In this issue of Anaesthesia, Luyet et al. report an elegant study examining the reliability of the classical ‘landmark’ approach to paravertebral blockade (PVB) . While this report will no doubt be of interest to thoracic anaesthetists, we urge all general anaesthetists not to be too hasty in turning the page, as this work has broader application.
In 2006, Exadaktylos et al. published a retrospective study analysing the medical records of 129 consecutive patients undergoing surgery for breast cancer. The findings suggested that paravertebral anaesthesia and analgesia for breast cancer surgery may reduce the risk of tumour recurrence or metastasis . While confirmation from prospective randomised controlled trials is awaited, the volume of in vitro data to support such a hypothesis is growing . Should such an approach be convincingly demonstrated to confer survival benefits, the delivery of effective regional anaesthesia and analgesia will become an established standard of care in cancer surgery. This increases the drive to develop a safe, effective and reliable technique for unilateral regional blockade.
In thoracic surgery, debate surrounding the ideal analgesic technique for thoracotomy has intensified in recent years with the suggestion that PVB has a better side-effect profile and has been associated with a reduction in complications compared with thoracic epidural analgesia (TEA) [4, 5]. The current practice of thoracic anaesthesia in the UK is represented by an approximately 2:1 split in favour of TEA over PVB , a split that has remained consistent over several years, perhaps suggesting that to many, either the reliability or the benefits of PVB over TEA are yet to be sufficiently proven to change practice.
Luyet et al. report an observational study aiming to explore the association between the location of paravertebral catheters placed using the classical ‘landmark’ technique, the distribution of contrast dye delivered through the catheter, and the extent of subsequent somatic block . In doing so, the authors encounter one of the many methodological difficulties that challenges studies of this sort: what constitutes ‘failure’ of regional anaesthetic blockade? In one recent, high-profile, meta-analysis comparing the analgesic efficacy and side effects of PVB versus TEA for thoracotomy, the included studies defined ‘failed technique’ as anything from ‘failure to catheterise the epidural space’ and ‘inadequate analgesia’ (technique failure) to unavailability of an infusion pump (system failure) . When appraising the regional anaesthetic literature, clinicians must pay careful attention to definitions of failure and decide what constitutes a clinically meaningful ‘failure’ in their practice.
It has been said that regional anaesthesia “always works provided you put the right dose, of the right drug, in the right place” . The pertinent question is: how do you deliver local anaesthetic to the ‘right place’ within the paravertebral space? It has been demonstrated that the analgesic effect of single-shot PVB lasts approximately six hours, and that as such, in order to provide prolonged postoperative analgesia via PVB, a catheter technique is required . In Luyet et al.’s study, the authors attempted to place PVB catheters by a percutaneous landmark technique in 31 patients. In one patient, ‘difficulties were encountered’ during paravertebral catheter placement, leading the investigators to switch to an ultrasound-guided approach (undoubtedly a case of technique failure). The authors define the spread of contrast dye within the paravertebral space either close to the intervertebral neural foramen, extrapleural laterally at the level of the ribs, extrapleurally at the level of the vertebral bodies or anterior to the vertebral bodies, as a successful radiological endpoint; the ‘right place’. In nine out of 30 catheters, dye was not seen in these locations. Including the failed catheterisation, the authors suggest this represents a failure rate of 32%; that is, they were unable to place the catheter such that dye was delivered to the anatomical destination they desired in 10 of the 31 cases.
Distribution of dye, however, proved to be a poor correlate for clinical effect (r = 0.05, p > 0.7). Four patients in whom the desired radiological spread was demonstrated received insufficient analgesia, yet six out of the nine patients in whom catheter position was not as desired had adequate analgesia. That is to say, that with the intention of providing adequate analgesia to all 31 patients, the authors failed in eight (26%) cases. As has been demonstrated before , the distribution of local anaesthetic block following a paravertebral injection is unpredictable; concentrating (in Luyet et al.’s study) only on the catheters that were radiologically deemed to be in the ‘right place’, the area of loss of sensation from the same standardised local anaesthetic bolus ranged from two to 12 dermatomal segments.
This leads to the question of whether a failure rate (in experienced hands) in excess of 25% is acceptable, and how could it be improved. A failure rate over 25% would clearly be unacceptable for a regional anaesthetic technique; authors describing brachial plexus blockade for anaesthesia of the arm describe satisfactory surgical anaesthesia in over 90% of patients . In contrast, following open thoracotomy, PVB is rarely relied upon to provide adequate analgesia without requiring systemic opioids ; as such it is often considered to be an opioid-sparing analgesic technique. Luyet et al. set the ambitious target of defining inadequate analgesia for patients undergoing thoracoscopic surgery as a morphine consumption of more than 20 mg in the first two hours postoperatively. Consequently, it may be reasonable to question whether the observed failure rate may partially reflect the authors’ asking too much of their technique. For cancer surgery, much of the purported benefit of PVB may be in avoiding the inhibitory effects of strong opioids on natural killer cell function [2, 3]. In such circumstances, a success rate of 75% or less would clearly be unacceptable.
The use of ultrasound-guided regional techniques has been heralded as one of the most important developments that the field of regional anaesthesia has seen ; the potential advantages for the practitioner to guide the needle under real-time observation are self-evident. Luyet et al. surmise that an ultrasound-guided technique may reduce the failure rates observed in their study. Since ultrasound-guided thoracic paravertebral blockade was first described by Hara et al. in a letter to this journal , there have been a number of promising reports of ultrasound-guided techniques both in cadavers [13–15] and in patients undergoing breast surgery [16, 17]. In 80–100% of cases, the paravertebral space can be satisfactorily imaged and though there are no direct comparisons of ultrasound versus landmark techniques reported in the literature, it would seem reasonable to suggest that ultrasound guidance of thoracic PVB offers the potential, in experienced hands, to allow consistent positioning of the needle’s tip within the paravertebral space. Unfortunately, ensuring correct needle tip positioning may only be half the battle. Several reports have demonstrated that catheters placed through needles ultrasonographically confirmed to be in the paravertebral space can be subsequently discovered to be misplaced (prevertebrally, epidurally, dorsally in the musculature of erector spinae or intrapleurally) in up to 40% of occasions [13, 14]. To this end, some of Luyet et al.’s previous work is worthy of mention. In a proof-of-concept cadaveric study they have described the use of a novel coiled catheter for PVB . In that study, none out of 60 catheters inserted were misplaced, suggesting that such technology may aid in preventing paravertebral catheters from straying too far from the site of needle insertion.
For thoracic surgery, a PVB catheter can be inserted under direct or thoracoscopic guidance. The placement of a catheter under direct vision was first popularised by Sabanthan et al. . Anecdotally, it would seem that attention to detail is important in the conduct of this technique, which has proven to be highly operator-dependent. The low failure rate of PVB quoted in the aforementioned meta-analysis  of TEA vs PVB was atributed to the common practice of intra-operative surgical placement of a PVB catheter under direct vision. In three studies where the PVB catheter was inserted surgically, there were no ‘technique failures’ although definition of failure was loose. Recent evidence would suggest that the technique is not so infallible; a randomised, double-blinded, placebo-controlled trial failed to demonstrate a morphine-sparing effect of surgically inserted PVB when compared with placebo .
What then of the safety of PVB? The classical description by Eason and Wyatt of thoracic PVB involves making needle contact with the transverse process and ‘walking off’ the cephalad surface to obtain a loss of resistance to air as the needle traverses the costotransverse ligament and enters the paravertebral space . Sonographic studies suggest that the parietal pleura lies, on average, 14 mm beyond the transverse process  and just 7 mm beyond the internal intercostal membrane (the lateral extension of the costotransverse ligament) , a margin of error thought by some to be too slim and offering ‘considerable potential’ for pneumothorax . The incidences of pleural puncture and pneumothorax after percutaneous ‘landmark’ PVB have been reported to be 1.1% and 0.5%, respectively . Though consequent pneumothoraces are said to be of minor significance, rarely requiring more than conservative treatment , it is questionable whether these figures derived from a denominator of 367 blocks performed in three expert centres can be extrapolated to general practice. While it is of course intuitive to suggest that an ultrasound-guided technique, during which the practitioner can see and thus avoid the parietal pleura, should reduce the incidence of pleural puncture, as yet there is no evidence to confirm this is the case. A number of cadaveric studies of ultrasound-guided techniques have observed catheters to be subsequently located in the pleural cavity [13–15]. Once again, it seems that whilst needle position can be controlled, catheter position is a different issue.
Proponents of PVB claim that the incidence of major neurological complications is significantly lower with PVB than with TEA [9, 25, 26]. The evidence base for such an assertion is, however, somewhat weak. The long history and sheer number of procedures performed, in combination with initiatives such as the Third National Audit Project on major complications of central neuraxial block , mean that the complications of TEA are well known and the incidences accurately described. The same cannot be said for PVB. Often, PVB is suggested as an alternative to TEA when the latter is contraindicated by coagulopathy or local infection [9, 25, 28]. With this in mind, it is concerning to note that in Luyet et al.’s study (as has been reported previously ), on occasion, catheters inserted via a paravertebral approach are subsequently demonstrated to be located in the thoracic epidural space. In addition to spinal nerve roots (the therapeutic target of PVB), the paravertebral space also contains an extension of the dura mater as it projects through the intravertebral foramen, as well as blood vessels to and from the spinal cord; clearly, all of these structures are vulnerable to damage from the anaesthetist’s needle. Major neurological complications have been described after PVB [29–31], possibly occurring following damage to thoracic radiculo-medullary arteries as they pass through the intervertebral foramen. It should be emphasised that the risks of percutaneous PVB catheter placement in coagulopathy are unknown. We suggest that until evidence of comparative safety is available, the same considerations and strict contraindications applied to TEA should also apply to PVB.
After thoracic surgery, advocates of PVB suggest that when compared with TEA, patients may be transferred directly to ward areas without the need for level-2 (high dependency unit) care. Whilst this approach may result in financial savings and is attractive in the current financial climate, it should be recognised that early postoperative complications are not all TEA related. Hypotension, bleeding and cardiorespiratory complications related to opioids and surgery must be recognised and treated expeditiously. Robust pain protocols supported by acute pain services are required, and rescue interventions for inadequate pain relief must be available.
Use of thoracic PVB is increasing within modern anaesthetic practice and may result in improved outcomes. Further work is required to refine the technique in order to enable consistent placement of both needle and catheter in the desired location, and to provide reliable analgesia whilst minimising complications. Such refinement will require randomised controlled comparisons with carefully defined endpoints of landmark, surgically assisted (videoscopic or direct) and ultrasound-guided techniques. Whilst evidence to date suggests that PVB is a safe technique, reported case series are small; prospective audit of outcomes is required to confirm safety.