DISCUSSION
Optimum surgical treatment of PSS arguably requires both endovascular intervention and adequate surgical decompression of the SV surrounding extrinsic structures. This is often further augmented via post-decompression secondary endovascular intervention or vascular reconstruction to ensure long term patency of the venous wall and lumen. Endovascular treatments such as mechanical thrombectomy performed in conjunction with CDTL, when combined with decompression and reconstruction as needed, seems to be particularly promising.
The clinical data from the series included in this literature review has shown that most patients undergoing thrombolysis and decompression had a definite venous wall stenosis secondary to the longstanding extrinsic compression: 42.3% (n=11) underwent venoplasty at some point. The durability of open surgery is impressive where adequate venolysis or early venoplasty seems to maintain patency of the SV and reduce rates of recurrent stenosis. Three developed SV occlusion post surgery of which one occurred on the first post operative day (symptomatic) and the second on routine venography four months later (asymptomatic). The third developed symptoms of arm swelling and pain thirty-six months post initial TA decompression which was lysed and underwent a further venoplasty successfully. The majority of patients in the series also had underlying SV stenosis following thrombolysis and decompression – possibly due to longstanding extrinsic compression of the SV leading to transmural fibrosis and ultimately thrombosis.
It must be highlighted that all patients in the cohort underwent thrombolysis via CDTL only. As outlined previously, CDTL is associated with several clinical limitations, such prolonged treatment time and key adverse events including pulmonary and cerebral embolism and entry site bleeding.4 Our review of current literature suggests that there exist viable alternatives to CDTL for the management of PSS. Schneider et al reported an average thrombolysis time of 12 hours overall (compared to 22 hours in the included series) when the AngioJet mechanical thrombolysis system was used for thrombus debulking prior to CDTL.11 Similarly, Shah et al. reported that use of the AngioJet system yielded successful thrombolysis in 2-3 hours for three patients.14 Results from Hileman et al. also seem to suggest the viability of mechanical thrombectomy as an alternative to CDTL in PSS: 93% of patients included had more than a 50% clot reduction when treated with mechanical thrombectomy, compared to 79% of patients treated with CDTL only.15Furthermore, O’Sullivan and colleagues reported an average thrombolysis time of 91 minutes in using the Trellis thrombolysis catheter.16 This alternative to traditional CDTL, also conventionally used for treating LEDVT, was associated with 50-95% clot removal in 82% of patients, and > 95% removal in 3 patients. It was also associated with no major complications.16 The extent to which the residual SV stenosis experienced by the patients in the included series can be attributed to CDTL being performed over other thrombolysis methods is unclear. Though more research into the use of such endovascular devices would be prudent, it seems that approaches typically used to tackle lower-limb DVTs offer a promising step forward in the endovascular management of PSS.
Data from Wooster et al. also seems to suggest that endovascular intervention can help improve overall clinical outcome in patients with upper extremity DVT in general. Sixty seven percent of patients in their cohort underwent endovascular intervention, which included procedures such as PTA, patch PTA, and stenting. A 100% surgical success rate with symptomatic relief was observed in patients that underwent endovascular intervention in conjunction with decompression. Moreover, this approach was associated with low rates of SV reocclusion and symptom recurrence (9.4% and 11.3% respectively).6 The clinical outcomes reviewed from available literature are summarised in Table 2.
Other contributory factors for re-thrombosis were incomplete resection of the costoclavicular ligament (CCL) and/or tendon of the subclavius muscle. The CCL is a rigid structure medial to the SV which can be seen by the TA and PC routes and causes SV compression particularly when the insertion into clavicle is lateral (figure 2). Similarly, the anteriorly lying subclavius muscle may be a further source of external compression on the SV and is easily identified by the IC route requiring excision to expose the vein. Recent series comparing the SC to the IC route for SV decompression have shown improved patency rates with fewer postsurgery symptoms for the IC approach17. Additionally, Molina et al. emphasise the importance of surgical access to the medial aspect of SV to adequately decompress the vessel surgically.18-21 The importance of carrying out decompression conjunction with endovascular interventions is therefore clear – as emphasised by Zurkiya et al., the degree of SV stenosis immediately post-decompression is usually comparable to that of pre-decompression because anatomical resection does not deal with intrinsic lesions to the SV.7 This also suggests that prolonging the interval between thrombolysis and decompression, and indeed between decompression and adjunctive intervention, would prove only detrimental to clinical outcome: time from presentation to decompression > 14 days is a documented factor associated with worse clinical outcome, as it may lead to early re-thrombosis warranting re-intervention down the line.22
There is little current data concerning the approach surgeons use to manage PSS, both in terms of open decompression and endovascular intervention. A survey of 60 United Kingdom (UK) members of the Vascular Surgical Society showed 4% performed surgery for PSS by a PC incision with 55% opting for a TA approach and 28% for SC23. Hence, at least in the UK, there appears to be considerable variation in approaches adopted. This report dates to 2004 and clinical practice may have significantly changed since then.
Not a single randomised controlled trial has been conducted to determine the optimum surgical approach for PSS, perhaps due to its low incidence. Hence, it may not be feasible for a prospective randomised trial to be performed with sufficient power to gain a definitive answer as to even the optimum surgical approach to treat acute PSS - particularly if the TA and PC/IC outcome differences are minimal. What seems increasing clear, however, are the added benefits of undertaking endovascular intervention alongside surgical decompression, both in terms of preoperative thrombolysis and postoperative adjunctive venous repair.
A prospective randomised trial investigating the relative efficacies of the various approaches to preoperative endovascular intervention, for example mechanical thrombectomy and CDTL techniques, would therefore be prudent. Such a trial would help to elucidate whether the management of PSS, and indeed UEDVT in general, would benefit from treatment strategies traditionally employed to tackle LEDVT or even intraarterial pathologies. Indeed, the shift towards endovascular approaches has gained great popularity amongst surgeons managing complex cardiac and aortic pathologies – one needs only to look towards the advent of thoracic endovascular aortic repair as an example of the potential for endovascular procedures to successfully resolve complex cardiovascular surgical pathologies, without the need for large incisions and surgical trauma.24