Panniculectomy: A Useful Technique for the Obese PatientUndergoing Gynecological Surgery

Penelope I. Blomfield, Thao Le, David G. Allen, and Robert S. Planner

Department of Gynecological Oncology, Mercy Hospital for Women, Clarendon Street, East Melbourne, 3002, Australia

Received October 13, 1997

Objectives. To establish the safety and efficacy of panniculec-tomy at the time of surgery for benign and malignant gynecolog-ical disease.

Methods. Retrospective review of the course of 57 patientsundergoing radical gynecological surgery and panniculectomy be-tween January 1992 and January 1997 at the Mercy Hospital forWomen, Melbourne. Data were collected regarding indication fortreatment, operative details, and complications of surgery.

Results. Of 57 patients in the study, 32 had a primary gyneco-logical malignancy, 11 had benign gynecological disease, 3 hadcervical dysplasia, 5 had endometrial hyperplasia, and the remain-ing 6 had incisional hernia repair. The mean age of patients was 55years with a mean weight of 101 kg (range 70–145 kg). The meanoperative time was 2 h 24 min, and blood transfusion was under-taken in 23 (41%) patients. Four (7.1%) individuals had a minorwound infection and 3 (5.4%) a moderate wound infection. Onepatient experienced a nonfatal pulmonary embolus and 2 patientsexperienced a deep vein thrombosis. There were no postoperativedeaths. Long term, 6 patients developed an incisional hernia.

Conclusions. Panniculectomy is a useful technique in obesepatients. It improves surgical access facilitating radical surgeryand is cosmetically pleasing to the patient. It has acceptablemorbidity when compared to conventional midline vertical ortransverse incisions in comparable populations. © 1998 Academic Press

INTRODUCTION

It is well recognized that abdominal and pelvic surgery inobese women is associated with increased anaesthetic risk andsurgical difficulty. Increased postoperative morbidity is en-countered, particularly in relation to abdominal wound infec-tions [1, 2]. The incision chosen by most gynecologists in suchpatients is subumbilical, midline, or transverse suprapubic.These incisions are associated with high rates of wound com-plications because of the difficulty in keeping skin underneaththe abdominal panniculus dry and edema in the dependantportion leads to a peau d’orange effect and poor tissue perfu-sion. In addition access at the time of surgery is often subop-timal.

To avoid such difficulties panniculectomy was described asearly as 1910 by Kelly [3]. The technique has so far not gained

popularity and is not widely taught to trainees of today, al-though the literature contains favorable reports of its use[4–6]. Other surgical options in the obese patient include aperiumbilical transverse incision [7], a vertical midline incisionwith downward traction on the panniculus [8] or even a formalabdominoplasty [9]. Compared to panniculectomy the lattertechnique involves a much more major plastic procedure withconsiderable undermining and mobilization of the subcutane-ous tissues, often as far as the costal margin.

Obesity is commonly encountered by the gynecological on-cologist because of its association with endometrial carcinoma.Over the past 5 years panniculectomy has been used by theDepartment of Gynecological Oncology at the Mercy Hospitalfor Women in a variety of situations. This paper retrospectivelyreviews the units’ experience of 57 patients who underwentpanniculectomy in combination with gynecological surgery.

PATIENTS AND METHODS

Between 1992 and 1997, 57 patients underwent panniculec-tomy as part of definitive gynecological surgery. Patients wereusually offered either a midline incision or panniculectomy andthe procedure and end effect were discussed fully. All patientswere given perioperative antibiotics and wore antiembolicstockings. Heparin was used for thrombotic prophylaxis in30% of patients.

Surgical Preparation and Procedure

Prior to surgery the area of skin and subcutaneous tissue tobe removed was assessed with the patient both supine and thenstanding (see Fig. 1). The area to be excised was marked outwith an indelible marker pen with the patient standing. Thelower incision commonly followed the skin crease underneaththe panniculus superior to the mons and groins passing to theflank. The upper incision was marked out so as to remove allthe dependant redundant fat and skin, but allowing appositionof the flaps with out tension. This meant that the umbilicus wasshifted inferiorly, sacrificed, or transposed at the patient’srequest. Removal of ‘‘handles’’ from the flank was frequentlyrequired by following the incision well laterally and curving it

GYNECOLOGIC ONCOLOGY70, 80–86 (1998)ARTICLE NO. GO985043

800090-8258/98 $25.00Copyright © 1998 by Academic PressAll rights of reproduction in any form reserved.

cranially. In order to prevent ‘‘dog ears’’ a ‘‘W’’ incision inthe flanks has been described; however, squaring off the endsis a very simple alternative that we found preferable. Markingout the area to be removed was also helpful in demonstratingto the patient the large size of the scar.

General anaesthesia was used for all operations in conjunc-tion with epidural infusions intraoperatively. Epidural analge-sia was used postoperatively. Once anaesthetised the area to beremoved was reassessed by lifting the panniculus and occa-sionally the upper incision was adjusted (see Fig. 2). Skinsterilization included the whole of the abdomen and the flanks

to the operating table laterally. It was our practice to work withtwo surgeons and two assistants where possible, since thisspeeds up the operation. The skin and subcutaneous fat wereremoved as far down as the abdominal wall fascia and perfo-rating blood vessels were either ligated or diathermied asencountered. Cutting diathermy was frequently used. It is im-portant to cut straight down to the fascia along the line ofincision without undermining skin or leaving too much subcu-taneous fat.

The pelvic and abdominal cavities were approached througheither a midline or a lower transverse incision in the rectus

FIG. 1. Patient requiring surgery for the diagnosis of endometrial carcinoma.

81PANNICULECTOMY DURING GYNECOLOGICAL SURGERY

sheath. Choice of incision depended upon the type of surgeryanticipated, the exposure required, and also the presence ofumbilical or incisional herniae. If the upper limit of a midlineincision was compromised, undermining the fat at the level ofthe fascial sheath allowed the incision to be taken furtherupward. Exposure of the pelvic organs was excellent andsuperior to that gained through a midline or transverse incision.Either a Balfour doyen self-retaining or Buchwalter retractorwas used to fully expose the surgical field.

Closure of peritoneum and fascia was undertaken in thenormal manner dependant upon type of incision used and insome cases the sheath was plicated as required. Following thishemostasis was meticulously checked. Two large suctiondrains (No. 12 Redivac) were laid on the fascia laterally andbrought out medially and superior to the wound, so that theycrossed in the midline, and sutured in place. Scarpa’s fasciawas closed by some surgeons with interrupted sutures. Anintracutaneous 3/0 continuous absorbable suture was used toclose the skin.

Postoperative Care

Patients were mobilized and cared for in the same manner asany other patient undergoing laparotomy. If the wound was felt

to be under particular tension the patient was nursed for a fewdays in a jack-knifed position but this was rarely necessary.Wound dressings were changed on the third postoperative dayand drains were left at least 6 days on suction and removedonly when drainage was less than 25 ml per 24 h. Thiscommonly meant the drains were left for 7 to 10 days.

RESULTS

Patient Demographics

The mean age of patients was 55 years (range 31 to 84years). The indications for gynecological surgery are describedin Table 1. Forty-two percent of patients offered panniculec-tomy had a carcinoma of the endometrium or endometrialhyperplasia and 19% had pelvic masses. The procedure pri-marily performed was simple hysterectomy, bilateral salpingo-oophorectomy and bilateral pelvic lymphadenectomy in 19patients, radical hysterectomy with bilateral pelvic lymphade-nectomy in 9, and simple hysterectomy with or without bilat-eral salpingoopherectomy in 18 patients.

Six patients had an incisional hernia repair, two patients hada unilateral or bilateral oopherectomy, one patient underwent

FIG. 2. Patient anaesthetised and marked prior to cleaning the abdomen. It is often neccessary to clean under the panniculus with an assistant supportingit. Once achieved the rest of the abdomen can be cleaned as described.

82 BLOMFIELD ET AL.

bilateral oophorectomy and bilateral pelvic lymphadenectomy,one patient under went debulking surgery for a stage 3Covarian carcinoma which included a small bowel resection, andone had a laparotomy for a recurrent vaginal sarcoma.

The mean weight of patients was 101 kg (70 to 145 kg).Thirty-six patients did not have the weight of the lipectomyspecimen recorded. Of those remaining the mean weight of thelipectomy specimen was 2.7 kg, ranging from 0.9 to 8 kg, 22%having over 4 kg removed. The final weight did depend to someextent on whether the specimen was allowed to dry or not.

The average length of the operation was 2 h 24min, theshortest being 1 h and the longest 6 h. Blood transfusion wasrequired either intra- or postoperatively in 24 (42%) patients,17 of these having 2 units or less. The mean hospital stay was12 days; however, 2 patients had a prolonged postoperativecourse for complications unrelated to the panniculectomy (27and 37 days). Twenty-two (39%) patients went home by day 10and 47 (82%) had done so by day 14.

A postoperative pyrexia was defined as a temperature ofgreater than 38°C on at least two occasions more than 24 hfollowing surgery; 19 (33.3%) patients experienced this. Seven(12%) patients developed wound infections. A minor woundinfection was defined as purulent discharge from the woundwithout significant wound breakdown or fever. Only 4 (7.0%)patients had documented evidence of this. Three (5.2%) pa-tients had a moderate wound infection which was defined aspurulent discharge associated with breakdown, fever, or ac-companying spreading cellulitis. One of these patients requiredresuturing of a small segment of the wound under generalanaesthesia. Other postoperative complications included deepvein thrombosis in 2 patients and pulmonary embolus in 1.There were 5 readmissions postdischarge from hospital; 2 ofthese were for wound infections. There were no postoperativedeaths.

Long-term complications were few, with six patients havingan incisional hernia documented during follow-up.

TABLE 2Wound Infection Rates in Obese and Nonobese Populations of Women Undergoing Abdominal Hysterectomy

and Abdominal and Pelvic Surgery Using Conventional Incisions

InvestigatorPatient population and body

habitus (if documented) Antibiotic cover Drain No.Definition of wound

infections No. %

Soperet al., 1995,Virginia, U.S.A. [11]

Gynecological patientsundergoing abdominalhysterectomy.Average weight 79 kg.

No No n 5 150 Purulent or serous dischargeor pathogen cultured.

17 11.3

Kandulaet al., 1993,Iowa, U.S.A. [12]

Gynecological patientsundergoing abdominalhysterectomy 1985–1989.

Not documented Not documented N/A Purulent discharge orpathogen cultured.

N/A 10.5

Pitkin, 1976,Iowa, U.S.A. [2]

Obese gynecologicalpatients undergoingabdominal hysterectomy1948–1973.Average weight 104 kg.

Not documented Not documented n 5 300 Infection without or withoutseparation.

Evisceration.Total.

84

387

28.0

1.029.0

Gynecological patientsundergoing abdominalhysterectomy 1948–1973.Average weight 63 kg.

n 5 300 Infection without or withoutseparation.

Evisceration.Total.

12

012

4.0

4.0Garrowet al., 1988,

Middlesex, UK [1]Obese women undergoing

abdominal surgery,BMI $30.

Not documented Not documented n 5 42 Serous discharge.Stitch abscess.Pus discharge.

1216

28.62.4

14.3Nonobese women

undergoing abdominalsurgery, BMI,30.

n 5 252 Serous discharge.Stitch abscess.Pus discharge.

353

14

13.91.25.6

TABLE 1Primary Indication for Surgery in 57 Patients

Undergoing Panniculectomy

Indication for surgery No. % Study population

Carcinoma endometrium 19 33.3Endometrial Hyperplasia 5 8.8Carcinoma cervix 10 17.5Cervical intraepithelial neoplasia 3 5.3Benign pelvic masses, fibroids, and benign

ovarian cysts 9 15.8Borderline or malignant ovarian carcinoma 2 3.5Menorrhagia 2 3.5Vaginal sarcoma 1 1.8Incisional hernia 6 10.5

Total 57

83PANNICULECTOMY DURING GYNECOLOGICAL SURGERY

DISCUSSION

Abdominal and pelvic surgery in obese patients is associatedwith high complication rates. Such patients present technicaldifficulties intraoperatively because of poor exposure and thesurgery is often prolonged. Two case-control studies suggestthat the major postoperative morbidity relates to problems withwound healing [1, 2]. Panniculectomy is an underutilized tech-nique that improves surgical access to the pelvis, is cosmeti-cally acceptable to the patient, and has acceptable morbiditycompared to conventional midline and Pfannenstiel ap-proaches. We have described the use of panniculectomy in 57patients over 5 years and are satisfied that it is a safe and usefulprocedure in selected patients.

Panniculectomy was offered to obese patients with endome-trial carcinoma or hyperplasia, to selected patients with earlycervical carcinoma, patients with a pelvic mass that wasthought to be benign, and to a small subset of patients under-going incisional hernia repair, usually in previous midlineincisions. It must be emphasized that although one patient hadan ovarian carcinoma the procedure was not commonly offeredto patients suspected of having ovarian malignancy since upper

abdominal access may be compromised and morbidity in-creased if extensive undermining of fat is necessary. Theapproach described is solely suitable for patients with a pan-niculus and although patients are invariably delighted with theresult it was not used as a primary cosmetic procedure.

The operative time is prolonged although with two surgeonsthis can be minimized and is not excessive. The average timetaken was 2 h 24min, which was shorter than the 4 h 28 minreported by Kohornet al. [5] in his series of 16 patients withsimilar gynecological pathology. The 19 patients with endometrialcancer included in our study had a mean operative time of 2 h 21min. Upon review of a group of 20 patients of normal habitus whoalso had a diagnosis of endometrial carcinoma operated upon bythe same teams during the same time period, a mean operativetime of 1 h 46 min wasnoted. Hence panniculectomy added anaverage of 35 min to the operative time.

Blood loss is also increased but is short lived and occurs asthe surgeons carry out the initial incisions down to fascia. Itcan be reduced by good surgical technique and the use ofcutting diathermy. Blood transfusion was necessary in 41% ofindividuals. It should be noted that an aggressive transfusion

TABLE 3Wound Infection Rates in Obese Patients Undergoing Pelvic Surgery through a Variety of Incisions Considered

by Authors to be Preferable to Conventional Subumbilical Midline and Transverse Suprapubic Incisions

Investigator Patient population Antibiotic Drain No. Definition of wound infections No. %

Kohorn et al., 1995,Yale, Connecticut,U.S.A. [5]

Obese gynecological surgerypatients withPanniculectomy.Average weight 126 kg.

Yes Yes n 5 16 Stitch abscess.Significant infection requiring

suturing.Total.

21

3

12.56.3

18.8Gal, 1994,

New York, U.S.A. [8]Obese gynecological

patients, supraumbilicalincision.Average weight 110 kg.

Yes No n 5 23 Superficial wound separation.Deep subcutaneous separation.Total.

617

26.14.3

30.4

Morrow et al., 1977,California, U.S.A. [13]

Obese gynecologicalpatients, pelvic celiotomy.Average weight 105 kg.

Yes Yes n 5 39 Wound infections. 5 12.8

Prattet al., 1978,Minnesota, U.S.A. [4]

Panniculectomy used to aidabdominal surgery.Average weight not stated.

Not documented Not documentedn 5 84 Wound complications.Secondary closure required.Total.

617

7.11.28.3

Krebs and Helmkamp,1984, Virginia andNew York, U.S.A. [7]

Obese women undergoingpelvic surgery.Periumbilical transverseincision.Median weight 127.8 kg.

Yes Yes n 5 21 Superficial wound infection.Deep wound infections with

prolonged drainage.Evisceration.Total.

22

15

9.59.5

4.823.8

Greeret al., 1990,Washington, U.S.A. [14]

Obese patients undergoingpelvic surgery.Supraumbilical upperabdominal midlineincision.Average weight 129 kg.

Not documented Not documentedn 5 16 Incisional seromas.Wound infection.

30

18.80.0

Cosinet al., 1994,Massachusetts, U.S.A. [6]

Obese patients undergoingpelvic surgery andpanniculectomy.Average weight 116.3 kg.

Yes Yes n 5 9 Wound infection.Seroma.

32

33.322.2

84 BLOMFIELD ET AL.

policy was followed in our hospital during the time period withall patients with a postoperative hemoglobin of less than 10g/dl being transfused. Indeed, only 6 of the 23 patients givenblood had a pretransfusion haemoglobin less than 8 g/dl.

With increasing pressure upon health care providers themean inpatient stay of 12 days seen in this study population isproblematic. This reflects the high-risk nature of this group, thepractice in our unit, and patient expectations of a long inpatientstay. Patients with drainsin situ could be managed at home,once the acute postoperative phase has passed with improvedhome nursing support.

Although only 30% of patients received prophylactic heparin,this included the three patients who had clinical and radiographicevidence of thromboembolic disease postoperatively. Use of epi-durals, antiembolic stockings, early mobilization, and good hy-dration helped to minimize the risk of thromboembolism.

As previously mentioned two case-control studies suggest thatthe major postoperative morbidity encountered in obese patientsrelates to problems with wound healing [1, 2]. According to theNational Research Council Wound Classification Criteria pelvicsurgery involving hysterectomy is classified as a clean contami-nated procedure and wound infection rates of around 10% can be

FIG. 3. Postoperative appearance of patient after panniculectomy, total abdominal hysterectomy, bilateral salpingo-oophorectomy, and bilateral pelviclymphadenectomy. Over 8 kg of subcutaneous tissue was removed.

85PANNICULECTOMY DURING GYNECOLOGICAL SURGERY

expected. In an extensive review of morbidity following abdom-inal hysterectomy for benign disease the most commonly reportedwound infection rates were between 4 and 8%. Soperet al. [11]found a wound infection rate of 11.3% for a population of womenundergoing abdominal hysterectomy. They suggested that woundinfections rates are directly related to depth of subcutaneous tissueat the site of the incision rather than to actual body weight or bodymass index. The wound infection rate was 40% in patients with 6cm or more of subcutaneous tissue but was zero in patients witha depth of subcutaneous tissue less than 3 cm.

Use of prophylactic antibiotics appears to reduce the incidenceof postoperative infectious morbidity although few studies includea separate analysis of incisional wound infection, reporting infec-tious morbidity as combinations of vaginal cuff cellulitis, pelvicabscess, and wound infections. Quoted wound infection rates alsodepend on the population studied, the type of surgery undertaken,the definitions of infection used, the length of follow-up, and theuse of suction drainage. Table 2 illustrates reported wound infec-tion rates in variety of obese and nonobese populations of womenundergoing abdominal and pelvic surgery using conventional in-cisions. Table 3 presents the wound infection rates in populationsof obese women undergoing pelvic surgery through a variety ofpreferred incisions. In the present study a low rate of woundinfection was observed, 7.0% for mild cases and 5.2% for mod-erate cases. Only 1 of 57 patients required resuturing undergeneral anaesthesia. We accept that this study is a retrospectivereview of patients notes and that the true wound infection inci-dence may be underestimated because of poor documentation;however, our clinical experience also supports these findings.

We conclude that panniculectomy is an underutilized techniquethat allows excellent access to the pelvis for both radical and moresimple gynecological surgery. The wound heals extremely wellwith low rates of complications compared to conventional inci-sions. Although operative time may be prolonged this is notexcessive and the time saved because of adequate pelvic accessshould be appreciated. We found no evidence of increased post-operative complications. Problematic wound infections are kept toa minimum, and patients are pleased to be free of their unsightly

panniculus and the difficulties in personal hygene that they present(see Fig. 3). We would recommend that this technique be consid-ered both by the general gynecologist as well as gynecologicaloncologist when an obese patient with a panniculus requirespelvic surgery.

REFERENCES

1. Garrow JS, Hastings EJ, Cox AG, Thomas MT, Meade TW: Obesity andpostoperative complications of abdominal operation. BMJ 297:181, 1988

2. Pitkin RM: Abdominal hysterectomy in obese women. Surg GynecolObstet 142:532–536, 1976

3. Kelly HA: Excision of the fat of the abdominal wall—lipectomy. SurgGynecol Obstet 10:229–231, 1910

4. Pratt JH, Irons GB: Panniculectomy and abdominoplasty. Am J ObstetGynecol 132:165–168, 1878

5. Kohorn EI: Panniculectomy as an integral part of pelvic operation is anunderutilized technique in patients with morbid obesity. J Am Coll Surg180:279–285, 1995

6. Cosin JA, Powell JL, Donovan JT, Stueber K: The safety and efficacy ofextensive abdominal panniculectomy at the time of pelvic surgery. Gy-necol Oncol 55:36–40, 1994

7. Krebs HB, Helmkamp BF: Transverse periumbilical incision in the mas-sively obese patient. Obstet Gynecol 63:241–245, 1984

8. Gal D: A supraumbilical incision for gynecological neoplasms in themorbidly obese patient. J Am Coll Surg 179:18–20, 1994

9. Voss SC, Sharp H, Scott JR: Abdominoplasty combned with gynecolog-ical surgical procedures. Obstet Gynecol 67:181–186, 1986

10. Harris WJ: Early complications of abdominal and vaginal hysterectomy.Obstet Gynecol Surv 50:795–805, 1995

11. Soper DE, Bump RC, Hurt GH: Wound infection after abdominal hyster-ectomy: effect of depth of subcutaineous tissue. Am J Obstet Gynecol173:465–471, 1995

12. Kandula PV, Wenzel RP: Post operative wound infection after abdominalhysterectomy: a controlled study of increased duration of hospital stay andtrends in post operative wound infection. Am J Infect Control 21:201–204,1993

13. Morrow CP, Wilfredo LH, Townsend DE, Disaia PJ: Pelvic ceiotomy inthe obese patient. Am J Obstet Gynecol 127:335–339, 1977

14. Greer BE, Cain JM, Figge DC, Shy KK, Tamimi HK: Supraumbilicalupper abdominal incision for pelvic surgery in the morbidly obese patient.Obstet Gynecol 76:471–473, 1990

86 BLOMFIELD ET AL.