The Lumbar Spine and Back Pain
In 2008 the United States Agency for Healthcare Research and Quality (AHRQ) recorded 3.4 million hospital visits for back pain and related symptoms – an average of 9400 per day. In the same year there were over 663,000 inpatient stays in hospital principally for back surgery and other related treatments. Also recorded in their statistics is the finding that adults aged 18 – 44 were the most likely to require emergency department care, whilst seniors aged 65 to 84 had the highest rates of hospitalisation. The overall costs for the inpatient treatments was over US$9.5 billion. Just to give you a taste of how prevalent spinal conditions are – these figures are from the USA but on a per capita basis the same holds true for all modern countries.
Before an intelligent discussion on the origins of back pain can begin it is essential that a basic understanding of the anatomy of the spine and the mechanics of the spine exists. Ensure you read The Spine - Basic Anatomy and Mechanics first, so that the following will make real sense.
The aetiology (cause) of back pain is not complex. A great myth has arisen about the causes of back pain mostly driven by a historical lack of ability to treat it. It is easier for the medical profession to create a myth about the causes of back pain than to admit defeat. In many ways medical profession itself has made the situation vastly more complex than it really is to the detriment of patients and physicians alike. The best way for the layperson to understand the causes of back pain are to go back to very basic principles, break down some myths and start from the beginning.
Back pain and Sciatica are two very different entities – even though the cause is often the same structure. Before an intelligent discussion about back pain can ensue some lessons in terminology are appropriate. Bad words and phrases which cause confusion in this speciality are 'bulging disc', 'disc rupture', 'spinal stenosis' and 'sciatica'. If you've read The Spine - Basic Anatomy and Mechanics page you will understand just how important the intervertebral discs are for the normal functioning of the spine. It is an unfortunate fact that the intervertebral discs start to dehydrate and degenerate in the late teens and early twenties and this process is irreversible and a normal part of the aging process.
Phrases like 'bulging disc' and 'disc rupture' are best ignored because they can lead to inappropriate treatment and replaced simply by 'disc degeneration' – we will talk about all the stages of disc degeneration and the consequences of it later on. 'Spinal stenosis' is simply a radiographic observation of a 'narrowing' of a hole in the spine whether that hole is the vertebral canal or the neuroforamen – it is not a diagnosis or a symptom. Whilst often quoted as a diagnosis, probably no phrase in the spinal literature has been more responsible for inappropriate surgery than 'spinal stenosis' – it is nothing more than something seen on an MRI or CT scan. 'Sciatica' is a term to describe pain in the distribution of the sciatic nerve and means nothing outside of its descriptive value. Pain coming from the nerves in the lumbar or cervical spine is best-termed radiculitis (if the nerve is not damaged) or radiculopathy (if the nerve is actually damaged). Radiculitis / radiculopathy can affect any of the lumbar nerves from L1 to S1. The sciatic nerve only starts at L5 (and maybe part of L4) and so this descriptive term doesn't apply to any nerves above this and is best avoided. The word radicular comes from the Latin word for root, the suffix -itis meaning inflammation, the suffix -opathy meaning disease.
In terms of a diagnosis – in my opinion, patients either have back pain (the causes of which we will discuss below), radicular pain (the causes and mechanisms of which are dealt with on Radiculopathy page) or both. The terminology I prefer to use reflects this simplification and I try to adhere to it consistently.
Here we are going to only talk about disc degeneration and the causes of back pain – not Radiculopathy.
Discogenic Back Pain
One of the commonest and most poorly understood mechanisms of back pain is pain originating purely from the disc itself – i.e. discogenic.
Figure 1 shows a schematic of what a normal disc looks like when sliced right through the middle.
Figure 2 shows the MRI scan counterpart for comparison.
Referring first to the schematic, you can see the pink nucleus constrained within the green annulus. NR stands for nerve roots in the cauda equina which are travelling down the spine to exit at levels below. The nerves labelled 'nerve roots' are the spinal nerves exiting at this level. Of particular importance to a discussion on discogenic back pain are the sinuvertebral nerves in the outer 1/3 of the annulus at the back of the disc space – labelled on the schematic.
Research has proven the ability of these nerves to both initiate and carry pain messages if irritated. These sinuvertebral nerves are highly sensitised to the nuclear material in the nucleosus pulposus if it bursts and leaks everywhere. When this happens, we term the situation discogenic back pain. We will see later on that a test called discography can look for and identify internal cracks and fissures in the disc from the nucleus to the sinuvertebral nerves. These sinuvertebral nerves cause intense back pain in sensitised individuals when they fire off. In 1991 researchers in Minneapolis, Minnesota performed over 700 surgeries on the lumbar spine using only local anaesthesia so that they could provoke certain areas of the spine and interrogate the patients in real-time about what pain was produced. When they provoked the posterior annulus in the region of the sinuvertebral nerve, central back pain ensued, and most patients reported that this was their usual back pain – we call this 'concordant reproduction'. Also, the researchers found that if they provoked more to the right then right sided back pain was produced and similarly to the left. Stimulating this area never reproduced the patient's radicular (nerve) pain thus confirming for the first time that the disc itself is a primary pain generator in this area and is a totally separate entity from radicular pain (nerve pain). Prior to this, and even sadly today in some centres, surgeons thought that back, as well as radicular, pain came from compression of the nerve roots leading to aggressive decompressive surgery with very little success. When the sinuvertebral nerve is irritated the pain can be felt in the middle of the lower back, sometimes radiating to the front and sides of the thighs and sometimes to the front of the lower abdomen depending on what level is affected. The discs at L4/5 and L5/S1 have a peculiar relationship with the discs at L1/2 and L2/3 in that special fibres called the gray rami communicantes (labelled Gray Ramus on the schematic) travel in a separate pathway up the side of the vertebral bodies explaining why many patients with a degenerate disc at L5/S1 experience not only back pain but pain in the groin – more often associated with radicular pain (nerve pain) from the L1 and L2 nerves.
Figure 3 below illustrates the nuclear material leaking out backwards and directly irritating the sinuvertebral nerves and gray rami communicantes which are now inflamed – coloured red.
If every single person's discs start to dehydrate and degenerate at such an early age you may be wondering why it is that this syndrome of discogenic back pain doesn't have a 100% prevalence in the community? The answer is very simple. It does have a 100% prevalence in the community, but some people are highly sensitised to this condition and some are not. It is only people who are sensitised that get significant symptoms. The normal mild aches and pains that are a natural part of getting older and which do not require treatment or come to a physician’s attention probably do represent the fact that this condition to a certain extent does affect everyone. A disc can become highly sensitised because of genetic factors and trauma – these are the two commonest reasons – both will be discussed later.
The depiction of discogenic pain in Figure 3 is actually the first stage of Internal Disc Disruption. Internal Disc Disruption will be discussed in detail and it describes the first stage in the development of a full-blown degenerate disc. It is intensely painful in sensitised individuals, very controversial in that many spine surgeons doubt the existence of this entity, and often reported as normal on MRI because no collapse is seen, and the nerves are not compressed. Internal Disc Disruption can also cause radicular pain (nerve pain) in whatever nerve root is exiting that disc level. In the same way that nuclear material can irritate the sinuvertebral nerves and gray rami, it can also irritate the exiting nerve root.
Figure 4 below illustrates the nuclear material (pink) leaking through a tear in the annullus, extending posteriorly (red arrow) to the sinuvertebral neve area and irritating the (yellow) exiting nerve root. Note that in this schematic the disc is not bulging, protruding or compressing anything at all. This is 'chemical radiculitis' of the sinuvertebral nerve and exiting nerve root from the nuclear material causing back and radicular pain.
So, it is absolutely possible to have an MRI scan reported as normal even though you have real cause for your back pain and leg pain. Despite this entity being clinically proven many clinicians and even some spine surgeons dismiss it.
Internal Disc Disruption (IDD) was first described in the 1970's. In 1995 the prevalence of IDD in back pain patients was calculated at between 30% and 50%. This same study also concluded that traditional examination findings were not able to predict the presence of IDD. A high degree of suspicion from someone who understands this phenomena and provocative discography are required to prove the diagnosis. As spine surgeons begin to understand more about the mechanisms of back pain, IDD is becoming more and more accepted. The 'North American Spine Society' and the 'International Society for the Study of Pain in its Taxonomy' both accept the existence of IDD. Recent articles published in Spine and respected texts (Bogduk, N et al "Pain Research and Clinical Management - volume 13 2002, Schwarzer, AC et al "The Prevalence and Clinical Features of IDD in Patients with Chronic Low Back Pain" Spine 1995; 20(17): 1878-88) describe IDD as "...the most common cause of chronic low back pain."
IDD can only really be diagnosed with a high degree of suspicion and provocative discography. Discography is a test performed by a skilled radiologist. The nucleus of the disc is injected with a mixture of contrast medium (so you can see it on CT scan) and antibiotics (because of the small risk of infection otherwise). The test has two components to it. Firstly, the visual component on the CT scan lets you see if there is a leak from the nucleus, and where the leak goes.
Figures 5 and 6 (below) are examples of a normal discography scan in the L3/4 disc. The bright white stuff is the contrast medium nicely contained within a nucleus which has maintained its height. There is no dye leaking into the surrounding annulus which has a nice, uniform appearance to it.
In contrast, in Figure 7 (below) see a post-discography CT scan though another patient, with a perfect example of a grade 5 full thickness annular tear. You can see the contrast medium leaking right back to the PLL and spreading out in what is often described as a 'ships anchor' appearance to the dye. All along the width of the base of the anchor is the very sensitive sinuvetebral nerve which when irritated, in sensitised individuals, causes severe back pain. This patient went on to have a total disc replacement and did extremely well.
Secondly, and probably more important, is the subjective part of the test which is a result of the injection of dye temporarily raising the pressure inside the nucleus. If on raising the pressure the symptoms of back pain are reproduced, then the disc being injected is identified as being a pain generator and responsible for the symptoms. Sometimes this part of the test is difficult to interpret, and the results are best left for analysis along with a detailed consultation and other radiology (e.g. MRI scan) but usually it is highly informative and accurate. Often provocative discography shows nuclear material leaking out not only around the exiting nerve root but into the epidural space. The epidural space is the black area on the schematics above and consists of veins, fatty tissue and ligaments that cushion the spinal cord or cauda equina.
In Figure 8 (below) you can see nuclear material in the epidural space now also irritating the lining of the spinal cord (called the dura – bright green on the schematic) and multiple nerve roots, not just the exiting one. IDD can thus cause pain in many nerve roots as well as the exiting one. Nerve conduction studies can define actual nerve damage (i.e. radiculopathy, not just radiculitis) in nerves affected this way. Nerve conduction studies need to be combined with electromyography to prove this.
Some patients are not suited to discography in which case the diagnosis can sometimes be made on an MRI scan with Gadollinium contrast injected – see figure 9 (below). This doesn't give you the subjective part of the test however and in my opinion suboptimal. Gadolinium is a contrast medium used in MRI scanning and is preferentially absorbed in areas of inflammation. On the scan on the right the red arrow is pointing to an area of increased signal where the gadolinium has concentrated in an annular tear.
IDD can cause discogenic back pain and radicular pain. As a disc continues to degenerate and progresses to a full-blown degenerate disc the discogenic components are still responsible but other painful situations arise and are discussed below.
Degenerative Disc Disease
If an IDD is left to progress the natural history is a fully degenerate disc. Degenerate Disc Disease (DDD) is the clinical syndrome of chronic pain from this entity. DDD causes back pain via 4 mechanisms. Most patients have pain as a result of a combination of some or all of these mechanisms existing simultaneously. Those mechanisms will each be discussed in detail – they are:
- Instability and Altered Load Mechanics.
- Discogenic Pain.
- Facet Arthropathy.
- Muscle Fatigue.
Instability and Altered Load Mechanics
In the normal spine the intervertebral disc has 'height' associated with it. In other words, the nucleus, by virtue of its hydrostatic pressure keeps the vertebral bodies apart a certain healthy distance. In this setting the posterior longitudinal ligaments and the anterior longitudinal ligaments (PLL and ALL) are taut and the vertebral bodies 'roll' over the nucleus giving a few degrees of flexion and extension (forward and backward bending) at each level.
Figure 10 (below) shows a normal – 'high' – disc acting as a primitive ball and socket joint. In fact, this movement is highly specialised and not primitive at all but beyond the scope of this description. When a disc loses its hydrostatic pressure and deflates the surrounding ligaments (ALL and PLL) lose their tension and the vertebral bodies come closer together. Normal, physiological, flexion and extension become wobbling and shaking everywhere – instability.
In much the same way as a flat tyre on a car wobbles and shakes all over the tarmac as a result of the loss of pressure in the tyre the vertebrae do the same as a result of loss of pressure in the disc. This instability is very hard to see on an MRI scan as when you're having the scan your spine is in a very neutral position with you lying on your back. Dynamic MRI scans may be able to pick this up by imaging your spine in a more 'loaded' position which mimics what happens when you stand up or sit upright but currently there is only one scanner such as this in Australia and no Medicare rebate exists for it.
Instability is a serious situation – one of the functions of the vertebral column is to protect the spinal cord and nerves and if this function is compromised by instability patients experience tremendous muscle spasm to protect the spinal cord as a secondary protector. This is painful. When the paraspinal muscles go into spasm like this often the patient is bedridden until they settle down. This represents the very acute situation often requiring treatment in the emergency department of the local hospital. All of the instability is experienced just behind the disc which is also where the nerve roots exit the spine, so the unstable movements also give the delicate nerve roots a hit every time the instability is experienced, producing nerve pain – radiculitis, or even radiculopathy. Most MRI scan reports are read by radiologists, not spine surgeons. An experienced spine surgeon can infer various facts from an MRI scan to determine whether or not instability exists.
The most common place for instability to catch a nerve root is the lateral recess of the vertebral canal – described on the Radiculopathy page – often reported a being normal. Whilst the commonest cause of this instability is the degenerate motion segment with collapse of the disc or subluxation of the facet joints, unfortunately the next commonest condition causing instability we treat nowadays is iatrogenic instability – in other words, surgeon created instability. In 1934 Mixter and Barr, two respected surgeons, coined the phrase 'disc herniation' and attributed all pain of spinal origin to the herniated disc. This led to an unprecedented amount of decompressive surgery to take the pressure away from the herniation. Unfortunately, decompressive surgery means taking away the posterior elements of the spine (lamina and facet joints) which are the stabilising elements. In the presence of an already unstable situation because of the degenerate disc this often leads to more problems than it fixes.
Figure 11 (below) depict a normal disc under load (i.e. standing up or sitting upright) – near right – while Figure 12 depicts a degenerate disc in the same position. In a normal motion segment, as a nucleus is loaded is absorbs the load and bulges into the annulus and into the endplates. The annulus and endplates, being tough structures resist this outward pressure and an equilibrium ensues whereby the pressure from the vertebral body above is equally distributed through the disc to the vertebral body below. The load is uniformly distributed. When a disc degenerates and loses the pressure in the nucleus (far right) the ability of the disc to evenly distribute the load is lost and pressure points ensue. The nucleus, instead of bulging into the surrounding structures has negative pressure and annular fibres are sucked into it. The endplates weaken and nuclear material bursts through into the bones above and below (called Schmorl's Nodes) – the pink arrow seen in Figure 12. The uneven distribution of load causes people to fidget and become restless in one position for much over 20 minutes and they are forced to move so as to shift the pressure points around. This is why patients with chronic lower back pain whose primary mechanism of pain is altered load transfer cannot get comfortable and fidget around or have to keep on the move.
Discogenic pain has been discussed above. In exactly the same way as patients with IDD get pain as a result of irritation of the sinuvertebral nerve and associated structures, the fully degenerate disc can also. Discography usually reveals complete nuclear disruption, not just discreet tears with contrast medium going everywhere. Sometimes the subjective part of the test is falsely negative as a result of no pressure being able to be exerted in the disrupted nucleus. Often these patients get a delayed strongly positive response from the discogram with dural irritation and headaches which can last a few days.
The facet joints are the two joints behind each motion segment. Rotation occurs at these joints in the lumbar spine and flexion in the cervical spine. That is why in the lumbar spine the facet joints are more sagitally oriented to allow for movement in this axis and in the cervical spine they are more coronally oriented to allow for flexion. Again, thinking like above, of the concept of the normal – 'high' – disc the facet joints are kept in a neutral position when a disc has height associated with it. Collapse the disc and the facet joints sublux – subluxation is a term that simply means slipped or non-congruent. In a subluxed position joints are painful and become arthritic. Very occasionally patients will present with primary facet joint osteo-arthritis. The discs can be normal-ish but with facet joint osteo-arthritis the discs always suffer and degenerate in a predictable manner.
Figure 13 is a CT slice demonstrating a grossly abnormal set of facet joints – defined as abnormal by the ratty appearance to the actual joint, and overgrowth of bone around the joint (osteophytes – the body's attempt to stabilise an abnormal joint by laying down more bone). Figure 14 is a CT slice that is much more normal looking. The gap between the joint is filled with cartilage which is not visible on a CT scan hence appearing as a gap. When the gap is gone you can assume the cartilage is also. When the cartilage is gone you have bone rubbing on bone and this is painful. Often the diagnosis of pain secondary to the facet joints is able to be diagnosed with a few simple questions. Sometimes the facet joints are blocked with local anaesthetic to work out which ones are the problem. The assumption being that if you block a joint and the pain disappears then the joint must be responsible.
If the diagnosis is still not clear, then a radionuclide bone scan can often discern a problematic facet joint, but this is a test involving a lot of radiation and time and not often used.
It has been mentioned in the page The Spine - Basic Anatomy and Mechanics how important a balanced spine is. In the sagittal plane the base of your neck should be through the middle or just in front of your sacrum. In the coronal plane your neck should balance over the middle of your pelvis. Patients who have lost their lumbar lordosis especially are tipped forward and have a positive sagittal balance.
Figure 15 illustrates how a spine gradually becomes further and further tipped forward (positive sagittal balance) as lumbar lordosis is lost, i.e. as the lumbar spine straightens out. The commonest reason for loss of the lumbar lordosis is disc degeneration. Some people are congenitally shaped with very little lumbar lordosis, but they still have a balanced spine because they also have very little thoracic kyphosis which balances the whole curve. People with this type of spine have what is known as a Type 1 spine and generally represent difficult patients to treat if they are symptomatic with DDD because it is not difficult to unbalance them.
A patient in significant positive sagittal balance has to spend a lot of energy to simply stay upright without falling over. Studies have been performed which demonstrate how fatiguing it is to live in positive spinal balance in terms of energy consumption and eventually the back muscles simply fail under the strain and patients can only walk a few hundred yards before fatiguing. Imbalance in the coronal plane is even worse in this regard and is detailed on the Scoliosis and Kyphosis page. The muscles down the sides of the spine (called the para-spinal muscles) act as a very strong tension band, working against large lever-arms to maintain an upright posture but they have a limit. With advancing age, as these muscles fail, the body compensates by tilting the pelvis to try and get the centre of gravity over the sacrum, flexing the knees and extending the hips. This gives rise to the classic depiction of the pensioner on road-signs:
About 60-70% of the lordosis in the lumbar spine comes from the 2 lowest motion segments L4/5 and L5/S1. Reconstructive surgery in the lumbar spine must be able to restore lordosis. The treatments that allow surgeons to best achieve this are discussed in the Lumbar Spine Reconstruction page. Lordosis is not difficult to put into a surgical reconstruction during primary surgery but if surgery fails to restore lordosis, trying to get it back with revision surgery is a very big deal and can rarely be achieved using minimally invasive techniques.
The Cervical Spine and Neck Pain
The mechanisms by which degenerate discs in the lumbar spine cause low back pain are very similar to those by which degenerate discs in the neck cause neck pain. The cervical spine is lordotic and bears load in a similar fashion to the lumbar spine although the forces are much smaller. Almost all rotation comes from the base of the skull and the C1/C2 motion segment. All levels of the cervical spine contribute to flexion and extension but most of it comes from the higher levels. Just as in the lumbar spine the most lordotic segments are the first to degenerate, and are the most sensitised, the most lordotic segments in the cervical spine are the first to suffer. These are C4/5, C5/6 and C6/7. Internal disc disruption in the neck probably exists and is treated but is much harder to diagnose. Discography is harder to perform in the neck and the results are not as reliably interpretable. Loss of lordosis in the neck is associated with severe headaches. Many patients with severe headaches have been incorrectly diagnosed as suffering from migraines whereas often the culprit is mechanical instability in the cervical spine.
Instability in the cervical spine causes tremendous muscle tension in the paraspinal muscles behind the head which inflames, traps and irritates the greater and lesser occipital nerves. These nerves supply a large portion of the scalp in a balaclava-type distribution and so patients with cervical DDD often describe headaches in this typical area. If the headaches seem to be related to mechanical activities where the neck is used, for example poor desk posture, overhead work, lifting above shoulder height etc. then it is likely these nerves responsible. One very good test to help work out if the headaches are due to DDD in the neck is to lie down with a towel rolled up under the neck and stay like this for 20 minutes or so. This artificially gives the neck some lordosis and if the headaches improve then in conjunction with a consultation and appropriate radiology the diagnosis is often obvious.
The facet joints in the cervical spine are strongly oriented in the coronal plane (i.e. front-on) which subjects them to greater shear forces than in the lumbar spine which are subject to more rotational forces. As a result of this when discs degenerate in the cervical spine the motions segments really stiffen up. Facet joint injections can really help this situation. Patients who have been involved in a whiplash injury, if there is significant degeneration pre-existing in the neck, often have a severe injury to the disc above this stiff segment producing a very painful situation where, as a result of the injury, the disc is irreversibly damaged and the forces involved in the whiplash produce a highly mobile and unstable segment which can be diagnosed on dynamic x-rays of the neck – where the x-ray beam is focused side-on and the patient flexes the neck forward.