- Email: [email protected]
Retrorectus prosthetic mesh repair of midline abdominal hernia
Retrorectus Prosthetic Mesh Repair Midline Abdominal Hernia
of
David McLanahan,
MD,
MD, L.T. King, MD, Charles Weems, MD, Michael Kathleen Gibson, MD, Seattle, Washington
BACKGROUND: Midline abdominal hernia is a common problem seen by the general surgeon. Recurrence rates are as high as 49% when an autogenous repair is performed, and as high as 11% when prosthetic mesh is used as a “bridge” or “onlay.” METHODS: This study analyzes results of midline abdominal hernia repair in 106 cases using prosthetic mesh, in the retrorectus position, as described by Stoppa and Wantz. Charts were reviewed, patient satisfaction determined by telephone interview, and recurrence rate by physician examination. RESULTS: Major systemic complications occurred in 17%. There were no deaths. Eighteen percent developed a wound complication, requiring a return to the operating room in 5%. There were three recurrences (3.5%). coNcLus10~S: Retrorectus placement of prosthetic mesh in the repair of midline abdominal hernia is effective and compares favorably with other methods. Significant complications are low, recurrence is rare, and patient satisfaction is high. Am J Surg. 1997;173:445-449. 0 1997 by Excerpta Medica, Inc.
M
idline abdominal hernia is a very common surgical problem. Although a few are primary, most of these are related to a previous surgical incision, in which the hernia incidence is 5% to 1 l%.’ These hernias usually enlarge and eventually become unsightly and symptomatic because of the continuing effects of intra-abdominal pressure and the lateral pull of the abdominal wall muscles. Complications include bowel incarceration, decreased respiratory efficiency, and ischemic skin ulcerations. Many of these patients are obese and have significant underlying medical problems such as diabetes, emphysema, and coronary artery disease. Surgical repair can be daunting, especially when need for emergency operation arises. Many different types of repair using native tissues and prosthetic materials have been devised to meet this challenge. Unfortunately, 31% to 49% of autogenous tissue repairs develop recurrence.’ The most common methods of From the Department of Surgery, Pacific Medical Center, Seattle, Washington. Requests for reprints should be addressed to David McLanahan, MD, Department of Surgery, Pacific Medical Center, 1200 12th Avenue South, Seattle, Washington 98144. Presented at the 83rd Annual Meeting of the North Pacific Surgical Association, Seattle, Washington, November 8-9, 1996.
0 1997 by Excerpta All rights reserved.
Medica,
Inc.
Novotney,
prosthetic repair use polypropylene mesh as a “bridge” or “inlay” attached to the margins of the aponeurotic defect, or as an “onlay” reinforcement over a primary repair. These methods may also have relatively high recurrence rates, up to ll%.j The placement of prosthetic material in a pocket beneath the rectus muscles and outside the peritoneum was devised in Europe by J. Rives and Rene Stoppa in the 1970~.~,~ Excellent results with huge, previously unrepairable hernias were reported. George Wantz has described the technique in the recent American literature” and in his beautifully illustrated book.7 Our study reports results in the repair of 106 midline abdominal hernias, using the Stoppa/Wantz retrorectus mesh placement technique. This operation is locally known as the “clock” repair, after the clockface-like appearance of the resulting circumferential ‘&inch abdominal skin incisions used to place the sutures that anchor the mesh to the abdominal wall muscles. Our purpose is to document this repair’s length of stay, complications, patient satisfaction, and recurrence rate.
PATIENTS
AND
METHODS
We reviewed 106 consecutive retrorectus mesh repairs of midline abdominal hernias, performed by five surgeons (44, 43, 8, 7, 4 cases each) at two Seattle hospitals during the j-year period from March 1991 to April 1996. All patients who could be located were questioned and examined at the end of the study period. To attempt an unbiased report of patient satisfaction, in most cases, i- to lo-minute telephone interviews were conducted by two University of Washington surgical residents, unknown to the patients. The physical examinations were performed by the original surgeons whenever possible, or by another physician when the patient was unable to come in or had moved out of the area. Patient characteristics are given in Table I. The hernia “collar size” was not often measured directly, but the hernias were variously described as “large,” “huge,” and “multiple,” and were often associated with rectus diastasis. The defects were evenly distributed among epigastric, mid-abdominal, and hypogastric sites, frequently involving the whole midline from xiphoid to pubis. Thirty-two percent of the patients had had their hernia previously repaired at least once, and 15% had had at least one previous mesh repair. The patients had the usual medical risk factors associated with age, obesity, cigarette smoking, and previous abdominal surgery, and these were not documented for the purposes of this study. Notable complicating factors included ascites (3 patients), small bowel obstruction (5), colon can0002-961 O/97/$1 PII SOOO2-9610(97)00067-6
7.00
445
REPAIR
T
OF
TABLE
MIDLINE
ABDOMINAL
HERNIA/MCLANAHANET=
I
I Patient
the mesh and intra-abdominal contents. In a few cases where there is insufficient tissue to close this layer, absorbable vicryl or dexon mesh, or incorporated omentum can he used to bridge the gap. (3) The mesh is placed underneath the full width of the rcctus muscles and overlapped above and helow the cephalad and caudad margins of the defect by 4 to 6 cm. It is anchored with interrupted absorbable sutures passed through the edge of the mesh, ahdominal wall, and out through the skin through ‘/ti-inch stab incisions. These are placed about 4 cm apart. The sutures are tied and cut, and then retract hack to the surface of the abdominal muscles. Attempt is made to make the mesh lay as wrinklefree as possible. Two Jackson Pratt drains are placed on the anterior surface of the mesh. (4) The remnants of anterior rectus fascia, linea alha, and scar tissue are then reapproximated to cover the mesh, isolating it from the subcutaneous tissue.
Characteristics
Gender(n) Male Female Age (years) Range Median Weight (pounds) Range Median Type of hernia Primary Incisional Number of repairs First hernia repairs One recurrence Two recurrences Three or more recurrences Previous mesh repair One mesh repair Two mesh repair Three mesh repair Urgency Elective Emergency/urgent Concurrent procedures None Bowel resection, etc.
43 63 33-89 56 111-550 226 7 99 72 16 9 9 16 8 7 1 95% 5% 84% 16%
cer (l), pacelnaker wires running through the hernia defect (2), and chronic renal failure (1). Five percent of the operations were performed on an emergency or urgent basis for small bowel obstruction. Preoperative pneumoperitoneum was not used. In addition to hernia repair, 17 patients (16%) had simultaneous procedures: lysis of adhesions (3 patients) or small bowel resection for obstruction (1); small bowel resection for enterotomy ( 1); excision of an enterocutaneous fistula with bowel resection in a patient with a previous inlay mesh repair (1); takedown colostomy, colectomy, and ileostomy for metachronous colon cancer ( 1); cholecystectomy (3); appendectomy (1); Meckel’s diverticulectomy (1); abdominoplasty (3); femoral heria plug repair (1); oophorectomy (1); and lung resection (1). The retrorectus mesh repair technique used was essentially the same as described and illustrated by Wantz,’ except that monofilament polypropylene mesh was used m most cases, rather than merselene (woven dacron) mesh as recommended by Wantz (Figure). All patients received prophylactic antibiotics, usually cephozolin. The essential features of the operation are as follows: (1) After opening the incision and excising excess skin and subcutaneous tissue, the redundant hernia sack is usually excised. (2) The posterior rectus fascia is incised close to the linea alba and a pocket created bluntly behind the full width of the rectus muscles and in front of the peritoneum and posterior rectus fascia cephalad to the linea semicircularis, and above the peritoneum caudad to it. In patients with previous transverse incisions, most commonly suhcostal, sharp dissection is necessary to create this plane. The peritoneum and posterior rectus fascia are closed with a running absorbable suture so as to interpose a barrier between
RESULTS In all, 104 patients with large midline abdominal hernias underwent 106 repairs using retrorectus prosthetic mesh. Hospital length of stay (mcluding the operation date) ranged from 2 to 12 days. Median stay was 4 days. There were no deaths. The onI\: intraoperative complication was an enterotomy requiring a small bowel resection. No blood transfusions were necessary. Sign&cant postoperative systemic complications occurred in 18 cases (17%). Major pulmonary complications occurred in 7 patients (7X), requiring an intensive care unit stay in 5 (5%). There were 4 cardiac complications requiring medication in 2, a pacemaker in 1, and a coronary an&ram in 1. There were no MIS. Six patients had early Gl complications including ileus or Olgilvie’s s);ndrome in 4, small bowel obstruction, not requiring reoperation in 1 . and persistent unexplained nausea requiring TPN in 1. One patient developed deep vein thrombosis. Nineteen patients (18%) developed a treated wound complication. Five (5%) required a return to the operating room, to excise necrotic skin ;Ind subcutaneous tissue (1 patient), sloughed umbilicus (2), evacuate a superficial hematoma (l), and repair a colocutaneous fistula in a patient who had had a previous inlay mesh hernia repair following several gastric operations. Five patients were diagnosed as having a superficial wound infection based on wound separation and drainage. They all healed with antibiotics and local care, although a prolonged period of time was necessary in 3 patients. Two patients developed an infection localized to a lateral suture site that healed after suture removal. Six other patients were given a course of antibiotics for a mildly errythematous or indurated wound without drainage. One patient with ascites, who had an ahdominoplasty tu excise ischemic skm ulcerations in association with her hernia repair, developed a superficial seroma that required three needle aspirations In no case was mesh exposed or was it necessary to remove mesh. Five patients required removal of one or two peripheral anchoring sutures, 2 to 8 months postoperatively because of significant local discomfort. This prahlem has not developed since we stopped using permanent suture material and have avoided incorporating rib or cartilage m these sutures. There were no known hernia recurrences within 6 months. For the purposes ofdetcrmming the true recurrence
L 446
THE
AMERICAN
JOURNAL
OF SURGERY@
VOLUME
173
MAY
1997
1 REPAIR
OF
MIDLINE
ABDOMINAL
HERNWMCLANAHANl
Figure. Stoppa/Wantz technique. A. Lateral muscle pull and intra-abdominal pressure enlarge defect. B. Pressure pushes fascia against mesh and mesh against muscle; solid “face-to-face” fusion follows. 1. Posterior rectus fascia and peritoneum closure beneath mesh. 2. Mesh. 3. Anterior rectus fascia and linea alba closure isolating mesh from subcutaneous tissue.
rate for this study, we have used the last documented physician physical examination, performed at least 6 months after the operation. Nine patients died of unrelated causes. We concluded the study period for 2 patients when they had subsequent, midline abdominal incisions for other inn-a-abdominal pathology. Eighty-six (84%) eligible patients were examined at least 6 months postoperatively. Followup on these patients ranged from 6 to 67 months. Median follow-up was 24 months. There were 3 documented recurrences (3.5%) in this group. Seventy-six patients (80% of those living) completed the patient satisfaction interview. Ninety-six percent of patients were highly satisfied or satisfied with their overall experience.
COMMENTS Few articles describing various repair techniques for midline abdominal hernia report method of follow-up, length of stay, and patient satisfaction. Many are simply reports of early complications, “known” recurrences, and surgeon perception as to the success of the procedure, and are not based on patient evaluation and physician examination at the end of the study period cited. Such information is notoriously unreliable. Our purpose was to document the retrorectus mesh repair’s length of stay, complications, patient satisfaction, recurrence rate, and perhaps, establish a benchmark for comparison with other techniques. Major postoperative systemic complications occurred in 17% of our patients. There were no deaths. Our wound complication rate of 18% is comparable to a reported rate of 15% to 45% for incisional hernia repair,s and 17.3%3 for prosthetic repair. Although 5 patients (5%) required a return to the operating room, most were handled on an outpatient basis. Patient satisfaction is very important in determining the usefulness of any SUrgicdl procedure. There is no randomized comparison of patient satisfaction with the various types of midline abdominal hernia repair, and such information is lacking in most reports, so that meaningful comparisons are impossible at this point. Most studies report the surgeon’s assessment of the particular operation, usually, simply whether a recurrence has developed. Detailed information THE
AMERICAN
about the patient’s own evaluation of his or her experience has not been reported. In our study, we wished to determine the patient’s assessment of the operation. This may help determine the retrorectus mesh repair’s ultimate value compared with other methods. In our series, 5 5% of the patients described their discomfort as moderate to severe in the first 2 months. Forty-five percent still had some discomfort 12 months or more after the operation. Although usually such symptoms were infrequent and mild, when present 11% of patients described them as moderate to severe, and 7% said they limited their activities. Thirty-four percent think they can feel the mesh. Wantz believes that patient discomfort is increased by using monofilament polypropylene mesh and that woven dacron (merselene) mesh has an advantage in that it conforms more easily to irregular body corm~urs, is softer and more supple, and may therefore be more comfortable. Despite theoretical advantages of monofilament material and marked preference of American surgeons for polypropylene, he points out that this material haa been used for years by European surgeons, without apparent increase in wound complications. We used merselene mesh in a few patients, mostly when the hernia involved the hypogastric area, and the mesh had to wrap into the space of Retzius. However, it is unknown how many patients with merselene mesh retrorectus repair have long-term discomfort caused by its presence or think they can feel their mesh. Ninety-six percent of our patients were highly satisfied or satisfied with their overall experience, although 19% were not satisfied with their scars or change in appearance. Three patients (4%) said they would not have had the operation if they knew what they now know about it. Patient satisfaction in the other retrorectus mesh series is not documented. Except for the study by Sugarman et al’ of mesh placement above the abdominal wall fascia using an onlay technique, in which 6% of the patients had “chronic pain,” it is unknown what percentage of patients with primary repair or other prosthetic techniques have prolonged discomfort, or what percentage of those with mesh think they can feel the prosthesis. The sine qua non in evaluating a hernia repair is its recurrence rate. Unfortunately, there is a well-known problem JOURNAL
OF SURGERY”
VOLUME
173
MAY
1997
447
TABLE
II Recurrence
Investigator Stoppa’o Wan@ Amid” Present Total
Number
series
* Above 01 beiow
of Cases 133 23 54 86 296
previous
Rate for
Midline
Recurrence
(n)
(1) (3)
THE
AMERICAN
Mesh Follow-Up
(25) 18.5%
Repair Interval
O-l 0 yr; 5.5 year mean Not stated Not stated 6 mo-5 yr 24 mo median
:.9%* 3.5%* 9.8%
Method
of Follow-Up
Physician examination Not stated Physician examination Physican examination
mesh.
with the validity of published statistics. Most surgeons do not learn of their recurrences and patients often do not bring them to attention, considering their prior experience with surgery. The method of determining recurrence, whether by “known recurrence,” patient report, or provider examination, is not stated by most authors. Some series do not count another hernia in a slightly different location as a recurrence. As we have done, others consider that another hernia is related to the first repair and could have been prevented if the initial repair had reinforced that area. Notoriously unreliable are patient questionnaires to determine recurrence rates.’ The only accurate means of determining recurrence rate is by direct physician examination. Recurrent incisional hernia after autogenous repair appears earlier than primary incisional hernia.’ The time course of appearance of recurrence after prosthetic repair is unknown. The longer the median follow-up, the more reliable the study is likely to he. Unbiased studies with careful follow-up physical examination have determined that autogenous tissue repair of midline abdominal hernia has recurrence rates as high as 31% to 49%. Even hernias smaller than 4 cm have a reocurrence rate of 25%.’ Complicated techniques using flaps, autogenous grafts, relaxing incisions, etc, may ultimately weaken the abdominal wall. Most surgeons now favor prosthetic material repair as its proven safety, decreased recurrence rate, and simplicity outweighs the disadvantages of autogenous repair. The most common methods of using prosthetic material are as an inlay graft attached to the collar of the aponeurotic defect in an edge-to-edge manner, or as an onlay reinforcement to an autogenous repair. These methods have certain disadvantages. The inlay graft attaches to the edge of the defect, where the tissue may already be weakened by previous surgery, scarring, and stretching. Intrahdominal pressure and muscle contraction pulling laterally act to distract the suture line, leading to the commonly seen reoccurrence between the lateral edge of the graft and the edge of the original defect. Onlay grafts may cover a repair that has too much tension and requires dissection of a subcutaneous flap, exposing the prosthetic to poorly vascularized fat, increasing the likelihood of seroma and wound infection. Ahdominal wall forces may break down the primary repair and push the graft off the outer fascial surface, predisposing to recurrence. Fifteen percent of our cases followed one or more failed prosthetic repairs. A less commonly used method of repair places the prosthetic intra-abdominally, hut may be associated with an increased incidence of bowel obstruction 448
Retrorectus
JOURNAL
OF SURGERY”
VOLUME
173
and fistula formation. Because of adhesions, another ahdominal operation may he more hazardous. An analysis of 16 series totaling 819 cases of various inlay and onlay prosthetic repairs found a combined recurrence rate of 6.4%.i However, the true recurrence rate may he much higher as the mean follow-up is less than 22 months or unstated for most of these series, and the method of follow-up evaluation is not stated by most authors. The retrorectus mesh repair has several advantages. It utilizes the forces of intrahdominal pressure to push the peritoneum and posterior rectus fascia against the mesh, and the mesh against the rectus muscle to achieve a solid face-to-face overlap. The mesh is rapidly incorporated, preventing recurrent stretching of the of the peritoneum and abdominal wall. Lateral recurrent hernia does not occur, and the only possihility of a midline recurrence is cephadad or caudad to an incompletely reinforced midline. The mesh is isolated from the abdominal contents, minimizing the possibility of bowel obstruction or fistula from bowel adhering to the mesh. Likewise, the mesh does not conract the subcutaneous space, minimizing the risk of serolna and wound infection. In the last 10 years, there have been three published reports of retrorectus mesh repair of midline abdominal hernia (Table II). Stoppa et al’” reported a series of 230 patients operated for incisional hernia, 67% of whom had repair with a large dacron mesh prosthesis. In the mesh group, they reported 17.6% wound complications. Of the 133 mesh repairs (approximately 81% midline) reexamined after a 5.5 year mean follow-up period, an 18.5% recurrence rate was found, with 81% “satisfactory” resuits.‘” The high recurrence rate probably reflects the exchallenge of many of the referred treme patients, need for emergency surgery, and careful followup examinations. Wantzh reports no recurrences, and that “no complications occurred and recovery was uneventful” in 23 large midline incisional hernia patients. Period of follow-up and method of determining recurrence was not stated. Amid et al” of the Lichtenstein Hernia Institute, report one recurrence above the edge of the mesh in 54 cases, 25 using a variation of the technique where the mesh is stapled to the undersurface of the anterior rectus sheath rather than using sutures through the muscle wall. Method of determining recurrence was by physician examination, but period of follow-up was not specified. In our study, we documented three recurrences (3.5%) in 86 patients examined 6 to 67 months postoperatively (median follow-up 24 months). Two were in the suhxiphoid area above the upper edge of the tnesh and one was in the MAY
1997
1 REPAIR
suprapubic area below the mesh, suggesting that the recurrence rate could have been lowered by being more aggressive in widely repairing the whole previous laparotomy scar, as recommended by Stoppa et al” and Amid et al.” All three were asymptomatic. Two have been repaired using another piece of mesh placed in the retrorectus position, adjacent to the original repair.
SUMMARY Retrorectus placement of prosthetic mesh in the repair of midline abdominal hernia is effective and compares favorably with other methods. Significant complications are low, recurrence rare, and patient satisfaction high.
REFERENCES 1. George CD, Ellis H. The results of incisional hernia repair: a twelve year review. Ann R Coil Surg. 1986;68:185-187. 2. Hesselink VJ, Luijendijk RW, deWilt JHW, et al. An evaluation of risk factors in incisional hernia recurrence. Surg Gynecol Obstet. 1993;176:238-234.
THE
AMERICAN
OF
MIDLINE
ABDOMINAL
HERNIA/MCLANAHAN
/
3. Sitzman JV, McFadden DW. The Internal retention repair of massive ventral hernia. Am Svrg. 1989;55:719-723. 4. Rives J. Major incisional hernia. In: Cherval Jr, ed. Surgery the Abdominal Wall. Paris: Springer-Verlag; 1987:116-144.
of
5. Stoppa RE. The treatment of complicared groin and incisional hernias. World .I Surg. 1989;13:545-554. 6. Wantz
GE. Incisional
hernioplasty
with
mersilene.
Surg Gynecol
Obstet. 1991;172:129-137. 7. Wantz GE. Atlas of Hemin Surgery. New York: Raven Press; 1991:179-216. 8. Houck JP, Rypins EB, Sarfkh J, et al. Repan of incisional hernia. Surg Gynecol Obstet. 1989;169:397-399. 9. Sugarman HJ, Kellum JM Jr, Remes HD, et al. Greater risk of incisional hernia with morbid117 &se than steroid-dependent patients and low recurrence with prefasclal polypropylene mesh. Am J Surg. 1996;171:80-84. 10. Stoppa R, Louis D, Verhaeghe I’, et al. Current surgical treatment of postoperative eventrations. Int Surg. 1987;72:42-
44. 11. Amid PK, Shulman technique for prosthetic Surg. 1994;60:936-937.
JOURNAL
OF
AG, Lichtenstem repair of mass~vr
SURGERY”
VOLUME
IL. A simple stapling incisional hernias. Am
173
MAY
1997
44
of
David McLanahan,
MD,
MD, L.T. King, MD, Charles Weems, MD, Michael Kathleen Gibson, MD, Seattle, Washington
BACKGROUND: Midline abdominal hernia is a common problem seen by the general surgeon. Recurrence rates are as high as 49% when an autogenous repair is performed, and as high as 11% when prosthetic mesh is used as a “bridge” or “onlay.” METHODS: This study analyzes results of midline abdominal hernia repair in 106 cases using prosthetic mesh, in the retrorectus position, as described by Stoppa and Wantz. Charts were reviewed, patient satisfaction determined by telephone interview, and recurrence rate by physician examination. RESULTS: Major systemic complications occurred in 17%. There were no deaths. Eighteen percent developed a wound complication, requiring a return to the operating room in 5%. There were three recurrences (3.5%). coNcLus10~S: Retrorectus placement of prosthetic mesh in the repair of midline abdominal hernia is effective and compares favorably with other methods. Significant complications are low, recurrence is rare, and patient satisfaction is high. Am J Surg. 1997;173:445-449. 0 1997 by Excerpta Medica, Inc.
M
idline abdominal hernia is a very common surgical problem. Although a few are primary, most of these are related to a previous surgical incision, in which the hernia incidence is 5% to 1 l%.’ These hernias usually enlarge and eventually become unsightly and symptomatic because of the continuing effects of intra-abdominal pressure and the lateral pull of the abdominal wall muscles. Complications include bowel incarceration, decreased respiratory efficiency, and ischemic skin ulcerations. Many of these patients are obese and have significant underlying medical problems such as diabetes, emphysema, and coronary artery disease. Surgical repair can be daunting, especially when need for emergency operation arises. Many different types of repair using native tissues and prosthetic materials have been devised to meet this challenge. Unfortunately, 31% to 49% of autogenous tissue repairs develop recurrence.’ The most common methods of From the Department of Surgery, Pacific Medical Center, Seattle, Washington. Requests for reprints should be addressed to David McLanahan, MD, Department of Surgery, Pacific Medical Center, 1200 12th Avenue South, Seattle, Washington 98144. Presented at the 83rd Annual Meeting of the North Pacific Surgical Association, Seattle, Washington, November 8-9, 1996.
0 1997 by Excerpta All rights reserved.
Medica,
Inc.
Novotney,
prosthetic repair use polypropylene mesh as a “bridge” or “inlay” attached to the margins of the aponeurotic defect, or as an “onlay” reinforcement over a primary repair. These methods may also have relatively high recurrence rates, up to ll%.j The placement of prosthetic material in a pocket beneath the rectus muscles and outside the peritoneum was devised in Europe by J. Rives and Rene Stoppa in the 1970~.~,~ Excellent results with huge, previously unrepairable hernias were reported. George Wantz has described the technique in the recent American literature” and in his beautifully illustrated book.7 Our study reports results in the repair of 106 midline abdominal hernias, using the Stoppa/Wantz retrorectus mesh placement technique. This operation is locally known as the “clock” repair, after the clockface-like appearance of the resulting circumferential ‘&inch abdominal skin incisions used to place the sutures that anchor the mesh to the abdominal wall muscles. Our purpose is to document this repair’s length of stay, complications, patient satisfaction, and recurrence rate.
PATIENTS
AND
METHODS
We reviewed 106 consecutive retrorectus mesh repairs of midline abdominal hernias, performed by five surgeons (44, 43, 8, 7, 4 cases each) at two Seattle hospitals during the j-year period from March 1991 to April 1996. All patients who could be located were questioned and examined at the end of the study period. To attempt an unbiased report of patient satisfaction, in most cases, i- to lo-minute telephone interviews were conducted by two University of Washington surgical residents, unknown to the patients. The physical examinations were performed by the original surgeons whenever possible, or by another physician when the patient was unable to come in or had moved out of the area. Patient characteristics are given in Table I. The hernia “collar size” was not often measured directly, but the hernias were variously described as “large,” “huge,” and “multiple,” and were often associated with rectus diastasis. The defects were evenly distributed among epigastric, mid-abdominal, and hypogastric sites, frequently involving the whole midline from xiphoid to pubis. Thirty-two percent of the patients had had their hernia previously repaired at least once, and 15% had had at least one previous mesh repair. The patients had the usual medical risk factors associated with age, obesity, cigarette smoking, and previous abdominal surgery, and these were not documented for the purposes of this study. Notable complicating factors included ascites (3 patients), small bowel obstruction (5), colon can0002-961 O/97/$1 PII SOOO2-9610(97)00067-6
7.00
445
REPAIR
T
OF
TABLE
MIDLINE
ABDOMINAL
HERNIA/MCLANAHANET=
I
I Patient
the mesh and intra-abdominal contents. In a few cases where there is insufficient tissue to close this layer, absorbable vicryl or dexon mesh, or incorporated omentum can he used to bridge the gap. (3) The mesh is placed underneath the full width of the rcctus muscles and overlapped above and helow the cephalad and caudad margins of the defect by 4 to 6 cm. It is anchored with interrupted absorbable sutures passed through the edge of the mesh, ahdominal wall, and out through the skin through ‘/ti-inch stab incisions. These are placed about 4 cm apart. The sutures are tied and cut, and then retract hack to the surface of the abdominal muscles. Attempt is made to make the mesh lay as wrinklefree as possible. Two Jackson Pratt drains are placed on the anterior surface of the mesh. (4) The remnants of anterior rectus fascia, linea alha, and scar tissue are then reapproximated to cover the mesh, isolating it from the subcutaneous tissue.
Characteristics
Gender(n) Male Female Age (years) Range Median Weight (pounds) Range Median Type of hernia Primary Incisional Number of repairs First hernia repairs One recurrence Two recurrences Three or more recurrences Previous mesh repair One mesh repair Two mesh repair Three mesh repair Urgency Elective Emergency/urgent Concurrent procedures None Bowel resection, etc.
43 63 33-89 56 111-550 226 7 99 72 16 9 9 16 8 7 1 95% 5% 84% 16%
cer (l), pacelnaker wires running through the hernia defect (2), and chronic renal failure (1). Five percent of the operations were performed on an emergency or urgent basis for small bowel obstruction. Preoperative pneumoperitoneum was not used. In addition to hernia repair, 17 patients (16%) had simultaneous procedures: lysis of adhesions (3 patients) or small bowel resection for obstruction (1); small bowel resection for enterotomy ( 1); excision of an enterocutaneous fistula with bowel resection in a patient with a previous inlay mesh repair (1); takedown colostomy, colectomy, and ileostomy for metachronous colon cancer ( 1); cholecystectomy (3); appendectomy (1); Meckel’s diverticulectomy (1); abdominoplasty (3); femoral heria plug repair (1); oophorectomy (1); and lung resection (1). The retrorectus mesh repair technique used was essentially the same as described and illustrated by Wantz,’ except that monofilament polypropylene mesh was used m most cases, rather than merselene (woven dacron) mesh as recommended by Wantz (Figure). All patients received prophylactic antibiotics, usually cephozolin. The essential features of the operation are as follows: (1) After opening the incision and excising excess skin and subcutaneous tissue, the redundant hernia sack is usually excised. (2) The posterior rectus fascia is incised close to the linea alba and a pocket created bluntly behind the full width of the rectus muscles and in front of the peritoneum and posterior rectus fascia cephalad to the linea semicircularis, and above the peritoneum caudad to it. In patients with previous transverse incisions, most commonly suhcostal, sharp dissection is necessary to create this plane. The peritoneum and posterior rectus fascia are closed with a running absorbable suture so as to interpose a barrier between
RESULTS In all, 104 patients with large midline abdominal hernias underwent 106 repairs using retrorectus prosthetic mesh. Hospital length of stay (mcluding the operation date) ranged from 2 to 12 days. Median stay was 4 days. There were no deaths. The onI\: intraoperative complication was an enterotomy requiring a small bowel resection. No blood transfusions were necessary. Sign&cant postoperative systemic complications occurred in 18 cases (17%). Major pulmonary complications occurred in 7 patients (7X), requiring an intensive care unit stay in 5 (5%). There were 4 cardiac complications requiring medication in 2, a pacemaker in 1, and a coronary an&ram in 1. There were no MIS. Six patients had early Gl complications including ileus or Olgilvie’s s);ndrome in 4, small bowel obstruction, not requiring reoperation in 1 . and persistent unexplained nausea requiring TPN in 1. One patient developed deep vein thrombosis. Nineteen patients (18%) developed a treated wound complication. Five (5%) required a return to the operating room, to excise necrotic skin ;Ind subcutaneous tissue (1 patient), sloughed umbilicus (2), evacuate a superficial hematoma (l), and repair a colocutaneous fistula in a patient who had had a previous inlay mesh hernia repair following several gastric operations. Five patients were diagnosed as having a superficial wound infection based on wound separation and drainage. They all healed with antibiotics and local care, although a prolonged period of time was necessary in 3 patients. Two patients developed an infection localized to a lateral suture site that healed after suture removal. Six other patients were given a course of antibiotics for a mildly errythematous or indurated wound without drainage. One patient with ascites, who had an ahdominoplasty tu excise ischemic skm ulcerations in association with her hernia repair, developed a superficial seroma that required three needle aspirations In no case was mesh exposed or was it necessary to remove mesh. Five patients required removal of one or two peripheral anchoring sutures, 2 to 8 months postoperatively because of significant local discomfort. This prahlem has not developed since we stopped using permanent suture material and have avoided incorporating rib or cartilage m these sutures. There were no known hernia recurrences within 6 months. For the purposes ofdetcrmming the true recurrence
L 446
THE
AMERICAN
JOURNAL
OF SURGERY@
VOLUME
173
MAY
1997
1 REPAIR
OF
MIDLINE
ABDOMINAL
HERNWMCLANAHANl
Figure. Stoppa/Wantz technique. A. Lateral muscle pull and intra-abdominal pressure enlarge defect. B. Pressure pushes fascia against mesh and mesh against muscle; solid “face-to-face” fusion follows. 1. Posterior rectus fascia and peritoneum closure beneath mesh. 2. Mesh. 3. Anterior rectus fascia and linea alba closure isolating mesh from subcutaneous tissue.
rate for this study, we have used the last documented physician physical examination, performed at least 6 months after the operation. Nine patients died of unrelated causes. We concluded the study period for 2 patients when they had subsequent, midline abdominal incisions for other inn-a-abdominal pathology. Eighty-six (84%) eligible patients were examined at least 6 months postoperatively. Followup on these patients ranged from 6 to 67 months. Median follow-up was 24 months. There were 3 documented recurrences (3.5%) in this group. Seventy-six patients (80% of those living) completed the patient satisfaction interview. Ninety-six percent of patients were highly satisfied or satisfied with their overall experience.
COMMENTS Few articles describing various repair techniques for midline abdominal hernia report method of follow-up, length of stay, and patient satisfaction. Many are simply reports of early complications, “known” recurrences, and surgeon perception as to the success of the procedure, and are not based on patient evaluation and physician examination at the end of the study period cited. Such information is notoriously unreliable. Our purpose was to document the retrorectus mesh repair’s length of stay, complications, patient satisfaction, recurrence rate, and perhaps, establish a benchmark for comparison with other techniques. Major postoperative systemic complications occurred in 17% of our patients. There were no deaths. Our wound complication rate of 18% is comparable to a reported rate of 15% to 45% for incisional hernia repair,s and 17.3%3 for prosthetic repair. Although 5 patients (5%) required a return to the operating room, most were handled on an outpatient basis. Patient satisfaction is very important in determining the usefulness of any SUrgicdl procedure. There is no randomized comparison of patient satisfaction with the various types of midline abdominal hernia repair, and such information is lacking in most reports, so that meaningful comparisons are impossible at this point. Most studies report the surgeon’s assessment of the particular operation, usually, simply whether a recurrence has developed. Detailed information THE
AMERICAN
about the patient’s own evaluation of his or her experience has not been reported. In our study, we wished to determine the patient’s assessment of the operation. This may help determine the retrorectus mesh repair’s ultimate value compared with other methods. In our series, 5 5% of the patients described their discomfort as moderate to severe in the first 2 months. Forty-five percent still had some discomfort 12 months or more after the operation. Although usually such symptoms were infrequent and mild, when present 11% of patients described them as moderate to severe, and 7% said they limited their activities. Thirty-four percent think they can feel the mesh. Wantz believes that patient discomfort is increased by using monofilament polypropylene mesh and that woven dacron (merselene) mesh has an advantage in that it conforms more easily to irregular body corm~urs, is softer and more supple, and may therefore be more comfortable. Despite theoretical advantages of monofilament material and marked preference of American surgeons for polypropylene, he points out that this material haa been used for years by European surgeons, without apparent increase in wound complications. We used merselene mesh in a few patients, mostly when the hernia involved the hypogastric area, and the mesh had to wrap into the space of Retzius. However, it is unknown how many patients with merselene mesh retrorectus repair have long-term discomfort caused by its presence or think they can feel their mesh. Ninety-six percent of our patients were highly satisfied or satisfied with their overall experience, although 19% were not satisfied with their scars or change in appearance. Three patients (4%) said they would not have had the operation if they knew what they now know about it. Patient satisfaction in the other retrorectus mesh series is not documented. Except for the study by Sugarman et al’ of mesh placement above the abdominal wall fascia using an onlay technique, in which 6% of the patients had “chronic pain,” it is unknown what percentage of patients with primary repair or other prosthetic techniques have prolonged discomfort, or what percentage of those with mesh think they can feel the prosthesis. The sine qua non in evaluating a hernia repair is its recurrence rate. Unfortunately, there is a well-known problem JOURNAL
OF SURGERY”
VOLUME
173
MAY
1997
447
TABLE
II Recurrence
Investigator Stoppa’o Wan@ Amid” Present Total
Number
series
* Above 01 beiow
of Cases 133 23 54 86 296
previous
Rate for
Midline
Recurrence
(n)
(1) (3)
THE
AMERICAN
Mesh Follow-Up
(25) 18.5%
Repair Interval
O-l 0 yr; 5.5 year mean Not stated Not stated 6 mo-5 yr 24 mo median
:.9%* 3.5%* 9.8%
Method
of Follow-Up
Physician examination Not stated Physician examination Physican examination
mesh.
with the validity of published statistics. Most surgeons do not learn of their recurrences and patients often do not bring them to attention, considering their prior experience with surgery. The method of determining recurrence, whether by “known recurrence,” patient report, or provider examination, is not stated by most authors. Some series do not count another hernia in a slightly different location as a recurrence. As we have done, others consider that another hernia is related to the first repair and could have been prevented if the initial repair had reinforced that area. Notoriously unreliable are patient questionnaires to determine recurrence rates.’ The only accurate means of determining recurrence rate is by direct physician examination. Recurrent incisional hernia after autogenous repair appears earlier than primary incisional hernia.’ The time course of appearance of recurrence after prosthetic repair is unknown. The longer the median follow-up, the more reliable the study is likely to he. Unbiased studies with careful follow-up physical examination have determined that autogenous tissue repair of midline abdominal hernia has recurrence rates as high as 31% to 49%. Even hernias smaller than 4 cm have a reocurrence rate of 25%.’ Complicated techniques using flaps, autogenous grafts, relaxing incisions, etc, may ultimately weaken the abdominal wall. Most surgeons now favor prosthetic material repair as its proven safety, decreased recurrence rate, and simplicity outweighs the disadvantages of autogenous repair. The most common methods of using prosthetic material are as an inlay graft attached to the collar of the aponeurotic defect in an edge-to-edge manner, or as an onlay reinforcement to an autogenous repair. These methods have certain disadvantages. The inlay graft attaches to the edge of the defect, where the tissue may already be weakened by previous surgery, scarring, and stretching. Intrahdominal pressure and muscle contraction pulling laterally act to distract the suture line, leading to the commonly seen reoccurrence between the lateral edge of the graft and the edge of the original defect. Onlay grafts may cover a repair that has too much tension and requires dissection of a subcutaneous flap, exposing the prosthetic to poorly vascularized fat, increasing the likelihood of seroma and wound infection. Ahdominal wall forces may break down the primary repair and push the graft off the outer fascial surface, predisposing to recurrence. Fifteen percent of our cases followed one or more failed prosthetic repairs. A less commonly used method of repair places the prosthetic intra-abdominally, hut may be associated with an increased incidence of bowel obstruction 448
Retrorectus
JOURNAL
OF SURGERY”
VOLUME
173
and fistula formation. Because of adhesions, another ahdominal operation may he more hazardous. An analysis of 16 series totaling 819 cases of various inlay and onlay prosthetic repairs found a combined recurrence rate of 6.4%.i However, the true recurrence rate may he much higher as the mean follow-up is less than 22 months or unstated for most of these series, and the method of follow-up evaluation is not stated by most authors. The retrorectus mesh repair has several advantages. It utilizes the forces of intrahdominal pressure to push the peritoneum and posterior rectus fascia against the mesh, and the mesh against the rectus muscle to achieve a solid face-to-face overlap. The mesh is rapidly incorporated, preventing recurrent stretching of the of the peritoneum and abdominal wall. Lateral recurrent hernia does not occur, and the only possihility of a midline recurrence is cephadad or caudad to an incompletely reinforced midline. The mesh is isolated from the abdominal contents, minimizing the possibility of bowel obstruction or fistula from bowel adhering to the mesh. Likewise, the mesh does not conract the subcutaneous space, minimizing the risk of serolna and wound infection. In the last 10 years, there have been three published reports of retrorectus mesh repair of midline abdominal hernia (Table II). Stoppa et al’” reported a series of 230 patients operated for incisional hernia, 67% of whom had repair with a large dacron mesh prosthesis. In the mesh group, they reported 17.6% wound complications. Of the 133 mesh repairs (approximately 81% midline) reexamined after a 5.5 year mean follow-up period, an 18.5% recurrence rate was found, with 81% “satisfactory” resuits.‘” The high recurrence rate probably reflects the exchallenge of many of the referred treme patients, need for emergency surgery, and careful followup examinations. Wantzh reports no recurrences, and that “no complications occurred and recovery was uneventful” in 23 large midline incisional hernia patients. Period of follow-up and method of determining recurrence was not stated. Amid et al” of the Lichtenstein Hernia Institute, report one recurrence above the edge of the mesh in 54 cases, 25 using a variation of the technique where the mesh is stapled to the undersurface of the anterior rectus sheath rather than using sutures through the muscle wall. Method of determining recurrence was by physician examination, but period of follow-up was not specified. In our study, we documented three recurrences (3.5%) in 86 patients examined 6 to 67 months postoperatively (median follow-up 24 months). Two were in the suhxiphoid area above the upper edge of the tnesh and one was in the MAY
1997
1 REPAIR
suprapubic area below the mesh, suggesting that the recurrence rate could have been lowered by being more aggressive in widely repairing the whole previous laparotomy scar, as recommended by Stoppa et al” and Amid et al.” All three were asymptomatic. Two have been repaired using another piece of mesh placed in the retrorectus position, adjacent to the original repair.
SUMMARY Retrorectus placement of prosthetic mesh in the repair of midline abdominal hernia is effective and compares favorably with other methods. Significant complications are low, recurrence rare, and patient satisfaction high.
REFERENCES 1. George CD, Ellis H. The results of incisional hernia repair: a twelve year review. Ann R Coil Surg. 1986;68:185-187. 2. Hesselink VJ, Luijendijk RW, deWilt JHW, et al. An evaluation of risk factors in incisional hernia recurrence. Surg Gynecol Obstet. 1993;176:238-234.
THE
AMERICAN
OF
MIDLINE
ABDOMINAL
HERNIA/MCLANAHAN
/
3. Sitzman JV, McFadden DW. The Internal retention repair of massive ventral hernia. Am Svrg. 1989;55:719-723. 4. Rives J. Major incisional hernia. In: Cherval Jr, ed. Surgery the Abdominal Wall. Paris: Springer-Verlag; 1987:116-144.
of
5. Stoppa RE. The treatment of complicared groin and incisional hernias. World .I Surg. 1989;13:545-554. 6. Wantz
GE. Incisional
hernioplasty
with
mersilene.
Surg Gynecol
Obstet. 1991;172:129-137. 7. Wantz GE. Atlas of Hemin Surgery. New York: Raven Press; 1991:179-216. 8. Houck JP, Rypins EB, Sarfkh J, et al. Repan of incisional hernia. Surg Gynecol Obstet. 1989;169:397-399. 9. Sugarman HJ, Kellum JM Jr, Remes HD, et al. Greater risk of incisional hernia with morbid117 &se than steroid-dependent patients and low recurrence with prefasclal polypropylene mesh. Am J Surg. 1996;171:80-84. 10. Stoppa R, Louis D, Verhaeghe I’, et al. Current surgical treatment of postoperative eventrations. Int Surg. 1987;72:42-
44. 11. Amid PK, Shulman technique for prosthetic Surg. 1994;60:936-937.
JOURNAL
OF
AG, Lichtenstem repair of mass~vr
SURGERY”
VOLUME
IL. A simple stapling incisional hernias. Am
173
MAY
1997
44