Use of formaldehyde gives a superior initial bonding, while glu-taraldehyde appears to have greater durability in vivo.12,13 A combination was, therefore, considered for optimal adhesive properties. The G/R/F glue, after disappearing from the surgical scene for some time, was re-introduced by a French group as GRFG (also then called the French glue) in 1979. The French group, Guilmet et al.15 described its use in aortic dissection. In 1982, a clinical series in aortic surgery involving the use of GRFG as an adjunct to prosthetic graft implantation was published by Bachet et al.70 Another series described the technique of complete re-fixation of the dissected layers in the ascending aorta and the aortic arch without any prosthetic material.71,72 Bellotto et al?3 studied pneumostasis of injured lung in rabbits with Gelatin-Resorcinol-Formaldehyde-Glutaraldehyde tissue adhesive. They evaluated the ability of the adhesive to seal incisional air leaks promptly in the lungs of rabbits during persistent ventilation and positive intratracheal pressure. They also assessed the shelf life of the non-proprietary formula by testing its efficacy at intervals and established a technique for application of the adhesive. They found it to be an effective pneumostatic giving a consistent decrease in the magnitude of air leak and complete pneumostasis in most of the cases, in the presence of clinically relevant positive pressure ventilation. In 2000, Nomori et al.74 used GRFG glue for sealing pulmonary air leakage during lung surgery. They mixed formaldehyde-glutaraldehyde jelly with 2.5% carboxymethylcellulose to increase its viscosity. They found the glue to be safe and effective in preventing pulmonary air ijr leakage from deeply cut lung. In 1999, Hata et al.75 concluded that the glue was good enough for improving long-term survival and to reinforce the diseased aortic wall after surgery for type-A aortic dissection, but it was not found to have sufficient haemostatic effect for use in anastomosed stitches from the outer side of the aortic wall.
In 1998, the authors, R. P. Singh et al., tested the efficacy of this promising glue as a haemostatic agent, sealant, and tissue adhesive in the liver of rat in an experimental setting in the Animal House of J. N. Medical College, Aligarh Muslim University, India. Gelatin/Resorcinol/Aldehyde (GRA) glue was prepared in two separate forms, i.e. the adhesive and the activator. The adhesive was prepared by mixing gelatin and resorcinol in 5:1 ratio weight/volume (w/v) and diluting the mixture with distilled water to a concentration of 60% w/v. A mixture of formaldehyde (9.25%) and glutaraldehyde (25%) in 9:1 ratio (volume/volume) was used as the activator to initiate cross-linking. The adhesive was packed in plastic syringes and the activator in medicine droppers. Both were to be mixed in equal volumes, just prior to application. Twenty-eight albino rats, weighing 125 to 150 grammes were used for this study. They were divided into four groups: A, B, C, and D, each comprising of seven rats. The rats were anaesthetised with ketamine. The abdomen was opened by a 5 cm long right subcostal incision, and the liver was exposed. While supporting it with a finger behind it, a wedge defect, about 1cm long and 0.5 cm deep was created in its most accessible part, usually the median lobe or the left lateral lobe. Profuse bleeding often ensued and was controlled by:
1. Applying freshly constituted and activated GRA glue on the cut surfaces.
2. Applying proximal manual pressure on the liver to control oozing, if it was persistent.
3. If this failed, haemostasis was achieved by re-application of the glue.
The haemostasis achieved was graded as excellent (1), satisfactory (2), and poor (3).
The wedge of liver was, then, apposed with the defect in the liver having glue on its surface and held in place for some time. The pressure was released every 15 seconds, and the time taken for the adhesion of the two cut surfaces was noted. The liver was released and the abdomen closed in layers. Harvesting of the rats was undertaken on specific days for each group (group A-7 days, group B-14 days, group C-21 days, and group D-28 days). During autopsy, the site of repair was evaluated for its visibility, disruption, haematoma formation, adhesions, necrosis signs, and residual glue amount. Samples of the repaired segment of the liver were obtained for histopathological examination. During operations, haemostasis was excellent in most cases, except one where two applications of the glue were needed. Haemostasis was graded as excellent in 71.4%, satisfactory in 25.0%, and poor in 3.6% cases. Mean (± SD) time for tissue adhesion was 2.6 minutes (Range: 2.4 to 2.7 minutes). All except one rat maintained good general condition in the postoperative period. This rat, belonging to group A died six hours after operation, presumably due to anaesthetic complications. Five rats developed wound infection that was managed with systemic antibiotics and local povidone iodine. The site of repair was visible in 50.0% of cases on day 7, 28.5% each on days 14 and 21, and 14.3% on day 28. Visible adhesions decreased from 66.6% on day 7 to 28.5% on day 28. Residual glue was seen in 33.3% cases on day 7 and, in none, on day 28. In the rat that died at six hours, small haematoma and fissure at the site of repair were noted. No other rats showed signs of disruption of wound or haemorrhage at the wound site. No necrosis was seen in any case. There was no bile leakage in any of the cases studied, showing good sealant property of the glue. Residual glue was present even up to 28 days, although its amount gradually decreased from 83.3% on day 7 to 43.0% on day 28. Subcapsular haemorrhage was seen only in the cases seen early. The rat that died on day 0 had subcapsular haemorrhage; this was also seen in 83.3% cases on day 7, but in none of the cases on subsequent days. An acute inflammatory response with predominantly neutrophils, eosinophils, and RBCs was noted on days 0 to 7. Mononuclear cells from day 14 gradually replaced the cellular response onwards. There was no evidence of fibrosis in any of the cases on day 7. It was present in 28.6% cases on day 14 and in 85.7% cases on day 28. Necrosis was not seen in any of the cases, and there were no foreign body giant cells. On statistical analysis, a significant decrease in subcapsular haemorrhage was observed. The authors experienced initial difficulties commensurate with the learning curve in the study. Therefore, two applications of glue were required in some of the initial experiments. In all remaining cases, however, haemostasis achieved was excellent. The decreasing trend in visibility of site of repair and residual glue was seen with passage of time. Despite the fact that no residual glue was seen upon naked eye examination, it could be identified microscopically in 43% of cases on day 28. Evidently, longer follow-up is needed to determine the time frame within which the glue is completely absorbed. Other authors have reported microscopic evidence of glue up to six to 16 weeks in experiments involving organs in different animals. Koehnlein et al.,10 however, reported the absence of glue in four to six weeks. Ennker et al?% using GR-DIAL glue in rabbit lungs found good adaptation of surfaces and progressive disintegration of glue with time, with good bio-resorption, when the incision was closed with a thin layer of glue. These differences maybe explicable on differences in experimental methods, animals, and organs involved. In this experiment, adhesions were found to decrease with time. This was in keeping with the gradual absorption of the glue. A disproportionate amount of adhesions could be attributed to experimentalists' initial unfamiliarity with the procedure. Some adhesions in group A were also noted to be distant from the defect, perhaps due to spillage of glue. Barring these few initial cases, adhesions were limited to the site of repair. Furthermore, most of the adhesions were flimsy and could easily be separated with finger, causing no bleeding on separation. Several other authors have similarly reported no adhesions or a few flimsy adhesions at various sites in different experimental settings.12 A few, however, noted extensive sub-pericardial adhesions that were difficult to separate with finger.77 The divergent results are likely due to differences in experimental subjects and methods. GRA was found to provide excellent haemostasis and tissue adhesion. It had good sealant property. Mean time for tissue adhesion was 2.6 minutes. The bond formed was pliable flexible and not hard in consistency, as was in the case of plastic adhesives. No prior sterilisation of the GRA mixture was needed, perhaps on account of the bacteriostatic properties of resorcinol and the aldehydes. It did not produce any tissue necrosis in the liver. The glue is cheap and easily prepared and stored. It also has a good shelf life. Prior to its use in humans, however, further studies and long-term follow-up are needed to evaluate its carcinogenicity and teratogenicity.
Was this article helpful?