Lumbar Interspinous Devices: Indications, Surgical Aspects, Clinical Considerations

Abstract

Interspinous devices (or spacers) are currently used in lumbar spine surgery, but their use is still controversial, mainly due to confusion between so-called first and second-generation spacers, and unrespect of formal indications to implant. Our first aim is trying to offer an interpretation for a correct indication of surgical implantation, paying additional attention to those pathological conditions in which these devices must not be used; second, to identify the right clinico-surgical method to avoid potential errors in this surgery; third, to stress those surgical tricks directed to perform a successful implant of these spacers.

1. Introduction

Nowadays, surgical implantation of lumbar interspinous devices (ISDs) has a widely accepted role in the management of spinal degenerative disorders, and this is confirmed by the current high number of scientific papers dealing with their use. Nevertheless, the issue of ISDs still remains quite controversial, and this is mainly due to two reasons: first, experimental data are often misinterpreted and therefore not considered strong enough to justify a clinical application; second, sometimes, indication to an ISD implant isn’t done in a correct way and the subsequent potential poor surgical outcome being then attributed to the ISD itself (and not to the surgeon!); third, the technique of implantation of an ISD is too often deemed ‘easy’ and, hence, those basic surgical steps not respected, so opening the way to poor postoperative results.

2. Historical remarks

We are indebted to Prof. Jacques Sénégas, who first elaborated the concept of ‘lumbar dynamic fixation’ about forty years ago and developed the ISD named Wallis [1]. Since then, a deluge of ISDs, often having different experimental and clinical supports, have been marketed and scientific papers concerning their use go on to be published; this context led to an unfavourable confusion about their correct use and, of course, careful and free from prejudices evaluation of clinical application and outcome.

3. Experimental data

Biomechanical studies have shown that all of ISDs act grossly in the same way: they enlarge spinal canal dimensions, increase neuro-foramina height, reduce intradiscal pressure in neutral and extended position [2]; moreover, they have a ‘primary stabilizing effect’ [3]. The ISDs designed ‘for fusion’ additionally show to promote bone fusion at the instrumented level.

4. Interspinous devices

ISDs may be grouped into two categories: the first group includes the so-called ‘simple’ spacers (first-generation), and the second one is constituted by the spacers ‘for fusion’ (second-generation); ‘simple’ spacers maybe static or dynamic (the term ‘elastic’ is sometimes used to indicate some dynamic devices). We’ll here take in consideration some of the most frequently implanted (and studied) devices, remembering that some of them are no more marketed.

4.1 Spacers ‘simple’

Wallis (Abbott, USA). It’s made of polyetheretherketone (PEEK) and two flat woven Dacron bands; experimentally these latter resist traction of 200 decaNewton (daN) and stretch approximately 20% before failure by overloading; biomechanical studies demonstrated that, after a surgical decompression, the Wallis implant produces a significant restriction of segmental range of motion (ROM), limiting the flexion–extension plane, having no effect on lateral bending and minimal impact on axial rotation [4]; moreover it’s able to markedly release the intradiscal pressure in extension (Figure 1).

Figure 1.

Lumbar stenosis. Sagittal CT scan at 15-year follow-up: the Wallis spacers, correctly placed, are visualized at L3-L4 and L4-L5 (black arrows).

DIAM (Medtronic, USA). It’s a dynamic stabilizer, made of a silicone core with polyester sleeve, providing segmental stability and restoring foraminal height, particularly in those disorders accompanied by hyperlordosis [5].

4.2 X-STOP (taken off the market)

SUPERION (Vertiflex, USA). This device may be a valid alternative to laminectomy in a well-restricted population of patients affected by one- or two-level lumbar spinal stenosis (LSS); however, patients should be informed of the high rate of possible reintervention [6].

COFLEX (Paradigm Spine, USA). It was the first of the so-called titanium ‘U-devices’, also due to its appearance as a ‘U-shape’ on lateral radiographs; it received FDA approval for implant in one- or two-level lumbar stenosis; later there were some observations concerning its use at the superior stenotic level of a decompression/fusion operation, to reduce the incidence of adjacent segment disesase (ASD) (the ‘topping-off’ technique) [7].

BACJAC (Surgalign, USA). It’s made of PEEK, a bioplastic which has an elastic modulus close to that of bone [e.g., human femur has a flexural modulus of about 20 gigaPascal (GPa), and PEEK Optima has about the same value]; its advantages are a self-deploying way of insertion, tissue-sparing technique, minimal (or absent) risk of subsidence, no MR/CT artefacts [8]. Stand-alone implant or following uni- or bilateral decompression provides mid-term relief of low-back and radicular pain, and improvement of neurogenic intermittent claudication (NIC), in the following lumbar spine disorders: central and foraminal stenosis, recurrent disc protrusion associated with foraminal stenosis, mild antero- or retrolisthesis, degenerative disc disease (DDD), synovial cysts associated with pseudospondylolisthesis. When considering these disorders, this device seems to provide a ‘primary stabilizing effect’; a 2-year follow-up confirms that this effect is still active, with no need for subsequent more invasive fixation surgical procedures [3].

APERIUS (Medtronic, USA). It was the first percutaneous titanium ISD available on the market; it was designed to be used as a stand-alone implant [9]; a good outcome at six months follow-up in LSS patients, with no migration of the device has been reported [10].

4.3 Spacers ‘for fusion’

ASPEN, ALPINE (Zimmer, USA). The ASPEN was the first device projected to achieve bone fusion, so revolutioning the market of existing spacers, which were simple distances. Experimentally it showed a reduced range of motion (ROM) in flexion/extension about equal to anterior lumbar interbody fusion (ALIF) [11]. Moreover, it remains one of the few devices which may also be implanted at the thoracic segment (Figure 2), and it’s worth that a ‘flared’ model is available to be best adapted to the L5-S1 level. The ALPINE differs from the ASPEN device because it’s adaptable in length and may be inserted in compression or extension.

Figure 2.

Thoracic degenerative disc disease. Lateral x-ray film: the ASPEN device is correctly inserted at the mid-thoracic indexed level.

ROMEO 2 PAD (Spineart, CH). It’s composed of two plates, one polyaxial, and a central core, made in PEEK or titanium; compared to the other devices, it has the advantage of being applied in a pre-assembled format, so making the surgical procedure faster and easier.

MINUTEMAN G3 (Spinal Simplicity, USA). This device for fusion is probably one of the fewest to be implanted percutaneously.

PITBULL (BMK, Korea). This device probably has one of the largest central cores to be filled with bone (to promote fusion).

SEASPINE (SeaSpine Orthopedics Corporation, USA). This spacer may also be applied at the thoracic segment (Figure 3), and a dedicated model for L5-S1 is available.

Figure 3.

Thoracic degenerative disc disease. Lateral x-ray film: the SEASPINE device is well visualized at the mid-thoracic indexed level.

BACFUSE (taken off the market). It derived from the BacJac spacer (see above) and was designed to act as a device for fusion; despite its somewhat bulky design, which probably contributed to its scarce diffusion, our long-term follow-up showed optimal bone incorporation without loss of regional lordosis (Figure 4).

Figure 4.

Spacers ‘for fusion’. Sagittal MR T1-weighted image at 2-year follow-up: the BACFUSE spacer is seen at L4-L5 (black arrow); the regional lordosis is unchanged.

5. Indications

Main indications for an ISD implant are grouped under the wide term of lumbar spine degenerative disorder; although this comprehensive definition may represent a protective shield for the surgeon, we strongly recommend trying to obtain an exact preoperative diagnosis, not only because ‘the more correct the indication, the higher the rate of success of the intervention’, but also because, concerning the term of lumbar degenerative conditions, instructions for use (IFUs) of different ISDs may vary. Therefore, we’ll here describe different degenerative lumbar disorders amenable to ISD implants.

5.1 Central stenosis

Central lumbar canal stenosis probably represents the most important indication of an ISD implant, and this has been extensively confirmed. Having this said, the main relevant aspects are two: performing or not decompression of neural structures before ISD implant, and coexistence of a listhesis which could be not adequately treated by ISD. About the first point, the decision to perform a decompressive step may be taken simply based on preoperative magnetic resonance (MR) or computed tomography (CT) scans and this is unequivocally true for severe stenosis causing bilateral NIC (Figure 5). Milder degrees of central stenosis could not necessarily require pre-ISD neural decompression; in such an occurrence, intraoperatively we behave as follows: after having skeletonized the indexed level, the application of traction on spinous processes (using two Backhaus forceps) is made; if frank hypermobility is appreciated, the ISD is inserted without delivering a flavo-laminotomy, because we know that it will realize an enlargement of the central canal and restore the bi-foraminal height, such to obtain a sufficient neural decompression (as we say, ‘we let the ISD make its dirty job’). It goes without saying that paracentral canal stenosis causing ‘intractable’ ipsilateral symptoms always needs a unilateral decompression before ISD insertion.

Figure 5.

Severe lumbar central stenosis. Axial MR T2-weighted image: preoperative study.

5.2 Foraminal stenosis

This is probably one of the most correct and valuable indications (Figure 6). Intraoperatively, after having performed a plain microsurgical foraminotomy on the affected side, only minimal contralateral skeletonization will be necessary before applying the ISD. We here must remember that in these cases, the choice of the ISD could favour those spacers, which can be inserted by sparing the supraspinous ligament.

Figure 6.

Lumbar foraminal stenosis. A preoperative axial CT scan shows the marked narrowing of the left neuroforamen (black arrow).

5.3 Recurrent disc herniation

Operated lumbar disc herniations may recur in different ways; clinical presentation is often persistent back pain, exacerbated by movements, associated with radicular irradiation, and the MR scan may show a small extruded disc fragment, surrounded by scar tissue, just at the entry of the neuroforamen; such condition, usually refractory to medical and physical therapy, is the ideal target for a micro-surgical decompression of the involved nerve root(s), associated with a selective discectomy, followed by ISD implant.

5.4 Isolated disc herniation

Implantation of an ISD has no role in treating an isolated primary lumbar disc herniation; however, some empirical evidence delineates a long-term reduced rate of recurrent disc herniation when a spacer is applied at the end of the removal of the extruded disc. In those selected cases in which the herniated fragment is associated with a frank degenerated disc, causing intense low-back pain, the option of the additional insertion of an ISD may be considered [12].

5.5 DDD

As already stated in Section 3, ISDs reduce intradiscal pressure and have a stabilizing effect [which is quantitatively comparable to ALIF, meaning about 50% reduction of range of motion (ROM)]; basing upon these experimental data, we implanted ISD in several cases of single DDD, obtaining good results in terms of long-lasting back pain relief (Figure 7). We wish to stress the occurrence of a hyperintense signal of interspinous ligament in MR T2-weighted images, which must be considered an additional element of disc disease, hence supporting the indication of an ISD implant (Figure 8).

Figure 7.

Lumbar DDD. Sagittal MR T1-weighted scan at 6-year follow-up: the L5-S1 degenerated disc is well seen, and the spacer is correctly inserted between the two spinous processes (black arrow).

Figure 8.

Lumbar degenerative disc disease (DDD). Preoperative sagittal MR T2-weighted image: the L5-S1 degenerative disc is shown, as well the hyperintense signal of the interspinous ligament (black arrow), which may be considered an additional element of the degenerative process.

5.6 Post-surgical instability

This is a well-known condition, which may follow one or more lumbar surgical procedures: refractory pain is mainly located at the low-back and flexion–extension x-ray films usually show minimal or no listhesis; these patients may be adequately treated by an ISD implant, which is safer than other invasive fixation procedures.

5.7 Spondylolisthesis

During a procedure aiming to stabilize a lumbar anterolisthesis, the matter of reducing the slipped vertebrae occurs, considering that reduction may potentially result in permanent nerve root damage; our experience led us to the conclusion that spinous fixation by means of a second-generation ISD plus additional bone fusion provides a valid result when used in those cases of spondylolisthesis up to grade-1 (Figure 9).

Figure 9.

Lumbar spondylolisthesis. Preoperative sagittal MR T2-weighted scan: the L4–L5 stenosis and listhesis grade-1 (white arrow) is well seen.

5.8 Synovial cyst

Lumbar synovial cysts are frequently diagnosed as the cause of lumbar and sciatic pain; except for rare instances of spontaneous resorption, most of these cysts require surgical removal; because of the association of satellite spondylolisthesis, which in our experience occurs in almost all cases, it may be advisable to correct the vertebral slippage contextually to the excision of the cyst. Associated anterolisthesis is never grade-2, so we routinely insert an ISD once the cyst is removed; our long-term follow-up x-ray controls demonstrate no need for additional stabilization procedures.

5.9 Transitional vertebra

Vertebral anomalies at the lumbosacral junction, named as sacralization of L5 or lumbarization of S1, are frequent and may give rise to refractory, mechanical back pain; in these cases, x-ray films usually show mild instability, which may appear as L5-S1 retrolisthesis. ISD implants proved to be effective in providing long-lasting, excellent back pain relief (Figure 10).

Figure 10.

Transitional vertebra. Preoperative sagittal MR T2-weighted scan: typical appearance of vertebral bodies and discs at the lumbosacral passage (white oval); L5-(lumbarized)S1 retrolisthesis may be appreciated too.

5.10 Adjacent segment disease

As already stated before (see Section 4), the implant of an ISD at the rostral level of a multilevel fixation may be useful to reduce the future risk of ASD. We never adopted this technique in a prophylactic way; however, we implant a second-generation ISD to treat those cases of ASD, in which a lumbar stenosis w/wo associated listhesis occurs rostrally or caudally to a previous posterior pedicular fixation (Figure 11). Of course, a careful evaluation of preoperative neuroradiological examination is mandatory to check the shape and volume of the spinous process to be included in the ISD anchor; intraoperatively, the terminal part of rods and the surgical scar may hinder skeletonization of the indexed spinous processes, but we never had to temporarily remove one of them to insert the ISD.

Figure 11.

Adjacent segment disease (ASD). Lateral x-ray film at 2-year post-ISD implant: pedicular fixation L2-L3-L4 is seen; the spacer at the caudal ASD level is well inserted between the spinous processes (black arrow).

6. Contraindications

ISDs may differ from each other in several aspects; however, all of them share two important features: first, they cannot be implanted at the level of a pars defect; second, they cannot be used in case of a spondylolisthesis >grade-1.

Concerning the first occurrence, it must be stressed that sometimes preoperative neuroradiological exams may scotomize an isthmic defect and the related mild listhesis be wrongly attributed to a degenerative disorder; if so, intraoperatively, the surgeon will have to face an unexpected pathoanatomical finding and be obviously addressed to change surgical strategy timely.

Anterolisthesis grade-2 may involve different scenarios. A surgeon may incorrectly interpret preoperative images and judge a borderline slippage as grade I instead of grade II, often due to a misleading neuroradiological report; this is why a flexion–extension x-ray film is mandatory in the preoperative setup of a lumbar instability. Or, a frank listhesis, deemed by the experienced surgeon when unexpectedly found out at surgery, will discourage him from implanting the scheduled ISD and possibly shift to a pedicle fixation.

Of course, several additional disorders (such as osteoporosis, coexisting infective states, bone anomalies of the posterior vertebral arch, etc.) constitute a contraindication to the implant of an ISD, but they are easily ascertained at the time of preoperative consultation.

7. Surgical tricks

The interspinous space may present multiple anatomic variations, and this must be kept in mind by the surgeon, particularly the shape of the space and its matching with the shape of the spacer [13].

The L5-S1 level. Anatomy of the passage L5-S1 may be quite variable, and this is mainly due to the shape and height of sacrum protuberance; when an ISD implant at L5-S1 is planned, great attention must be paid to the preoperative study, and if the MR scan is not clear about bone aspects of sacrum, a CT examination must be done before surgery.

Regardless of the type of ISD, the basic rule to implant it successfully is to insert it as deep as possible, and the right placement is at the line of articular processes (Figure 12); suboptimal insertion of the spacer does not necessarily involve a poor clinical outcome, but x-ray follow-up controls, at least over the first postoperative year, are strongly recommended.

Figure 12.

Depth of ISD implant. Postoperative axial CT scan: the base of lateral plates of the spacer (black arrows) are correctly placed at the level of articular processes (black dotted line).

The surgical technique to implant an ISD may differ, and it’s also due to the type of spacer and the strumentarium to implant. Some spacers are to be applied through the interspinous ligament, assembling the two parts once they are in situ; this implies that the supraspinous ligament has not to be sacrificed (sparing technique) offering an additional barrier to potential superficial dislocation of the device. Alternatively, if you use a ‘U’-spacer or opt for a not-sparing technique, attention must be paid not to over-decorticate the facing surfaces of spinous processes because this manoeuvre does not enhance bone incorporation and may result in the spinous process weakening.

The last step of implantation is clamping the lateral plates of the spacer against spinous processes; this is probably the most hazardous manoeuvre in terms of risk of rupture of spinous process and must be done cautiously, squeezing the dedicated compressors in a step-by-step alternate way. Once you have compressed the device in situ, we suggest checking its final mechanical seal: the spacer is grabbed by a needle-holder and pulled superficially and horizontally, so observing that the lumbar segments move as a unique block.

Multilevel stenosis of the lumbar canal, though usually symptomatic at just one level, may be encountered in the daily practice and the matter of operate upon two or three levels is not trivial; are ISDs insertable at multiple levels? Yes, they are. However, much attention must be paid to the preoperative CT scan: the decisive factor is represented by the height of the middle of the three indexed spinous processes, which must be high enough to receive plates of both spacers (Figure 13). Based on our experience, the implant of two spacers having a metal-to-metal final construct at one or both sides of the middle spinous process may result in a poor clinical outcome in terms of back and/or radicular pain.

Figure 13.

Two-level lumbar spinal stenosis. Flexion x-ray film at 2-year follow-up: the spacers are correctly placed, and the construct is no metal-to-metal.

8. Complications

Wrong level. Although bizarre to comprehend, an ISD may be implanted at the wrong level (Figure 14); this may occur when ‘kissing’ spinous processes are found at surgery, often in an obese patient, so making intraoperative fluoroscopic controls incorrectly interpreted.

Figure 14.

‘Wrong level’. Sagittal reformatted CT scans: this patient came to our attention after having performed an “interspinous stabilization L4-L5” elsewhere (as reported in medical discharging chart); soon after surgery her bilateral claudicatio radicolaris started to worsen; on the left, the DIAM device applied at the L3-L4 (wrong level), and, on the right, the COFLEX device correctly inserted at L4-L5; the DIAM at L3-L4 was intentionally left in place to avoid late compromise of segmental stability.

Spinous process rupture. In most of cases in which this complication happens, this occurs intraoperatively; factors leading to such a complication are incorrectly sized device selection and poor surgical technique; we never encountered early or late postoperative rupture of a spinous process in an ISD-implanted patient, and in those rare cases in which we found out this complication, retrospective review of postoperative x-ray and/or CT exams indeed revealed that the rupture had occurred at surgery.

Regardless of the type of ISD used, most of the fractures involve the junction of the inner and middle thirds of the cranial spinous process. When we started inserting a PEEK ISD, the first bone fracture case was discovered only through CT scan, which we routinely performed on postoperative day 1, and it passed clinically unnoticed (e.g., postoperative pain was the same that uncomplicated patients referred). How to manage this complication? It must be remembered that the lateral plates of the ISD (especially those of second-generation) act as fixators of the ruptured spinous process, usually making revision surgery unnecessary; the use of a lumbar corset for two months (instead of one) is advisable.

Dislocation of ISD. When a spacer migrates into the spinal canal, urgent revision surgery, aimed to remove it and repair the dural sac, if lacerated, is mandatory. Migration of ISD in the paravertebral soft tissues is far more frequent and may be accompanied by recurring pain or claudication; in such cases, an election revision operation is planned, and two options are available: once the displaced spacer has been removed, a new, different-type ISD is inserted; alternatively, transpedicular fixation is performed.

Miscellaneous. Scientific literature reports different complications: dural tear, nerve root injury, late lumbar/radicular recurrent pain, neurological impairment, and chronic lumbar pain.

9. Personal considerations

Dutch school of neurosurgery goes on to keep a leading role in the management of spinal stenosis; over 70 years ago, in a tribute to another giant of spinal surgery (Clovis Vincent), Prof Verbiest first described the pathophysiological correlation between symptoms of cauda equina compression and spinal stenosis, which, curiously and interestingly, he defined as ‘rare’ [14]. Nowadays, the group of Moojen goes on sharing elegant works concerning the outcome of patients receiving an ISD implant: in their very recent randomized control trial paper, a 5-year follow-up good outcome in ISD implant, if compared to conventional lumbar decompression was reported [15]. Moreover, it’s worth that they also state that the ISD group seems to have less back pain during long-term follow-up; this constitutes a very important issue: our experience, using different types of ISD, supports the same conclusion: ISD patients, implanted for treating lumbar stenosis (and, above all, for a DDD), early in the postoperative course refer that back pain is of different quality and far milder than preoperatively, and this substantial pain relief maintains over time.

We started to implant ISDs in 2005, and since then, we have used different types of spacers; on time, we realized that each of them has its own pros and cons, which must be well-known by the surgeon to best adapt to any patient elected for this kind of surgery.

In my daily activity, patients affected by spondylolisthesis grade 2, in which implanting an ISD represents a formal contraindication, are very few; on the contrary, the number of cases harbouring a listhesis up to grade- 1 is far larger, and almost all of them are ideal candidates to ISD implant; this means that at least as first surgical management step, more invasive procedures (such as transpedicular fixation, anterior lumbar interbody fusion, XLIF, etc.) are not necessary.

Lumbar spinal stenosis affects about one-third of people aged 70, and those of them requiring a surgical treatment are offered different options, the first of whom being a decompressive one- or more-level laminectomy. Based upon our 30-year experience, we think that: a] standard one- or multilevel laminectomy [e.g., with the removal of spinous process(es)] for the management of lumbar stenosis has to be abandoned; b] of course, minimally invasive decompressive procedures [with sparing of spinous process(es)] are still highly valid and have to be correctly proposed to the ‘stenotic’ patients as a first surgical option; c] ISD implant, coupled or not with a decompressive step, actually represents the ideal treatment for one-level lumbar stenosis w/wo listhesis up to grade1; d] empirically, if compared, the group of decompressed-only and that of ISD-implanted patients share about the same high rate of improvement of NIC, but the latter seems to fare largely better in terms of relief of lumbar pain.

Most of the scientific papers dealing with the outcome of different surgical operations, performed to achieve lumbar stabilization take bone fusion rate into account as the main indicator of success. This methodological approach was developed and consolidated more than 50 years ago when global aspects concerning this kind of surgery were different: hardware, age of patients (usually less than 80 years), and diagnostic tools (grossly limited to x-ray films). Nowadays, we feel that the concept of bone fusion, adopted to evaluate the outcome of surgical options for DDD and lumbar stenosis, should be replaced by that of ‘bone synthesis’, which is more helpful and adherent to global (functional, clinical, and radiological) analysis of operated patients.

One of most frequent criticisms made to ISDs is the absence of long-term follow-up studies; this may be partially true and it’s also due to the replacement of the first-generation spacers with the newest ones, probably making ongoing studies obsolete and, therefore, abandoned. Moreover, recent covid-19 epidemics determined a tremendous gap in the follow-up studies, especially concerning many aged patients operated upon for lumbar stenosis, who deceased or were lost over last three years. Currently, our documented experience with hundreds of implanted ISDs allows us to affirm that: (a) a first-generation PEEK spacer demonstrates no neuroradiological migration or bone spinous process rupture at a 13-year follow-up (Figure 15); (b) a second-generation device shows the same good result at a 10-year follow-up (Figure 16).

Figure 15.

Lumbar DDD. Three-dimensional reformatted coronal CT scan at 13-year follow-up: the BACJAC spacer is well inserted between the spinous processes of L5 and S1 (dotted black oval); a spinal cord stimulation electro-catheter is also seen at L2-L3.

Figure 16.

Lumbar DDD. Preoperative (left) and 10-year follow-up (right) sagittal MR T2-weighted scans: the L4-L5 degenerated disc (left, white arrow) and the long-term correctly placed ROMEO 2 PAD spacer (right, white arrow) are well seen.

10. Final reflections

Of course, most of the spinal disorders described in Section 5 may be treated by means of other surgical techniques; concerning this statement, some observations help both the surgeon and the patient. First, a ‘(at least) second opinion’ will help the patient to take the right decision to manage his/her problem; second, ISD implant may be considered as an ‘intermediate step in the continuum of care for patients with NIC associated with lumbar spinal stenosis’ [16].

In some western countries ISDs are widely implanted by physiatrists and interventional pain specialists; recently, a small series of complications related to ISD implants, in particular, the migration of the spacer in the spinal canal, performed by pain specialists, was reported [17]. Over the last two decades, the industry has promoted the ISD surgical implant and, based upon its presumed ‘easiness of technique’, it forced its use by ‘non-spine surgeons’; the matters of incorrect surgical indication, poor knowledge of the biomechanics of the spinal column, inability to manage potential complications, have rocketed. It has been strongly recommended that ‘spinal instrumentation, however minimally invasive, should be performed by fellowship-trained spine surgeons’ and, in the same way, a consultation with an experienced spine surgeon before an ISD implantation should be performed [17].

As for most lumbar spine surgical procedures, the problem of costs exists for ISD implant too; this directly leads to our (as surgeons) formidable task to conciliate budget-based medicine and the goal of ‘best treatment for the ideal patient’; therefore, settings of clinical assessment, surgical experience, profound knowledge of available spacers, must be managed with skill and simultaneously by the experienced surgeon. This goal will revitalize the somewhat abandoned ‘ancient doctor-patient relationship’ and decrease the high rate of litigation in spinal surgery [18].

Very recently, the experience of ISD implantation was reported as being very satisfying in terms of minimal surgical invasiveness and good clinical outcome [19]; based upon our 20-year ISD-dedicated surgery, we truly share these observations: early in the postoperative course, most of our ISD-implanted patients refer to stand and walk with a straight back after a long time of pain and claudication, and this excellent result usually maintains over time.

Acknowledgments

Our huge thanks run to Dr Susanna USAI (Acting Director, Neurologia 3, Istituto Neurologico ‘Carlo Besta’, Milan, Italy): her acute and stimulating observations are always precious to us.

Conflict of interest

The Author declares no conflict of interest.