During the preoperative planning of a UBE, it is essential to thoroughly examine the architecture of the spinous process (SP), laminae, and facets. Proper reading of the axial sections of the magnetic resonance imaging (MRI) images can help us plan the surgical trajectory and location of the portals for UBE.

Let’s start with the SP. SP is the posterior continuation of bilateral laminae which rises posteriorly and caudally marking the palpable midline of the vertebral column. Unlike the common belief, the SP base sometimes extends anteriorly beyond the ventral margins of the laminae and ligamentum flavum (LF) (Figure 1). Knowing this becomes specially important if you are considering visualization of the contralateral exiting nerve root. Careful identification and burring of the SP base is essential to establish sublaminar approach to exiting nerve root on the opposite side (Figure 2).

If the SP has an enlarged or hypertrophic base, the primary blunt tip obturator may reach the spino-laminar junction but the serial dilators (Figure 3) may be unable to reach the base of the spino-laminar junction. Only with lateral x-ray imaging can we discern that the dilators are superficial. This could pose challenges when introducing instruments from the lower portal, since the soft tissues may not be well separated and visibility could be impaired. In such instances, we recommend utilizing a T-Handle (Figure 4) to remove the soft tissue from the base to create the working space. Another point to remember here is that a bulkier SP base will need more bony work to get across the midline for contralateral decompression.

Orientation of the SP: many times, it will be seen that the SP will be tilted towards right or left side (Figure 5). Such tilt of the SP will shadow the spino-laminar junction and effectively reduce the spino-laminar to lamino-facetal junction. This is an important finding while planning the location of the working portal, which normally lies approximately 1 cm away from the midline. But in such tilted SP cases if the tilt is towards the same side as that of the surgical approach, then the working portal needs to be shifted further away from the midline (1.5–2 cm), so that the tilted SP does not interfere with the smooth passage of the instruments from the working portal to the spino-lamiar junction.

A previous anatomical investigation has demonstrated that the length of the SP, measured from the midpoint of the superior end plate to the posterior tip of the SP, is greatest in the L3 vertebra. It appears to be similar but slightly shorter in the L1, L2, L4, and L5 vertebrae [8].

The lamina has undergone thorough previous examination, and we are all knowledgeable with its anatomy. The lamina is connected to the SP on one side and the medial facet on the other side. But the spino-laminar junction lies more cranial in coronal and dorsal in sagittal axis as compared to the lamino-facetal junction. (Figure 6A and B). This is important to remember while creating the working space over the caudal edge of the cranial lamina of the target level, as we are moving the obturator with dilators over the laminar edge we need to move laterally, ventrally and caudally so that the edge of the cranial lamina can be traced correctly from spino-laminar to lamino-facetal junction. Wrong movement can lead to slipping of the dilators over the pars, which can lead to excessive intraoperative bleeding or going over the facetal caps which again can lead to bleeding and may result in iatrogenic instability of the joint. Whereas when we are exposing the cranial edge of the caudal lamina we need to move laterally and ventrally till the tip of inferior facet, which overhangs the base of superior facet.

The lamina is broadest at the L5–S1 level and typically narrows as we travel towards the head. This is because the lamina is connected to the facet joint on the side, which is more sagittal orientated in higher levels and more coronal at the L5–S1 level [9]. The relationship of interlaminar-window to that of the disc plays an important role in determining how much of laminar removal is needed to reach the disc level. The lower boundary of the L4 and L5 lamina corresponds to the level of the disc, and less bone removal is needed to reach the disc space in lower levels compared to the L2 and L3 lamina, which require more removal of the lamina. The point of convergence between the SP and lamina is the primary anatomical reference point in all forms of minimally invasive operations, corresponding to the location where the instruments are targeted.

The lamino-facetal junction becomes visible when the color change caused by the capsular fibers of the facet joint becomes apparent. This juncture is significant due to the need for meticulous planning in nonfusion surgeries to determine the maximum amount of bone that may be removed without compromising stability. During microscopic and open conventional surgery, the pars interarticularis is always visible when performing a laminectomy. However, with endoscopy, the pars or the lateral extent of decompression may be overestimated due to the more concentrated field of vision. In cases of significant facet hypertrophy without lysthesis or instability, the occurrence of this is considerably more probable particularly when an isolated decompression procedure is intended. In many cases, the facetal hypertrophy extends over and conceals the lamino-facetal junction. The orientation of the facet joints of the lumbar spine is primarily sagittal, and this angle reduces more as we progress towards the lower lumbar region [10]. This provides the joint with increased flexibility in bending and straightening, but limited ability to rotate. Multiple studies have demonstrated that the facet joints are useful for locating the disc space, as there is often a distance of 1–2 mm between the disc space and the highest part of the superior articular process [11].

The LF consists of 2 distinct layers, namely the superficial layer and the deep layer. The layers span from the cranial to the caudal laminae, occupying the interlaminar area. From a surgical standpoint, LF acts as a protective barrier between the dura and the high-speed burr used to remove the laminae. The midline cleft in LF (Figure 7), which is the neck of the butterfly indicates the midline and also determines the cranial extent of the bony cutting for decompression. Beyond this the LF extends horizontally in a pattern and adhere to the bottom of the superior articular process laterally, and merge with the contralateral flavum in midline and fuse with the Interspinous ligament posteriorly (Figure 8). There is also a postero-lateral extension of the superficial flavum that extends caudally under the tip of inferior facet and above the caudal lamina and superior facet junction. Knowing this helps in isolating the superficial layer of flavum, one can simply move an angled curette over the cranial edge of the caudal lamina and pass it below the tip of inferior facet and medial between lava and Interspinous ligament to loosen the flavum and peel off the superficial flavum. Along the medial aspect go the superior facet the LF dips down and gets attached to facet, this marks the roof of the lateral recess and is essential to remove to achieve decompression. The classical finding here is that the LF turns whitish in color along its bony attachments including the medial border of the superior facet, which is clearly seen under endoscopic vision and helps operating surgeon to identify the lateral extent of the decompression (Figure 9A and B). Foraminal ligament, which can be seen on the sagittal sections on MRI at the foraminal levels (Figure 10), is nothing but superior extension of LF from tip of superior facet to base of the cranial pedicle. This foraminal ligament forms the roof of the exiting nerve root as the nerve leaves the epidural space and goes laterally under the cranial pedicle. The foraminal ligament can be seen as the offending pathology in cases of lysthesis where the foraminal ligament crumples on itself due to the forward displacement of the superior facet tip and impinges on the exiting root. Kindly note that in cases of stable lysthesis where the surgeon is planning to do only decompression without stabilization, cutting of the foraminal ligament can help relieve the pressure over the exiting nerve root. The LF extends beyond the facet laterally as intertransverse membrane, which is one of the anatomical landmarks while doing a para-UBE surgery (Figure 11).

LF along the caudal lamina is attached to the posterior part of the upper surface of the caudal laminae. Naffzinger et al. [12] observed variations in the degree of cranial laminae covering by the flavum at the L1–2 level, which is 50%, whereas at the L3–5 levels, it increases to 70%.

The use of 4K+ resolutions in new cameras and endoscopy technologies has significantly enhanced the clarity of visualizing the dura, nerve roots and peridural structures. As soon as the midline split in LF is exposed, the irrigating fluid enters the epidural cavity and starts compressing the thecal sac, this potentially increases the space between the under surface of LF and the thecal sac, which is useful for us for moving the operating instruments with safety. Under the LF there is a layer of fat followed by a layer of thin peridural membrane that covers the dura. This peridural membrane contains majority of the epidural vessels. Careful isolation of this peridural membrane helps in restricting epidural bleeding. The authors recommend using a hooked radio frequency (RF) probe for engaging, lifting and dissecting the layers of fat and peridural membrane (Figure 12) for clear isolation of the dura. Once the peridural membrane is opened, the irrigation fluid again enters between the membrane and the dura resulting in hydrodissection of the membrane off the dura.

When operating near the dura and nerve roots, it is important to maintain a lower RF setting to avoid causing thermal damage to the spinal cord. The potential repercussions of such injuries are currently under investigation.

The dura, under any type of irrigation medium surgery, remains stout in the midline and gets pressed on the sides because of the pressure of the irrigating fluid. This gives the dura a typical of bilateral scalloped shape (Figure 13), which the endoscope spine surgeon needs to get used to otherwise one may get confused and may consider the midline as a dural tear or injury. The reason for this shape is that the dura in the middle line is attached to the under surface of the cranial and the caudal lamina, whereas it is free on either side, which gets pressed down by the pressure of the irrigating fluid, giving it that typical shape.

Solaroglu et al. [13] has spoken about the presence of 1–2 ligaments in the dorsal aspect of the dura. The authors have observed multiple suspensory ligaments but are inconsistent in their presence. The very first prominent suspensory ligament is located in the midline (Figure 14) that connects the dura with the under surface of the LF. These midline suspensory ligaments could be multiple in number, and a careful visualization of the under surface of LF is necessary to make sure that the dura does not get pulled while removing the LF. The authors therefore, recommend a blunt probe to carefully visualize the under surface of LF and removal of the LF in piecemeal manner, if such suspensory ligaments are present. A 30° scope become specially useful in such scenarios where by rotating the lens of the scope helps the operating surgeon to visualize the under surface of the LF. When present these suspensory ligaments can be cut by using the hooked probe of the RF cautery or by using micro scissors. The second prominent suspensory ligament is seen attaching from the medial aspect of traversing nerve root on each side (Figure 15), travelling below under the LF and attaching to the caudal lamina. This is a very flimsy structure and does not need any special dissection, but in cases of a tight lateral recess, this ligament can help the operating surgeon determine the lateral extent of decompression. Shi et el. [14] Has also mentioned in an endoscopic cadaveric study that these ligaments get thicker and stronger as we move more caudally to the sacrum. Despite our efforts to study the architecture of the nerve root in a specific manner, numerous deviations in the path of the anterior and posterior nerve roots within the intervertebral foramen have been documented [15,16]. Once the transverse root has been identified, the axilla and shoulder can be discovered by locating the axillary fat pad, which typically shows noticeable blood vessel development during dissection. All patients had high vascularity in the axillary region as a result of the development of the axillary venous plexus (Figure 16) [17]. The authors were unable to find any documented published material about this axillary venous plexus and would like to propose the name Heo’s plexus of veins (after our teacher Dr Heo Dong Wha, Harrison Spinartus Hospital Chungdam, Seoul, South Korea) to describe the same. The Heo’s plexus plays a very crucial role in identification of the traversing root specially in cases of down-migrated disc which are located in the axillary region of the traversing root. These discs push the traversing nerve root laterally making the axilla more wider and under magnified endoscopic vision can be mistaken as shoulder of the nerve root; in such scenario Heo’s plexus along with axillary fat pad can help with the exact location of nerve and prevent potential damage to the nerve. Extreme precaution should be executed to extend the lateral extent of decompression and careful de-roofing of the nerve should be done to identify the shoulder in such cases before the RF cautery can be safely used for hemostats in this region.

As we explore the foramen from medial to lateral, structure seen are in the order fat, vessels and then the nerve. Hence, one needs to be careful of overzealous use of instruments at this point.