Case Report

Laparotomy wound recurrence of endometrial carcinoma

Victor Macias,a,b,* Barbara Baiotto,c JosePardo,a Fernando Mun˜oz,b and Pietro Gabrieleb

a Department of Radiotherapy, Hospital General de Catalunya, c/Gomera s/n, 08190 Sant Cugat del Valles, Barcelona, Spainb Department of Radiotherapy, Istituto per la Ricerca e la Cura del Cancro, 10060 Candiolo, Turin (Italy)

c Department of Medical Physics, Istituto per la Ricerca e la Cura del Cancro, 10060 Candiolo, Turin (Italy)

Received 5 March 2003

Abstract

Background. Most endometrial cancer relapses are either pelvic or distant metastases. Recurrences in the laparotomy scar are extremelyrare, as to our knowledge this is only the sixth reported case and the first treated with brachytherapy. We also present a review of theliterature with an emphasis on radiation therapy treatment.

Case. A 64-year-old woman underwent total abdominal hysterectomy and bilateral salpingo-oophorectomy for IIIA endometrialadenocarcinoma, followed by systemic chemotherapy and pelvic external-beam radiotherapy. Three years later, a recurrence in thelaparotomy scar was resected and additional chemotherapy was administered. A solitary mass in the abdominal wound was then found 4years after that and was reirradiated by combining external-beam radiation therapy and interstitial brachytherapy after complete resection.Histology of the abdominal wall implants was very similar to that of the primary tumour.

Conclusion. Laparotomy wound recurrence (LWR) for endometrial cancer usually appears as a solitary mass that grows slowly withoutconcomitant failure sites a few years after hysterectomy. In these six patients there is no common histology pattern that could suggestpredictive factors for LWR. Common postoperative irradiation does not deliver sufficient a dose to the anterior abdominal wall to preventthis kind of failure.© 2003 Elsevier Inc. All rights reserved.

Keywords: Endometrial cancer; Recurrence; Radiotherapy; Brachytherapy

Introduction

Endometrial cancer is the most commonly diagnosedgynecologic neoplasm [1]. Most endometrial cancerrelapses are either pelvic or distant metastases and occurwithin the first 2 years after hysterectomy [2]. Recur-rences in the laparotomy wound are extremely rare as,to our knowledge, only five cases have been previouslyreported [3–7]. This case is presented with a reviewof previous similar cases, possible mechanisms for thistype of spread and an analysis of radiation therapy treat-ment.

Case report

R.M., a 64-year-old white female, gravida 1, with non-insulin-dependent diabetes and hypercholesterolemia, wasoperated on for eventration in November 1994, when a leftovarian mass was suspected. An abdominal CT-scanshowed increased uterus dimensions with hypodense con-tent and a 5� 4 � 5 cm anexus mass, without increasedlymph nodes. Chest X-ray was normal. Ca 125 and Ca 19.9levels were 54 Ul/ml (normal� 35 UI/ml) and over 240UI/ml (normal � 37 UI/ml), respectively. The patient un-derwent total abdominal hysterectomy and bilateral sal-pingo-oophorectomy through midline infraumbilical inci-sion in January 1995. Pelvic lymphadenectomy was notperformed. Pathology examination revealed a uterus weigh-ing 100 g with a 7.5 cm uterine cavity completely occupied

* Corresponding author. Fax:�34-93-589-26-18.E-mail address: maciasrt@yahoo.com (V. Macias).

R

Available online at www.sciencedirect.com

Gynecologic Oncology 91 (2003) 429–434 www.elsevier.com/locate/ygyno

0090-8258/$ – see front matter © 2003 Elsevier Inc. All rights reserved.doi:10.1016/j.ygyno.2003.07.001

by an exophytic necrosed mass involving the cervical canal.On microscopy, a G2 endometrial adenocarcinoma withfocal squamous metaplasia was seen to have infiltrated morethan half the myometrium with the involvement of thecervix and the left Fallopian tube os. Focal ovarian endo-metriosis in the right parametrium was observed. Vaginaland parametrial resection margins were free of disease.Peritoneal washings were positive to neoplastic cells. It wasstaged as pT3a N0 M0, stage IIIA (FIGO). Serum tumoralmarkers returned to normal levels. Systemic chemotherapyconsisting of three cycles of adryamicin with cisplatinumwas administered between April and July. Six weeks later,the pelvis was irradiated to a dose of 50 Gy in 2 Gy/dayfractions using four 10 � 15 cm obliqued fields of 1.5 MVphotons. The patient was free of disease until February1998, when an abdominal CT scan detected two cysticnodes localized at the level of the top of the laparotomy scarmeasuring 7 � 5 � 1.5 cm (Fig. 1). There was no evidenceof intraabdominal recurrence or distant metastases. Ca 19.9serum level increased to 113 Ul/ml, whereas Ca125 was

normal. The histological examination of the resected nodesconfirmed the presence of adenocarcinoma with similarcharacteristics to the initial tumour (adenocarcinoma withsquamous metaplasia) infiltrating subcutaneous fat and scarconnective tissue. Serum markers decreased to normal lev-els. The patient was treated with adjuvant systemic chemo-therapy using the same schema as before.

In November 2001, the patient was referred to our insti-tution when diagnosed with a second abdominal scar recur-rence. The mass, measuring 10 � 12 cm, had infiltrated theabdominal wall without involving adjacent intraperitonealorgans such as the colon or the small intestine (Fig. 2). Noabdominal lymph nodes or metastases were detected. Ca19.9 and Ca 125 increased to 683 and 43 UI/ml respectively.The tumour was completely removed, although resectionmargins were considered microscopically affected. Histo-logical examination showed adenocarcinoma similar to thatpreviously described. Serum markers decreased to normallevels. Four months later, the patient underwent external-beam radiotherapy in the scar using a direct field of 12 MeV

Fig. 1. Abdominal CT scan showing the first laparotomy wound recurrence (small node) in 1998.

430 V. Macias et al. / Gynecologic Oncology 91 (2003) 429–434

electrons produced by a linear accelerator. The dose deliv-ered was 36 Gy in 2-Gy/day fractions over 4 weeks. Ninedays afterwards, an implant of interstitial brachytherapy(BT), CT-guided, was performed (Fig. 3). Under sedationand local anaesthesia 11 vectors were placed into the fat ofthe anterior abdominal wall crossing the scar. Then a secondCT scan was made for dosimetry purposes and a 3D isodoseplan was prepared. Finally, plastic implant catheters wereconnected to an afterloading computer-controlled machinewith a high-dose-rate iridium-192 source. The radiationtime was 14 min for each fraction. The delivered dose was20 Gy in four daily fractions of 5 Gy over 1 week. Totalprescribed dose (EBRT � BT) to the scar and 4.5 cmsurrounding tissues was estimated to be 64 Gy. Catheterswere gently removed immediately after the last session,without anaesthesia. Treatment was very well toleratedwithout analgesics or other drugs and interference withnormal activities was minimum. To date, 7 months aftercompletion of the BT, skin hyperpigmentation is the only

sequel of the two radiotherapy treatments. Abdominal CTscan and ultrasounds show no abnormality.

Discussion

The characteristics of endometrial cancer patients withreported laparotomy wound recurrence (LWR) are summa-rized in Table 1.

How cancer cells implant themselves in the laparotomywound and grow in situ is not known. Despite the commonpresence of cancer cells intraperitoneally, as proved in pos-itive peritoneal washings, the implantation and growth ofcells in the abdominal wound is very rare. The long disease-free interval and the solitary recurrence in the abdominalscar without other concomitant metastases in most casessuggest that the recurrence could be due to cells beingimplanted directly in the wound and then growing slowly,rather than being spread via blood or lymphatic vessels fromthe initial tumour. Unlike LWR, skin metastases of endo-

Fig. 2. Second recurrence in the laparotomy scar (2001).

431V. Macias et al. / Gynecologic Oncology 91 (2003) 429–434

metrial cancer outside the laparotomy wound appear asmultiple firm nodules that grow quickly in various bodysites with coexistent metastasis in different organs. Thisreflects a generalised spread through lymphatics and bloodvessels. There are two possible mechanisms to explain howmalignant cells reach the abdominal wound: (1) Penetratingthe thickness of the uterine wall or the Phallopian tube tospread intraperitoneally through the laparotomy wound.Deep myometrial invasion or positive peritoneal washingsto atypical cells were assessed in all patients except one,whose data were not available. (2) Spilling through thecervical os at the moment of hysterectomy and implantingdirectly in the incision during the surgical act. They willremain viable or not depending on factors such as theimmune status of the patient.

Two of six patients with LWR had concomitant failures,in vagina [3] and paraaortic lymph nodes [4]. The presenceof cancer in these sites shows an unusual dissemination forendometrial cancer. It may be explained, as suggested by

Espinos et al. [4], by blockage of lymph vessels with thepossibility of retrograde flow or formation of alternativechannels, both caused by the patient’s age or the surgicaltreatment.

With regard to overall survival, the patient with LWRand concomitant abdominal lymph nodes died 3 monthsafter diagnosis. There is no available data from the patientwith concomitant vaginal metastasis. Three of four patientswith isolated LWR, including the present case, had noevidence of disease 3 years (16–47 months) after implantresection. The other patient developed lung metastases 2years after the primary treatment with no abdominal scarrecurrence. These data suggest that when the laparotomyscar is the unique initial failure site, a disease-free intervalof at least a few years can be expected and treatment mustbe realized with curative intent. All of the 6 patients weretreated in a different way, including mass resection, exter-nal-beam radiotherapy (alone or postoperative), BT, chemo-therapy, and hormonal therapy. Due to the rarity of this

Fig. 3. Abdominal CT scan showing isodose curves of the brachytherapy treatment. Prescribed dose was 20 Gy to 85% isodose curve (light blue). Scar iscompletely involved in the 50% isodose curve (yellow). Notice that small parts of the small intestine are placed between the 50 and 30% (green) isodosecurves.

432 V. Macias et al. / Gynecologic Oncology 91 (2003) 429–434

clinical presentation it has not been possible to establish theoptimal treatment combination. In the absence of metasta-sis, surgery may be performed as initial treatment if theabdominal mass seems resectable.

Postoperative EBRT can be administered in solitaryLWR to avoid a second abdominal scar recurrence as in thecase we are reporting. We suggest a dose of about 70–75 Gyfocused on the abdominal wound, which is similar to thecurative treatment of vaginal cuff scar recurrence [8]. Giventhe documented reduction of tumour volume, even withmoderate doses [4], EBRT with concomitant hormonal ther-apy (medroxyprogesterone acetate) may be used in pallia-tive and preoperative treatments.

Irradiation of the pelvis after hysterectomy is a commonpractice to prevent locoregional failures of endometrial can-cer. A dose to 45–50 Gy is considered sufficient to treatmicroscopic disease [8]. As shown in Table 1, three patientswere irradiated after hysterectomy, but developed thewound implant 1.5–5 years afterwards. Possible reasons forthese failures are as follows:

1. The aim of postoperative radiation therapy of endo-metrial cancer is to “sterilise” the possible microscopicresidual disease at the level of the tumour bed, pelvic lymphnodes, and vaginal cuff scar. At the same time, to avoidsignificant toxicity, rectum, bladder, skin, and femoralheads have to be protected as much as possible. As a resultof these goals, the usual standard technique consists of fourisocentric beams at 0°, 90°, 180°, and 270° gantry angles to

the pelvis using megavoltage photons. With this beam ar-rangement, higher doses are placed in the centre of thepelvis while sensitive organs are irradiated under tolerancelimits. Reviewing dose distribution in a random group ofpatients with endometrial cancer treated with the describedtechnique in our institution, the laparotomy wound is un-derdosed, receiving around 60% or less of prescribed dose(Fig. 4).

2. On the other hand, the cranial border of the radiationfields are usually positioned at the level of the fourth andfifth lumbar vertebrae to permit the irradiation of externaliliac lymph nodes while minimising the exposure of thesmall intestine. However, as reported by Khalil et al. [7], thecranial part of the laparotomy incision is sometimes aboveL5 level, so it is not included in the irradiation fields.

In our opinion, the current radiation therapy plan shouldnot be modified because the acute and chronic toxicity,mainly intestinal, to be expected would be greater than thetheoretical benefit of preventing this rare type of spread.

When the laparotomy scar has to be irradiated afterprevious radiotherapy of the pelvis, as in the present case,the technique explained above cannot be used again, be-cause the small intestine has already received 50 Gy. Thisdose corresponds to the tolerance dose to keep the proba-bility of severe chronic toxicity from therapeutic irradiationunder 5% [9]. Electron beams or BT, thanks to its short dosefalloff, allow this goal to be obtained. In the case we arereporting the combination of a direct electron beam and then

Table 1Characteristics of endometrial cancer patients with laparotomy wound recurrence

Author Agea

(years)Treatment of PT pT stageb

(Grade)P washings DFI

(years)Concomitantrecurrencesites

Treatment of LWR RT dose/DPF(Gy)

Follow-upc

(months)

Chapman (1988) 65 RT NA NA 7 Vagina CT � PG NA NA

Espinos (1993) 77 TAH-BSO �EBRT� � BT

pT1c(G2-3)

� 1month

Paraaorticlymph nodes

EBRT� � PG 43�24�

Deceased(3)

Curtis (1994) 50 TAH-BSO �P32W � PG

pT1b(G1)

� 1.5 None Resection � EBRT 55.8/1.8 Lungmetastasis(24)

Kotwall (1994) 65 TAH-BSO pT1c(G1-2)

� 7 None Resection 0 NED(�15)

Khalil (1998) 58 TAH-BSO �EBRT

pT1c(G1-2)

� 5 None Resection � PG 55/NA NED (48)

Present case 56 TAH-BSO �CT � EBRT

pT3a(G2)

� 3 None Resection � CT (1strecurrence)

50/NA NED (11)

6.5 Resection � EBRT � BT� PG (2nd recurrence)

36/2 (EBRT)

20/5 (BT)

a Age at diagnosis of primary endometrial carcinoma; PT, primary tumour; RT dose, radiation therapy dose either in initial treatment or in recurrencetreatment; TAH-BSO, total abdominal hysterectomy with salpingo-oophorectomy; EBRT, external-beam radiation therapy; BT, Brachytherapy; P32W,intraperitoneal 32P washings; CT, chemotherapy.

b T-stage, primary tumour extent, according to UICC, 5th edition; NA, not available; DFI, disease-free interval from the primary treatment; AWR,abdominal wall recurrence; PG, progestin; DPF, dose per fraction.

c Follow-up, follow-up since scar implant resection; NED, no evidence of disease.

433V. Macias et al. / Gynecologic Oncology 91 (2003) 429–434

interstitial BT achieved the best dose distribution. To date,neither acute toxicity (except a mild skin reaction) norchronic toxicity has been assessed.

Considerations about serum level of Ca 19.9 in LWR arenot available. In the present case, Ca 19.9 increased clearlywhen recurrence was present and diminished to normallevel after treatment. This marker is being used in thepatient’s follow-up in order to obtain an early diagnosis ofrecurrence.

References

[1] American Cancer Society (ACS). Cancer facts and figures—1998.Atlanta, GA: ACS, 1998.

[2] Perez CA, Bedwinek JM, Breaux SR. Patterns of failure after treatmentof gynecologic tumours. Cancer Treat Symp 1983;2:217.

[3] Chapman GW, Fabacher P, Thompson H. Incisional recurrence ofendometrial carcinoma. J Nat Med Assoc 1988;80:350.

[4] Espinos J, Garcia-Patos V, Guiu XM, Delgado E. Early skin metastasisof endometrial carcinoma: case report and review of literature. Cutis1993;52:109.

[5] Curtis MG, Hopkins MP, Cross B, Tantri MD, Jenison EL, Rehmus E.Wound seeding associated with endometrial cancer. Gynecol Oncol1994;52(3):413–5.

[6] Kotwall CA, Kirkbride P, Zerafa AE, Murray D. Endometrial cancerand abdominal wound recurrence. Gynecol Oncol 1994;53(3):357–60.

[7] Khalil AM, Chammas MF, Kaspar HJ, Shamseddine AI, Seoud MA.Case report: endometrial cancer implanting in the laparotomy scar. EurJ Gynaecol Oncol 1998;19(4):408–9.

[8] Chao KSC, Perez CA, Brady LW 1999. Radiation oncology: manage-ment decisions. Philadelphia: Lippincott-Raven p. 513–4.

[9] Emami B, Lyman J, Brown A, Coia L, Goitein M, Munzenrider JE,Shank B, Solin LJ, Wesson M. Tolerance of normal tissue totherapeutic irradiation. Int J Radiat Oncol Biol Phys 1991;21(1):109 –22.

Fig. 4. Abdominal CT scan showing isodose curves of a standard beam arrangement in the postoperative irradiation of endometrial carcinoma patients. Gantryangle beams are 0° (beam 1), 90°, 180°, and 270° (beam 4). Common prescribed dose is 45–50 Gy to the ICRU point, i.e., the intersection of the four beams,which corresponds to 100%. Lymphatic regions are included to the 95% isodose curve (yellow). The laparotomy scar receives less than 60% of the prescribeddose (60% isodose curve in light blue, outer elliptic isodose line).

434 V. Macias et al. / Gynecologic Oncology 91 (2003) 429–434