2. Subarachnoid anatomy

subarachnoid space is defined by its relation to the arachnoid mater, which is one of the three meningeal layers. cranial and spinal meninges. The spinal subarachnoid space communicates freely with the ventricular system of the brain.

Dura mater: this is the strongest of the meningeal coverings and consists of fibroelastic connective tissue. The cranial dura has two layers: an outer endosteal layer which lines the skull, and a meningeal layer which invests the brain. The inner layer continues downwards as the spinal dura. The width of the dura varies with the spinal level; in the lumbar region it is said to be between 0.3 and 0.5 mm thick Tarlov’s cysts and may (rarely) explain the apparently impeccable spinal which then completely fails to work. It is usually not possible to inject into these structures.

Arachnoid mater: this is a fine non-vascular membrane, which is closely applied to the dura. The subdural space between these two layers is a potential capillary space, containing a small amount of lubricant serous fluid. It is widest in the cervical region and, laterally, adjacent to the nerve roots themselves.

Pia mater: this is a fine vascular membrane which invests the spinal cord. Its lateral projections form the denticulate ligament, which attaches to the dura and supports the cord. The filum terminale is the terminal extension of the pia mater which runs from the end of the spinal cord to attach to coccygeal periosteum

Subarachnoid space: this contains CSF, and the anterior and posterior roots of the 31 pairs of spinal nerves. The subarachnoid space extends laterally as far as the dorsal root ganglion.

ultrafiltrate of plasma, which is found in the spinal and cranial subarachnoid spaces and within the cerebral ventricles. It is formed by secretion and ultrafiltration from the choroid arterial plexus in the lateral third ventricles and the fourth ventricle. Its rate of production is constant at around 0.4 ml min1 (575 ml per day). The total volume in adults is between 120 and 150 ml, around 25–35 ml of which is found in the spinal subarachnoid space and most of which is distal to the cord in the area of the cauda equina

The PCO2 is higher than that of blood, and the pH of CSF is slightly below arterial pH at 7.32. concentrations are similar (but not identical) to plasma. The protein concentration is less, but levels are not uniform and demonstrate a gradient between the ventricles,

Thoracic kyphosis and lumbar lordosis: the adult spine has a number of natural curves, the high points of which (in the supine position) are the fifth cervical (C5) and the second or third lumbar (L2/3) vertebrae, and the low points of which are the fifth and sixth thoracic (T5/6) and the second sacral (S2) vertebrae. This has relevance for the spread of intrathecal hyperbaric solutions.

The spinal cord in the adult ends at the level of the intervertebral disc at L1/L2. There is some variation, and in up to 10% of subjects the cord may end as high as T12/L1 or as low as L2/L3. (In the neonate the cord ends at the lower border of L3.) It is of obvious importance to identify the vertebral level as accurately as possible. line drawn between the highest points of the iliac crests (the intercristal or Tuffier’s line) passes across either the spinous process of L4 or the L4/L5 interspace. (Some textbooks say L3/L4, but examination of any plain PA X-ray will show this in most cases to be wrong.) This is the technique that is most commonly used by anaesthetists, but various studies have compared palpation with MR imaging to demonstrate that its accuracy is limited The lowest rib (which is palpable only in very thin subjects) is at the level of T12. The first spinous process which is clearly palpable is C7, which is the vertebra prominens (although the spinous process of T1 below it is actually more prominent still). The inferior angle of the scapula in the neutral position is at the level of T7 or T8.

Drug dose: one prime determinant of spread is the mass of drug. The greater the amount of drug, the higher and more prolonged the block Level of injection: in the supine patient with a normal spine the maximum height of the lumbar lordosis is at L2/L3. Baricity of drug:

this is another important determinant. Plain solutions of local anaesthetic are approximately isobaric relative to CSF Hyperbaric (‘heavy’) solutions are made so by the addition of glucose. ‘Heavy bupivacaine contains glucose 8% Tend to pool in the thoracic kyphosis at T5/6, and produce blocks which are generally higher but which are claimed to be more predictable than those produced by isobaric solutions. Solutions which pool in the lumbosacral area may have a relatively enhanced effect because the nerves of the cauda equina have a large surface area and only a thin layer of pia mater. This appears to increase their sensitivity to local anaesthetic.

Patient position: this is linked to baricity. If the patient is in the decubitus position, the curves of the spine have no influence. Trendelenberg positioning will clearly increase the rostral spread of a hyperbaric solution Patient height: there may be reduced cephalad spread in taller subject; the relationship is not reliable enough to allow accurate prediction. Patient age: there may be increased cephalad spread with advancing age, although again the block height cannot reliably be predicted

Pregnancy: term pregnancy is said to be associated with greater block height, which is made higher still with multiple pregnancy. The mechanism may relate to the relatively smaller volume of the dural sheath because of encroachment in the epidural space by the engorged venous plexus. In clinical practice this is not reliably true.

Speed and direction of injection: forceful injection shortens the onset time but does not usually influence the final height of block. There are some data to suggest that if the side hole of a pencil point needle is directed rostrally, then block height may be increased.
Barbotage, weight of patient, gender of patient, adjuvant drugs, vasoconstrictors: none of these factors has any significant effect on block height.

Needle misplacement: pencil-point needles have side ports of varying dimensions, and there are some in which these can be 1.0 mm in length. It is theoretically possible therefore to puncture the dura and to obtain CSF but to straddle the dura (which is around 0.3 mm thick) with the side port and inject some of the solution proximally into the epidural space. Slower injection will increase the volume which escapes proximally. Incorrect injectate: this is a possibility – typically, although not that commonly, with the substitution of bupivacaine with the lidocaine used for infiltration anaesthesia. Suboptimal anaesthesia is sometimes blamed on the intrathecal drug itself, with the claim that it may be part of a faulty batch. In most instances this is unlikely: of Lumbar CSF volume Innate resistance to local anaesthetics this to some form of mutation of the sodium channel. However, although numerous mutations of the voltage-gated sodium channel have been described in other contexts and which can have profound effects on function, domain D-IV, subunit S6

General or spinal anaesthesia for hip fracture fixation: this is an issue of perennial debate, but the data tend to disappoint proponents of each technique. One large recent observational study in the UK of 30,130 patients receiving general anaesthesia and 22,999 given spinals showed no difference in mortality at either 5 or 30 days It may be, however, that more sophisticated subgroup analysis is needed, because a large American series reported improved survival in patients undergoing fixation of intertrochanteric fractures under neuraxial block. In respect of femoral neck fractures there was no difference So the standard mantra remains: there is no overwhelming evidence to support one or other of the techniques, but it is more important to ensure cardiovascular stability throughout. Intraoperative hypotension is associated with worse outcomes.