Definition of HZLDH
It has been estimated that roughly 10–20% of all disc herniations migrate in a craniolateral direction and may hence be located in the preforaminal and foraminal regions of the “hidden zone” . Some authors specifically describe the hidden zone as an area bordered laterally by the pedicle, ventrally by the dorsal part of the vertebral body and covered dorsally by the pars interarticularis of the hemilamina [5, 6]. However, the description of this type of disc herniation varies from author to author. Soldner et al.  use “canalicular” instead of “foraminal” and termed the HZLDH as “canalicular and cranio-posterolateral” lumbar disc herniation. Schulz et al.  termed sequestration in the hidden zone as “craniolateral lumbar disc herniation”. Papavero et al.  described this pathology as “fragment extruded cephalad into the spinal or root canal impinged the exiting root”. Despite the wide variety of descriptions, all authors mentioned above recognize the difficult surgical exposure in this clinical scenario. (Fig. 4)
Current surgical strategies
The standard surgical procedure for HZLDH is the interlaminar approach, during which removal of a major portion of the pars interarticularis is necessary [7, 9]. Donaldson et al.  reported that the resection of more than 50% of the facet joint is required in approximately two thirds of cases using the microsurgical interlaminar approach, which is likely to cause postoperative instability. For this reason, the interlaminar approach has gradually lost its popularity . Since Di Lorenzo introduced the microscopic translaminar approach, the concept of preserving the bony borders of the lamina and sparing facet joints has been widely accepted in the management of HZLDH and has since been further described by several authors [11–14]. Several further novel approaches have been described recently. Wang et al.  described an approach called endoscopic transpedicle fenestration, with a bony hole drilled on the pedicle, sparing the lamina and facet joints. Reinshagen et al.  introduced the microscopic translaminar crossover approach. An angled fenestration is created in the contralateral hemilamina, whereby the medial portion of the hemilamina, just at the base of the spinous, is targeted to the hidden zone on the symptomatic (ipsilateral) side. This approach offers advantages when treating recurrent patients who previously underwent extended laminotomy. Here, we introduced the MELS for the first time to provide a novel option for the HZLDH.
Considerations of surgical anatomy
Papavero et al.  compared a disc fragment that is extruded cephalad underneath the lamina to a fish underneath the surface of a frozen lake, with there being two methods to hook the “fish”. The first is to cross the surface with an icebreaker and to catch the fish. The second option is to cut a small hole in the ice surface targeting the fish and to cast the rod. The translaminar approach remains popular for the treatment of HZLDH, however some authors have argued that this technique has its limitations . Daghighi et al.  reported that disc fragments cranially migrated into the hidden zone are more commonly seen in higher lumbar levels. Due to segment-dependent changes of vertebral anatomy, the translaminar approach must be located very laterally in the upper lumbar levels in order to reach the medial hidden zone [11, 15, 20] (Fig. 5). Disruption of the lateral hemilamina and pars interarticularis have furthermore been correlated to an increased risk of stress fracture and instability [6, 12].
In our technique, the sequestered nuclear pulposus impinging the exiting nerve root can be compared to a melon on a vine, thus the “MELS” can be described using the Chinese saying “follow the vine to get the melon” (tracking along the stem). Along the lateral border of the lamina, the vine can be easily found, thus the melon near the vine can be explored and dragged out with a hook. When managing the higher level with relative slender lamina, an undercut of the lateral hemilamina is enough to find the nerve root and the fragment. When managing fragments at the hidden zone of the L5 level, as the above-mentioned segment-related anatomy features, the wide and short lamina of L5 lead to a relatively small operating space, thus small parts of antero-superior S1 articular process were sometimes needed to explore the deep located L5 nerve root. Ivanov et al.  reported that the lateral half of the pars has the largest thickness and removing one fourth of the lateral aspect of the isthmus has minimal influence on the stresses in the remaining neural arches. In the MELS, we removed only 2–3 mm lateral margin of lamina and the isthmus was maintained almost intact. No radiological instability was observed at the final follow-up in all patients and no patient suffered severe postoperative low back pain-induced disability.
Fragmentectomy without discectomy
There is still a dispute regarding whether to deal with the intervertebral space or not in the treatment of HZLDH. Faulhauer et al.  proposed that fragment excision is superior to conventional disc removal due to a lower rate of postoperative spinal instability complications, while Kotil et al.  reported an increased recurrence rate for fragmentectomy compared with discectomy. Ebeling et al.  reported that cranio-lateral disc herniations commonly appear as a complete sequestration, thus management of intervertebral space is seldom needed. Moreover, Barth et al.  reported that discectomy did not result in additional benefits in the treatment of lumbar disc herniation, and the sequestrectomy demonstrated significantly less postoperative disc degeneration than standard microdiscectomy. In view of these considerations, we chose not to manage the intervertebral space in most of our patients. However, we performed disc removal in two cases because we found that the migrated fragments connected to the inferior disc closely, which indicated a tendency that more fragments would come out along the path. The recurrence rate of 4.8% (1/21) within the 2 years follow-up was relatively low compared with the study by Papavero et al. (7%) and was comparable with the study by Soldner et al. (3.3%) [11, 15, respectively]. However, larger patient groups and longer follow-up periods are needed to clarify the necessity of discectomy after fragmentectomy.
There are several limitations in this study. Firstly, our sample size was small, involving only 21 patients. Although the main purpose of this study was to introduce this novel approach, and a satisfactory outcome was obtained in the present study, larger sample sizes will be needed to provides stronger evidence for our conclusions. Secondly, management of L5 may be slightly more challenging, as the operation space is relative small. Finally, it is difficult to access the superior intervertebral space via this extralaminar approach, though the inferior discectomy is feasible. However, if superior discectomy is required, we would recommend the use of other approaches.