Sheath Tip Radial Artery Disruption as a Mechanism for Forearm Hematoma: Insights from the Distal Radial Artery Approach

The Achilles’ heel of radial artery (RA) access is its high rate of radial artery occlusion (RAO), with compartment syndrome being the most feared but rare complication, typically beginning with forearm haematoma (FH) [1]. The distal radial artery (dRA) approach, where the puncture is beyond the RA bifurcation, appears to have less RAO [2]. A recent meta-analysis, which included 18 randomized controlled trials with 8,205 patients, found an RAO rate of 1.8% for dRA versus 6.6% with RA, p<0.001, thus reducing RAO by nearly four-fold [3]. On the other hand, where hematoma was tabulated in 8 studies, there was no significant difference between approaches (3.2% versus 4.2%, p=0.12). As dRA is an emerging technique, undoubtedly these trials included multiple learning curves. A more distal dRA puncture beyond the anatomical snuffbox is associated with hand hematoma [4]. Over the last five years, we have defaulted to dRA for our coronary intervention [5]. To date, we have undertaken over 1,500 dRA cases in allcomers, and it is our experience that FH is uncommon in practice. Nevertheless, we present a case of FH with the dRA approach and postulate its potential mechanism.

A 49-year-old man was admitted with an ST-elevation myocardial infarction and underwent an emergency percutaneous coronary intervention (PCI). This was undertaken via his right dRA with a single pass needling, followed by Seldinger insertion of a 10cm 6Fr Glidesheath Slender^™ (Terumo Corporation, Tokyo, Japan) and 200ug of nitroglycerine administered upfront via the sheath. He was preloaded with aspirin and ticagrelor, treated with heparin, and was given tirofiban as a bolus followed by an infusion. The left coronary artery was imaged first with a 5Fr JL3.5 diagnostic catheter that demonstrated tight lesions in the circumflex artery. The catheter was withdrawn with a “short” 150cm J-tipped 0.035” wire left in the aortic root. A 6Fr JR4 guiding catheter was next advanced through the sheath; there was initial resistance but with forward pressure, there was a sudden “give”, and the catheter was threaded with ease into the descending aorta. Right coronary artery angiogram was performed, revealing distal artery occlusion that was recanalized successfully. The sheath was removed, and a secure pressure dressing was placed over the dRA puncture site.

When the patient was transferred to the ward, his right forearm was noted to be tense. Tirofiban was discontinued, and additional pressure dressing was applied around the forearm. Interestingly, the dRA access site remained flat, and there was no extension of swelling into the hand, which would always be the case if the bleeding was from the dRA puncture site. Fortunately, compartmental syndrome did not ensue. He had staged coronary intervention to the left circumflex artery five days later. Extensive bruising around the forearm spread above the elbow, hence grade 4/5 on the EASY Hematoma Scale (Figure 1A) [6]. The right dRA was re-accessed with a 7cm 4Fr Micropuncture Pedal Access sheath (Cook Medical, Bloomington, IN, USA) [5]. A radial arteriogram was performed following 200ug of nitroglycerine given intra-arterially, and this showed a “napkin ring” lesion corresponding to the shape of the tip of the previous 6Fr sheath (Figure 1B). A guidewire was advanced down the artery, and an 11cm 4Fr Prelude IDeal sheath (Merit Medical, Utah, USA) was swapped in, followed by a 4Fr JR4 diagnostic catheter that was used to intubate the right coronary artery, revealing a patent artery. Next, a “long” 260cm J-tipped 0.035” exchange length wire was advanced into the aortic root, and the sheath was removed. A 90cm 5Fr hydrophilic coated guiding sheath (Cook Medical, Bloomington, IN, USA) was then introduced into the ascending aorta, and a short hemostatic valve (Terumo Corporation, Tokyo, Japan) was attached to the other end of the catheter forming an extra-long radial sheath. This allowed a 5Fr EBU3.5 guiding catheter to be easily advanced down the ascending aorta, and the left coronary artery was engaged. Coronary intervention to the circumflex artery was completed uneventfully. Before the guiding sheath was removed, a repeat radial arteriogram was performed, showing no contrast extravasation to suggest rebleeding (Figure 1C). Figure 1D shows the set-up of the radial sheath used in this case.

Figure 1

FH following RA access coronary intervention occurs in 3-4% of cases in randomized controlled trials [3, 7] but has been reported to be as high as 10% in a contemporaneous allcomers’ consecutive series, using the 6Fr sheath [8]. Catheter-induced RA perforation is usually recognized peri-procedurally when the catheter fails to advance with pain and swelling [7]. Conversely, a J-tipped non-hydrophilic 0.035” (0.89mm) exchange wire is barely large enough to fit within a radial perforator artery with a mean diameter of 0.86mm [9] that has an antegrade acute angle take-off, making a floppy loop retrograde advance an impossibility to access it, as opposed to one that has a hydrophilic coating which has the habit of seeking small arteries. Puncture of such a vessel leads to bleeding into the muscles, potentially causing compartmental syndrome when the pressure within it rises above 30mmHg [6]. FH with subcutaneous tracking is assumed, at least initially, to be caused by ineffectual manual pressure at the puncture site. This is logical as FH always encompasses the spot where the sheath enters the forearm. In our case, the absence of swelling at the dRA puncture site and hand hematoma point immediately to a forearm “non-puncture site” RA bleeding. With the 6Fr sheath, FH does not usually spread beyond grade II on the EASY Hematoma Scale, and higher grades of FH are generally due to wire trauma to small perforator arteries [6]. Since only an atraumatic non-hydrophilic 0.035” J-tipped wire was used, we can exclude wire puncture as an etiology.

When dRA puncture is done within the “triangular” snuff box area bounded by extensor pollicis brevis and longus, floored by scaphoid and trapezium, bleeding is unusual with localized pressure on that spot. However, the dRA in some patients is more palpable distally, beyond the extensor pollicis longus between the metacarpals, and some beginners are tempted to access this segment of the artery. Hand hematoma can occur, as there is no bony structure against which to compress the artery, in addition to retrograde bleeding via the deep palmar arch branch. This is analogous to RA punctured too proximally, where compression is ineffective before the radial head, or too distally where the compression device is liable to dislodge with wrist movements.

Our supply of soluble verapamil was curtailed at the time of the COVID-19 pandemic. As a result, we used only nitroglycerine as a radial spasmolytic. We suspect RA spasm developed during the first catheter use; when this was removed, the RA “clamped” onto the tip of the sheath. The sheath also slipped out by 2cm to a segment of RA that was more tortuous. The sheath tip dissected the RA when it was inadvertently pushed back in. This is plausible as the “napkin ring” lesion in the RA was about 1cm from the tip of the 4Fr sheath that is 7cm long, as shown in Figure 1B, whereas the offending 6Fr sheath is 10cm long. This paper-thin radial sheath released in 2013 with a thickness of 0.12mm (the same thickness as the razor blade, versus 0.20mm for a standard 6Fr sheath) has a sharp edge designed to be flushed with the introducer for friction-free and piercing entry through the skin into the vessel. When the second catheter was advanced down the sheath, the tip of this catheter, as it exited the sheath, was impeded by some prolapsed RA tissue. Pushing this catheter forward had a “scissors-like effect, “ further dissecting and perforating the artery at the sheath tip (Figure 2). Indeed, there has been three reported cases from 2015 whereby the dissected RA vessel wall tissue, including both endothelium and media with smooth muscle cells (we believe at the sheath tip), was extended, detached, carried forward, and embolized down the coronary artery by the catheter [10–12]. This is a different phenomenon from eversion endarterectomy with non-hydrophilic sheath, where when the sheath is pulled out, a tubular layer of “thin” endothelium is detached from the media due to sheath friction, compared to the “clean cut” RA tissue that includes different layers of the vascular structure [13, 14].

Figure 2

Hydrophilic 70-90cm guiding sheaths, with a supplied tapering introducer, have been used as long radial sheaths to overcome RA spasm and subclavian tortuosity, providing direct access to the ascending aorta. But care must be exercised to avoid the small high take-off radial recurrent artery from the axillary artery because entrapment and avulsion of this artery can occur [15, 16]. However, our setup is novel, and we have previously used it to cross a severe 360° radial loop to facilitate subsequent painless multiple catheter exchanges [17]. In this case, we used a 5Fr system from the dRA to avoid exacerbating the disrupted RA caused by the 6Fr sheath. In our unit, FH with dRA is very rare. Also, in our dRA case series, 20% are diagnostic coronary angiograms undertaken without heparin with a 4Fr system [16]. FH is neither seen with this approach nor RAO because dRA is patently hemostatic when compressed following the procedure. If PCI is needed in selected cases, an off-the-shelf 5Fr guide is advanced sheathlessly and seamlessly with a 125cm 4Fr MPA2 diagnostic catheter as a tapering railtrack introducer. This “virtual 3Fr” PCI practice minimizes FH and is just about slender enough to cross the small radial recurrent artery [16]. This is an elegant, simple, and low-cost workaround for dRA when an ad hoc PCI is planned, with added reinforcement in overcoming RA spasm, tortuosity, and subclavian grip. The thicker 0.035” J-tipped exchange wire better supports traversing the tortuous dRA course with this “4Fr-diagnostic-in-5Fr-guide” catheter set-up. A 6Fr guide can also be introduced with the same 4Fr diagnostic catheter, but pre-dilation of the dRA track with a 5Fr femoral sheath introducer is necessary. A 0.014” PCI guidewire using balloon-assisted tracking (BAT) is not advisable with dRA; first, it is challenging to push the guide with a balloon inflated at its tip into the dRA through the skin, and second, the guidewire is liable to kink, prolapsing the guide outside the artery [5]. Alternatively, Asahi Intecc®’s all-in-one sheathless Eucath 6.5Fr guide can be used, or a purpose-made tapered-tipped introducer, RAILWAY^™ Sheathless Access System from Cordis Corporation to facilitate 6/7Fr guide introduction for complex PCI [18].

We believe that “sheath tip radial artery disruption” or STRAD phenomenon can be precipitated by a mismatch between sheath size and RA caliber, particularly in the presence of RA spasm, atheroma, and tortuosity. The delayed FH in our case also meant that while the sheath was in place, it covered the dissection. Akin to the commonly used strategy dealing with traumatic RA perforation intraprocedurally by occluding the RA with a guiding catheter. But, when the “tamponading” sheath was removed at the end of the procedure, with full anticoagulation from heparin on board and glycoprotein IIb/IIIa infusion, the artery started to extravasate blood. Furthermore, the slender sheath used in this case, which creates a smaller puncture hole in the RA, has not reduced FH, nor has it reduced RA spasm or RAO [19], lending further support to our hypothesis. RA injury at the puncture site does not predict RAO [20], and neither does the ubiquitous minor RA endothelial injury due to stretching by the 6Fr sheath body [21], but multiple needling attempts do, as well as increased catheter exchanges [8]. Other predictors of FH include old age, low BMI, non-clopidogrel dual antiplatelet therapy, and anticoagulation [8].

The sheath tip can disrupt the RA with the purported STRAD phenomenon, causing FH. This may be a more common complication than we realize hitherto. As the RA access site bleeding cannot normally be differentiated from that occurring in the RA remote from the RA puncture wound, this is a novel bleeding cause. What we observed is generalizable to the RA approach, raising interesting new prospects for research into RAO and FH, as these are sequelae of RA vessel wall injury. The dRA method allows for easy re-access for radial arteriography in the event of FH to determine the precise bleeding etiology. Finally, manual pressure at the forearm section corresponding to the sheath tip can be a temporalizing measure.