Abdominal Wall Hernias in Patients With Abdominal Aortic Aneurysmal Versus Aortoiliac Occlusive Disease Kevin A. Hall, MD, Brian Peters, MD, Stephen H. Smyth, MD, James A. Warneke, MD, William D. Rappaport, MD, Charles W. Putnam, MD, Glenn C. Hunter, MD, Tucson,Arizona
BACKGROUND: This study was undertaken to determine the incidence of ventral incisional hernias (VlHs) and inguinal hernias (IHs) in patients with abdominal aortic aneurysmal (AAA) versus those with aortoiliac occlusive disease (AIOD). PATIENTS AND METHODS: The medical records of 193 patients (128 with AAA and 65 with AIOD) who had undergone elective aortic reconstruction were reviewed to determine the number and location of abdominal wall hernias (AWHs). RESULTS: Forty-one AWHs (28 IHs and 13 VIHs) were detected in patients with AAA compared to 13 (11 IHs and 2 VIHs) in patients with AIOD. There was a significantly greater incidence of VlHs in patients with AAA versus patients with AIOD (10% versus 3%, P ~0.05) and recurrent AWHs (28% versus 19%, P
number of causal factors, including obesity, age, malnutrition, preexisting systemic or malignant disease, postoperative distension, site of incision, suture material, nerve injury, wound dehiscence, and infection, have been implicated in the etiology of ventral incisional hernias (VIHS).*-~ A history of wound infection can be elicited in 39% to 75% of patients diagnosed with VIHs within the first year of surgery+3-5 There is increasing evidence that inguinal hernias (IHs) and VIHs occur more commonly in patients with abdominal aortic aneurysms. 6-8While many of the patient-related
A
From the Departments of Surgery (KAH, BP, JAW, WDR, CWP, GCH) and Radiology (SHS), University of Arizona Health Sciences Center, Tucson, Arizona. Requests for reprints should be addressed to Glenn C. Hunter, MD, Associate Professor of Surgery, University of Arizona Health Sciences Center, Department of Surgery, Room 5409,150l North Campbell Avenue, Tucson, Arizona 85725. Presented at the 47th Annual Meeting of the Southwestern Surgical Congress, San Antonio, Texas, April 28-26, 1995.
risk factors for VIH may be present in patients undergoing bypass grafting for aortoiliac occlusive disease (AIOD) and abdominal aortic aneurysmal disease @AA), the most common, wound infection, is infrequent, occurring in approxi, mately 5% of patients undergoing aortic recon.struction.9 This suggests that etiologic factors other than infection may contribute to ventral hernias in aneurysm patients. This study was undertaken to determine the incidence of abdominal wall hernias (AWHs) in patients with AA4 versus those with AIOD. PATIENTS AND METHODS The medical records of 193 consecutive patients undergoing elective aortic resection for aneurysmal or bypass grafting for occlusive disease were reviewed. Patient demographics and risk factors associated with hernias were recorded. Only patients in whom a midline incision was performed were included. Additional review of all AWH repairs over the most recent 5year period was undertaken to more accurately describe the number of abdominal hernias in patients with AAA. All wounds were closed using a mass suture technique. In obese patients, the skin and subcutaneous tissue were dissected from the fascia and the running sutures placed 1 cm apart and approximately 2 cm from the margins of the incision. During the first few years of the study, the fascia was closed with no. 2 Dexon (polyglycolic acid; Davis & Geck, Danbury, Connecticut) or Vicryl (polygalactin 910; Ethicon, Somerville, New Jersey) suture. Thereafter, fascial closures were performed using either no. 2 Prolene (polypropylene; Ethicon) or Novafil (polybutester; Davis & Geck) suture. A sample of abdominal computed tomography (CT) scans, performed as part of our routine surveillance of aortic grafts, were reviewed in order to assist in identifying the morphologic characteristics of the ventral abdominal wall defects. Periumbilical hernias were repaired primarily. The larger diffuse defects were repaired using Marlex mesh (CR Bard, Billerica, Massachusetts) or a polytetrafluoroethylene (PTFE) patch. Closed-suction drainage was usually employed for 24 to 48 hours, and all patients received perioperative antibiotics. Statistical Analysis Data are expressed as mean * standard error of the mean and were analyzed using analysis of variance and chi-square testing. RESULTS One hundred ninety-three patients, 128 with AAAs and 65 with AIOD were evaluated. Except for a higher incidence of diabetes in patients with occlusive disease and a predomI
572
THE AMERICAN
JOURNAL
OF SURGERY@
VOLUME
170
DECEMBER
1995
ABDOMINAL WALL HERNIAS/HALL ET AL TABLE Patient-Related Risk Factors for Abdominal Wall Hernlas AM (K) (n=128) Sex (male) Hypertension Diabetes mellitus Myocardial infarction COPD Smoking
102 70 6 41 33 96
(80) (55) (5) (32) (26) (75)
AIOD (%) (n=85) 37 (57) 37 (57) 12 (18) 23 (35) 15 (23) 51 (78)
AAA = abdom,na/ aortic aneurysmal disease; AlOD = aortoiliac occlusive disease: COP0 = chronic obstructive pulmonary disease.
inance of aneurysms in male patients, there were no significant differences in risk factors between patients with AAA and AIOD (Table). Only 3 patients developed wound infections (1.6%); another 2 patients were operated on for repair of early wound dehiscence ( 1%). Forty-one AWHs (28 IHs and 13 VIHs) were detected in the 128 patients with AAA compared to 13 (11 IHs and 2 VIHs) among the 65 patients with AIOD (32% versus 20%, P ~0.05). Patients with AAA had a significantly greater (P ~0.05) number of VIHs than those with AIOD (10% versus 3%); there was no difference in the incidence of IHs (22% versus 17%) between the two groups. The mean follow-up was significantly greater for patients having AAA repair (562 + 79 days) compared to those with AIOD (442 i 60 days, P ~0.03). Clinically, two distinct varieties of VIHs were observed. Focal, circumscribed defects in the region of the umbilicus were present in 5 patients (3 with AAA and 2 with AIOD). Most patients, however, presented with a diffuse bulging of the abdominal wall extending from the xiphistemum to the umbilicus. Discomfort localized to the region of the hernia was the most common presenting symptom. None of the patients presented with obstruction or incarceration. At surgery, the fascia in the subgroup of patients with diffuse bulging defects was markedly attenuated. The running suture was almost invariably found to be intact but loosened. Three patients have had more than two attempts at repair of their VIH. During the last 5 years, 28.5% (53 of 186) of all AWH repairs (37 ventral, 14 inguinal, 2 umbilical) were performed in the 147 individuals who had undergone aortic resection for aneurysmal disease. Twenty-seven percent (10 of 37) of the ventral and 36% (5 of 14) of IHs repaired were recurrent. The 28.3% (15 of 53) recurrence rate for AWHs in patients with AAA was statistically greater than the 18.8% (25 of 133) for the remaining 133 patients (P ~0.01). Computed Tomography
Findings
Incisional hernias containing bowel or omentum, all located above the umbilicus, were detected in 10% (6 of 61) of CT scans. Whereas the rectus muscles were closely approximated in patients with well-healed incisions, one or both rectus muscles were displaced from the midline with the intervening space occupied principally by attenuated fascia and skin in those patients with hernias. A well-
Figure 1. Abdominal computed tomography scan demonstrating a well-healed incision in a patient with an end-to-side bypass for aortoiliac occlusive disease.
Figure 2. Abdominal computed tomography scan demonstrating a typical ventral incisional hernia containing a loop of bowel in a patient following abdominal aortic aneurysm repair.
healed incision in a patient with occlusive disease is shown in Figure 1, and a characteristic VIH in an AAA patient is demonstrated in Figure 2.
COMMENTS Midline and oblique incisions are the most commonly used approaches for exposure of the abdominal aorta. Although transverse incisions are associated with the lowest complication rate, their use is often limited to those patients with pulmonary insufficiency.5 Despite the benefits of oblique incisions, a significant number of late wound complications have been reported; for example, diffuse bulging in 11% to 23% of cases and incisional hernias in ~%.~x~,~JO These complications usually occur without a history of wound infection or dehiscence. Presumably, the diffuse bulging is the result of degenerative atrophy resulting from injury to the subcostal and intercostal nerves or their branches. Gardner et al8 were
THE AMERICAN JOURNAL OF SURGERY@ VOLUME 170 DECEMBER 1995
I
573
\BDOMINAL
WALL HERNIASMALL
ET AL IJ
able to reduce the incidence of bulging from 11% to 0.03% by preserving the branches of the 1 lth intercostal nerve. Unlike oblique incisions, where the occurrence of abdominal wall bulging and hemiation has been directly related to injury of the nerves supplying the muscles of the abdominal wall, injury to cutaneous nerves cannot readily be implicated in the occurrence of VIH in patients with midline incisions. There are few reports in the literature addressing the incidence of VIHs in patients with AAA. In this report, VIHs were detected in 10% of AAA patients compared to 3% in those with AIOD. This incidence is considerably lower than that reported by Stevick et al7 (37% versus 10%); however, the hernias reported in the latter study were small epigastric hernias, 3 to 8 cm in diameter, which is a different morphology from what we found. In both studies, patients with AAA were at increased risk for developing VIH. Two distinct types of VIH were identified. Focal defects, adjacent to the umbilicus, were present in only 5 patients and diffuse bulging in the remainder. The less frequent, focal periumb&al defects appeared to be the result of poor technique when using a running closure. The diffuse defects, manifest by attenuation of the intervening fascia and displacement of the rectus muscles, are more difficult to prevent and more likely to recur. Hernias evident within the first year after surgery are usually associated with the well-established causal factors3 Why patients undergoing AAA repair should have an increased predisposition to VIH is not fully understood. Wound infection, the most frequently associated risk factor in patients without aneurysms, was present in only 3 patients (1.6%). Experimentally, fascial wounds reach their maximum tensile strength at 200 to 300 days compared to 14 to 21 days for skin1iJ2 Therefore, sutures of appropriate durability and tensile strength should be selected when closing these w~unds.~~ Although we used braided absorbable suture (Dexon and Vicryl) in the ‘early part of the study, it appears unlikely that the type of suture material alone was a major contributing factor. Both braided absorbable and monofilament nonabsorbable sutures have been shown to be equally efficacious in closing midline incisions in prospective studies.14J5 Furthermore, the incidence of VIH has remained unchanged despite the use of nonabsorbable monofilament suture. There continues to be a debate regarding the use of single or layered closure of abdominal wall incisions. In experimental animals, fascial wounds closed in a single layer have less tensile strength at 8 days than those closed in 2 layers; however, this difference was no longer apparent once the wounds were fully healed.12 Although mass suturing of the musculoaponeurotic layers of the abdominal wall is associated with a low incidence of wound dehiscence, the incidence of late incisional hernias remains at about 10%.15J6It is commonly believed that the tension used to approximate rhe layers of the abdominal wall, using the mass closure technique, may cause mechanical or ischemic injury to the tissues contributing to the development of AWHs. Mayer et all5 have reported a 10% incidence of incisional hernias in wounds closed under normal tension compared to 5.5% in wounds tightly sutured, suggesting that some tension may be necessary for primary healing. Primary repair with interrupted nonabsorbable monofilament suture may be appropriate in patients with small de574
THE AMERICAN
JOURNAL
OF SURGERY@
VOLUME 170
fects. Although none of the periumbilical defects closed using this technique in our study have recurred, recurrence rates of 2% to 10% have been reported.’ Although some larger defects may be closed primarily using fascial release incisions as in the Keel repair, prosthetic material such as Marlex mesh or PTFE is usually necessary to reinforce defects too large for primary closure.16-20 Recurrence rates after primary repair of incisional hernias are reported from 30% to 50%; Sitzmann and McFadden2’ have reported a 2.5% recurrence rate in patients using internal retention sutures. The recurrence rates after repairs using mesh is approximately 10%.16 There was no statistical difference in the incidence of II-Is in patients with AAA compared to those with AIOD (22% versus 17%). This finding is contrary to that of Cannon et al6 who found a greater number of IHs (26% versus 15%) in patients with AAA. The higher incidence of IHs in patients with AAA observed by Cannon et al6 may possibly be attributed to the larger number of patients evaluated in their study. The increased incidence of incisional and recurrent hernias in patients with AAA is difficult to attribute to the usual causal risk factors. Defects in collagen synthesis and degradation have been reported in patients with IHs and AAA.22~23 Friedman et alZZhave demonstrated an increase in type III collagen and a corresponding lower al (l):al(l 11) collagen ratio in patients with IH compared to controls. Cannon et al6 have demonstrated an increase in proteolytic activity in AAA patients who smoke compared to those with Leriche’s syndrome, and they postulate that the increase in proteolytic activity may contribute to the greater frequency of IHs in patients with AAA. Ventral hernias were detected on CT scans in 10% of the patients undergoing graft surveillance. While we do not advocate the use of CT scans to diagnose incisional hernias, review of the scans did allow us to define the anatomic features of these defects. The etiology of IHs and incisional hernias in patients with AAA is undoubtedly multifactorial. While operative technique and patient-related risk factors certainly contribute to the development of these defects, the available evidence sug gests that the increased incidence of such hernias in patients with AAA may reflect a systemic defect in collagen synthesis, degradation, or both. Further prospective studies are necessary to more clearly define the underlying mechanisms predisposing patients with AAA to AWHs. REFERENCES 1. Santora TA, Rosyln JJ. Incisional hernia. Surg CIin North Am. 1993; 73:557-571. 2. Liakakos T, Karanikas I, Panagiotidis H, Dendrinos S. Use of Marlex mesh in the repair of recurrent incisional hernias. Br_I Surg. 1994;81: 248-249. 3. Ellis H, Gajraj H, George CD. Incisional hernias: when do they occur? Br .I Surg. 1983;70:290-291. 4. Buchnall TE, Cox PJ, Harold E. Burst abdominal and incisional hernia. A prospective study of 1129 major laparotomies. BMJ. 1982;284: 931-933. 5. Larson GM, Vandertoll DJ. Approaches to repair of ventral hernia and full-thickness losses of the abdominal wall. Surg CIin North Am. 1984;64:335-349. 6. Cannon DJ, Lasteel L, Read RC. Abdominal aortic aneurysm, DECEMBER
1995
-
1 ABDOMINAL
WALL HERNIAS/HALL
ET AL
I
Leriche’s syndrome, inguinal hemiation, and smoking. Arch Surg. 1984;119:387-389. 7. Stevick CA, Long JB, Jamasbi B, Nash M. Ventral hernia following abdominal aortic reconstruction. Am Surg. 1988;54:287-289. 8. Gardner GP, Josephs LG, Rosca M, et al. The retroperitoneal incision. Arch Surg. 1994;129:753-756. 9. Szilagyi DE, Smith RF, Elliott JP, Vrandecic MP. Infection in tierial reconstruction with synthetic grafts. Ann Surg. 1972;176:321-333. 10. Honig MP, Mason RA, Giron F. Wound complications of the retroperitoneal approach to the aorta and iliac vessels. J Vast Surg. 1992;15:28-34. 11. Douglas DM. The healing of aponeurotic incisions. Br ] Surg. 1952;40:79-84. 12. Dudley AF. Layered and mass closure of the abdominal wall: a theoretical and experimental analysis. Br J Surg. 1970;57:664-667. 13. Goligher JC, Irvin TT, Johnston D, et al. A controlled clinical trial of three methods of closure of laparotomy wounds. BrJ Surg. 1975; 62~823429. 14. Irvin TT, Koffman CG, Duthi HL. Laye; closure of laparotomy wounds with absorbable and non-absorbable suture materials. BrJ Surg. 1976;63:793-796. 15. Mayer AD, Ausobsky JR, Evans M, Pollock AV. Compression suture of the abdominal wall: a controlled trial in 302 major laparotomies. Br J Surg. 1981;68:632-634. 16. Langer S, Christiansen J. Long-term results after incisional hernia repair. Actn Chir SC&. 1985;151:217-219. 17. Pless J, Lontoft E. Giant ventral hernias and their repair. Scami .f Pht Reconsn Surg Hand Surg. 1984;18:209-213. 18. Bauer JJ, Salky BA, Gelemt IM, Kreel I. Repair of large abdominal wall defects with expanded polytetrafluoroethylene (PTFE). Ann Surg. 1987;206:765-769. 19. Rubio PA. New technique for repairing large ventral incisional hernias with Marlex mesh. Surg Gynecol O&t. 1986;102:275-276. 20. Read RC, Yoder G. Recent trends in the management of incisional hemiation. Arch Surg. 1989;124:485-+88. 21. Sitzmann JV, McFadden DW. Internal retention repair of massive ventral hernia. Am Surg. 1989;55:719-723. 22. Friedman DW, Boyd CD, Norton P, et al. Increases in type III collagen gene expression and protein synthesis in patients with inguinal hernias. Ann Surg. 1993;218:754-760. 23. Powell J, Greenhalgh RM. Cellular, enzymatic and genetic factors in the pathogenesis of abdominal aortic aneurysms. J Vast Surg. 1989;9: 297-304.
DISCUSSION James Thomas, MD (Kansas City, Kansas): As I am sure most of you are aware, several years ago Dr. Raymond Reed, former president of this organization, suggested that there was a relationship between cigarette smoking, abdominal aortic aneurysms, incisional hernias, and COPD (chronic obstructive pulmonary disease) or emphysema. These authors today have reported, although in a retrospective manner, a higher incidence of ventral and recurrent ventral hernias in patients with abdominal aortic aneurysms compared to the incidence of these entities in patients with aortic occlusive disease. These data substantiate those theories suggested by a number of clinicians, including our former President, Dr. Reed. There has been a suggestion in the literature that the relationship between abdominal aortic aneurysms and the development of hernias is based upon abnormalities of collagen, elastase, and alpha,-antitrypsin. Since, in our practice at the University of Kansas, we have not noted this high incidence of incisional hernias, I have a number of questions for the authors. I am particularly conTHE AMERICAN
cemed about the methodology of closure of the abdomen. We continue, at this time, to use interrupted nonabsorbable materials to close the abdomen. In contrast, I noted that a number of people in this paper had running closures with ahsorbable material. I would ask the authors what number of people? What percentage of those individuals who had recurrent or primary hernias had absorbable materials used in a running manner for closure of the abdominal wall? And, in addition, I would ask the authors to describe in some detail their technique of closing the abdominal wall, citing a number of complications associated with its use. I wonder, in this particular group of patients, was cautery typically or routinely used to open the abdominal cavity? Finally, were any tissue or other levels of collagen, elastase, or alpha,-antitrypsin measured in those individuals in particular who had recurrence of their ventral hernias! Lastly, I would like to note that this is a retrospective study, and it has been repeatedly noted a number of times earlier that it is difficult to confirm these data in that manner. I would suggest that the data we can really rely on would best be collected in a prospective, controlled manner in which the methodology for abdominal wound closure was standardized. Jim Chandler, MD (Boulder, Colorado): It is a bit unfair to comment with reference to a comment, but I am going to do it anyway. It may sound like heresy, but a retrospective assessment of this particular subject may be less bias vulnerable than a prospective study. We can be confident that these wounds were closed with equivalent care, since there was no motivation to do otherwise. In the prospective setting, knowledge that the boss has a study underway to test his pet hypothesis that aortic aneurysm patients have a systemic defect making them prone to wound hernias might engender extra care in the wound closures of certain patients. Maybe the wound closures would be more carefully done in aneurysm patients because of their putative hernia proclivity, or more care might be lavished in closing aortoiliac occlusion disease patients to avoid excessive umbrage from having a hernia develop in the “wrong group.” So, I like this particular retrospective study. Harrison Lazarus, MD (Salt Lake City, Utah): With the association between hernias and aneurysms, is there any value in screening elderly patients with hernias to see if, in fact, they have aneurysms? I have a second question. In seeing that you have had this instance of hernias postoperatively, what do you plan to do in terms of modifying your technique to prevent them with open surgical repair or are you going to go to endovascular repair of all of these aneurysms? Arlo H emueck, MD (Kansas City, Kansas): I have a question about technique. Often our residents, when making a midline abdominal incision, fail to clean off the fat from the fascia. When these wounds are re-approximated, a fat-to-fat closure results instead of fascia-to-fascia. I’ve always felt, without any data for support, that this increased the incidence of midline incisional hernias, and I would like your comment on this matter. Barry Fisher, MD (Las Vegas, Nevada): I had only one difficulty in understanding and drawing any conclusions from this paper at all. That is because of the definition of hernia. Hernia to me is a hole in the fascia, and it seems to me that this paper mixes apples and oranges when it lumps diastasis of the rectus with true ventral hernias. JOURNAL OF SURGERY@’ VOLUME 170
DECEMBER
1995
575
Kevin Thomas, MD (Wichita Falls, Texas): 1 was just wondering if acute abdominal compartmental syndrome may be an etiology since it has been observed in trauma patients, and if there might be a way of measuring intra-abdominal pressures in a prospective manner with this problem?
CLOSING Glenn C. Hunter, MD: I thank the discussants for their cogent remarks. Dr. Thomas, in the present climate of managed care, the routine follow-up of patients undergoing general surgical procedures is difficult to justify. Therefore, we chose to evaluate the incisions of patients with aortic aneurysmal and occlusive disease who are seen annually in our vascular clinic. In the early 198Os, we used Vicryl and Dexon. I cannot tell you exactly how many patients who developed incisional hernias had their wounds closed with braided absorbable suture. All the hernias we have repaired to date have had their primary incisions closed with Prolene or Novafil. There are a number of studies in the literature that show that there are no differences in the incidence of incisional hernias in wounds closed with braided absorbable sutures versus those closed with nonabsorbable sutures. When closing midline wounds, we take bites approximately 2 cm from the edge of the fascia and 1 cm apart.
576
THE AMERICAN
JOURNAL OF SURGERY@
VOLUME 170
We have assayed alpha,-antitrypsin levels in tissue from patients with aneurysmal and occlusive disease because of the known association with COPD. The results have been too variable to demonstrate any statistical relationship. I knew the cautery papers would come back to haunt us some day. In the cautery studies, we addressed wound infection rather than healing. Dr. Lazarus, you asked whether we should screen patients with hernias to look for aneurysms? I would not use the presence of a hernia as an indicator for screening patients for aneurysms. As for endovascular repair, all I can tell you is that the one study I am aware of has been put on hold because of mechanical problems with struts. Dr. Hermreck, I agree fully. I believe that you need to dissect the skin and subcutaneous tissue off the fascia, especially in obese patients, so that the fascial margins can be clearly defined. Dr. Fisher raises a very intriguing question. When you examine these patients, their hernias are somewhat different from the usual incisional hernia. There is a diffuse bulging that is almost always confined to the upper part of the incision. One or both rectus muscles are usually distracted from the midline, with bowel or omentum adherent to the defect. I do not believe that it is simply divarication of the recti. If you have any other term we can use, we would gladly appreciate it. Thank you for the privilege of the floor.
DECEMBER
1995