CONCLUSIONS:
Barriers to prevent the formation of PPAs are heterogeneous with respect to composition and effectiveness. The present systematic review of the literature identified a total of 13 barriers, including two barrier product combinations. These models were analyzed with respect to their safety and efficacy in the clinical setting for the cardiac surgery patient.
The safety of adhesion barriers varies considerably with respect to rates of infection, bleeding events, and mortality. No barrier had reported rates of 0.00% for all three variables; however, Seprafilm demonstrated the lowest combined rates (infection 0.00% vs control 0.00%; bleeding event 0.00% vs control 0.00%; mortality 0.57% vs control 1.22%). Cova CARD, Porcine and Polyester Gelatin Sheet, and PTFE + Seprafilm also demonstrated rates of 0.00% for at least two categories with no reported data on the third. The heterogeneity in reporting was found to be prevalent, as few barriers had reported control data. Only four of the 13 barriers had reported data for all three safety variables. PTFE and COSEAL were two of these barriers, and they also demonstrated relevant safety concerns. PTFE had the second greatest mortality rate (4.89% vs control 1.22%), third greatest infection rate (1.14% vs control 4.88%), and third greatest bleeding event rate (0.75% vs control 1.22%). Although less than the control population, PTFE may influence infection and bleeding because it requires both sutures for placement as well as re-operation for removal. Another explanation for these findings is that PTFE was assessed in the greatest number of studies with some dating back to over three decades ago; general surgical techniques have likely improved considerably since that point in time. Silcone rubber had the second greatest bleeding event rate (6.86%); however, it also was assessed in the earliest study identified (1981).13
COSEAL had the greatest infection rate (6.58%) and bleeding event rate (10.53%). COSEAL is a sprayable synthetic polymeric hydrogel that is bioresorbable. No possible reasons for the infection or bleeding event rates have been proposed; however, infection and bleeding were also among the greatest adverse events in a randomized control trial of COSEAL in vascular surgery.14 REPEL-CV also demonstrated significant safety concerns with not only the greatest mortality rate (17.05%; control 13.04%) but also the second greatest infection rate (4.55%; control 1.45%). The primary study of REPEL-CV was in infants undergoing initial sternotomy for eventual staged palliative cardiac operations with no significant difference identified in mortality rates between the barrier and control group (p=0.6405).15 The immune system function of these infants was likely limited as well and may have influenced the response to a foreign body, resulting in increased infection rates relative to the control group. Overall, the ability to draw comparisons between the safety of the adhesion barriers is limited due to the heterogeneity with respect to study design and reporting bias. As noted, many studies did not provide control groups for comparion. The varied procedures performed in differing patient populations suggest further limit the ability to draw definitive conclusions between the groups. No major studies have also identified the overall rates of infection, bleeding event, and mortality in cardiac surgery without adhesion barriers. A detalied understanding of these topics would be of great value for those attempting to weigh the risks and benefits of adhesion barrier use in cardiac surgery.
The efficacy of adhesion barriers in preventing PPA formation and limiting tenacity scores varied as well. Unlike the safety data, most barriers did have reported data for both efficacy variables. Only one barrier had an adhesion formation rate of 0.00% with a TSS of 0.00: Polyisoprene Blue Band Strips. The study regarding Polyisoprene Blue Band Strips did not identify a control group for comparison though and only had nine patients in the barrier group with a mean placement time of 31 days.16 The strips were applied to the major vessels surrounding the heart and were not used to cover the heart itself. Excluding Polyisoprene Blude Band Strips, PTFE demonstrated the lowest adhesion formation rate (37.31% vs control 100.00%) and the third-lowest TSS (26.50 vs control 80.26%). PTFE was the most frequently used barrier among all studies (67% of patients) with the widest range of years (1988-2012). PTFE has historically been effective because it is physiologically inert, has low adhesiveness with cells/tissues, separates damaged surfaces without degradation, and is biocompatible.3 PTFE is commonly used in cardiac surgery today due to its demonstrated effectiveness in reducing adhesions; however, the safety concerns discussed previously should be considered. Applying PTFE in combination with other barriers is a particular area of interest that may address these concerns, as PTFE + Seprafilm resulted in limited infections and bleeding events.17
Cova CARD was also among the most effective barriers with the third lowest adhesion formation rate (78.95%) and the second lowest TSS (15.00). Cova CARD is a relatively new barrier that acts as a resorbable, malleable porcine collagen membrane, promoting tissue regeneration.3, 18 While the human data is limited, collagen sheets have been shown to resemble native pericardial membranes at 24 weeks after operation in animal models.3, 19 The adhesion formation rate may be greater relative to PTFE; however, the improved TSS suggests that this barrier may provide easier dissection than PTFE. Easier dissection may also explain the improved safety identified regarding infection and bleeding events relative to PTFE. Adhesion formation occurred with almost every patient when using Seprafilm (95.83% vs control 94.12%), COSEAL (100.00%), REPEL-CV (100.00%; control 100.00%), Silicone Rubber (100.00%), Polyglycolic Acid Mesh (100.00% vs control 100.00%), and Porcine and Polyester Gelatin Sheet (100.00%). The study assessing Polyglycolic Acid Mesh should be discussed though, as it used computerized tomography (CT) imaging instead of re-operation to identify adhesion formation.20 Polyglycolic Acid Mesh was compared against PTFE and a no barrier control group that both had adhesion formation rates of 100.00% as well. The increased rate of adhesions for PTFE in this study relative to its overall adhesion formation rate (37.31%) suggests that CT imaging may allow for more detailed identification of adhesions that may not necessarily be clinically relevant in the re-operation setting. Nevertheless, the study did note that PTFE was a more effective substitute than Polyglycolic Acid Mesh with respect to the reported “total adhesion scores” (p < 0.001). While the adhesion formation rate was high for nearly all barriers, the TSS demonstrated noteworthy variance. The similar TSS scores in the control groups also suggest that comparison may be appropriate among the barriers’ TSS scores. Nevertheless, the wide variance in reported values for safety and efficacy likely relates to the subjective approach used to report these variables and suggests a need for standardization moving forward.
This study has limitations given its study design as a systematic review. These limitations relate to the evidence included in the review as well as the review process itself. With respect to the evidence included, not all studies reported the same safety and efficacy variables. Some of the relevant variables discussed in this review were not mentioned in some studies; furthermore, other variables included in a few studies were not assessed here to ensure appropriate comparisons. Statistical analyses comparing the barriers was also limited due to the heterogeneity of data reported. A major contributor was the variance in study design and presence of control groups, which contributed to the risk of bias in the studies. Future adhesion barrier studies should address these limitations by providing detailed safety and efficacy data that includes the variables discussed here for both the intervention and comparison cohorts. Other potentially valuable variables include structural injuries, visibility, dissection duration, and ease of use. With respect to the review process, the search strategies were unable to assess all publications to date on adhesion barriers in the clinical setting of cardiac surgery. Although the search strategies used were broad and included two separate databases, other relevant articles may have not been identified. Publication bias likely influenced the results as well in that only published studies were used in this systematic review. Future systematic reviews should be mindful of these limitations and pursue a more inclusive approach as additional PPA barrier studies are published.
In conclusion, this is the first systematic review of adhesion barrier safety and efficacy in cardiac surgery. The findings suggest that no ideal adhesion barrier currently exists for preventing PPAs. While the barriers assessed in this study are commonly used in other forms of surgery, future barrier development must focus on the requirements unique to operating in and around the heart during cardiac surgery. PTFE has historically been used, but the relevant safety and efficacy concerns identified here suggest areas for improvement. Recent adhesion barriers have been developed that demonstrate improvements in infection rates, bleeding event rates, mortality, adhesion formation, and tenacity scores. In particular, Cova CARD may provide better outcomes. Combinations of adhesion barriers, such as PTFE + SprayGel and PTFE + Seprafilm, may also provide synergistic improvements in safety and efficacy. However, further validation is required before drawing any conclusions. PPAs ultimately pose a major burden to patients and providers in terms of morbidity, mortality, and surgical ease. Reducing their formation is vital to improving outcomes for all cardiac surgery candidates.