Problematic leaking tissue surfaces are unusual but not rare in surgical patients. Understanding the sealing strength of these surfaces in terms of tissue adhesive sealant (TAS) strength–elasticity properties is limited. As a result, standards for these materials have not been established. The data in this report show that the sealing efficiencies (SEs) of a TAS with greater cohesive and adhesive strengths are higher than those produced by use of a similar TAS with lower cohesive and adhesive strengths when tested on leaking tissue surfaces in anticoagulated pigs. These data may be useful in establishing standards for TASs that are proposed for use in sealing similar leaking tissue surfaces and are consistent with a hypothetical model for sealed leaking tissue surfaces that is presented in the appendix of this report.

Tissue Adhesive Sealants. L.C. TAS has 3 components: a cross-linking agent produced by reactions of glutaraldehyde and amino-group containing agents (L-glutamic acid), ultrasound-processed porcine albumin, and ultrasound-processed porcine-skin collagen extract.^* Glutaraldehyde TAS (GA TAS) has 3 components: 14% glutaralde-hyde, porcine albumin, and a commercially available collagen topical-hemostatic agent. The TASs were prepared for use either by the addition of the cross-linking agent to a mixture of processed porcine albumin and collagen (L.C. TAS) or by the addition of the 14% glutaraldehyde to a mixture of porcine albumin and commercial collagen (GA TAS). As tested in vitro on the adventitial surfaces of porcine aortic strips and on solid materials, L.C. TAS had an adhesive strength of ~>1.6 kg/cm² and a cohesive strength of >6.0 kg/cm², whereas GA TAS had an adhesive strength of ~<0.6 kg/cm² and a cohesive strength of <1 kg/cm².

Method of Application. After sponging the leaking tissue surfaces with gauze sponges, we applied TASs and kept these sealants in place by packing them with gauze sponges. These were held by hand for 3 minutes by applying counterpressure of ~1 to 2 lbs/in². For each application, the TAS consisted of 4 × 4-cm² collagen pads, each impregnated with 2 to 3 cc of the albumin and cross-link solutions. The resulting thickness of the adhesive layer was ~2 to 4 mm. After removal of the packing sponges with counter-force methods (if the layers adhered to the sponges used to exert compression), we inspected the treated surfaces for residual fluid leakage. Any areas of residual fluid leakage through or along the edges of previously applied TAS were treated by reapplication of TAS, as described for 1st-time applications. For leaking tissue surfaces with substantial residual leakage, we applied L.C. TAS to complete the seal.

Experimental Models of Leaking Tissue Surfaces. Experimental animals were 12 mixed-breed pigs (weight range, 40–60 kg) assigned in an alternating manner to 2 groups: Group I (L.C. TAS), 6; and Group II (GA TAS), 6. All animals received humane care in compliance with the Principles of Laboratory Animal Care as formulated by the National Institutes of Health, and in compliance with the Guide for the Care and Use of Laboratory Animals as prepared by the Institute of Laboratory Animal Resources. Each pig was premedicated with intramuscular ketamine and anesthetized with isoflurane and oxygen, after which endotracheal intubation was done. Assisted ventilation was performed by hand with an Ambu bag. After midline laparotomy, 16G and 18G intravenous (IV) catheters were placed in the upper abdominal vena cava and aorta, respectively, and were secured by 5–0 polypropylene sutures. Each pig was supported by the use of IV saline and dopamine. This was done to maintain a mean aortic pressure of 60 to 70 mmHg, as measured by an aneroid manometer connected to the aortic catheter by saline-filled pressure tubing. Heating of the IV fluids and warming of the laboratory itself maintained rectal temperatures at >35 °C. Each pig was anticoagulated with porcine heparin (IV loading dose, 300 IU/kg) with repeated IV doses of 100 to 200 IU/kg to keep HEMOCHRON® activated clotting times (ITC, a subsidiary of Thoratec Corp.; Edison, NJ) at >250 seconds as measured every 30 to 40 minutes on blood obtained from the aortic line. At the end of the experiment, each pig was sacrificed by intracardiac injection of 150 mEq of potassium chloride.

Five-mm Aortic Punch Holes. In each pig, in sequence, two 5-mm aortic punch holes were produced by use of a 5-mm surgical aortic punch, and these were treated with the test TAS in such a manner as to produce an overlap length of ~5 mm and a seal thickness of ~2 to 3 mm. Compression of the TAS was performed in such a manner as to prevent visible leakage but not interrupt aortic blood flow. In sequence, each punch hole was treated with a TAS, as described above.

Splenic Leaking Tissue Surfaces. In each pig and in sequence, 2 splenic leaking tissue surfaces were produced by use of a large knife and scissors, and these were treated with TAS. Amputation laceration injuries were placed more or less at right angles to the long axis of the spleen in order to produce leaking tissue surface areas of ~16 cm² located >3 cm from the splenic capsule. In sequence, each splenic leaking tissue surface was treated with TAS pads, as described above.

Hepatic Leaking Tissue Surfaces. In each pig and in sequence, 2 leaking tissue surface areas (~16 cm² located >3 cm from the hepatic capsule) were produced by means of a large knife and scissors. In sequence, each hepatic leaking tissue surface was treated with TAS pads, as described above.

Lung Leaking Tissue Surfaces. After each pig was repositioned, a left thoracotomy was performed. In sequence, 2 left-lung leaking tissue surfaces (~16 cm² located >3 cm from the visceral pleura) were produced by means of a large knife and scissors. In sequence, each leaking tissue surface was treated with TAS and evaluated for residual bleeding and air leakage during lung hyperinflations for 10 to 15 seconds at 25 to 35 cm. Water was monitored by Ambu bag manometer during surface irrigations with saline. Residual leakage was treated as out-lined above.

None of the experimental animals became so unstable as to require premature sacrifice before test completion. Results were recorded as complete sealing after 1st or 2nd application and any required 3rd application. These results were expressed as sealing efficiencies (SE% = surfaces sealed/surfaces treated) after 1st and 2nd TAS applications. The results in Group I were compared with those obtained in Group II by use of 2-tailed t tests and 95% confidence intervals (CIs).¹ The correlation coefficients of TAS strengths and of all SEs were calculated.² This was done for the results of 1st applications for each type of leaking tissue surface and for the overall SEs of Group I and Group II (Fig. 1; Tables I and II).

Fig. 1 Tissue adhesive sealant (TAS) sealing efficiencies (SE% = surfaces sealed/surfaces treated) for 1st applications and all 1st and 2nd applications.

TABLE I. Sealing Efficiencies (SEs)

TABLE II. Group I versus Group II: First-Application Sealing Efficiencies; Statistical Analysis: 2-tailed t tests, 95% Confidence Intervals, and Correlation Coefficient*

Aortic Punch Holes. First-application SEs in Group I were 91.6%, and in Group II they were 16.6% (P <0.001; 95% CI, 49%–100%). One aortic punch hole in Group I required a 2nd L.C. TAS application. In Group II, 4 of 10 aortic punch holes were sealed by 2nd applications of GA TAS. All Group II aortic punch holes with residual bleeding (that is, bleeding after 2 attempts at sealing with GA TAS) were closed with L.C. TAS applications.

Splenic Leaking Tissue Surfaces. First-application SEs in Group I were 83.6%, and in Group II they were 0 (P <0.001; 95% CI, 78%–88%). In Group I, 2 splenic leaking tissue surfaces required a 2nd L.C. TAS application for sealing. In Group II, 2nd GA TAS applications were successful in 33.3%. In Group II, 5 of 8 splenic surfaces with residual bleeding required sealing with L.C. TAS to prevent excessive blood loss.

Hepatic Leaking Tissue Surfaces. First-application SEs in Group I were 83.6%; in Group II, they were 16.3% (P <0.001; 95% CI, 37%–97%). Two hepatic tissue surfaces in Group I with residual bleeding were sealed with L.C. TAS 2nd applications. In Group II, 3 of 10 hepatic tissue surfaces were sealed with GA TAS 2nd applications. In this group, 4 liver hepatic tissue surfaces were sealed with L.C. TAS applications to prevent excessive blood loss.

Lung Leaking Tissue Surfaces. First-application SEs in Group I were 75%; in Group II, they were 0 (P <0.001; 95% CI, 50%–99%). In Group I, 2 lung tissue surfaces were sealed with 2nd L.C. TAS applications, and 1 lung tissue surface required a 3rd L.C. TAS application for complete sealing. In Group II, 7 lung tissue surfaces after 2 GA TAS applications required L.C. TAS applications for substantial air leaks, and 1 lung tissue surface after 2 GA TAS applications required 1 L.C. TAS application for both residual bleeding and air leakage. In 4 Group II pigs, 2 GA TAS applications stopped bleeding completely and air leakage almost completely.

The correlation coefficient² was r = 0.79 for L.C. TAS adhesive strength 1.65 kg/cm² and GA TAS adhesive strength 0.6 kg/cm² versus SEs, 1st and 2nd applications.

L.C. TAS is a true composite material with cohesive and adhesive strengths that are greater than the sums of its components, calculated on a weight/volume basis. Experimental TAS does not have these characteristics. L.C. TAS has a complex chemistry that is under study. Because the results of its biocompatibility tests have been satisfactory, L.C. TAS is undergoing commercial development.

The obvious explanation for the differences in the SEs of these TASs is that the cohesive and adhesive strengths of L.C. TAS are greater than those of GA TAS. As tested on the external surfaces of porcine aortic strips, the adhesive strengths of L.C. TAS and GA TAS are >1.6 kg/cm² and ~0.61 kg/cm², respectively. Tested as single materials, the cohesive strengths of these materials are L.C. TAS, >6.4 kg/cm² and GA TAS, ~<1 kg/cm² (unpublished data, obtained by use of an Instron 3360 material strength test device [Instron; Norwood, Mass]).

In the absence of standards and specifications for TASs, selection of animal models for testing leaking tissue surfaces is a matter of judgment. We elected to use anticoagulated animals, because our experience had been that mixed-breed pigs can erratically “auto-seal” leaking tissue surfaces—assisted by counterpressure packing—over periods of 3 to 5 minutes, thereby complicating SE evaluations of TASs. The tissue surfaces used in this study were chosen on the basis of 4 considerations: these surfaces are reproducible; they resemble problematic leaking tissue surfaces encountered in surgical patients; they can be used in testing sterile TASs during long-term survival of the animal models; and use of these tissue surfaces did not exceed our limits in terms of available materials and personnel. We had not been successful in the use of similar test protocols in smaller animals, because they could not tolerate multiple testing.

It is thought that TAS effectiveness is best evaluated by determining the presence or absence of residual fluid leakage. This endpoint, which is quickly and easily verified, is in most circumstances the clinically desired endpoint. In experiments that test pig survival, this endpoint has been reliable in avoiding postoperative bleeding or air leaks that might result in postoperative animal morbidity or mortality. A disadvantage of this endpoint is the requirement, by protocol, that persistent but possibly “minor” leakage sites be closed. This can result in the possible waste of materials. Further, the use of this endpoint can complicate evaluations of treated leaking tissue surfaces insofar as leakage can result from misapplications of TASs, as well as from true failures. That is, placement of TASs on these leaking tissue surfaces in such a manner as to obtain sufficient amounts of overlap (for aortic punch holes) or minimal amounts of overlap (for other surfaces) requires both skill and practice.

At present, there are no generally accepted standards and specifications for TAS material strength properties and performance.

On the basis of these data and the use of hypothetical sealing joint structures, we have proposed an “Analysis of Tissue Adhesive Sealant Efficiencies in Relation to Leaking Tissue Surface Sealing Strength Requirements” (available as an Appendix to this report). This analysis suggests that the required TAS cohesive strength for a lapseal joint with a relatively flat profile (TAS extensibility, ~1%–5%); round tissue defect diameter, 0.5 cm; TAS thickness, 0.3 cm; and disruptive pressure, 100 mmHg is ~3.6 kg/cm². If the overlap ratio for this lapseal joint is 1:3 and the surrounding tissue substrate contains no defects, this analysis predicts that the required TAS adhesive strength equals ~>0.550 kg/cm²; were 50% of the substrate surface to consist of substrate defects, the predicted required TAS adhesive strength would equal ~>1.1 kg/cm². This analysis is consistent with the observation that a TAS that can seal an aortic punch hole may not be efficient in sealing tissue substrates that harbor multiple tissue defects. For this reason, we do not think it appropriate to use aortic punch hole testing as a “stand-alone” standard for TASs, although we do find this test useful as a supplement. Furthermore, when aortic punch hole testing is used, it is important that aortic pressure and manner of application, TAS thickness, degree of overlap, and anticoagulation status be specified, in addition to defect size.

A limitation of the leaking tissue surfaces used in this study is that these were normal tissue substrates in which the leaks were produced by lacerating injuries. These surfaces were easily débrided with moist gauze sponges, and loose tissue fragments were so removed. These surfaces did not require the removal of adherent clots and were widely exposed. Therefore, the degree to which these results will predict performance in patients who have substantial tissue substrate alterations—by disease processes or by mechanical injury that cannot easily be corrected by débridement—remains in question.

The SEs of 2 TASs were evaluated in anticoagulated pigs on 4 types of leaking tissue surfaces. L.C. TAS (cohesive strength, 6.5 kg/cm²; adhesive strength, >1.65 kg/cm²) SEs were higher than those of GA TAS (cohesive strength, ~1.0 kg/cm²; adhesive strength, ~0.6 kg/cm²). This is consistent with the hypothesis that TAS SEs are functions of their cohesive–adhesive strengths when tested on leaking tissue surfaces in anticoagulated pigs. In this experimental model, the SEs of TASs that were applied in a specified manner on aortic punch holes were similar to those of TASs that were applied to other leaking tissue surfaces as test substrates.

This study was performed with the invaluable assistance of James Tilton, PhD, Arthur Weibel, MS, and the students and staff of the Departments of Animal Physiology and Veterinary Medicine, North Dakota State University, Fargo, North Dakota; and the faculty, students, and staff of the Department of Chemistry, Minnesota State University Moorhead, Moorhead, Minnesota; George Pomonis, MD, Department of Chemistry, U.S. Department of Agriculture Laboratory, Fargo, North Dakota; and David Todd, MD (deceased), Robert Agnew, MD, and Richard Bernstein, MD, of the Department of Surgery, School of Medicine, University of North Dakota, Fargo, North Dakota, who assisted the authors in the design and performance of this study.

The Appendix to this study was prepared with the assistance of Michael Page, PhD, of the Department of Chemistry, Physical Chemistry Division, North Dakota State University, Fargo, North Dakota.

Fig. 1 “Generic” lapseal joint as a part of a hypothetical pressurized sphere.

For the lapseal joint in which component extensibility is 1%–5%, with a 1° (~0.0174 radian) I d (2a) (R) and (R) (d/2a)(1):

Since DFI = (DP) (R)/2, DFI 0.25 (DP)(D)/a radians 15 (DP)(d)(2). DF II = (DFI) (cosine a), and since (cosine a) → 1 as (a) → zero, (DFII) DFI for (a) 1°.

The total DFII is thus (DFI) (IId), where (πd) is the defect circumference.

The region I strength is (CFS) (t), where CFS is the seal material cohesive failure stress.

Region II strength is (AFS)(AO), where AFS is the adhesive failure stress and AO the area of overlap (0.78) (D²−d²).

Combining these disruptive force–strength relationships for this joint model gives the following: for region I: DP ~> 0.06 (CFS) (t/d); for region II: DP ≤ 0.03 (AFS) (D²/d² − 1). For example, a lapseal joint with (a) = 1°, (t) = 0.3 cm, (d) = 0.5 cm, (D) = 1.5 cm, and (DP) = 100 mmHg ( 0.132 km/cm²) will have a “required” CFS 3.6 km/cm² and a “required” AFS 0.550 km/cm².

For the case in which the substrate contains multiple closely placed defects occupying ~50% of the surface, the “required” AFS will be 1.1 km/cm², which is approximately the CFS of skeletal muscle.