Introduction

Low back pain is the primary reason for patients visiting physicians in Japan. This condition ranks second among the complaints in the United States, following upper respiratory tract complaints [1]. Herniated discs (HD) account for 4 % of the total cases of mechanical low back pain, and occur in approximately 2.8 million patients annually [2]. Patients with lumbar HD experience acute onset unilateral or bilateral lower extremity pain and numbness associated with the low back pain [3]. However, 70 % of lumbar HD patients recover from sciatica within 6 weeks of its onset. Thus, considering the natural history of HD, the overall patient prognosis is good. However, an estimated 10 % of patients will experience continued pain and neurological deficits, and surgical intervention should be considered for these patients [1]. A systematic review comparing surgical intervention and conservative management indicated that surgical intervention enables faster pain relief, compared to prolonged conservative treatment, during short-term follow-up, although no marked differences are noted during long-term follow-up [4, 5].

Lumbar HD patients are primarily between 20 and 40 years of age, employed, and play an active role in society. To reduce the quality of life issues caused by surgical intervention, including microdiscectomy, more effective and less invasive treatments—which require less treatment time—need to be developed.

Molecular genetics and biology

Several previous studies have described the potential gene markers for lumbar disc diseases, including collagen 9A2 [6, 7], vitamin D receptor [8], matrix metalloproteinase (MMP)-3 [9], cartilage intermediate layer protein [10], collagen 11A1 [11], thrombospondin (THBS2) [12], sickle tail (SKT) [13], MMP-9 [12], asporin (ASPN) [14], and carbohydrate sulfotransferase [15]. Systematic reviews demonstrated that there is moderate evidence of correlation of ASPN, COLXIA1, SKT, THBS2, and MMP-9 with HD [16]. However, further studies are needed to identify the gene that is strongly correlated with the disease. Assessment of the upregulation and downregulation patterns of specific candidate genes in animal models may facilitate the identification their precise roles in disc degeneration.

Regenerative medicine techniques for the treatment of disc degeneration have recently been developed for clinical use. Co-culture of autologous mesenchymal stem cells with the patient’s nucleus pulposus cells or annulus fibrosus cells may be a good alternative for regenerating the degenerated disc matrix [17]. In addition, the administration of autologous platelet-rich plasma may promote a reparative effect on degenerative disc tissues [18]. These methods suggest the potential for regeneration of degenerated disc tissues in the near future.

Development of image examination techniques

The resorption process of HD was demonstrated using sequential magnetic resonance imaging (MRI), and this resorption process may be the reason for the relatively good prognosis in cases of HD (Fig. 1) [19]. Vroomen [20] demonstrated that 70 % of patients with HD indicated the disappearance of sciatica within 6 weeks of its onset. In addition, the non-contained classification types of HD, such as transligamentous extrusion and sequestration, as well as the enhanced contrast noted around HD, indicated a high tendency for resorption, thus suggesting that vascularization around the HD would be an important factor for HD resorption [21]. This HD resorption phenomenon was demonstrated in the lumbar, thoracic, and cervical regions of the spine [22, 23].

Fig. 1
figure 1

Sequential magnetic resonance imaging of a 66 year-old man demonstrating resorption of the herniated disc

Full size image

A recent study indicated that it was impossible to assess good or unfavorable outcomes using MRI at the 1-year follow-up for patients who had been treated for lumbar HD [24]. However, with recent developments in MRI, the observation of tract fibers with diffusion tensor tract images [25] as well as the identification of symptomatic nerve roots (due to spinal disorders) with diffusion-weighted imaging, is currently possible [26]. Through these new technological advancements in MRI, the identification of symptomatic nerve tissues with HD will be possible in the near future, thus facilitating more accurate investigations for HD patients.

Surgery for HD

A study on HD due to disc rupture or cartilaginous tumor was first reported in 1934. The study concluded that ruptures of the disc were more common than tumors. The authors recommended that the primary mode of treatment should be surgery. In 1939, Love described a surgical technique that involved the identification of an HD mass through myelography and subsequent removal of the HD via partial laminectomy. A study on the long-term outcomes of discectomy, including a follow-up period of more than 10 years, showed relatively favorable results with an average improvement rate of 73.5 % [27]. A prospective randomized multicenter study was performed on 1,244 cases at 13 medical institutes in the United States to compare whether surgical or conservative treatment showed favorable outcomes [28]. When the Short Form 36 and Oswestry Disability Index (ODI) assessment was performed at 3 months, 1 year, and 2 years after intervention, surgical treatment was found to be more effective. However, the findings of that study should be cautiously interpreted, as the crossover rates between the surgery and conservative group were 40 and 45 %, respectively.

The first case of the use of the microendoscopic discectomy system in Japan involved a patient with lumbar HD in 1998. Since then, the use of minimally invasive spinal endoscopic surgery has increased rapidly, according to an annual report by the Japanese Orthopaedic Association. A prospective study was performed on 120 consecutive patients with lumbar HD treated by the lumbar MED system, with a 5-year follow-up period; the authors indicated that this less invasive and efficient approach using a small skin incision ensured minimal tissue damage and a shorter length of hospital stay, as well as excellent clinical outcomes (ODI and lumbar visual analogue scale scoring), as compared to conventional discectomy [29]. Cases of intraoperative dural tears, surgery at the wrong disc level, and contralateral symptoms owing to the lack of complete decompression of the nerve root were mainly reported for operations performed by inexperienced surgeons, although the frequency decreased with increasing experience of the operator [30]. However, cases of postoperative epidural hematomas, nerve root anomalies, and mechanical failures of surgical instruments were noted even in operations performed by experienced surgeons [30].

Mechanism of HD resorption

The classical definition of HD is a protrusion or extrusion of a degenerated disc, which results in the presence of a low amount of disc matrix components such as proteoglycan and type II collagen in the spinal canal or neural foramen, where there is abundant vascular supply. Surgically removed tissues indicated tissue granulation, along with the marked infiltration of macrophages and some lymphocytes, concomitant with neovascularization [31]. We developed a co-culture model of disc chondrocytes and peritoneum macrophages derived from MMP-3- or MMP-7-deficient mice to reproduce the acute phase of HD and investigate the role in MMPs in HD resorption. The MMP-3 derived from disc chondrocytes plays a crucial role in the generation of a macrophage chemoattractant, which results in the subsequent infiltration of active macrophages into disc tissues [32]. In addition, the MMP-7 released by macrophages is essential for the release of soluble TNF-α from these macrophages [33]. Thus, extensive communication between disc chondrocytes and extrinsic infiltrated macrophages is important for HD resorption.

Vascular endothelial growth factor (VEGF), a potent angiogenic factor, was found to be strongly expressed in surgical samples of lumbar HD. We previously reported that the upregulation of VEGF under co-culture conditions strongly induced neovascularization [34]. TNF-α induces the release of VEGF from disc cells through NF-kB signaling, which results in CD31 expression on endothelial cells and formation of an anastomosing network [35]. Interestingly, the degree of angiogenic activity was found to be closely related to aging.

Monocyte chemotactic protein (MCP)-1 is a CC chemokine that plays an important role in the activation and recruitment of macrophages. MCP-1 was found to be expressed in both infiltrated macrophages and disc cells from surgical HD samples [36]. Moreover, TNF-α acts as the initiator of inflammation, following contact between macrophages and disc chondrocytes [37]. TNF-α induces the release of thymic stromal lymphopoietin (TSLP) from disc cells through the NF-kB pathway [38]. The disc cells then express the TSLP receptor and produce MCP-1 through the phosphatidyl-inositol 3-kinase/Akt pathway. Interestingly, MCP-1 expression in murine intervertebral discs showed age-related decreases, whereas its response to inflammation showed age-related changes.

Tumor necrosis factor-like weak inducer of apoptosis (TWEAK) is a member of the TNF-α superfamily of cytokines. The TWEAK and its receptor, Fn14, were expressed in disc tissues [39]. The TWEAK-induced disc cells to generate MMP-3 via a c-Jun N-terminal kinase, resulting in disc matrix degradation. TWEAK also stimulated MCP-1 through the NF-kB pathway (Fig. 2) [40, 41].

Fig. 2
figure 2

Schematic model demonstrating the mechanism of herniated disc resorption. Activated macrophages generate tumor necrosis factor (TNF)-α. Matrix metalloproteinase (MMP)-7 derived from macrophages releases the soluble TNF-α from these macrophages, which induces disc cells to generate monocyte chemotactic protein (MCP)-1 and MMP-3. Thymic stromal lymphopoietin derived from disc cells promotes MCP-1 production from disc tissues. MMP-3 derived from disc cells releases chemoattractant and degrades the disc matrix. Vascular endothelial growth factor induces neovascularization

Full size image

Development of a less invasive treatment for HD

Lumbar HD commonly occurs in patients aged 20–40 years with active lifestyles. These patients show acute onset of low back and/or lower extremity pain, which results in absenteeism from daily work. Thus, a less invasive treatment compared to microdiscectomy with microscopy or endoscopy is required.

Chemonucleolysis is a treatment that involves the administration of enzymes into the HD, and has been proposed as an alternative and less invasive approach to avoid surgery. Chemonucleolysis causes the degradation of aggrecan and/or collagens, which results in the decrease of the HD matrix and water content, as well as a reduction in the pressure exerted by the HD on nerve tissues. Smith previously reported on the administration of chymopapain, derived from Carica papaya, into lumbar HD. Although its therapeutic effects have been well documented, chymopapain contains a wide range of substrates, such as aggrecan and collagens, which can lead to certain adverse side effects, including anaphylactic shock, subarachnoid hemorrhage, transverse myelitis, and discitis. Purified bacterial collagenase, derived from Clostridium histolyticum, has also been used for chemonucleolysis. A prospective randomized study using chymopapain and collagenase showed good and excellent outcomes 5 years after treatment in 72 % of the chymopapain group and in 52 % of the collagenase group [42]. In addition, a prospective randomized trial of 100 consecutive lumbar HD patients was performed to determine whether chemonucleolysis with chymopapain or standard discectomy showed better or improved outcomes. It was found that there were no differences between the two treatments at 1, 10–13, and 24–27 years after the treatment [43]. Thus, chymopapain—which had been recognized as a favorable chemonucleolysis reagent—was subsequently withdrawn as a treatment option owing to the associated complications.

We demonstrated that MMP-7, which is strongly expressed in human HD material, plays a crucial role in the HD resorption process [32]. Therefore, we developed recombinant human (rh) MMP-7, which may be an ideal candidate as a chemonucleolysis drug. rhMMP-7 degraded human surgical samples of HD in a concentration-dependent manner [44]. Moreover, this effect is not significantly correlated with patient age, HD degeneration grade, and interval between the onset of symptoms and surgery. The aggrecan cleavage rates of rhMMP-7 exhibited a 1,000-fold increase when compared to that of type 1 or type 2 collagens [44]. The intradiscal administration of rhMMP-7 was found to decrease the proteoglycan and water content in a canine vivo model [44]. Moreover, epidural injections of rhMMP-7 did not show any adverse effects at both the injection site and nerve tissues [44]. Thus, rhMMP-7 may be a novel chemonucleolysis agent (Fig. 3). We are currently performing clinical trials using rhMMP-7 on lumbar HD patients in the United States and are carefully monitoring the patients’ conditions.

Fig. 3
figure 3

Hypothetical mechanism of chemonucleolysis with rh MMP-7

Full size image