Accepted Manuscript Results of a prospective, randomized, controlled study of the use of carboxymethylcellulose sodium hyaluronate adhesion barrier in trauma open abdomens Stanislaw P. Stawicki, MD FACS John M. Green, MD FACS Niels D. Martin, MD FACS Raymond H. Green, DO James Cipolla, MD FACS Mark J. Seamon, MD FACS Daniel S. Eiferman, MD FACS David C. Evans, MD Joshua P. Hazelton, DO Charles H. Cook, MD FACS Steven M. Steinberg, MD FACS PII:

S0039-6060(14)00100-7

DOI:

10.1016/j.surg.2014.03.007

Reference:

YMSY 3694

To appear in:

Surgery

Received Date: 2 January 2014 Accepted Date: 9 March 2014

Please cite this article as: Stawicki SP, Green JM, Martin ND, Green RH, Cipolla J, Seamon MJ, Eiferman DS, Evans DC, Hazelton JP, Cook CH, Steinberg SM, Results of a prospective, randomized, controlled study of the use of carboxymethylcellulose sodium hyaluronate adhesion barrier in trauma open abdomens, Surgery (2014), doi: 10.1016/j.surg.2014.03.007. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

Results of a prospective, randomized, controlled study of the use of carboxymethylcellulose sodium hyaluronate adhesion barrier in trauma open abdomens

John M. Green MD FACS 2 Niels D. Martin MD FACS 3,6 Raymond H. Green, DO 4 James Cipolla MD FACS 5,6

SC

Mark J. Seamon MD FACS 4,6

RI PT

Stanislaw P. Stawicki MD FACS 1,6

Daniel S. Eiferman MD FACS 1 David C. Evans MD 1

M AN U

Joshua P. Hazelton DO 4

Charles H. Cook MD FACS 1,6

Steven M. Steinberg MD FACS 1,6 1

Department of Surgery, Division of Trauma, Critical Care, and Burn, The Ohio State University

College of Medicine, Columbus, Ohio;

2

Department of Surgery, Carolinas Medical Center,

3

Thomas Jefferson University and the University of Pennsylvania,

Philadelphia, Pennsylvania;

4

Department of Surgery, Cooper University Hospital, Camden, New

Jersey;

5

TE D

Charlotte, North Carolina;

St. Luke’s University Health Network, Bethlehem, Pennsylvania, U.S.A.; for 6 OPUS 12

Foundation, Inc. Multi-Center Trials Group & OASIT Investigators.

EP

Acknowledgment: The OASIT Study Investigators would like to thank the following team members for their valuable contributions to the project: Paul R. Beery II, MD, FACS (OSU); Brian A. Hoey,

AC C

MD, FACS and Christy Stehly, BS (SLUHN); Lisa M. Capano-Wehrle, MPH (CUH); Alec C. Beekley, MD, FACS and Melissa Witt, BS (TJU); Jill Story, RN (CMC) Presentations: Presented at the 9th Annual Academic Surgical Congress in San Diego, CA, February 4-6, 2014.

Funding: The OASIT (Open Abdominal Surgery In Trauma) study was funded by Sanofi. The project was conceived by OASIT Investigators and executed independently. Study data remain the property of OASIT Investigators. Sanofi received the final manuscript for courtesy review, but did not perform any data or contextual editing. Keywords: Open abdomen; Damage control surgery; Abdominal adhesions; Adhesion barrier; Clinical trial

ACCEPTED MANUSCRIPT

ABSTRACT Introduction: The open abdominal (OA) approach is a management strategy used in the most severely injured trauma victims. It prevents the development of abdominal compartment syndrome and the lethal triad of hypothermia, acidosis, and coagulopathy. In addition to the high morbidity and

RI PT

mortality inherent to the need for leaving the abdominal fascia open, a major challenge faced by surgeons who treat patients with OAs is the gradual development of dense adhesions that make reoperations progressively more difficult. This randomized, prospective proof-of-concept study was conducted to determine the effect of carboxymethylcellulose sodium hyaluronate adhesion barrier

SC

(CMHAB) on abdominal adhesions and wound characteristics in trauma open abdomens.

Methods: A prospective, randomized, controlled study of wound and adhesion characteristics with/without CMHAB was conducted at five Level I trauma centers. Consenting patients were

M AN U

randomized to either CMHAB or “no adhesion barrier” (NAB) group. We evaluated patient demographics, injury characteristics/severity, reason for OA management, wound sizes (transverse/longitudinal), Zuhlke adhesion score, abdominal contamination score, hospital/ICU lengths of stay, morbidity and mortality.

Results: Thirty patients were enrolled (17 randomized to CMHAB; 13 randomized to NAB) with mean age of 40.3 years, ISS 29.6, AIS abdomen 3.68, APACHE II of 14.4, and 67% blunt trauma

TE D

mechanism. The groups were well-matched with regards to age, gender, ISS/abdominal AIS, penetrating/blunt injury rates, initial lactate/base deficit, mortality, OA indications, and contamination scores. There were no differences in non-abdominal or abdominal complications (i.e., fistula, abscess, wound-related) between the groups. Patients with CMHAB had significantly shorter

EP

intensive care unit lengths of stay (14.5 vs 22.3 days, p9 days or requiring 5 or more operations. There was no difference in wound sizes, overall or abdominal complications, or mortality between the CMHAB and NAB groups. Further research is warranted to better delineate potential benefits of CMHAB, especially in the setting of reoperative surgery in post-OA patients.

ACCEPTED MANUSCRIPT

INTRODUCTION The damage control/open abdominal (DC/OA) approach is a management strategy used in the most severely injured trauma victims (1, 2). This strategy prevents the development of abdominal compartment syndrome and helps attenuate the lethal triad of hypothermia, acidosis, and

RI PT

coagulopathy (2). In addition to high morbidity and mortality inherent to the magnitude of injury severity and the physiologic insult associated with DC/OA management, a major challenge faced by surgeons who treat patients with open abdomens is the gradual loss of abdominal domain and the concurrent development of dense adhesions that make consecutive reoperations progressively more

SC

difficult (2-4). Commonly encountered during DC/OA management, mechanical injury and tissue desiccation of the exposed bowel and peritoneal surfaces likely play a role in adhesion development and propagation (5). Additional risk factors for adhesion formation include postoperative surgical

M AN U

complications and hemoperitoneum, both very common in the DC/OA population (6, 7). Evidence suggests that tissue precoating with dilute hyaluronic acid solution may reduce damage to serosal surfaces during surgery and may limit formation of postsurgical adhesions (5). The primary objective of this proof-of-concept study was to determine the effect of carboxymethylcellulose sodium hyaluronate adhesion barrier (CMHAB) on abdominal adhesions and wound characteristics in trauma patients undergoing damage control surgery and open abdominal management. We hypothesized that

TE D

the application of CMHAB in trauma open abdomens would result in reduced adhesion severity, faster wound closure, and smaller wound sizes.

METHODS

EP

A prospective, randomized, controlled, patient-blinded study of adhesion and wound characteristics in the presence or absence of CMHAB (Seprafilm, Genzyme Biosurgery, Bridgewater, New Jersey,

AC C

U.S.A.) was conducted at five Level I trauma centers in the continental United States. After obtaining Institutional Review Board approval at each participating institution, the study was registered with www.ClinicalTrials.gov (identifier NCT01594385). Consenting patients were randomized to either “carboxymethylcellulose sodium hyaluronate adhesion barrier” (CMHAB) or “no adhesion barrier” (NAB) group. Descriptive representation of study enrollment and randomization is outlined in Figure 1. In order to facilitate patient recruitment, enrollment was permitted at any time prior to the second take-back operation. Alternatively, patients could be enrolled within 48 hours of admission to the hospital if the first take-back operation was delayed beyond that period. Consent by legally authorized representatives was permitted due to the fact that most patients in the study were expected

ACCEPTED MANUSCRIPT

to be intubated and sedated during the initial post-injury period.

Patient randomization was

performed using sealed envelopes with group assignments pre-determined in blocks of 5 patients at a time. Assignments were revealed at the time of the first operation, after consent was obtained. Patient exclusion criteria included age 89, pregnancy, prisoner status, abdominal closure

RI PT

prior to patient enrollment, or anticipated mortality within 48 hours of initiation of DC/OA management. Eligible for inclusion were all patients between ages 18-89 who did not meet exclusion criteria.

Clinical parameters collected during the study included patient demographics (age, gender),

SC

injury characteristics (blunt versus penetrating trauma) and severity (injury severity score [ISS] and region-specific abbreviated injury scale [AIS] assessments), clinical reason for DC/OA management, wound sizes (area in cm2 approximated from transverse and vertical wound dimensions), Zuhlke

M AN U

adhesion scores (8), abdominal contamination scores (compiled from Mickevicius, et al. (9) and Morales, et al (10)), hospital and intensive care unit (ICU) lengths of stay (LOS), morbidity and 28day mortality. Physiologic acuity assessment was performed using both the Acute Physiology and Chronic Health Evaluation (APACHE II) score (11, 12) and Simplified Acute Physiology Score (SAPS II) (13). Glasgow Coma Scale (14) determination was performed on admission, at the time of hospital discharge, and during all available follow-up appointments. In addition, we measured

TE D

Glasgow Outcome Scores (15) and functional outcome measures (FOM) (16) at the time of hospital discharge and then during each outpatient visit. Specific FOM components included locomotion, expression/communication, and self-feeding (16). Summary of pertinent scoring systems is presented in Table 1.

EP

Intraoperative adhesion scores and contamination scores were determined independently by two surgeons at the beginning of each operative procedure. The two independent scores were then

AC C

averaged. Inter-rater agreement (kappa) was also calculated for paired adhesion measurement scores. Wound sizes were determined by the attending surgeon at the beginning and at the end of each procedure. The measurement procedure consisted of wound size assessment at the level of the fascia in the abdominal midline as well as at the widest portion of the lateral extent of the open abdominal wound. The wound area was then approximated by multiplying the two measurements. The CMHAB intraoperative application protocol was standardized across all sites. Prior to participating in the study, all site investigators, study coordinators, and participating surgeons were trained in the application of CMHAB according to the procedure outlined in Figure 2. For procedures involving initial CMHAB application, pieces of adhesion barrier were placed as per protocol at the conclusion of the respective case, after a thorough abdominal washout. For procedures

ACCEPTED MANUSCRIPT

involving repeat application of the CMHAB material, any remaining adhesion barrier from the preceding operation was washed out and/or removed, the intended operative procedure performed, and the CMHAB then re-applied after a thorough abdominal washout. Minor protocol deviations were reported in 5 operative cases, consisting of (a) the use of fewer than required pieces of CMHAB

RI PT

(2 operations); (b) lack of coverage of all required anatomic areas (3 procedures). These deviations did not warrant excluding any of the patients from the study. Patients in the control group underwent abdominal operations as intended initially (i.e., prior to revealing randomization assignments), without placement of CMHAB at the conclusion of their operative procedures.

SC

Although no specific directive was imposed on participating sites with regards to the clinical management of DC/OA patients, most patients (27/30) underwent abdominal dressing changes using a commercially available negative pressure wound therapy (NPWT) device (KCI, San Antonio,

M AN U

Texas, U.S.A.). The remaining three patients had either an improvised negative pressure device placed or underwent a hybrid approach, with some procedures utilizing an improvised NPWT device and other procedures involving the commercially available device.

The study was initially powered to demonstrate significance at 20% difference in Zuhlke adhesion scores. However, a planned interim analysis performed after reaching the 25th enrollment, demonstrated that the observed adhesion score difference was much greater than the initially

TE D

anticipated 20%. Consequently, the goal study sample size was adjusted downward to approximately 10-15 patients in each group (total 25-30 patients). Given this updated information, study enrollment was stopped after reaching 30 patients. Statistical analysis included Student’s t-test for normally distributed continuous data, Mann-Whitney U-test for non-normally distributed variables. Chi-Square

EP

and Fisher’s Exact tests were utilized for categorical data, as appropriate. We utilized PASW Statistics 18 (IBM Corporation, Armonk, New York, U.S.A.) to perform study-related analyses and

AC C

Microsoft® Excel™ (Microsoft Corporation, Redmond, Washington, U.S.A.) to design scientific figures. Statistical significance was set at alpha = 0.05.

RESULTS

Thirty patients were enrolled (17 randomized to CMHAB; 13 randomized to NAB) with mean age of 40.3±16.3 years, ISS 29.6±10.3, AIS abdomen 3.68±0.94, APACHE II of 14.4±7.37, and 67% with blunt trauma mechanism. The two groups were well matched, with no differences in patient age, gender, ISS, abdominal AIS, penetrating/blunt injury rates, or initial laboratory assessments (including initial lactate/base deficit). Demographic and laboratory characteristics of the two study groups can be found in Table 2.

ACCEPTED MANUSCRIPT

Indications for damage control/open abdominal management are outlined in Table 3. Massive resuscitation (>12 liters over 24 hours) and hemodynamic instability/shock were the most common indications.

Severe abdominal contamination and planned re-operation for definitive

repairs were among less common reasons to leave the abdomen open. Of note, contamination scores

RI PT

were similar for the CMHAB and NAB groups (Figure 3).

Overall study 28-day mortality was 3/30 (10%), with 2 cases (11.8%) in the CMHAB group and 1 case (7.7%) in the NAB group (p=1.00). Of note, both mortalities in the CMHAB group were related to advanced directives/withdrawal of care and were not due to any study-related

SC

complications. There were no differences between the groups in terms of non-abdominal or abdominal (i.e., fistula, abscess, wound-related) complications. Detailed morbidity profile is presented in Table 4.

M AN U

For composite intraoperative adhesion scores, we noted no differences between the two groups during the first four operations (Figure 4a) or for approximately 1 week following the index operation (Figure 4b). Adhesion scores diverged at the 5th operative intervention or after ~1 week of OA therapy. Adhesion scores in the NAB group starting with the 5th operation were approximately 1 Zuhlke point or 67% higher than in the CMHAB group. The inter-rater agreement for paired Zuhlke scores was moderate to good at 0.71 (linear-weighted method, Std Err 0.07; 95% CI 0.57-0.84).

TE D

There were no statistical differences between wound sizes over time, closure types, or wound closure characteristics between CMHAB and NAB (Table 5). Although the CMHAB group had shorter mean hospital length of stays than the NAB group (24.7±16.6 versus 32.5±17.6) this difference was not statistically significant. The CMHAB group had

EP

significantly shorter ICU lengths of stay (14.5±8.8 versus 22.3±12.0 days, p0.05) in the NAB group. Table 7 demonstrates known adhesion scores for operations performed during the 1-year follow-up period (median, 1.5 Zuhlke, range 1-1.5). The only bowel obstruction requiring operation occurred in the NAB group. The information regarding longterm adhesion characteristics is provided for descriptive purposes only because of lack of comparison groups of meaningful size.

ACCEPTED MANUSCRIPT

Abdominal closure characteristics, including techniques and wound sizes are outlined in Table 7. Due to the lack of statistically significant differences in wound sizes or closure rates between the CMHAB and NAB groups, we combined the two groups in order to outline the

RI PT

chronologic progression of a “typical” trauma open abdomen (Figures 5a and 5b).

DISCUSSION

Trauma damage control/open abdominal (DC/OA) population continues to pose one of the greatest clinical challenges for the acute care surgeon (17, 18). Although it is associated with high injury

SC

severity, morbidity, and mortality, trauma DC/OA is known to reduce mortality in the highest acuity trauma patients (1, 19). Common sequelae of DC/OA management include abscesses, fistulae, and abdominal adhesions (2). Several factors may promote progressively intense intra-abdominal

M AN U

adhesions in this patient population, including the exposure of bowel and peritoneal surfaces to repeated manipulation and desiccation (5-7, 20), as well as abdominal contamination and intraperitoneal blood present as a result of the initial injury (6, 7). Therefore, any strategy to help reduce adhesions and allow for earlier abdominal closure is beneficial. Our results reflect previous reports on trauma damage control (1, 17, 19).

In addition, there

were no statistically significant demographic or physiologic differences between the CMHAB and

TE D

NAB groups, suggesting that any observations regarding abdominal adhesions are less likely to be influenced by factors such as injury severity or surgical indication differences between the two groups.

The severity of abdominal adhesions, the main outcome measure in our study, displayed

EP

several important trends over time. First, during the initial week of CMHAB application, there was no significant difference between the CMHAB and NAB groups. Starting after approximately 7 days

AC C

of therapy, the two groups diverge, due to significantly greater increase in adhesions in the NAB group (Figure 4b). This translates roughly into the first 4 consecutive open abdominal procedures not being associated with any noticeable increase in adhesion severity, and statistical differences between the two groups emerging only after the 4th consecutive operation (Figure 4a). Given that contamination scores were essentially identical in both groups (Figure 3), it is unlikely that abdominal contamination differentially influenced adhesion formation in either of the two groups. All of the above observations fit well with the established pathophysiology of peritoneal adhesions (20) and previously described behavior of abdominal adhesions in an experimental model (21). However, due to limited scope and resources of the current study we were unable to further determine underlying factors associated with the observed temporal behavior of adhesions. Of

ACCEPTED MANUSCRIPT

importance, we found no statistical difference in abdominal abscess formation between the two groups (Table 4), suggesting that the use of CMHAB increase risk of intra-abdominal infection or abscess. It has to be noted that the adhesion barrier used in the current study has previously been associated with sterile intra-abdominal fluid collections (22). However, it has been shown to be safe

RI PT

in the setting of abdominal and pelvic abscesses, both before and after CMHAB placement (23, 24). The limited long term results of CMHAB application are worth noting. First, adhesion severity noted during procedures performed within the 9-month follow-up period was modest in CMHAB patients (median Zuhlke score 1.5, time period range 41-244 days). In addition, the only

SC

bowel obstruction requiring operative intervention was noted in the NAB group (Zuhlke score of 3). Although the Zuhlke adhesion score (8) represents an objectivized subjective method of assessing surgical adhesions, a generalization can be made that each additional point on that adhesion severity

M AN U

scale translates to an escalation in intervention(s) needed in order to achieve adhesiolysis (Table 1). Previously published data indicate that although the overall bowel obstruction rate does not change following CMHAB use, obstructive disease requiring reoperation was significantly reduced by the use of adhesion barrier (25). Salum, et al., corroborate this finding, noting a trend toward greater success of non-operative management of adhesive bowel obstructions following the use of CMHAB (24).

TE D

We observed no statistically significant differences in rates of primary fascial closure, time to fascia closure, or wound sizes between the two study groups (Table 6). Given the similar temporal wound characteristics for both groups, we combined them in order to establish a much-needed picture of what happens with wounds in DC/OA patients over time. Such composite representation

EP

can be seen in Figure 5a, and demonstrates that approximately half of the abdominal wounds were still >50 cm2 in size (indicating incomplete wound closure) at the end of the 32-week follow-up period. Figure 5b shows the average wound size (in cm2), demonstrating that the mean wound size

AC C

decreases rapidly (i.e., from ~400 cm2 to ~125 cm2) during the first 6-8 weeks, followed by a gradual decrease in size to approximately 40-60 cm2. Average times between consecutive procedures, procedures to fascial closure, and the total number of operations (Table 7) were consistent with previously reported studies (26). The time to definitive abdominal coverage (i.e., primary fascial closure, immediate abdominal wall reconstruction, synthetic mesh closure, biologic mesh coverage) did not differ significantly between the two groups, but was shorter than previously reported ranges (26). There are important limitations of this study.

Among those is the small sample size,

reflective of the decreasing incidence of prolonged open abdominal management in this population,

ACCEPTED MANUSCRIPT

likely from more effective resuscitation strategies that evolved over the last decade. The small size of the study also limits the authors’ ability to effectively examine secondary outcomes that may ultimately prove to be significant given enough statistical power to demonstrate such differences. Although most patients in this study continued to follow-up as requested by their providers, patients

RI PT

were not required to report back for follow-up beyond what was medically necessary. This, along with three patients (one following discharge to a remote institution and two following initial outpatient clinic appointments) lost to follow-up may have affected long-term wound measurement results. In addition, while adhesion score differences between the CMHAB and NAB groups are

SC

promising, further research is needed to determine if these short-term observations of CMHAB behavior translate into longer-term benefit. This latter point is at the center of our research group’s proposed investigation into intermediate- and long-term behavior of adhesions in open abdomens.

M AN U

Open abdomens associated with penetrating abdominal trauma represent a unique patient population. In this study, one-third of all patients had penetrating injuries. Overall penetrating abdominal injury severity for this group was significant, with PATI scores >31 in both groups. Due to small sample size of the penetrating trauma group, meaningful comparisons were not possible.

CONCLUSIONS

TE D

Although CMHAB did not completely eliminate adhesions in this preliminary proof-of-concept study, it limited their severity in the setting of repeat laparotomy for damage control in open abdominal management. The effect became noticeable following 1 week of therapy or after the 4th consecutive OA procedure. There was no difference in final wound sizes, overall or abdominal

EP

complications, or mortality between patients randomized to CMHAB and NAB. Further research is warranted to better delineate potential benefits of CMHAB, especially in the setting or reoperative

AC C

surgery in post-OA patients.

ACCEPTED MANUSCRIPT

REFERENCES 1.

Cipolla, J., Stawicki, S. P., Hoff, W. S., McQuay, N., Hoey, B. A., Wainwright, G., and

Grossman, M. D. A proposed algorithm for managing the open abdomen. Am Surg 71: 202-207, 2005. Smith, B. P., Adams, R. C., Doraiswamy, V. A., Nagaraja, V., Seamon, M. J., Wisler, J.,

RI PT

2.

Cipolla, J., Sharma, R., Cook, C. H., Gunter, O. L., and Stawicki, S. P. Review of abdominal damage control and open abdomens: focus on gastrointestinal complications. J Gastrointestin Liver Dis 19: 425-435, 2010.

Lin, C. Y., Chen, W. P., Yang, L. Y., Chen, A., and Huang, T. P. Persistent transforming

SC

3.

growth factor-beta 1 expression may predict peritoneal fibrosis in CAPD patients with frequent peritonitis occurrence. Am J Nephrol 18: 513-519, 1998.

Baillie, D. R., Stawicki, S. P., Eustance, N., Warsaw, D., and Desai, D. Use of human and

M AN U

4.

porcine dermal-derived bioprostheses in complex abdominal wall reconstructions: a literature review and case report. Ostomy Wound Manage 53: 30-37, 2007. 5.

Burns, J. W., Skinner, K., Colt, J., Sheidlin, A., Bronson, R., Yaacobi, Y., and Goldberg, E.

P. Prevention of tissue injury and postsurgical adhesions by precoating tissues with hyaluronic acid solutions. J Surg Res 59: 644-652, 1995.

Duron, J. J., Silva, N. J., du Montcel, S. T., Berger, A., Muscari, F., Hennet, H., Veyrieres,

TE D

6.

M., and Hay, J. M. Adhesive postoperative small bowel obstruction: incidence and risk factors of recurrence after surgical treatment: a multicenter prospective study. Annals of surgery 244: 750-757, 2006.

Fometescu, S. G., Costache, M., Coveney, A., Oprescu, S. M., Serban, D., and Savlovschi, C.

EP

7.

Peritoneal fibrinolytic activity and adhesiogenesis. Chirurgia (Bucur) 108: 331-340, 2013. Zuhlke, H. V., Lorenz, E. M., Straub, E. M., and Savvas, V. [Pathophysiology and

AC C

8.

classification of adhesions]. Langenbecks Archiv fur Chirurgie. Supplement II, Verhandlungen der Deutschen Gesellschaft fur Chirurgie. Deutsche Gesellschaft fur Chirurgie. Kongress: 1009-1016, 1990. 9.

Mickevicius, A., Valeikaite, G., Tamelis, A., Saladzinskas, Z., Svagzdys, S., and Pavalkis, D.

Morbidity predicting factors of penetrating colon injuries. Acta chirurgica Iugoslavica 57: 55-58, 2010. 10.

Morales, C. H., Villegas, M. I., Villavicencio, R., Gonzalez, G., Perez, L. F., Pena, A. M.,

and Vanegas, L. E. Intra-abdominal infection in patients with abdominal trauma. Arch Surg 139: 1278-1285; discussion 1285, 2004.

ACCEPTED MANUSCRIPT

11.

Rutledge, R., Fakhry, S., Rutherford, E., Muakkassa, F., and Meyer, A. Comparison of

APACHE II, Trauma Score, and Injury Severity Score as predictors of outcome in critically injured trauma patients. American journal of surgery 166: 244-247, 1993. 12.

Rutledge, R., Fakhry, S. M., Rutherford, E. J., Muakkassa, F., Baker, C. C., Koruda, M., and

RI PT

Meyer, A. A. Acute Physiology and Chronic Health Evaluation (APACHE II) score and outcome in the surgical intensive care unit: an analysis of multiple intervention and outcome variables in 1,238 patients. Crit Care Med 19: 1048-1053, 1991. 13.

Le Gall, J. R., Lemeshow, S., and Saulnier, F. A new Simplified Acute Physiology Score

American Medical Association 270: 2957-2963, 1993. 14.

SC

(SAPS II) based on a European/North American multicenter study. JAMA : the journal of the

Pal, J., Brown, R., and Fleiszer, D. The value of the Glasgow Coma Scale and Injury Severity

15.

M AN U

Score: predicting outcome in multiple trauma patients with head injury. J Trauma 29: 746-748, 1989. Teasdale, G. M., Pettigrew, L. E., Wilson, J. T., Murray, G., and Jennett, B. Analyzing

outcome of treatment of severe head injury: a review and update on advancing the use of the Glasgow Outcome Scale. Journal of neurotrauma 15: 587-597, 1998. 16.

Ohio trauma registry public records request packet. In.

17.

Rotondo, M. F., Schwab, C. W., McGonigal, M. D., Phillips, G. R., 3rd, Fruchterman, T. M.,

TE D

Kauder, D. R., Latenser, B. A., and Angood, P. A. 'Damage control': an approach for improved survival in exsanguinating penetrating abdominal injury. J Trauma 35: 375-382; discussion 382-373, 1993. 18.

Stawicki, S. P., Brooks, A., Bilski, T., Scaff, D., Gupta, R., Schwab, C. W., and Gracias, V.

93-101, 2008.

Stawicki, S. P., Cipolla, J., and Bria, C. Comparison of open abdomens in non-trauma and

AC C

19.

EP

H. The concept of damage control: extending the paradigm to emergency general surgery. Injury 39:

trauma patients: A retrospective study. OPUS 12 Scientist 1: 1-8, 2007. 20.

Buckman, R. F., Woods, M., Sargent, L., and Gervin, A. S. A unifying pathogenetic

mechanism in the etiology of intraperitoneal adhesions. The Journal of surgical research 20: 1-5, 1976. 21.

Reijnen, M. M., Meis, J. F., Postma, V. A., and van Goor, H. Prevention of intra-abdominal

abscesses and adhesions using a hyaluronic acid solution in a rat peritonitis model. Arch Surg 134: 997-1001, 1999. 22.

Tyler, J. A., McDermott, D., and Levoyer, T. Sterile intra-abdominal fluid collection

associated with seprafilm use. Am Surg 74: 1107-1110, 2008.

ACCEPTED MANUSCRIPT

23.

Beck, D. E., Cohen, Z., Fleshman, J. W., Kaufman, H. S., van Goor, H., and Wolff, B. G. A

prospective, randomized, multicenter, controlled study of the safety of Seprafilm adhesion barrier in abdominopelvic surgery of the intestine. Diseases of the colon and rectum 46: 1310-1319, 2003. 24.

Salum, M. R., Lam, D. T., Wexner, S. D., Pikarsky, A., Baig, M. K., Weiss, E. G., Nogueras,

RI PT

J. J., and Singh, J. J. Does limited placement of bioresorbable membrane of modified sodium hyaluronate and carboxymethylcellulose (Seprafilm) have possible short-term beneficial impact? Diseases of the colon and rectum 44: 706-712, 2001. 25.

Fazio, V. W., Cohen, Z., Fleshman, J. W., van Goor, H., Bauer, J. J., Wolff, B. G., Corman,

SC

M., Beart, R. W., Jr., Wexner, S. D., Becker, J. M., Monson, J. R., Kaufman, H. S., Beck, D. E., Bailey, H. R., Ludwig, K. A., Stamos, M. J., Darzi, A., Bleday, R., Dorazio, R., Madoff, R. D., Smith, L. E., Gearhart, S., Lillemoe, K., and Gohl, J. Reduction in adhesive small-bowel obstruction

M AN U

by Seprafilm adhesion barrier after intestinal resection. Diseases of the colon and rectum 49: 1-11, 2006. 26.

Stawicki, S. P., and Grossman, M. "Stretching" negative pressure wound therapy: Can

dressing change interval be extended in patients with open abdomens? Ostomy Wound Manage 53:

AC C

EP

TE D

26-29, 2007.

ACCEPTED MANUSCRIPT

Table 1. Scoring systems utilized in this study, including: (a) the Zuhlke classification of adhesion severity; (b) the abdominal contamination score; and (c) the Glasgow Outcome Score. Zuhlke Adhesion Score 1 2 3 4

SC

RI PT

Filmy adhesions, easy to separate by blunt dissection Stronger adhesions; blunt dissection possible, partly sharp dissection necessary; beginning of vascularization Strong adhesions; lysis possible by sharp dissection only; clear vascularization Very strong adhesions; lysis possible by sharp dissection only; organs strongly attached with severe adhesions; damage to organs hardly preventable

Peritoneal/Abdominal Contamination Score* 0 1

M AN U

No gross contamination observed Cloudy fluid seen in the peritoneal cavity upon opening the abdomen, no particulate matter, no fistulae seen Criteria as in #1 plus enteric fistula, leak and/or abscess with localized contamination confined to no more than two quadrants of the abdomen Criteria as in #1 plus enteric fistula, leak and/or abscess with diffuse contamination of the peritoneal cavity

2 3

Glasgow Outcome Score

TE D

*Adapted, modified & compiled from Mickevicius, et al (2010) and Morales, et al (2004).

Term

1

Dead

No life

2

Vegetative state

Unaware of self and environment

4 5

AC C

3

EP

Level

Definition

Severe disability

Unable to live independently

Moderate disability

Able to live independently

Mild disability

Able to return to work/school

ACCEPTED MANUSCRIPT

Table 2. Characteristics of Seprafilm and Control groups at the time of hospital presentation

(n=17)

(n=13)

13/17 40.4 ± 16.7 28.3 ± 10.4 3.69 ± 0.79 2.60 ± 1.26 3.13 ± 1.13 2.83 ± 1.17 6/17 37 (12-49) 11.2 ± 4.4 13.8±7.33 30.1 ± 13.5

10/13 40.2 ± 16.4 31.3 ± 10.3 3.67 ± 1.15 2.67 ± 0.82 2.88 ± 0.64 3.00 ± 1.53 4/13 31.5 (13-50) 11.8 ± 4.3 15.2±7.62 33.5 ± 14.5

AC C

EP

TE D

Physiologic & Laboratory Parameters Initial pH 7.25±0.10 Initial Serum Lactate 5.00 ± 2.10 Base Excess -7.33 ± 5.48 WBC Count 10.9 ± 5.32 Hemoglobin 12.1 ± 2.03 Hematocrit 36.0 ± 6.34 Platelet Count 177 ± 88.2 Serum Sodium 141 ± 3.21 Serum Chloride 108 ± 4.43 Serum Potassium 3.89 ± 0.50 Serum Bicarbonate 21.6 ± 4.59 Blood Urea Nitrogen 12.6 ± 4.68 Serum Creatinine 1.07 ± 0.41 AST / ALT 310 / 183 T Bili / D Bili 1.10 / 0.45 Alkaline Phosphatase 57.5 ± 15.4

P-value

n/s (p=1.00) n/s (p=0.98) n/s (p=0.44) n/s (p=0.96) n/s (p=0.91) n/s (p=0.59) n/s (p=0.83) n/s (p=1.00) n/s (p=0.46) n/s (p=0.68) n/s (p=0.60) n/s (p=0.51)

RI PT

NAB

SC

General Characteristics Gender (Male) Age (Years) Injury Severity Score Abdominal AIS Extremity AIS Chest AIS Head & Neck AIS Penetrating Injury Median PATI Score GCS Score APACHE II SAPS II

CMHAB

M AN U

Parameter

7.20 ± 0.12 3.81 ± 2.00 -9.22 ± 5.24 9.72 ± 4.86 10.7 ± 3.05 31.8 ± 8.79 153 ± 64.5 140 ± 5.44 109 ± 5.85 4.03 ± 1.14 21.1 ± 0.90 12.2 ± 3.79 1.07 ± 0.30 766 / 531 1.27 / 0.50 52.3 ± 10.7

n/s (p=0.26) n/s (p=0.32) n/s (p=0.51) n/s (p=0.56) n/s (p=0.14) n/s (p=0.14) n/s (p=0.43) n/s (p=0.52) n/s (p=0.78) n/s (p=0.68) n/s (p=0.72) n/s (p=0.82) n/s (p=0.99) n/s n/s n/s (p-0.57)

ACCEPTED MANUSCRIPT

Table 3. Clinical indications for initiation of damage control/open abdominal management. CMHAB

NAB

(n=17)

(n=13)

Hemodynamic instability/Shock

7

9

Lethal triad

5

5

Severe abdominal contamination*

4

1

Inability to definitely repair all injuries 3

4

Massive resuscitation/Tissue edema

9

5

Planned reoperation

4

5

Abdominal compartment syndrome

6

3

3

0

n/s (p=0.24)

7

4

n/s (p=0.71)

RI PT

Reasons for Damage Control Listed etiology

Bleeding requiring packing

n/s (p=0.16) n/s (p=0.71)

SC

n/s (p=0.35)

M AN U

Abdominal tissue loss

P-value

n/s (p=0.67)

n/s (p=0.48) n/s (p=0.44) n/s (p=0.69)

AC C

EP

TE D

* Defined as abdominal contamination score of 2 or greater (see Table 1 for scoring details)

ACCEPTED MANUSCRIPT

Table 4. Complications noted during the course of damage control/open abdominal management. CMHAB (n=17)

NAB (n=13)

P-value

n/s (p=0.71) n/s (p=0.36)

TE D

M AN U

Abdominal Biliary drain dislodgement 1 Abdominal abscess 3 1 Hepatic necrosis, segmental 1 Omentalischemia/necrosis 1 2 Leak/fistulab 3 6 b Includes biliary, pancreatic, gastric, small bowel, large bowel Ostomy complication 1 Abdwall/wound complicationc 5 3 c Includes abscess, necrosis, other infection, or dehiscence 2*

Mortality

1d

n/s (p=0.18) n/s (p=1.00) n/s (p=1.00) n/s (p=1.00) n/s (p=1.00)

SC

Non-abdominal Respiratory failure 5 5 Other pulmonarya 2 4 a Atelectasis, pleural effusion, pneumonia Cardiac arrest (injury-related) 2 Sepsis 4 3 Neurologic 2 2 End-organ failure 3 2 Deep vein thrombosis 1 -

RI PT

Inpatient Complications

n/s (p=0.43) n/s (p=0.61) n/s (p=0.43) n/s (p=0.56) n/s (p=0.12) n/s (p=1.00) n/s (p=1.00)

n/s (p=1.00)

EP

*Both cases involved withdrawal of care per previously stated patient/family wishes; One patient with devastating neurologic injury was transferred to hospice d Patient died following abdominal closure

AC C

Outpatient Complications† Severe/persistent pain 4 3 Leak/fistulaa 2 2 a Includes biliary, pancreatic, enteric Malnutrition/Weight loss 2 3 Bowel obstruction 1* *Required laparotomy and lysis of adhesions Constipation 1 Wound complication 2 1 Incisional hernia 1 Physical deconditioning 1 1 Nausea/Vomiting 1 1

n/s n/s n/s n/s n/s n/s n/s n/s n/s

†Data limited by patient compliance with follow-up visits; Incisional hernias only declared if definitive (primary fascial or mesh-based) closure performed (n=22); Wounds remaining open were termed “planned ventral hernias”

ACCEPTED MANUSCRIPT

Table 5. Abdominal closure characteristics, including wound sizes and closure techniques NAB (n=13)

425 281 -144

408 171 -237

p-value

RI PT

Wound Sizes (cm2) Mean Initial Wound Size Mean Final Wound Size Mean Wound Size Decrease

CMHAB (n=17)

n/s (p=0.78) [u-test] n/s (p=0.32) [u-test] n/s (p=0.38) [u-test]

M AN U

SC

Operative Details Abdominal operations 4.18 ± 1.67 3.92 ± 1.66 n/s (p=0.68) Procedures to fascial closure 2.86 (n=7) 3.78 (n=9) n/s (p=0.20) Days to definitive coverage† 9.62 ± 5.87 8.82 ± 5.32 n/s (p=0.71) Median [Range] 7 [2-23] 8 [2-18] Mean days between procedures 2.83 ± 1.46 2.78 ± 2.20 n/s (p=0.89) Combined groups mean ± StDev 2.81 ± 1.83 Combined groups median [Range] 2.00 [1-8] † Included primary fascial closures, immediate abdominal wall reconstructions, synthetic mesh closures, and biologic mesh coverage procedures

Primary fascial closure

TE D

Abdominal wound closure 7

9

3

-

Immediate abd wall reconstruction

1

1

Vicryl mesh ± STSG

2

2

2

1

2

-

339 113.9 ± 92.3 89 (23-339)

215 94.3 ± 60.6 73.5 (23-215)

Bioprosthetic material

Abdomen not closed*

EP

Combined modality approach

n/s (p=0.56)

AC C

*Due to withdrawal of care Follow-up Period Last reported visit (days) Mean ± S.D. Median (Range)

n/s (p=0.56) n/s (p=0.47)

ACCEPTED MANUSCRIPT

Table 5. Discharge and disposition data CMHAB (n=17)

NAB (n=13)

P-value

Glasgow Coma Scale

14.4 ± 1.55

13.5 ± 3.29

n/s (p=0.32)

Glasgow Outcome Score

4.07 ± 0.80

3.40 ± 0.97

n/s (p=0.07)

FOM-Self Feeding

2.93 ± 1.44

3.25 ± 1.14

FOM-Locomotion

2.60 ± 1.24

2.25 ± 0.87

FOM-Expression/Communication

3.47 ± 1.13

3.45 ± 1.21

Home

6

2

Rehab

6

6

Long-term acute care

EP

n/s (p=0.54)

n/s (p=0.42) n/s (p=0.98)

n/s (p=0.41) n/s (p=0.71)

-

2

n/s (p=0.18)

3

2

n/s (p=1.00)

TE D

Nursing home

AC C

SC

M AN U

Discharge disposition

RI PT

Discharge parameters

ACCEPTED MANUSCRIPT

Table 6. Abdominal operations reported during late follow-up period Description of operation

Average Zuhlke score (post-injury day)

CMHAB group 1 (41)

Laparotomy, lysis of adhesions, takedown of enterocutaneous fistula, primary fascial closure

1.5 (216)

Laparotomy, lysis of adhesions, colostomy revision, creation of ileal conduit

1.5 (244)

NAB group

EP

TE D

M AN U

Exploratory laparotomy, lysis of adhesions

1.5 (145)

SC

Laparotomy, lysis of adhesions, colostomy take-down

AC C

RI PT

Laparotomy, takedown of colo-cutaneous fistula

3 (76)

ACCEPTED MANUSCRIPT

Figure 1. Diagram outlining patient screening and recruitment process during the active enrollment period (March 5, 2010 – September 30, 2013).

RI PT

Screened Trauma Open Abdomen Cases (March 2010-September 2013)

76

Not Approached

SC

Abdomen Closed Prior to Eligibility Determination Completed

12

M AN U

10 Prisoner Status

7

Age 48 Hours

AC C

EP

6

9

Enrolled in Study

30

Randomized to CMHAB

Randomized to NAB

17

13

ACCEPTED MANUSCRIPT

Figure 2. Standardized protocol for application of carboxymethylcellulose sodium hyaluronate adhesion barrier (CMHAB). 1. The CMHAB is applied prior to concluding the operative procedure in the following anatomic areas: (a) colic gutters; (b) the pelvis; (c) perihepatic area; and (d) anterior surfaces of exposed

RI PT

small bowel loops.

2. It is critically important that CMHAB material not be placed directly over anastomotic or bowel repair areas. Previous studies suggest that placement of CMHAB directly over anastomotic areas (i.e., wrapping) may lead to increased rates of anastomotic breakdown.

3. If there is an anastomosis or a bowel repair in an area that would normally be covered with

SC

CMHAB as per protocol, then the surgeon should place the CMHAB piece so that its closest extent is at least 1 inch away from the bowel anastomosis and/or repair.

M AN U

4. In terms of general principles of anatomic placement of CMHAB, as well as precise handling instructions pertaining to the research protocol, please refer to Figure 2b below. 5. Intraoperative pre-processing and application of the CMHAB material: [A] Each CMHAB sheet supplied for this study will be approximately 5 x 6 inches in size.

AC C

EP

TE D

For study purposes, each sheet will be cut in half




ACCEPTED MANUSCRIPT

[B] Following that, each smaller sheet should be approximately 3 x 5 inches in size. These

M AN U

SC

RI PT

pieces will then be applied to the following anatomic areas

TE D

(a) Two CMHAB pieces between the liver and the anterior abdominal wall; (b) Four pieces over the exposed bowel surfaces anteriorly; (c) Two slightly staggered pieces of CMHAB in each colic gutter; and (d) Two pieces in the pelvic area. If any of the above areas involve an anastomosis or bowel repair, then the CMHAB should be placed at least 1 inch away from the

AC C

EP

anastomosis and/or bowel repair.

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Figure 3. Contamination scores for CMHAB and NAB groups. Red line represents CMHAB group and the black line represents NAB group. Values are shown as Mean ± StdErr. None of the differences are statistically significant.

ACCEPTED MANUSCRIPT

Figure 4. Zuhlke adhesion severity scores for CMHAB and NAB groups. [A] Adhesion scores grouped by consecutive procedure number; [B] Adhesion scores grouped by chronology. Values shown represent Mean ± StdErr; *Denotes statistically significant difference.

TE D

M AN U

SC

RI PT

[A]

AC C

EP

[B]

ACCEPTED MANUSCRIPT

RI PT

Figure 5. Composite wound size characteristics up to 32+ weeks for the combined DC/OA sample from the current study. [A] Red represents proportion of wounds with area >30 cm2 and black represents proportion of wounds with area >50 cm2; [B] Composite representation of approximate wound areas (cm2) for the combined DC/OA patient group. Values shown represent Mean ± StdErr. Data based on all available follow-up visit measurements.

TE D

M AN U

SC

[A]

AC C

EP

[B]