Evaluation of a Gelatin Matrix as a Topical Hemostatic Agent for Hepatic Bleeding in the Dog
New generation topical hemostatic agents containing thrombin have been developed for use in surgical procedures when control of bleeding by conventional methods is either ineffective or impractical. The authors compared the safety, hemostatic efficacy, and handling characteristics of a thrombin-containing topical surgical hemostatic agent (a gelatin matrix) to a hemostatic gelatin sponge for treatment of parenchymal bleeding after liver biopsy. Fourteen dogs were enrolled in this prospective clinical study. Paired 1.5 cm × 1.5 cm and 0.5 cm deep liver biopsies were obtained via laparotomy for each dog. One bleeding liver biopsy lesion was treated with the gelatin matrix and the other with a gelatin sponge. The treated liver biopsy sites were compared for bleeding severity, time to hemostasis, cumulative blood loss, and hemostatic agent handling characteristics. Median time to hemostasis was significantly shorter (P = 0.034) and median cumulative blood loss was significantly lower (P = 0.033) for the lesions treated with the gelatin matrix than the gelatin sponge. Adverse reactions were not observed within the first 24 hr postoperatively. When used to control parenchymal bleeding from liver biopsy sites in the dog, the evaluated gelatin matrix was safe and more effective than the gelatin sponge.
Introduction
Due to the parenchymal nature of the liver, hepatic bleeding following liver resection, biopsy, or trauma may be difficult to control using conventional means of hemostasis such as ligatures, clips, electrosurgery, and manual pressure.1,2 In patients with an underlying coagulopathy, hepatopathy, or hepatic neoplasia, hemostasis can be even more difficult to achieve, and there is an increased risk of postoperative hemorrhage.2–8 In human hepatic surgery, delayed hemostasis results in increased operative times, operative costs, blood product transfusions, and higher rates of postoperative morbidity and mortality.1
To address those concerns, topical hemostatic agents have been developed for use in hepatic surgery (and a variety of other surgical procedures) when control of bleeding by conventional methods is either ineffective or impractical.9–11 Traditionally, those agents included topical hemostatic agents composed of gelatin, collagen, or oxidized cellulose that rely on the patient’s functional coagulation system to exert their hemostatic effects. Although conventional hemostatic agents are commonly used in veterinary surgery, they have demonstrated variable efficacy and may be of limited benefit for use in aggressive bleeding, patients with coagulopathies, irregular wound geometries, and in locations that are difficult to access.4,5,12–15 This has led to the development of a new generation of topical hemostatic agents that are able to exert their hemostatic effects in the absence of a fully functional coagulation system and have improved handling characteristics. Those include hemostatic agents that mimic the final stages of the coagulation cascade (gelatin-based or collagen-based hemostatic agents containing thrombin) or function independently of the coagulation cascade (fibrin sealants, glutaraldehyde-based glues, polyethylene glycol-based sealants). To the authors’ knowledge, there are no studies evaluating the safety and efficacy of commercially available new generation topical hemostatic agents in dogs.
A new generation topical hemostatic agent that is composed of a bovine-derived gelatin matrix component and a human-derived thrombin component is available. Those components are combined in the sterile surgical field resulting in 5mL of a highly viscous, granular, flowable gel that is applied using a single syringe and supplied plastic applicator tip. The mechanism of action of that product is related to the physical properties of the gelatin matrix and the activity of thrombin in the conversion of fibrinogen to fibrin. When applied to the actively bleeding tissue surface, the thrombin-coated gelatin granules fill the wound bed, absorb blood, and swell, providing a tamponade effect that physically restricts blood flow. Furthermore, blood passing through the spaces between the gelatin granules is exposed to high concentrations of thrombin that converts the patient’s circulating fibrinogen to fibrin monomers that polymerize to form a fibrin clot. The fibrin polymers physically entrap the gelatin granules and other cellular elements, resulting in a structurally stable gelatin-fibrin matrix that seals the bleeding site.
Although the use of that gelatin matrix in a variety of human surgical settings has been previously reported, the safety and efficacy of the gelatin matrix for surgical hemostasis in dogs have not been reported.4–6,9–12,16–19 The purpose of this prospective study was to evaluate the safety and efficacy of the thrombin containing, gelatin-based topical surgical hemostatic agenta, referred to as the gelatin matrix in this report, compared with a conventional gelatin-based surgical hemostatic agent, referred to as the gelatin spongeb, for treatment of parenchymal bleeding after liver biopsy in the dog. The primary objective was to compare hemostatic efficacy by evaluating time to hemostasis, blood loss volume, and immediate hemostasis rate in liver biopsy lesions treated with either the gelatin matrix or gelatin sponge. Differences in hemostatic agent handling characteristics and potential modes of failure were also identified, as well as any postoperative complications and adverse reactions in relation to their use. The hypotheses were that the gelatin matrix would be more effective than the gelatin sponge in achieving hemostasis after liver biopsy and in achieving hemostasis in liver biopsy lesions with a higher bleeding severity. The authors also hypothesized that the gelatin matrix would have acceptable handling characteristics and would be safe to use for hemostasis for parenchymal bleeding following liver biopsy in the dog.
Materials and Methods
Study Population
Large-breed, client-owned dogs undergoing elective prophylactic gastropexy and dogs requiring splenectomy for spontaneously occurring splenic disease were included. All dogs weighed > 10 kg and were > 6 mo of age. Exclusion criteria included prior reaction to either bovine or porcine gelatin products, prior exposure to human blood products, hypofibrinogenemia, and previously diagnosed active bacterial hepatitis.
All pet owners were advised that two liver biopsies would be obtained and that hemostasis would be achieved in one biopsy site using a conventional hemostat (the gelatin sponge) and the other site using a thrombin-containing hemostat (the gelatin matrix). Clients were advised that if at any time the surgeon determined that continuing with the hemostat testing was not in the patient’s best interest (i.e., the patient became unstable under anesthesia), the hemostat testing procedure would be aborted and the primary surgical procedure(s) completed.
All precautions were taken to ensure that hemostasis of the liver biopsy sites had been successfully achieved. Prior to abdominal closure, liver biopsy sites were inspected and hemostasis was confirmed. For the first 24 hr, all patients were continuously monitored for any clinical signs of either intra-abdominal hemorrhage or adverse reactions. As described below, blood pressure, mucous membrane color, body temperature, heart rate, respiratory rate, packed cell volume (PCV), and total solids (TS) were measured immediately after surgery and 2 hr, 4 hr, 12 hr, and 24 hr postoperatively.
Informed consent was obtained from all owners of the 14 dogs participating in this study. For the 12 dogs presenting for elective prophylactic gastropexy surgery, in exchange for participation in this study, all services (including diagnostic tests, anesthesia, surgery, hemostatic agents, and hospitalization) were provided at no cost by the Animal Surgical and Emergency Center, which funded this study. For the two dogs presenting for splenectomy, the hemostatic agents used to treat bleeding from the liver biopsy sites were provided at no cost by the Animal Surgical and Emergency Center.
Preoperative Evaluation
A complete physical exam, serum biochemical profile, complete blood count, PCV, and urinalysis were performed for each dog. Buccal mucosal bleeding time, manual platelet count, quantitative fibrinogen, prothrombin time (PT) and partial thromboplastin time (PTT) were also determined.
Anesthesia Protocol
Dogs were premedicated with glycopyrrolatec (0.01 mg/kg subcutaneously [SC] or intramuscularly) and hydromorphoned (0.1 mg/kg IV or SC). Anesthesia was induced with diazepame (0.2 mg/kg IV) and propofolf IV to effect (mean, 3.0 mg/kg; range, 1.7–4.0 mg/kg). One dog was premedicated with atropineg (0.01 mg/kg SC) and hydromorphone (0.1 mg/kg SC) and induced with a combination of diazepam (0.5 mg/kg IV) and ketamineh (4.0 mg/kg IV). All dogs were maintained with sevofluranei in oxygen.
Liver Biopsy Technique
Mean arterial blood pressure (MAP) was recorded immediately prior to the creation of each liver biopsy lesion. If intraoperative hypotension (MAP ≤ 60 mm Hg) was identified, additional crystalloid fluid therapy was administered until the MAP was > 60 mm Hg. Using a No. 10 scalpel blade, a 1.5 cm × 1.5 cm and approximately 0.5 cm deep biopsy was obtained from the center of the left lateral liver lobe (Figure 1A). To create equal sized, paired liver biopsy lesions, a premeasured square plastic template was used to trace the margins of the liver biopsy site. The depth was visually estimated. Liver biopsies were weighed using a precision scale j, and the exact dimensions (measured in cm) were recorded. Bleeding severity score at the biopsy site was estimated on a scale from 1 to 3 where 1 was mild/oozing, 2 was moderate/flowing, and 3 was severe/heavy flowing. For each dog, one liver biopsy lesion was treated with the gelatin matrix and the other with the gelatin sponge. The order of hemostatic agent selected to treat the first liver biopsy lesion was randomized by coin toss. If the first bleeding liver biopsy lesion was treated with the gelatin matrix then the second biopsy lesion was treated with gelatin sponge. Once hemostasis of the first treated liver biopsy lesion was achieved, the second liver biopsy lesion was created approximately 2 cm from the first. The location of the second liver biopsy lesion with respect to the first was not standardized. Because of the relatively large size of the left lateral liver lobe, both liver biopsy lesions were placed near the center of the lobe and were visually judged to be equidistant from the hilus and margins of the liver lobe.
Citation: Journal of the American Animal Hospital Association 49, 5; 10.5326/JAAHA-MS-5927
Following hemostasis of both biopsy sites, a prophylactic muscle flap gastropexy was performed in all dogs. Additional surgical procedures including ovariohysterectomy, cryptorchid castration, and splenectomy were performed in some dogs.
Topical Hemostatic Agent Application
The gelatin matrix was prepared at least 1 min prior to use, resulting in 5 mL of matrix with a thrombin concentration of 400 IU/mL. The gelatin matrix was delivered from a single barrel syringe with the supplied plastic applicator tip to the base of the bleeding liver biopsy lesion. According to the manufacturer’s instructions, a saline-moistened gauze sponge was held against the gelatin matrix to keep it approximated to the bleeding surface for 1 min then removed (Figure 1B).
The gelatin sponge was cut to a size slightly larger than the wound dimensions (1.8 cm × 1.8 cm × 0.7 cm). Because gelatin sponge may shrink slightly when initially exposed to blood, this ensured that there was sufficient material to completely cover the liver biopsy lesion surface and edges. Gelatin sponge was applied dry to the bleeding liver lesion. According to the manufacturer’s instructions and to prevent inadvertent disruption of the clot upon removal, a gauze sponge moistened with saline was held against the gelatin sponge to keep it approximated to the bleeding surface for 1 min then removed.
Time to Hemostasis
Hemostasis was defined as the complete cessation of the flow of either blood or serum from the margins of the treated liver biopsy lesions. Time elapsed (in sec) between hemostatic agent application and hemostasis, including reapplication if required, was determined using a digital timerk. If hemostasis was achieved on removal of the gauze sponge, the time to hemostasis was recorded as 1 min. Immediate hemostasis was defined as hemostasis occurring within 3 min of initial hemostatic agent application.
Blood Loss
Using a precision scale, hemorrhage was quantified by comparing the weight of gauze sponges before and after absorbing blood from the treated liver biopsy lesions. The dry gauze sponge was held against the liver surface adjacent to, but not contacting, the bleeding lesion and blood was collected. The gauze was reweighed, and total blood loss was converted from g into mL (1 g of blood = 1 mL of blood). Because a saline-moistened gauze sponge was held against each hemostatic agent for 1 min after application, the quantification of blood loss was not started until 1 min after hemostatic agent application. The volume of blood collected over a 1 min interval was recorded at the following time points after hemostatic agent application: min 1–2, min 2–3, min 3–4, min 6–7, and either min 10–11 or until hemostasis was achieved. Cumulative blood loss was defined as the total amount of blood collected (mL) over those time points for each treated liver biopsy lesion.
Reapplication
The hemostatic agent was reapplied if blood loss was subjectively considered to be excessive. For the gelatin sponge, an additional piece was applied to the actively bleeding area. If the gelatin sponge became saturated and dislodged completely from the bleeding liver lesion, the gelatin sponge was replaced with a new piece. For the gelatin matrix, the applicator tip was placed through the existing gelatin matrix to the base of the tissue bed in the actively bleeding area and additional gelatin matrix was applied.
Intraoperative Evaluation
Immediately following hemostasis, for some dogs, both treated liver biopsy lesions were irrigated with saline. A 60 cc catheter tip syringe with 20–45 cc of saline was held approximately 8 cm above the treated liver biopsy lesion and the hemostatic agent was gently irrigated. The irrigated liver biopsy lesion was blotted with a dry gauze sponge. For some dogs, the amount of hemostatic agent remaining in the liver lesion bed was visually estimated and recorded as a percentage of the original amount of hemostatic agent present prior to irrigation (Figure 1C).
Treated biopsy sites were re-examined prior to abdominal closure to confirm hemostasis and the presence of the hemostatic agent. Subjective observations of the handling characteristics of the hemostatic agents were made including the following: the ability of the hemostatic agent to adhere and conform to the liver biopsy lesion during application, the ease of reapplication of the hemostatic agent (if required), and the ability to irrigate and remove excess hemostatic agent not incorporated in the clot once hemostasis was achieved. Modes of failure, defined as observations of the possible causes of ongoing bleeding following hemostatic agent application, were also subjectively evaluated for both hemostatic agents. Treated liver biopsy lesions were observed to determine if bleeding was occurring either through the hemostatic agent or from an exposed margin of the lesion. If bleeding recurred following hemostasis, observations were made to determine if this was caused by hemostatic agent dislodgement, manipulation of the liver lobe, or following saline irrigation of the hemostatic agent.
Postoperative Monitoring
Dogs were monitored for clinical signs of either intra-abdominal hemorrhage or adverse reactions through visual patient assessment and observations of blood pressure, mucous membrane color, body temperature, heart rate, respiratory rate, PCV, and TS immediately after surgery and at 2 hr, 4 hr, 12 hr, and 24 hr postoperatively.
Statistical Analysis
Treated lesions were compared for MAP, liver biopsy area, liver biopsy weight, bleeding severity score, blood loss volume, time to hemostasis, and time to final biopsy site inspection using the exact Wilcoxon signed rank test for paired data. The data were expressed as median and range, and the level of significance was P ≤ 0.05. The rates of immediate hemostasis for treated lesions were compared using an exact χ2 test. Statistical analyses for those comparisons were performed using statistical softwarel. The effect of bleeding severity score and hemostatic agent on time to hemostasis was determined by linear regression with robust variance estimation using statistical softwarem. The Kaplan-Meier method was used to compare the times to hemostasis between the treated lesions. The log-rank test was used to compare the probability curves.
Results
Study Population
Fourteen dogs (four females, four spayed females, one male, and five castrated males) were included. Included breeds were the Great Dane (n = 2), American Staffordshire terrier (n = 2), boxer (n = 2), golden retriever (n = 1), Great Pyrenees (n = 1), Saint Bernard (n = 1), standard poodle (n = 1), Belgian Malinois (n = 1), Siberian husky (n = 1), German shepherd dog (n = 1), and mixed-breed (n = 1). Median body weight was 28.9 kg (mean, 35.2 kg; range, 18.2–70 kg). Median age was 2.3 yr (mean, 3.7 yr; range, 6 mo to 10 yr). Twelve dogs had normal physical exams and two dogs presented for hemoabdomen with splenic mass. Paired liver biopsies with hemostatic agent testing and prophylactic muscle flap gastropexy were performed in all dogs. Other surgical procedures included ovariohysterectomy (n = 4), splenectomy (n = 2), and cryptorchid castration (n = 1).
Preoperative Evaluation
All dogs had a normal buccal mucosal bleeding time, PT, platelet count, and TS. Four dogs had an elevated fibrinogen, ranging from 263 mg/dL to 538 mg/dL (reference range, 90–255 mg/dL). One dog with a splenic mass had a prolonged PTT at 17.7 sec (reference range, 10.6–16.8 sec). The median PCV for all dogs was 42% (range, 21–56%). Both dogs with splenic masses were anemic with a PCV of 31% and 21% (reference range, 37–55%). The dog with the PCV of 21% received 2 units of packed red blood cells in the perioperative period.
Liver Biopsy
There were no significant differences in liver biopsy dimensions, liver biopsy weight, bleeding severity score, and MAP prior to liver biopsy for the treated lesions (Table 1). Median volume of the gelatin matrix used was 4.1 cc (range, 2.0–5.0 cc). Median volume of gelatin sponge used was 2.21 cc (range, 1.58–3.39 cc), which corresponded to a sponge dimension of 1.8 cm × 1.8 cm × 0.7 cm.
Data are presented as median (range).
Bleeding score was based on a scale from 1 to 3 where 1 was mild/oozing, 2 was moderate/flowing, and 3 was severe/heavy flowing.
MAP, mean arterial blood pressure.
Time to Hemostasis
Median time to hemostasis was significantly shorter (P = 0.034) for the liver biopsy lesions treated with the gelatin matrix (median, 136 sec; range, 60–583 sec) than the lesions treated with the gelatin sponge (median, 373 sec; range, 60–660 sec). Hemostasis was achieved for all liver lesions by 11 min. At each time point, the proportion of lesions with hemostasis was greater for the gelatin matrix-treated biopsy lesions than the gelatin sponge-treated lesions (Figure 2). Immediate hemostasis was achieved in a significantly higher percentage (P = 0.046) of gelatin matrix-treated than gelatin sponge-treated lesions (57% versus 14%, respectively).
Citation: Journal of the American Animal Hospital Association 49, 5; 10.5326/JAAHA-MS-5927
Time to hemostasis was significantly longer in liver lesions with higher bleeding severity scores (P = 0.001). In lesions with higher bleeding severity scores, treatment with the gelatin matrix resulted in significantly shorter times to hemostasis than treatment with the gelatin sponge (P = 0.003).
Blood Loss
Median cumulative blood loss was significantly lower (P = 0.033) for the gelatin matrix-treated biopsy lesions (median, 0.12 mL) than the gelatin sponge-treated lesions (median, 1.18 mL) as shown in Table 2.
Data are presented as median (range).
Intraoperative Evaluation
Saline irrigation of the treated liver biopsy lesion was performed in 12 of 14 dogs. It was not performed in the first two dogs in the study. The amount of saline used for irrigation was the same for both the matrix- and sponge-treated lesions (median, 35 cc; range, 20–45 cc). A visual estimation of the amount of hemostatic agent remaining following saline irrigation was performed in 8 of 12 dogs. It was not performed in the first four dogs that received saline irrigation. The estimated median amount of hemostatic agent remaining following saline irrigation was 75% for the gelatin matrix (range, 50–85%) and 100% for the gelatin sponge.
In 6 of 14 gelatin sponge-treated lesions, the gelatin sponge failed to either conform or adhere well to the cut tissue surface during application, which resulted in bleeding from an exposed portion of the liver lesion. The gelatin matrix conformed and adhered well to all treated lesions.
In one gelatin sponge-treated lesion, reapplication was required due to bleeding resulting in complete saturation and dislodgement of the gelatin sponge 5 min after application. The gelatin sponge was replaced with a new piece and hemostasis occurred in 4 min. In one gelatin matrix-treated lesion, an insufficient amount of product was inadvertently used during the initial application, which resulted in failure to cover the entire bleeding surface of the liver biopsy lesion. At 2 min after the initial application, an additional 1 cc of gelatin matrix was applied from the same syringe to the bleeding uncovered portion of the liver biopsy lesion. Hemostasis was achieved 1 min after reapplication. Out of 5 cc of the prepared gelatin matrix, a total of 4 cc was used to treat the liver biopsy lesion.
In 2 of 14 gelatin sponge-treated lesions, mild bleeding occurred upon manipulation of the liver lobe after hemostasis had already been achieved. Bleeding ceased within 0.5–3 min of gauze sponge compression. Rebleeding was not observed in the gelatin matrix-treated lesions. None of the liver biopsy sites were bleeding at the time of abdominal closure. Time elapsed between hemostatic agent application and final biopsy site inspection was not significantly different (P = 0.24) for gelatin matrix- (median, 33.5 min; range, 18–50 min) and gelatin sponge-treated lesions (median, 26.5 min; range, 14–67 min).
Postoperative Monitoring
By 2 hr postoperatively, 12 of 14 dogs had a normal PCV. Both splenectomy dogs remained anemic, but the PCV was not significantly different than the immediate postoperative value. At 24 hr, TS were decreased in those two dogs and three others (range, 4.4–5.2 g/dL). In the 24 hr postoperative monitoring period, none of the dogs either demonstrated a fever (i.e., a body temperature > 39.2°C) or hypotension (MAP ≤ 80 mm Hg). Clinically apparent intra-abdominal hemorrhage or adverse reactions were not identified.
Discussion
Larger liver biopsies are recommended to increase the accuracy of histopathologic diagnosis but can result in increased bleeding severity.20,21 The liver biopsy method used in this study provided challenging bleeding conditions to compare the hemostatic efficacy of a gelatin matrix and gelatin sponge and is a highly reproducible method of quantifying blood loss and time to hemostasis.13 Negative control liver biopsy lesions were not created for this study because a previous canine liver biopsy study showed that digital pressure alone for 10 min was ineffective for hemostasis.22
In the current study, the median time to hemostasis was significantly shorter for the gelatin matrix-treated lesions (136 sec) than gelatin sponge-treated biopsy lesions (373 sec). This was consistent with studies of human patients undergoing cardiac, vascular, or spinal surgery that were treated with either gelatin matrix or thrombin-soaked gelatin sponge. For each cohort, treatment with the gelatin matrix resulted in significantly shorter times to hemostasis and higher hemostatic success rates, which (in that human study) was defined as bleeding cessation within 10 min.4,12,17
The authors of this report found that immediate hemostasis, defined as hemostasis occurring within 3 min of hemostatic agent application, was achieved in a significantly higher percentage of gelatin matrix-treated lesions (57%) than gelatin sponge-treated lesions (14%). The time point was chosen because, by virtue of their individual mechanisms of action, in a clinical setting, both the gelatin matrix and sponge require adequate time for absorption of, and interaction with, blood to form an effective fibrin clot. In the aforementioned human studies, gelatin matrix-treated sites also had significantly higher rates of bleeding cessation within 3 min than thrombin-soaked gelatin sponge treated sites for the cardiac (72% versus 23%) and spinal (97% versus 71%) cohorts.4,17 The presence of high levels of exogenous thrombin in gelatin matrix may account for the faster hemostasis times observed in this study. That suggestion does not fully explain the faster hemostasis times observed in other studies when the gelatin matrix was compared with thrombin-soaked gelatin sponge. The swelling and tamponade effect of the gelatin granules (as they absorb blood) may also contribute to the faster hemostasis times observed with the gelatin matrix.
Due to the high vascularity and sinusoidal architecture of the liver, bleeding severity after liver biopsy can vary between biopsy locations.3,20 The authors of this study were able to evaluate the relative efficacy of gelatin matrix and gelatin sponge under conditions of varying bleeding severity. Because each hemostatic agent was applied in a manner recommended by the manufacturer for clinical use, requiring complete coverage of the hemostatic agent with a gauze sponge, the study authors did not attempt to quantify blood loss over the first min following application. Both hemostatic agents require time for interaction with blood to exert their hemostatic effects. Attempting to record blood loss sooner than 1 min postapplication was not considered to be clinically relevant.
The median cumulative blood loss was significantly lower for gelatin matrix-treated lesions (0.12 mL) than gelatin sponge-treated lesions (1.18 mL). The liver biopsy lesions with higher bleeding severity scores took longer to achieve hemostasis. For more aggressive bleeding, treatment with the gelatin matrix resulted in significantly shorter times to hemostasis than gelatin sponge. Similarly, in human cardiac surgery, treatment with gelatin matrix resulted in higher hemostatic success rates, defined as cessation of bleeding within 10 min, than thrombin-soaked gelatin sponge for both “oozing” and “heavy” bleeding conditions.4 In canine models of hemostasis, gelatin sponge alone was effective for diffuse oozing, but either ineffective or of variable efficacy for more aggressive bleeding.13–15 The findings of this study and others suggest that although conventional hemostatic agents may be used effectively for hemostasis in mild bleeding, thrombin-containing topical hemostatic agents, like the gelatin matrix used in this study, are more effective and can be used for bleeding conditions ranging from oozing to heavy bleeding.9
Because 13 of 14 dogs in this study had normal coagulation parameters, it is not possible to make conclusions regarding the efficacy of the gelatin matrix for hemostasis in dogs with a coagulopathy. The authors of this study expect that the gelatin matrix would be effective in dogs with either an inherent or induced coagulopathy because the gelatin matrix contains thrombin that acts in the last stage of the clotting cascade to convert the patient’s circulating fibrinogen to a fibrin clot, eliminating the need for a fully functional coagulation cascade or adequate functional platelets. In animal studies, conventional topical hemostatic agents were less effective under conditions of anticoagulation than thrombin-containing hemostatic agents.3,23 The gelatin matrix used in this study has been shown to be highly effective in human patients with an acquired coagulopathy undergoing hepatic resection surgery and was more effective than thrombin-soaked gelatin sponge in patients with heparin-induced coagulopathy undergoing cardiac surgery.4–6
The authors of this study identified several handling advantages of the gelatin matrix over the gelatin sponge during application, reapplication and removal of excess material. Due to its flowable, granular consistency the gelatin matrix conformed better to the wound resulting in more complete coverage of the bleeding area. Conversely, in 6 of 14 of gelatin sponge-treated sites, the gelatin sponge did not adhere or conform well to the wound during application. In those cases, the interface between sponge and tissue became slippery, causing motion of the sponge across the cut liver surface and bleeding from an exposed portion of the biopsy lesion. In the one patient where reapplication was required, the authors were able to precisely deliver additional gelatin matrix to the bleeding area without disrupting the existing clot. In one patient, reapplication of gelatin sponge required removal and replacement of new sponge, which may disrupt the clot that had already formed at the tissue surface. In human studies, the gelatin matrix was also judged by the surgeon to be better than thrombin-soaked gelatin sponge in its ease of application and ability to conform to the surgical site.4,12,16,17
For both the gelatin matrix and gelatin sponge, once hemostasis is achieved, the manufacturers recommend gentle saline irrigation to remove excess hemostatic agent, leaving behind only the material that is incorporated in the hemostatic clot. Because the gelatin sponge functions as a single unit, it cannot be partially removed by irrigation. Therefore, once hemostasis has been achieved, the surgeon must choose between either leaving all of the gelatin sponge in the wound or risk rebleeding by removing it. In the current study, due to its granular nature, the study authors were able to remove excess gelatin matrix that was not incorporated in the clot without disrupting the hemostatic seal. This is advantageous because removal of excess foreign material is more desirable for wound healing and allows for improved visualization of the surgical site.19
Following hemostasis, rebleeding was observed in 2 of 14 gelatin sponge-treated lesions when the liver lobe was manipulated for a final inspection of the treated biopsy lesion. With the gelatin sponge, the fibrin clot mainly forms at the interface between the sponge and tissue, which may make it more fragile and prone to disruption when the overlying sponge is manipulated.4,15,17 The individual gelatin granules in the gelatin matrix are surrounded by fibrin strands and become integrated within the fibrin clot, which may result in a stronger clot. Until the gelatin matrix clot fully matures, the gelatin granules subjectively appear to remain loosely connected to each other. This may result in a more flexible clot that is not easily disrupted by manipulation of either the gelatin matrix clot or surrounding tissues.
Although not evaluated in this study, an important feature to consider when choosing a topical hemostatic agent is the timing and magnitude of swell of the hemostatic agent after implantation. When exposed to saline, gelatin sponge reportedly swells by 320% for up to 3 days after implantation.16 The manufacturer recommends removing the gelatin sponge (once hemostasis is achieved) in situations where there is the possibility of dislodgement or compression of nearby anatomic structures. In contrast, the gelatin granules of the gelatin matrix reportedly swell by only 10% to 20%, which occurs within 10 min of application.16 The rapid and minimal swell of the gelatin matrix may make it a better choice than gelatin sponge for use in neurosurgical applications and surgeries where compression of surrounding structures might result in complications.
The clinical advantage of using a thrombin-containing topical hemostatic agent such as the gelatin matrix used in this study (costing $204) must be weighed against its increased cost compared with a conventional hemostatic agent such as gelatin sponge (costing only $11). In human surgeries, the use of gelatin matrix has resulted in increased cost savings due to faster hemostasis times, shorter operative times, decreased procedural and hemorrhagic complications, fewer blood transfusions, and shorter hospital stays.18,19 In clinical veterinary practice, gelatin matrix use may be most justified in either aggressive bleeding conditions or when its handling properties are preferred.
Using the aforementioned postoperative monitoring parameters, the authors of this study did not identify any adverse reactions in relation to either gelatin matrix or gelatin sponge use. In this study, there was no attempt to either identify or characterize any potential immune response to the topical human thrombin component of the gelatin matrix. To the authors’ knowledge, no studies have been performed to determine whether repeated exposure to topical human thrombin in dogs increases the risk of adverse reactions. Until more is known about the immunogenicity of human thrombin in dogs, it is important to weigh the potential risks versus benefits of repeated administration of the gelatin matrix.
Due of the difference in appearance of the hemostatic agents used in this study, the participating surgeons could not be blinded to the treatment group, allowing for the possible bias of results. Other limitations of this study include the small study population as well as the lack of dogs with naturally occurring coagulopathies.
Because the thrombin in the gelatin matrix used in this study converts the patient’s fibrinogen to fibrin, for optimal hemostasis, adequate circulating levels of fibrinogen are required. Dogs with hypofibrinogenemia were excluded from this study. Four dogs in the current study had fibrinogen levels above the reference range that could have created a bias in favor of the gelatin matrix. A study of coagulation assays in healthy dogs demonstrated a high degree of individuality and analytical variation for PT, PTT, and fibrinogen, making population-based reference ranges alone insensitive interpretation criteria for either hypercoagulability or hypocoagulability.24 Due to the limitations of routine plasma-based assays, the inclusion of thromboelastography in future studies would be beneficial to more accurately identify patients with an underlying coagulopathy and predict hemorrhagic potential.24,25 Thromboelastography provides data about the entire coagulation system, including cellular and soluble components, and is highly correlated with fibrinogen concentration.25
Additional studies are warranted to evaluate the efficacy of the gelatin matrix under conditions of coagulopathy and in comparison with other thrombin-containing hemostatic agents such as fibrin sealants, thrombin-soaked gelatin sponges, and other commercially available gelatin-thrombin and collagen-thrombin based products. In canine models of hemostasis, fibrin sealants were effective in controlling bleeding following liver biopsy using a 15 gauge needle in anticoagulated dogs and in a study of liver tip excision biopsy in combination with digital compression.22,26 In other studies, fibrin sealants have been shown to be effective for “oozing” type bleeding but ineffective for more aggressive bleeding due to the tendency for it to be washed away from the application site by the pressure from arterial bleeding.27,28 Optimal application of fibrin sealant requires a dry surgical field, and once it is applied additional material cannot be effectively reapplied through the existing material.28,29 This is in contrast to the gelatin matrix used in this study, which can be applied through a pool of blood to achieve hemostasis in conditions of mild to severe bleeding and, if necessary, additional gelatin matrix can be applied through the existing material.
Conclusion
The use of a gelatin matrix topical surgical hemostatic agent to treat bleeding after liver biopsy in the dog resulted in faster times to hemostasis and lower cumulative blood loss than with a gelatin sponge. Immediate hemostasis was achieved in a significantly higher percentage of gelatin matrix-treated lesions compared with gelatin sponge-treated lesions. For liver biopsy sites with higher bleeding severity scores, treatment with the gelatin matrix resulted in shorter times to hemostasis. The authors conclude that when used to control parenchymal bleeding from liver biopsy sites in the dog, the evaluated gelatin matrix is a safe and more effective topical hemostatic agent than the gelatin sponge.
A: Appearance of a 1.5 cm × 1.5 cm and 0.5 cm deep bleeding liver biopsy lesion in the center of the left lateral liver lobe. B: Appearance of the same liver biopsy lesion after hemostasis was achieved following treatment with a gelatin matrix. C: Appearance of the same gelatin matrix-treated liver biopsy lesion following saline irrigation and subsequent removal of excess gelatin granules that were not incorporated within the clot.
Comparison of time to hemostasis for gelatin matrix-treated and gelatin sponge-treated bleeding liver biopsy lesions using Kaplan-Meier estimates.
Contributor Notes
D. Polidoro's updated credentials since article acceptance are DVM, DACVS-SA.
P. Kass’ present affiliation is the Department of Population Health and Reproduction, School of Veterinary Medicine, University of California, Davis, CA.
Since article acceptance the company name has changed to Animal Specialty & Emergency Center.
