with associated disc herniation, but without segmental instability. The end points of the study were the patient reported clinical outcomes and the radiologically evident effects in the lumbar spine.
A total of 10 patients, 8 male and 2 female, took part in the present study. Their age at the time of operation ranged from 26 to 61 years (mean±SD=41.6±10.4). All of them suffered from long-standing low-back pain and sciatic pain from one-level disc degeneration. The Ethical Committee of our Institution approved the study. Ten patients were enrolled after being informed of the product and the procedure, as well as the intended outcomes and all associated risks. Inclusion criteria were a) lumbar and lower extremity pain caused by unilevel lumbar disc herniation, as proved by magnetic resonance imaging (MRI), and b) no response to adequate conservative treatment for a period of at least six months. All patients presented with severe symptoms interfering with daily activity. Exclusion criteria included a) severe degenerative disc disease with collapsed intervertebral space, b) disc space less than 5 mm, c) serious concomitant disease, d)previous surgical treatment in the lumbar spine, e) severe osteoporosis, and f) segmental instability. The study was initiated in August 1999 and enrolment continued until November 2000.
All operations were performed by the senior author (PS). The disc was approached posteriorly (posterior intralaminar approach) in a manner consistent with unilateral standard microdiscectomy, under the operative microscope, as previously described23, accompanied by flavectomy and trimming of the edges of the superior and inferior lamina only when necessary to visualize the nerve root. Before the PDN devices imlantation, preoperative disc-height measurements had been confirmed through the use of sizing instruments that corresponded to the different PDN sizes available. For all patients in this study, two devices were implanted, anteriorly and posteriorly, to the disc space and finally secured together with a knot. The use of fluoroscopy was essential for the identification of the affected disc space, the verification of the completeness of the enucleation with the help of a water soluble contrast agent, and the alignment of the devices. Patients remained in a recumbent position for twenty-four hours following surgery. They were mobilized thereafter and instructed from the physiotherapist to avoid prolonged sitting, extreme flexion and full physical activities during the first 4 weeks. They were discharged the 4th postoperative day and a lumbar brace was used for 6 weeks, in term to avoid excessive flexion.
The outcomes reported by the patients using self-administration supervised by the survey administrators (NF, AT) were pain, back specific degree of dysfunction, and work disability, assessed with the visual analogue scale (VAS)24, Oswestry scale25,26, and Prolo Functional Economic Scale27 respectively. In addition, generic health status was assessed with SF-36v228.
Along with patient reported outcomes, radiologically evident effects were also studied. Total lumbar range of motion (RoM) was measured in flexion-extension dynamic lateral radiographs with the Cobb method as the change in the angle between the upper surfaces of L1 and S1 vertebral bodies. The height of intervertebral space at the treated level and the level above was measured in midsagittal view of T1 lumbar spine MRI.
Preoperative and postoperative evaluation was performed one week before and 6 weeks, 3, 12, 48, and 96 months after the operation. It included physical examination, recording of patient reported outcomes (VAS, Oswestry scale, Prolo scale and SF-36v2), and radiological assessment (plain films of the lumbar region in anterior-posterior and lateral views, flexion and extension lateral views, and magnetic resonance imaging of the lumbar spine). VAS, Oswestry scale, and Prolo scale instruments were administered in all visits except for the follow-up of 6 weeks postoperatively, because not enough time had elapsed from surgery. The SF-36v2 was administered only in the visit of 96 months after the operation. All follow-up visits included physical examination and radiological assessment (plain films, functional films, and MRI). All operation-related complications were recorded.
The normality of all the analyzed variables was studied with the Shapiro-Wilk test. Friedman's analysis of variance (ANOVA) providing with exact significance levels was used, in order to assess the change of the median values of VAS, Oswestry scale, Prolo scale, and disc heights of the intervertebral space at the treated level and the level above. If the latter was statistically significant, it was followed by post hoc Wilcoxon signed ranks tests with the appropriate Bonferroni correction applied (the critical level of exact significance was set at. 005 for VAS, Oswestry, and Prolo scales and at. 00333 for operated disc height). Norm-based scoring (NBS) was applied on each of the health domain scales and summary measures of SF-36v2 and the group scores were compared with the average range for the group scores of the 1998 U.S. general population (47 to 53 NBS points)28, as well as with the respective mean scores of the back-pain/sciatica specific population, taking minimally important difference (MID) into account28. Evaluation of the association between the recorded variables in pairs was conducted with Pearson product moment correlation, when both variables followed the normal distribution, or with Spearman correlation in all other cases. All analyses were conducted with SPSS 15.0 (SPSS Inc., 2005)29.
Ten patients have been included in the study and completed follow up. One patient denied any radiological evaluation at 48 months after the operation and two patients were lost to follow-up at 96 months after the operation. The demographics, with accompanying history and diagnosis of each patient, are summarized in Table 1 and the clinical and radiological outcomes, with the exception of SF-36v2, are summarized in Table 2. The preoperative and postoperative radiological findings of one case with L4-L5 lateral herniation are illustrated in Figure 1.
There were statistically significant differences in the score of the visual analogue scale across time (Friedman's ANOVA, p: 1.47x10-7, df: 4). It appeared that the score of the visual analogue scale was significantly decreased between the preoperative time period and 3 months (Wilcoxon signed ranks test, p: .004), 12 months (p: .002), and 48 months postoperatively (p: .002). No other statistically significant differences were observed, although the score of the visual analogue scale at the 96 months visit still remained lower than the preoperative in all patients (Figure 2a; Table 3).
In addition, there were statistically significant differences in the Oswestry scale across time (Friedman's ANOVA, p: 1.41*10-7, df: 4). It appeared that the score of the Oswestry scale was significantly decreased between the preoperative time evaluation and 3 months (Wilcoxon signed ranks test, p: .002), 12 months (p: .002), and 48 months postoperatively (p: .002) and between the 3 months and 12 months visit (p: .004). No other statistically significant differences were observed, although the score of the Oswestry scale at the 96 months follow-up still remained lower than the preoperative in all patients (Figure 2b; Table 3).
Furthermore, there were statistically significant differences in the Prolo scale across time (Friedman's ANOVA, p: 1.47*10-9, df: 4). It appeared that the score of the Prolo scale was significantly increased between the preoperative evaluation and 3 months (Wilcoxon signed ranks test, p: .002), 12 months (p: .002), and 48 months postoperatively (p: .002) and between the 3 months visit and 12 months (p: .004) and 48 months follow-up (p: .004). No other statistically significant differences were observed, although the score of the Prolo scale at the 96 months visit still remained higher than the preoperative in all patients (Figure 2c, Table 3).
From the examination of the SF-36v2 Health Survey Profile at 96 months postoperatively (Figure 3), one could observe that the group scores of PCS (physical component summary, 46.3±6.8 NBS points) and MCS (mental component summary, 45.2±9.6 NBS points) were below the average range for the group score of the 1998 U.S. general population, but there was no clinically important difference from the respective group scores of the backpain/ sciatica specific population.
In particular, the group scores of PF (physical functioning, 50.5±5.1 NBS points), VT (vitality, 49.4±8.1 NBS points), and SF (social functioning, 50.7±6.1 NBS points) were within the average range for the 1998 U.S. general population, while there was a clinically important increase from the respective group scores of the backpain/ sciatica specific population.
On the other hand, the group scores of BP (bodily pain, 42.8±8.2 NBS points) and MH (mental health, 45.4±10.3 NBS points) were also below the average range for the 1998 U.S. general population, but there was no clinically important difference from the respective group scores of the back-pain/sciatica specific population.
Moreover, the group scores of RP (role physical, 42.8±7.1 NBS points), GH (general health, 42.6±4 NBS points), and RE (role emotional, 40.3±8.3 NBS points) were below the average range for the U.S. general population, while there was a clinically important decrease from the respective group scores of the back-pain/sciatica specific population.
Relationships between patients reported outcome variables were examined preoperatively, at 3 months, 12 months, 48 months, and 96 months follow-up visits.
At 12 months follow-up, significant relationships between the scores of the visual analogue and Oswestry scale (Spearman : .64, p: .045), visual analogue and Prolo scale (Spearman : -.67, p: .04), and Oswestry and Prolo scale (Spearman : -.86, p: .002) were observed.
At the 48 months visit, significant relationships were detected between the scores of the visual analogue and Prolo scale (Spearman : -.67, p: .04) and Oswestry and Prolo scale (Spearman : -.64, p: .045).
At the 96 months visit, significant relationships were detected between the scores of Prolo and SF-36v2 Norm-Based RP Health Domain Scale (Spearman :. 89, p: .003), time between the onset of back pain and the score of SF-36v2 Norm-Based GH Health-Domain Scale (Spearman : -.73, p: .04), and time between the onset of leg pain and the score of SF-36v2 Norm-Based PF Health Domain Scale (Spearman : -.73, p: 0.4).
No other statistically significant relationships between patients reported outcome variables could be traced at any time point.
Postoperative complications included: (1) pulmonary thrombosis in one patient, which developed one month postoperatively and was successfully treated, resulting in the patient's return to his previous strenuous employment one year after the operation, while retardation of re-enrolment was not relevant to implantation but to the thrombosis, and (2) posterior migration of one device outside the disc space, resulting in nerve-root compression and being subsequently explanted one month postoperatively. In Figure 4, the posterior migration of the device is illustrated in CT scan. The patient remained asymptomatic after the second surgery, and he is still observed closely at regular basis for delayed instability.
The range of motion at 48 months after the operation was 26.4±9.9 (mean±standard deviation) and at 96 months after the operation 20.2±11.8, which might be regarded as restricted in relation to the normal population according to the literature30–32, although it should be noted that a considerable amount of mobility was conserved.
There were statistically significant differences in the operated intervertebral disc height across time (Friedman's ANOVA, p: 6.01*10-6, df: 5). It appeared that the operated intervertebral disc height was significantly increased from the preoperative time period to 6 weeks postoperatively (Wilcoxon signed ranks test, p: 0.002). No other significant changes were observed (Figure 2d; Tables Tables33 and and44).
There were no statistically significant differences in the intervertebral disc height above the operated one across time (Friedman's ANOVA, p: 0.818, df: 5) (Table 3) and no statistically significant relationships between the height of the operated intervertebral disc and the level above at any time point.
Except for the case of implant extrusion, where the implant was displaced posteriorly through the posterior ligament opening and compressed the exiting nerve root, there were two more cases of device dislocation, the first presenting an intraspace rotation of 20° and the second presenting a partial subsidence into the inferior vertebral body. Both dislocations were detected in the 12 month postoperative MRI and remained stable till the end of the study without any clinical consequence at any time. It should also be noted that at the 96 months visit, in the patient with posterior migration of one device, asymptomatic recurrence of the remnant lumbar disc herniation was observed.
A stable radiological finding of all patients was the presence of high signal on T2 sequences of the end plates but again without any clinical sign.
At 96 months, there was a significant relationship between the operated intervertebral disc height and SF-36v2 Norm-Based GH Health Domain Scale (Pearson r: -.75, p: .03). No other statistically significant relationship could be traced between quality of life, pain or functional status variables, and radiological findings at any time point.
The present study suggests that pain, disease specific functional status, and work disability, in a group of patients with lumbar disc disease treated with PDN implantation, were significantly ameliorated between the preoperative period and the immediately postoperative period and remained stable or better until at least 48 months postoperatively. These findings are in accord with previous clinical studies19,21,22,33–35. On the other hand, no significant difference in these outcomes could be detected between the 96 months visit and the preoperative or the other postoperative follow-up, perhaps because of having two patients lost to follow-up, but still all outcomes remained better compared with the preoperative status in all patients.
In specific, it should be noted that PDN implantation resulted in a longer recovery period than simple microdiscectomy, with postoperative lumbar pain graver than that of simple microdiscectomy (as confirmed by the VAS at 3 months), probably due to disc height correction. However, the immediate postoperative lumbar pain was well controlled with usual painkillers, it was shortly decreased, and the long-term outcome justified the use of this technique. Soon lumbar pain and sciatica were satisfactorily controlled and from the 12 months follow-up until the 48 months visit only one patient experienced occasional remnant mild pain, while on the 96 months follow-up mild pain had relapsed in some of the patients and all but two of them had an excellent disease specific functional status and a good or excellent outcome in Prolo scale. Professional activity was allowed after 3 months and most of the patients were able to return to work within the first six months after surgery, while some returned sooner than advised (Table 1).
Regarding the SF-36v2 Health Survey Profile at 96 months postoperatively, the group scores of PCS and MCS indicated that, in general, the patients that comprised the study group did not have the physical and mental performance and capacity of the average population, but that of the lumbar disease population. In specific, the group scores of BP and MH were a sign that the patients' emotional status was characterized more frequently by anxiety and depression than the average general population and that, although they experienced little or no pain for years, they felt limited in their normal activities. Even more, the group scores of RP, GH, and RE suggested that the patients had a pessimistic view for their health in general and also that they had more health-related problems with work or other daily activities than the average population with lumbar disc disease.
On the other hand, the group scores of PF, VT, and SF showed that the patients were characterized by improved subjective well-being and felt they could perform physical and social activities without the health limitations of lumbar disc disease. Perhaps these, largely contradicting, aspects of the patients for their health reflected the avoidance of triggering the pain again and their caution about their health, as they were eight years older than when operated. In consequence, it could be regarded that, though the patient profile was altered as compared with the typical lumbar disc disease population, it still retained many aspects of their disease that should be further investigated.
The significant increase of disc height of the treated level since the first postoperative control and its stability during the study period could be expected given the hydrophilic nature of the PDN device and its concomitant ability to distract the vertebrae upon hydrating. Additionally, the maintenance of the disc height at the level above in all patients after 8 years may reflect a role of the PDN device in the normalization of lumbar spine function. Both findings have been previously reported19,21,22,33–35, but for shorter time of follow-up. Furthermore, this study also showed that the range of motion in the long-term had maintained a considerable amount of mobility. It is clear from this study that the restorative role of the PDN device in the lumbar region is stable at least for a period of eight years.
Another interesting radiological finding was the persistence of T2 high signal at the level of treated endplates, but without any clinical consequence, although present even after eight years. It is difficult to interpret this finding, but it has been suggested21 that it is probably a sign of remodeling due to uneven distribution of loads at the contact areas within the disc space.
An important piece of this investigation was the examination of the relationships between the outcomes of interest. For example, it should be noted that the significant correlations between the scores of the visual analogue, Oswestry, and Prolo scale at the 12 months and 48 months visits were probably due to the fact that nearly all patients had excellent scores in these scales at these follow-ups.
Moreover, at the long term visit, significant relationships axist between the scores of various SF-36v2 Norm- Based Health Domain Scales and other patient reported outcomes. In specific, the association between 96 months Prolo and RP could perhaps be explained by the fact that they express similar notions, while the association between the preoperative duration of back pain and the score of GH could probably imply that the patients' estimation of their health in general was related with the time they suffered with back pain. It should also be noted that the correlation between the preoperative duration of leg pain and the score of PF could probably be attributed to an association between physical performance as evaluated by the patients themselves and the time they experienced pain in their lower extremity.
In addition, the only significant correlation of the radiological findings with each other as well as with the other outcomes was an association between the operated intervertebral disc height and SF-36v2 Norm-Based GH Health Domain Scale at 96 months, which remains to be clarified by further research. Interestingly, though, no statistically significant relationship between the height of the operated intervertebral disc and the level above or the treated disc height and pain and the other patient reported outcomes could be traced.
Although amelioration in visual analogue, Oswestry, and Prolo scales was observed, this is not a comparative study that could render these results to PDN implantation or not. Instead, the aim of this study was to evaluate the feasibility and the long-term clinical results of PDN implantation. While most clinical studies have reported short-term results of PDN implantation19,33, only the Wiesbaden group21 reported results from a 2-year follow-up and Rosales-Olivares et al36 from a 4-year follow-up. It seems that this is the first publication reporting results of 8 years of follow-up on a single institution.
The major limitation of this study was its small sample size enhanced by the withdrawal of two patients, which could be rendered responsible for the low power of the statistical tests. The power was not only immediately limited by the sample size, but from the nature of the tests used as well. As most of the variables followed a sparse distribution, the statistics used in their analysis were mostly non-parametric, which are frequently underpowered to detect significant differences and associations. On the other hand, this fact results in adding more confidence to the statistically significant results of the current study.
In patients presenting with low-back pain and sciatic pain due to degenerated disc disease and disc herniation, microdiscectomy associated with artificial disc implantation has the theoretical advantage to eliminate on one hand the risks of long-term disc degeneration, disc collapse, and eventually instability of microdiscectomy alone and on the other hand the risk of adjacent level disease associated with fusion37. Currently, the surgical approaches used for PDN implantation are (1) posterior intralaminar, (2) paraspinal transforaminal, and (3) anterior-lateral transpsoatic approach (ALPA)33,35. In the present study, the intralaminar approach was preferred as less invasive and more familiar. Classic microdiscectomy represents a refinement of the standard hemilaminotomy with emphasis on magnification, lighting, and hemostasis.
It seems that the most crucial technical point is the minimal necessary opening of the posterior ligament / annulus complex in term to avoid implant extrusion (Figure 4), which may cause some difficulty in disc material removal. In the case of implant extrusion after 2 months postoperatively, immediate removal was easily performed for the posterior extruded device, but the anterior one was removed with difficulty. It should be noted that in this case disc height postoperatively increased significantly comparing to usual. Over-distraction might have been the cause of the extrusion, especially if it was associated with annulus incompetence. Appropriate size selection and limited laminar distraction are crucial for avoiding this complication.
It should also be noted that the two patients that experienced postoperative complications conducted a normal life thereafter at least until the last visit, and that, with the exception of the posterior migration of one device, all other cases of implant dislocation were asymptomatic.
Despite the restrictions of the limited number of the study group, the present study suggests that the clinical outcome of PDN implantation is satisfactory, providing significant pain relief and allowing the majority of the patients to return shortly to work. It is also suggested that the disease specific functional outcome is excellent and remains for long, although it could not be supported that the generic health related quality of life is that of the general population. A considerable amount of mobility is conserved, even though not that much as in the general population. No instability or adjacent level disease has been documented after 8 years of follow-up.
Although results from PDN implantation are very encouraging, further prospective studies comparing simple microdiscectomy with and without implantation are needed in order to establish more trustworthy evidence as regard to this much promising technique.
Conflict of Interest Statement
We state that there is no conflict of interest regarding this manuscript and any of its authors, as no funds were received in support of this study.