4 DISCUSSION
Ultrasound-guided BCV approach has gained increasing popularity for CVCs insertion in children11-13 and TIVAPs implantation in adults.20-22 Anatomically, the BCV approach has potential advantages during central venous cannulation compared with its counterparts of IJV and SCV approach. On one hand, the BCV is formed by the confluence of IJV and SCV posterior to the sternoclavicular joint with fewer anatomical variants, whose caliber is also not affected by patient’s hemodynamic and volume status, respiratory motion, and ultrasound probe compression;17 while puncturing the BCV under ultrasound guidance, the entire needle trajectory, which is parallel to the pleura, can be visualized in real-time by the operator, decreasing the risk of pneumothorax. On the other hand, the lumen size and blood flow of BCV are the largest among those available central veins including IJV and SCV,18 and the catheter-to-vein diameter ratio of the catheter (Babyport, 4.5F = 1.49mm) and BCV (about 3.2mm-diameter in infants)9 is less than 45%, which has been considered as a protective factor for CRT;11 meanwhile, the flow of BCV is close to that of the superior vena cava with low risk of flow disturbance, which further decreases the risk of CRT.12Moreover, the cannulation site of the catheter is located in the supraclavicular fossa,17 which may alleviate patient discomfort at the neck,7 reduce the risk for infection,12,18 and maintain a smoother catheter curvature.21
To our knowledge, no previous literature has evaluated using BCV approach for TIVAPs placement in pediatric oncology population. Right-sided approach was adopted universally in the present study group due to the operator preference and the theoretical risk of thoracic duct injury during left-sided BCV access. However, previous literature suggested that the left BCV approach could be safe and reliable in infants,14-16,23children19,24 and adults10 for CVCs catheterization. Some authors even considered the left BCV approach to be superior to the right15,16,19,24 because the left BCV courses horizontally, the ultrasound-guided manipulation is easier,16 with a higher successful cannulation rate by first attempt.15 On the contrary, Avanzini et al.11 believed that the left BCV approach may increase the risk of CRT due to the greater surface contact between the catheter and the vessel wall, so the right BCV should be recommended as the preferred mode for access.
The technical success rate was 100% in the present study group, consistent with previous literature on BCV approach for CVCs catheterization in children (94%-100%)11-13,16 and TIVAPs placement in adults (96.5%-100%).10,21 In the present study, the success rate for cannulation by first attempt was 91.42% (32/35), similar to previous reports in adult studies (90%-99.30%).10,21 But it was higher than those reported for children (65.4%-73.8%),9,12,16,17,25which can be attributed to differences in age and weight among different patient populations. The majority of the children in the present study were older than 12 months, while most of the patients included in most other studies were infants and even premature infants9,16,17,25 with weigh as low as 2.5kg.9,16,17 According to the finding reported by Breschan et al.,9the younger age and lower weight are associated with higher risk for repeated attempts in gaining BCV access.9
The correlation between proceduralist experience and patient outcome for pediatric TIVAPs insertion continues to be a subject of investigation. Recently, a study from Shilati et al.8 demonstrated that, individual surgery volume and specialty training might influence the incidence of early revision or replacement with an inverse correlation. Limited by the small number of cases in the single institutional children’s hospital, it is difficult to accumulate a large amount of experience in a short time. However, our surgeons were from the integrated medical union of people’s hospital, and we had extensive experience with hundreds of ultrasound-guided BCV punctures for adult TIVAPs procedures. In our children’s population, no intraoperative complications occurred despite the small sample cohort, the average procedural time of 44.6 mins was comparable with the 41.7-47 mins reported by Bawazir et al.,26 and the median fluoroscopy time of 10 seconds was obviously shorter than that in other study.8 Likewise, Oulego-Erroz et al.17 emphasized the necessity for specialists to receive training and gain experience in adults in advance, in order to shorten the learning curve in children as soon as possible.
Within a cumulative 19,723 catheter-days in the present study, the total and postoperative complication rates were both 11.43% (4/35) due to no occurrence of intraoperative complications, which translated into 0.20 complications per 1000 catheter-days. Such finding was in accordance with prior published literature on children reporting complication rates ranging from 7.46% to 30%7,27-29 and from 0.15 to 0.90 complications per 1000 catheter-days.30 However, the incidence of total complications in adult literatures on BCV approach ranged from 3.18% to 6.00%,21,22 which was lower than that of our pediatric population. Thus, such difference in spectrum of complications between adults and children is likely attributable to differences in body size, vessel caliber and vertical growth, which warrants further investigation.
Currently, there is no consensus regarding the safety of TIVAPs placement in children with small body sizes, especially infants less than 1 year of age. Two recent retrospective cohort studies showed that patients with low weight (less than 7 kg) might be associated with an increased risk of intra- or post-operative complications.29,31 An infant with 6.3-kg-weight from the present study developed local hematoma and catheter dysfunction postoperatively. The hematoma was treated by repeated wet compress and local pressure during the initial three days after surgery, then slowly resolved after nearly two months. Chemotherapy had to be carried out simultaneously due to disease progression. Generally, suturing is not required for the skin incision at the exit of the catheter, but suture was implemented in the present infant due to blood oozing to achieve hemostasis at the end of the procedure. It is hypothesized that, the blood from the venipuncture site seeped into the tunnel and accumulated around the catheter and port pocket, leading to subsequent local subcutaneous hematoma. In a challenging pediatric group with hemophilia, early pocket site bleeding was not associated with increased episodes of infectious complications.32 Therefore, noninvasive methods are preferred for managing the local hematoma, and surgical debridement should be a matter of prudence.
Catheter dysfunction, defined as inability of blood withdrawal with or without difficulty of fluid injection, can be a sequela of fibrin sheath formation, catheter thrombosis, or catheter tip adherence to the vascular wall.2 This complication occurred in two patients in the present study group, which were likely related to intraluminal thrombotic obstruction without ability of blood aspiration and fluid injection, one of whom was the infant presented with local hematoma as described above. Patency was restored in both catheters by thrombolytic treatment using urokinase (5,000 IU/ml). Previous studies on TIVAPs and central venous catheters in children suggested catheter dysfunction being the most common complications,6,29and the occurrence or recurrence of such complication might be a warning of the increased risk of CRT.33 Thus, once the event of catheter dysfunction occur, it warrants sufficient attention by clinicians.
While some authors reported that infection is most common and serve as the leading cause of unplanned device removal,34,35 no infectious complications such as CRBSI and local infection were found in the present pediatric series. For instance, in a large prospective investigation including more than 4,000 adult patients,36 port related infection was the most frequent complication, which was associated with neutropenia after high-intensity chemotherapy; the authors hypothesized that intravenous chemotherapy should be carried out at least 6 days after TIVAPs implantation to reduce infection risk. By contrast, in a multi-institutional study of 500 children under 5-year-old, the most common complications identified were mechanical in nature instead of infectious events.30 Notably, the proportion of infectious events was certainly not low, almost being the one of the top two. Similar findings were also reported by two retrospective cohort studies comprising infants less than 1-year-old.29,31By comparison, all our patients received chemotherapy within 3 days after surgery, though infection did not occur. Larger prospective study is warranted to exclude the possibility of the present study’s underpowering as the cause of low infection rate.
The present study should be interpreted with several caveats. Due to its retrospective, single-center, and small-sample design, further prospective, multi-institutional, large-sample sized study is required to elucidate whether the new vascular access can be widely applied to this specific patient population. And then, the outcome of the left BCV access option was lacking in our pediatric series, deserving further investigation. Furthermore, this study is noncomparative in nature, comparative studies with IJV and/or SCV access are warranted to determine the safety and efficacy of BCV access.