Case 18-2005 — A 45-Year-Old Woman with a Painful Mass in the Abdomen
http://www.100md.com
《新英格兰医药杂志》
Presentation of Case
A 45-year-old woman was admitted to this hospital with a painful abdominal mass.
She had first noticed a mass in her lower abdomen on the right side two and a half years earlier. There was intermittent severe pain at the site that occurred approximately once a month and lasted four to five days; it was worse during some months than others. She described the pain as feeling "like a hot brick." Five weeks before admission, she came to the surgical clinic of this hospital for further evaluation. A slightly tender mass, 4 cm in diameter, was palpated in the right lower abdomen. The remainder of the physical examination revealed no abnormalities.
An abdominal ultrasonographic study performed at this evaluation showed a highly vascular, mixed echoic mass, 3.5 cm by 3.5 cm by 1.0 cm, in the lower right anterior abdominal wall. Abdominal and pelvic computed tomographic (CT) scanning after the oral and intravenous administration of contrast material three weeks before admission confirmed the presence of an irregularly shaped, nonenhancing, heterogeneous lesion within the right lower anterior abdominal wall. The liver was fatty and also contained an enhancing lesion, 1.9 cm in diameter, in the lower right lobe, which showed early washout of contrast material. There were multiple, low-attenuation lesions in the liver that were thought to represent cysts. The other organs appeared to be normal. Magnetic resonance imaging (MRI) performed before and after the administration of gadolinium two weeks before admission to evaluate the liver lesion confirmed the presence of an early enhancing lesion, 1.7 cm by 1.8 cm, in the right lobe of the liver that had a minimally increased signal on a T2-weighted sequence and was isointense on a T1-weighted sequence. Multiple bright lesions on T2-weighted images were present within the liver, and they were thought to be cysts. No lymphadenopathy was seen. MRI of the other abdominal organs showed no abnormalities.
Scoliosis had developed in this patient during her adolescence, and Harrington rods were placed when she was 15 years of age. At 30 years of age, she was treated for a grade 1 cerebellar pilocytic astrocytoma, with two resections and postoperative irradiation with a dose up to 55.8 Gy. As a result of the astrocytoma, she had double vision, which required special corrective lenses, and ataxia, for which she had learned to compensate after rehabilitation therapy. A fibroadenoma of the left breast was excised when she was 28 years of age, as was a fibroadenoma of the right breast when she was 30 years of age. Three years before admission, bilateral intraductal papillomas of the breasts were excised, with the additional findings of focal atypical ductal hyperplasia and lobular neoplasia in situ. She had had two pregnancies, and both of her daughters were delivered by cesarean section. A hysterectomy because of uterine fibroids had been performed at another hospital three years before admission. At that time, pathological examination confirmed the presence of multiple leiomyomas, benign findings in the endometrium, and chronic cervicitis. The ovaries had not been removed.
She lived with her husband and daughters, who were all well. She did not smoke and drank three to four alcoholic drinks per month. Her only medication was a daily multivitamin.
On admission, the patient's vital signs were normal. The height was 167.5 cm and the weight 49.5 kg. The abdomen had a low, 6-cm scar and was soft, with no organomegaly. An irregular, indurated mass in the right lower quadrant, 4 cm in diameter, with no overlying skin retraction or pitting, was slightly tender to palpation. The remainder of the physical examination showed no abnormalities.
The results of routine laboratory tests performed on admission were within the normal ranges. An electrocardiogram showed a normal sinus rhythm and slight left atrial enlargement. A chest radiograph showed probable pulmonary emphysema with bullous changes in the upper lobes. No infiltrates, masses, effusions, or lymphadenopathy were seen. Harrington rods were noted in the thoracic spine.
A diagnostic procedure was performed.
Differential Diagnosis
Dr. Michael G. Muto: May we review the radiologic studies?
Dr. Mary Jane O'Neill: Sagittal and transverse views as visualized on the superficial ultrasonographic imaging of the anterior abdominal wall show that within the fibers of the rectus muscle there is a well-circumscribed, hypoechoic, solid-appearing mass lesion that is displacing the fibers of the right rectus muscle (Figure 1A). The CT examination shows asymmetry of the rectus; there is a high-density, moderately lobulated mass within the substance of the right abdominal rectus muscle (Figure 1B). On a CT scan obtained at the same session, there is a relatively ill defined enhancing mass near the surface of the right lobe of the liver (Figure 1C). The ovaries appeared to be normal.
Figure 1. Imaging Studies of the Abdomen.
An ultrasonographic image of the lower right abdominal wall (Panel A) shows a solid, hypoechoic lesion (arrows) in the right rectus muscle. The small arrowheads indicate the placement of the focal zone of the ultrasonographic beam. A CT image (Panel B) through the lower abdomen, which was obtained after the administration of contrast material, reveals an enhancing, solid mass (arrow) in the right rectus muscle. The same study shows a lesion (arrow) near the surface of the right lobe of the liver (Panel C). A T1-weighted axial MRI scan obtained after the administration of contrast material in the arterial phase shows a small arterially enhancing lesion (Panel D).
MRI was performed to evaluate the liver lesion. On the T1-weighted images, the lesion is dark, without evidence of any hemorrhage. It is ill defined on T2-weighted images, and it has prominent enhancement after gadolinium administration (Figure 1D). It is unclear whether this lesion is intraparenchymal or on the surface of the liver, and there are no specific features on imaging of either a benign or a malignant lesion. In a patient of this age and sex, the features that are apparent on MRI would be consistent with focal nodular hyperplasia.
Dr. Muto: This multiparous woman with a history of two cesarean sections and a hysterectomy had a mass of the abdominal wall that had been present for more than two years and was intermittently painful. The radiographic studies showed that the mass was heterogeneous and located in the substance of the right rectus abdominis muscle, 2 cm above the Pfannenstiel incision and 3 cm to the right of midline. Radiographic imaging also showed two normal-appearing retained ovaries in the pelvis and no suggestion of ventral hernia or bowel obstruction. The solitary lesion noted in the liver could not be characterized as benign in appearance. All the other abdominal viscera appeared normal.
The differential diagnosis of a painful abdominal-wall mass is shown in Table 1.1
Table 1. Differential Diagnosis of a Painful Abdominal-Wall Mass.
Hernias and Abscesses
A wide variety of ventral hernias may occur as a painful abdominal-wall mass. The location of the mass makes an inguinal or femoral hernia improbable in this case. A spigelian hernia occurs at the posterior lateral abdominal-wall fascia through the linea semilunaris, and although it is rare, a mass can appear in this location, particularly if the hernia were to cause incarceration or strangulation. Finally, an incisional hernia must certainly be considered, especially if there have been multiple laparotomies performed through the same incision. In order to understand how an incisional hernia can occur in a region that is cephalad and lateral to the skin incision, it is necessary to review how this incision is performed.
In the Pfannenstiel incision, the skin and fascia are incised transversely. The fascia is then dissected from the underlying rectus abdominis muscles in a cephalad and caudad direction, which creates subfascial spaces above and below the skin incision (Figure 2). This technique allows the rectus abdominis muscles to be separated along the midline. A vertical laparotomy is then performed. The Pfannenstiel incision is widely used in both obstetrics and gynecology, because it affords generous access to the pelvis and is stronger than a vertical incision and ensures excellent cosmetic results. One drawback to the incision is the creation of the subfascial spaces, which can increase the risk of infection or hematoma formation, as compared with a vertical midline incision. Also, although the subfascial space is readily created during a primary incision, when the incision is reopened, this subfascial space is often heavily scarred and difficult to recreate, resulting in trauma to both the underlying rectus muscle and the fascia. Such trauma may increase the risk of hematoma, infection, or fascial disruption. Therefore, although it is unusual for incisional hernias to develop through a Pfannenstiel incision, they may occur either beneath the skin incision or superolateral to it.
Figure 2. The Pfannenstiel Incision.
The incision begins with a transverse skin and fascial incision (inset). The fascia is dissected from the underlying rectus abdominis muscle in a cephalad and caudad direction, which creates subfascial spaces above and below the skin incision. The rectus abdominis muscles are separated along the midline, and a vertical laparotomy is then performed.
I do not know if any of this patient's three prior laparotomies were complicated by infection or hematoma formation; however, just having three procedures performed through the incision is risk enough for hernia formation to be a consideration. The cyclicity of her pain, the clinical presentation, and the imaging studies, however, are not consistent with this diagnosis.
A diverticular, appendiceal, or tubo-ovarian abscess can secondarily infect the rectus muscle and lead to the development of a chronic abdominal-wall abscess. There is no evidence, however, to support an infectious cause in this case. There is also no history of traumatic injury to the muscle, belly rupture, or prior instrumentation.
Tumors
Any benign or malignant tumor of muscle, integument, nerve, or blood vessels can involve the rectus muscle and occur as a painful mass. Desmoid tumors are benign but locally destructive lesions. These tumors often occur in young women after a pregnancy, and they may grow slowly in the abdominal wall or surgical scar. There is evidence that suggests that these tumors are hormone-sensitive, including the predominance of the lesion in women of reproductive age and regression in menopausal women and among those who are receiving tamoxifen therapy. This tumor would not, however, be expected to present with cyclic pain, and it usually appears as a homogeneously solid mass on CT images, rather than a complex mass.2
Endometriotic Implants
Endometriosis is the ectopic growth of endometrial glands and stroma outside the endometrial cavity. It affects 3 to 10 percent of all women of reproductive age. About 4 of every 1000 women between 15 and 64 years of age are hospitalized with endometriosis each year.3 The disease is a major cause of infertility and chronic pelvic pain.4 The most common extrauterine sites involved include the ovary, fallopian tube, and adjacent pelvic tissues. Pelvic spread is thought to occur by retrograde menstruation by way of the fallopian tube, during which viable endometrial cells are implanted in dependent portions of the pelvis. Rarely, the disease occurs outside the pelvis in sites such as the lungs, resulting in cyclic hemoptysis, hemothorax, or pneumothorax, and within retroperitoneal lymph nodes and liver parenchyma. These findings suggest lymphatic or vascular dissemination. Finally, the occurrence of endometriosis in men taking estrogen therapy suggests yet a third pathogenetic mechanism: endometriosis arising as a result of coelomic metaplasia.
Endometriotic implants in scars resulting from episiotomy and cesarean section most often occur at the time of vaginal or cesarean delivery and are the result of direct implantation of viable endometrial cells into the subcutaneous or subfascial spaces that are exposed by the surgical incisions on the abdominal wall or perineal body. The classic manifestations of endometriosis in surgical scars are focal cyclic pain and a slow-growing mass.5,6,7,8 The pain can often be timed to the menses in women who have not undergone oophorectomy. Progressive cycles of bleeding and local inflammatory response lead to the development of a complex vascular or fibrotic mass. The overall incidence of endometriomas occurring in cesarean-section scars has been estimated at 0.03 to 0.15 percent.9
Because of the clinical similarity between endometriosis in a surgical scar and an incisional hernia, general surgeons, rather than gynecologists, are often the first to evaluate patients with this disorder. Although the development of endometriosis by way of implanted cells is widely reported in the gynecologic literature, this entity is not well recognized among general surgeons. As a result, the diagnosis is often not considered. In particular, the telltale history of cyclic pain is either not elicited or, as in the case under discussion, not appreciated. As Nirula and Greaney report,10 among 10 cases of incisional endometriosis in caesarean-section scars, in only 2 cases was the diagnosis suspected preoperatively, and these were diagnosed by a single surgeon who was aware of the disease from prior experience.
The most probable diagnosis in this case is endometriosis of the anterior abdominal wall. Although endometrial cells can implant within a ventral hernia sac, it seems far more likely that implantation occurred beneath the upper flap of the Pfannenstiel incision after the patient's primary or repeated cesarean section or during the hysterectomy.
The Hepatic Lesion
There is a second finding that must be explained in this case — the suspicious hepatic lesion noted on CT scanning and subsequent MRI. This may be entirely unrelated to the patient's abdominal-wall mass; although it could not be characterized as benign on the basis of radiologic criteria, its features were thought to be consistent with focal nodular hyperplasia. Malignant degeneration of endometriotic implants has been widely reported in the gynecology and gynecologic-oncology literature.11,12 Clear-cell and endometrioid adenocarcinomas are by far the most common histologic types reported. Endometrioid and clear-cell carcinomas predominate in intraperitoneal and ovarian implants, whereas clear-cell cancers more frequently are associated with extraperitoneal implants.13
In conclusion, the most likely diagnosis for this 45-year-old woman's illness was an implant of endometriosis within her surgical scar. Given her ongoing symptoms, complete excision would be the most appropriate procedure for both diagnosis and therapy. If the mass contained evidence of an endometriosis-associated cancer, a liver metastasis should be suspected.
A Physician: Were all three radiologic studies necessary to make this diagnosis?
Dr. Muto: The ultrasonographic study was a reasonable choice for primary imaging of this ill-defined mass. Once a mass was clearly defined, abdominal and pelvic CT scanning was required to rule out an incarcerated hernia. When the CT scanning revealed the unexpected finding of a suspicious-appearing cystic liver lesion, her physicians were obliged to evaluate it further with an MRI scan.
A Physician: Is MRI useful for the diagnosis of endometriosis?
Dr. O'Neill: Pelvic endometriosis has characteristic increased signal intensity on T1-weighted images; however, when the endometriosis is located within fascial or muscular tissues, it often does not produce the characteristic imaging findings. In this patient, CT scanning would have been the single best imaging test because it can localize the palpable abnormality and allow the clinician to better assess the remainder of the abdomen and pelvis for findings that would affect the differential diagnosis.
Dr. Nancy Lee Harris (Pathology): There was no evidence of endometriosis on the uterus that was removed at hysterectomy. Is the risk of endometriosis in a surgical incision independent of the presence of pelvic endometriosis?
Dr. Muto: Many reported cases had no evidence of pelvic endometriosis.
Dr. Harris: This patient was seen by a general surgeon, whose differential diagnosis included endometriosis in a scar, but the primary diagnosis was a desmoid tumor.
Clinical Diagnosis
Desmoid tumor of the abdominal wall.
Dr. Michael G. Muto's Diagnosis
Endometriosis of the abdominal wall.
Pathological Discussion
Dr. Esther Oliva: A specimen of soft tissue measuring 9.0 cm by 6.0 cm by 1.2 cm was received in the pathology department. Sectioning revealed areas of hemorrhage and cyst formation. On microscopical examination, there was a biphasic growth of glands and stroma in a background of fibrous tissue. Some of the glands were cystically dilated (Figure 3A). The glands were coiled with abundant luminal secretions and were lined by cuboidal and columnar cells with basally located nuclei (Figure 3B). Cytologic atypia and mitotic activity were absent. The glands were surrounded by a cellular stroma that was composed of small cells with scant cytoplasm and oval-to-elongated nuclei, which were whorled around small blood vessels (Figure 3B). Edema, focal elastosis, and areas of recent hemorrhage were evident (Figure 3A).
Figure 3. Histologic Sections of the Mass in the Abdominal Wall (Hematoxylin and Eosin).
An endometrial gland, cystically dilated (Panel A), is embedded in an edematous stroma, adjacent to areas of scarring and elastosis. A higher-power view (Panel B) shows endometrial glands lined by cells with tall cytoplasm and basally located nuclei, with no cytologic atypia.
Endometriosis is defined as the presence of endometrial-type glands and stroma outside the endometrium and myometrium. It most commonly involves the ovaries, the uterine ligaments, the rectovaginal septum, the cul-de-sac and peritoneum of the uterus, the fallopian tubes, the rectosigmoid, the ureter, and the bladder.14 Endometriosis of the skin and soft tissues makes up 3.5 percent of the cases of extrapelvic endometriosis, and most cases occur in surgical scars.15 Because endometriosis frequently occurs after gynecologic or obstetrical procedures, most commonly after delivery by cesarean section, it typically involves the lower abdominal wall and, less frequently, the umbilicus. In one series, the incidence of endometriosis in a scar after hysterotomy of a gravid uterus was approximately 1 percent.16 Although the true incidence of endometriosis in cesarean-section scars is difficult to determine, it has been estimated to be between 0.03 percent and 0.4 percent.17,18 Endometriosis can also involve the lower genital tract in areas of obstetrical or surgical trauma; the most common are episiotomy scars.19 Cases of cutaneous or soft-tissue endometriosis have been described after an appendectomy or repair of inguinal hernia.20 Finally, a small number of cases of spontaneous cutaneous endometriosis have been reported, typically involving the umbilicus and, less commonly, the inguinal and perineal regions.21,22,23 Cutaneous or soft-tissue endometriosis is only rarely associated with pelvic endometriosis, a much more common disease.
On gross examination, soft-tissue endometriotic lesions range in size from microscopic to more than 10 cm in largest dimension, and there may be areas of hemorrhage or even cystification, as there were in this case. Frequently there is prominent fibrosis and variable degrees of recent or old hemorrhage associated with the lesions, as was seen in this case; in some cases prominent myxoid change may be seen.24 Malignant transformation of endometriosis is a well-known phenomenon.25 The frequency of malignant transformation is estimated to be 1 percent or less in ovarian endometriosis; approximately 75 percent of cancers arising from endometriosis originate in the ovary. The incidence of malignant transformation in cutaneous or soft-tissue endometriosis is not known, with only isolated cases reported.26,27,28,29 Areas of cancer typically appear as discrete, firm nodules within the endometriotic tissue; no suspicious foci were seen in this case.
Scar endometriomas are believed to result from spillage of endometrial tissue during surgery with secondary implantation in the abdominal fascia or subcutaneous tissues. However, even though retrograde menstruation occurs in 75 to 90 percent of women and spillage into surgical incisions is probably quite frequent during gynecologic or obstetrical surgery, endometriosis occurs with a much lower frequency, suggesting that additional factors, including environmental and genetic factors (cellular and humoral immune system abnormalities, as well as estrogen abnormalities, among others), may confer susceptibility to the development of endometriosis in a small group of patients.30,31,32,33
Dr. Isaac Schiff (Obstetrics and Gynecology): Would irrigating the wound extensively before closing a Pfannenstiel incision reduce the risk of endometriosis?
Dr. Muto: There is no literature on the subject, but it is reasonable to think that irrigation might help. In addition, it is important to avoid tearing muscle fascicles or fascia, particularly on the second or third entry through the incision. Using electrocautery to define the planes also helps maintain hemostasis. In this particular case, the clinical time course suggests implantation at the time of hysterectomy. This seems counterintuitive, since endometrium is not usually exposed during a total hysterectomy, and suggests a mode of spread other than direct implantation.
Dr. Harris: Dr. O'Neill, was there any follow-up on the liver lesion?
Dr. O'Neill: Repeated CT scanning of the abdomen and pelvis with contrast material, four months after the operation, showed that the liver lesion was unchanged in size. It may well turn out to be a benign lesion that is unrelated to the endometriosis.
Dr. Harris: Dr. Robert Scully, the former editor of the Case Records and former head of gynecologic pathology, is with us today. Do you have any comments?
Dr. Robert E. Scully (Pathology): It is interesting from a historical viewpoint that when Pfannenstiel was practicing in the 19th century, most gynecologic pathology was done by gynecologists instead of surgical pathologists. Pfannenstiel actually wrote a very nice paper on the epithelial tumors of the ovary in which he illustrated the serous borderline tumors. He was the first one to use the term "borderline" for this group of tumors. It is conceivable that someday someone will report a serous borderline tumor of the ovary that presented in a Pfannenstiel scar, so it will be a Pfannenstiel tumor in a Pfannenstiel incision.
Anatomical Diagnosis
Endometriosis associated with a surgical scar.
Source Information
From the Division of Gynecologic Oncology, Gillette Center for Women's Cancers, Dana–Farber Cancer Institute and Brigham and Women's Hospital (M.G.M.); the Departments of Radiology (M.J.O.) and Pathology (E.O.), Massachusetts General Hospital; and the Departments of Obstetrics, Gynecology and Reproductive Biology (M.G.M.), Radiology (M.J.O.), and Pathology (E.O.), Harvard Medical School — all in Boston.
References
Suleiman S, Johnston DE. The abdominal wall: an overlooked source of pain. Am Fam Physician 2001;64:431-438.
Shields CJ, Winter DC, Kirwan WO, Redmond HP. Desmoid tumours. Eur J Surg Oncol 2001;27:701-706.
Speroff L. Endometriosis. In: Speroff L, Glass RH, Kase NG, eds. Clinical gynecologic endocrinology and infertility. 6th ed. Baltimore: Lippincott Williams & Wilkins, 1999:1057-73.
Olive DL, Schwartz LB. Endometriosis. N Engl J Med 1993;328:1759-1759.
Seydel AS, Sickel JZ, Warner ED, Sax HC. Extrapelvic endometriosis: diagnosis and treatment. Am J Surg 1996;171:239-239.
Franklin RR, Navarro C. Extragenital endometriosis. In: Chadha DR, Buttram VC Jr, eds. Current concepts in endometriosis. New York: Liss, 1990:289-308.
Wolf Y, Haddad R, Werbin N, Skornick Y, Kaplan O. Endometriosis in abdominal scars: a diagnostic pitfall. Am Surg 1996;62:1042-1044.
Firilas A, Soi A, Max M. Abdominal incisional endometriomas. Am Surg 1994;60:259-261.
Wolf GC, Singh KB. Cesarean scar endometriosis: a review. Obstet Gynecol Surv 1989;44:89-95.
Nirula R, Greaney GC. Incisional endometriosis: an underappreciated diagnosis in general surgery. J Am Coll Surg 2000;190:404-407.
Ness RB. Endometriosis and ovarian cancer: thoughts on shared pathophysiology. Am J Obstet Gynecol 2003;189:280-294.
Brinton LA, Gridley G, Persson I, Baron J, Bergqvist A. Cancer risk after a hospital discharge diagnosis of endometriosis. Am J Obstet Gynecol 1997;176:572-579.
Stern RC, Dash R, Bentley RC, Snyder MJ, Haney AF, Robboy SJ. Malignancy in endometriosis: frequency and comparison of ovarian and extraovarian types. Int J Gynecol Pathol 2001;20:133-139.
Clement PB. Pathology of endometriosis. Pathol Annu 1990;25:245-295.
Honore GM. Extrapelvic endometriosis. Clin Obstet Gynecol 1999;42:699-711.
Chatterjee SK. Scar endometriosis: a clinicopathologic study of 17 cases. Obstet Gynecol 1980;56:81-84.
Roncoroni L, Costi R, Violi V, Nunziata R. Endometriosis on laparotomy scar: a three-case report. Arch Gynecol Obstet 2001;265:165-167.
Blanco RG, Parithivel VS, Shah AK, Gumbs MA, Schein M, Gerst PH. Abdominal wall endometriomas. Am J Surg 2003;185:596-598.
Isbister WH. Endometriosis in an episiotomy scar preceding pelvic endometriosis. ANZ J Surg 2002;72:314-315.
Steck WD, Helwig EB. Cutaneous endometriosis. Clin Obstet Gynecol 1966;9:373-383.
Singh KK, Lessells AM, Adam DJ, et al. Presentation of endometriosis to general surgeons: a 10-year experience. Br J Surg 1995;82:1349-1351.
Michowitz M, Baratz M, Stavorovsky M. Endometriosis of the umbilicus. Dermatologica 1983;167:326-330.
Ahn GH, Scully RE. Clear cell carcinoma of the inguinal region arising from endometriosis. Cancer 1991;67:116-120.
Ying AL, Copeland LJ, Hameed A. Myxoid change in nondecidualized cutaneous endometriosis resembling malignancy. Gynecol Oncol 1998;68:301-303.
Sampson JA. Endometrial carcinoma of the ovary arising in endometrial tissue at that site. Arch Surg 1925;10:1-72.
Markopoulos C, Gogas H, Eleftheriou G, Floros D. Endometrioid carcinoma arising in a scar of caesarean section: case report. Eur J Gynaecol Oncol 1996;17:520-521.
Gucer F, Pieber D, Arikan MG. Malignancy arising in extraovarian endometriosis during estrogen stimulation. Eur J Gynaecol Oncol 1997;19:39-41.
Park SW, Hong SM, Wu HG, Ha SW. Clear cell carcinoma arising in a Cesarean section scar endometriosis: a case report. J Korean Med Sci 1999;14:217-219.
Todd RW, Kehoe S, Gearty J. A case of clear cell carcinoma arising in extragonadal endometriosis. Int J Gynecol Cancer 2000;10:170-172.
Seli E, Berkkanoglu M, Arici A. Pathogenesis of endometriosis. Obstet Gynecol Clin North Am 2003;30:41-61.
Redwine DB. Was Sampson wrong? Fertil Steril 2002;4:686-693.
Campbell IG, Thomas EJ. Endometriosis: candidate genes. Hum Reprod Update 2001;7:15-20.
Vinatier D, Orazi G, Cosson M, Dufour P. Theories of endometriosis. Eur J Obstet Gynecol Reprod Biol 2001;96:21-34.(Michael G. Muto, M.D., Ma)
A 45-year-old woman was admitted to this hospital with a painful abdominal mass.
She had first noticed a mass in her lower abdomen on the right side two and a half years earlier. There was intermittent severe pain at the site that occurred approximately once a month and lasted four to five days; it was worse during some months than others. She described the pain as feeling "like a hot brick." Five weeks before admission, she came to the surgical clinic of this hospital for further evaluation. A slightly tender mass, 4 cm in diameter, was palpated in the right lower abdomen. The remainder of the physical examination revealed no abnormalities.
An abdominal ultrasonographic study performed at this evaluation showed a highly vascular, mixed echoic mass, 3.5 cm by 3.5 cm by 1.0 cm, in the lower right anterior abdominal wall. Abdominal and pelvic computed tomographic (CT) scanning after the oral and intravenous administration of contrast material three weeks before admission confirmed the presence of an irregularly shaped, nonenhancing, heterogeneous lesion within the right lower anterior abdominal wall. The liver was fatty and also contained an enhancing lesion, 1.9 cm in diameter, in the lower right lobe, which showed early washout of contrast material. There were multiple, low-attenuation lesions in the liver that were thought to represent cysts. The other organs appeared to be normal. Magnetic resonance imaging (MRI) performed before and after the administration of gadolinium two weeks before admission to evaluate the liver lesion confirmed the presence of an early enhancing lesion, 1.7 cm by 1.8 cm, in the right lobe of the liver that had a minimally increased signal on a T2-weighted sequence and was isointense on a T1-weighted sequence. Multiple bright lesions on T2-weighted images were present within the liver, and they were thought to be cysts. No lymphadenopathy was seen. MRI of the other abdominal organs showed no abnormalities.
Scoliosis had developed in this patient during her adolescence, and Harrington rods were placed when she was 15 years of age. At 30 years of age, she was treated for a grade 1 cerebellar pilocytic astrocytoma, with two resections and postoperative irradiation with a dose up to 55.8 Gy. As a result of the astrocytoma, she had double vision, which required special corrective lenses, and ataxia, for which she had learned to compensate after rehabilitation therapy. A fibroadenoma of the left breast was excised when she was 28 years of age, as was a fibroadenoma of the right breast when she was 30 years of age. Three years before admission, bilateral intraductal papillomas of the breasts were excised, with the additional findings of focal atypical ductal hyperplasia and lobular neoplasia in situ. She had had two pregnancies, and both of her daughters were delivered by cesarean section. A hysterectomy because of uterine fibroids had been performed at another hospital three years before admission. At that time, pathological examination confirmed the presence of multiple leiomyomas, benign findings in the endometrium, and chronic cervicitis. The ovaries had not been removed.
She lived with her husband and daughters, who were all well. She did not smoke and drank three to four alcoholic drinks per month. Her only medication was a daily multivitamin.
On admission, the patient's vital signs were normal. The height was 167.5 cm and the weight 49.5 kg. The abdomen had a low, 6-cm scar and was soft, with no organomegaly. An irregular, indurated mass in the right lower quadrant, 4 cm in diameter, with no overlying skin retraction or pitting, was slightly tender to palpation. The remainder of the physical examination showed no abnormalities.
The results of routine laboratory tests performed on admission were within the normal ranges. An electrocardiogram showed a normal sinus rhythm and slight left atrial enlargement. A chest radiograph showed probable pulmonary emphysema with bullous changes in the upper lobes. No infiltrates, masses, effusions, or lymphadenopathy were seen. Harrington rods were noted in the thoracic spine.
A diagnostic procedure was performed.
Differential Diagnosis
Dr. Michael G. Muto: May we review the radiologic studies?
Dr. Mary Jane O'Neill: Sagittal and transverse views as visualized on the superficial ultrasonographic imaging of the anterior abdominal wall show that within the fibers of the rectus muscle there is a well-circumscribed, hypoechoic, solid-appearing mass lesion that is displacing the fibers of the right rectus muscle (Figure 1A). The CT examination shows asymmetry of the rectus; there is a high-density, moderately lobulated mass within the substance of the right abdominal rectus muscle (Figure 1B). On a CT scan obtained at the same session, there is a relatively ill defined enhancing mass near the surface of the right lobe of the liver (Figure 1C). The ovaries appeared to be normal.
Figure 1. Imaging Studies of the Abdomen.
An ultrasonographic image of the lower right abdominal wall (Panel A) shows a solid, hypoechoic lesion (arrows) in the right rectus muscle. The small arrowheads indicate the placement of the focal zone of the ultrasonographic beam. A CT image (Panel B) through the lower abdomen, which was obtained after the administration of contrast material, reveals an enhancing, solid mass (arrow) in the right rectus muscle. The same study shows a lesion (arrow) near the surface of the right lobe of the liver (Panel C). A T1-weighted axial MRI scan obtained after the administration of contrast material in the arterial phase shows a small arterially enhancing lesion (Panel D).
MRI was performed to evaluate the liver lesion. On the T1-weighted images, the lesion is dark, without evidence of any hemorrhage. It is ill defined on T2-weighted images, and it has prominent enhancement after gadolinium administration (Figure 1D). It is unclear whether this lesion is intraparenchymal or on the surface of the liver, and there are no specific features on imaging of either a benign or a malignant lesion. In a patient of this age and sex, the features that are apparent on MRI would be consistent with focal nodular hyperplasia.
Dr. Muto: This multiparous woman with a history of two cesarean sections and a hysterectomy had a mass of the abdominal wall that had been present for more than two years and was intermittently painful. The radiographic studies showed that the mass was heterogeneous and located in the substance of the right rectus abdominis muscle, 2 cm above the Pfannenstiel incision and 3 cm to the right of midline. Radiographic imaging also showed two normal-appearing retained ovaries in the pelvis and no suggestion of ventral hernia or bowel obstruction. The solitary lesion noted in the liver could not be characterized as benign in appearance. All the other abdominal viscera appeared normal.
The differential diagnosis of a painful abdominal-wall mass is shown in Table 1.1
Table 1. Differential Diagnosis of a Painful Abdominal-Wall Mass.
Hernias and Abscesses
A wide variety of ventral hernias may occur as a painful abdominal-wall mass. The location of the mass makes an inguinal or femoral hernia improbable in this case. A spigelian hernia occurs at the posterior lateral abdominal-wall fascia through the linea semilunaris, and although it is rare, a mass can appear in this location, particularly if the hernia were to cause incarceration or strangulation. Finally, an incisional hernia must certainly be considered, especially if there have been multiple laparotomies performed through the same incision. In order to understand how an incisional hernia can occur in a region that is cephalad and lateral to the skin incision, it is necessary to review how this incision is performed.
In the Pfannenstiel incision, the skin and fascia are incised transversely. The fascia is then dissected from the underlying rectus abdominis muscles in a cephalad and caudad direction, which creates subfascial spaces above and below the skin incision (Figure 2). This technique allows the rectus abdominis muscles to be separated along the midline. A vertical laparotomy is then performed. The Pfannenstiel incision is widely used in both obstetrics and gynecology, because it affords generous access to the pelvis and is stronger than a vertical incision and ensures excellent cosmetic results. One drawback to the incision is the creation of the subfascial spaces, which can increase the risk of infection or hematoma formation, as compared with a vertical midline incision. Also, although the subfascial space is readily created during a primary incision, when the incision is reopened, this subfascial space is often heavily scarred and difficult to recreate, resulting in trauma to both the underlying rectus muscle and the fascia. Such trauma may increase the risk of hematoma, infection, or fascial disruption. Therefore, although it is unusual for incisional hernias to develop through a Pfannenstiel incision, they may occur either beneath the skin incision or superolateral to it.
Figure 2. The Pfannenstiel Incision.
The incision begins with a transverse skin and fascial incision (inset). The fascia is dissected from the underlying rectus abdominis muscle in a cephalad and caudad direction, which creates subfascial spaces above and below the skin incision. The rectus abdominis muscles are separated along the midline, and a vertical laparotomy is then performed.
I do not know if any of this patient's three prior laparotomies were complicated by infection or hematoma formation; however, just having three procedures performed through the incision is risk enough for hernia formation to be a consideration. The cyclicity of her pain, the clinical presentation, and the imaging studies, however, are not consistent with this diagnosis.
A diverticular, appendiceal, or tubo-ovarian abscess can secondarily infect the rectus muscle and lead to the development of a chronic abdominal-wall abscess. There is no evidence, however, to support an infectious cause in this case. There is also no history of traumatic injury to the muscle, belly rupture, or prior instrumentation.
Tumors
Any benign or malignant tumor of muscle, integument, nerve, or blood vessels can involve the rectus muscle and occur as a painful mass. Desmoid tumors are benign but locally destructive lesions. These tumors often occur in young women after a pregnancy, and they may grow slowly in the abdominal wall or surgical scar. There is evidence that suggests that these tumors are hormone-sensitive, including the predominance of the lesion in women of reproductive age and regression in menopausal women and among those who are receiving tamoxifen therapy. This tumor would not, however, be expected to present with cyclic pain, and it usually appears as a homogeneously solid mass on CT images, rather than a complex mass.2
Endometriotic Implants
Endometriosis is the ectopic growth of endometrial glands and stroma outside the endometrial cavity. It affects 3 to 10 percent of all women of reproductive age. About 4 of every 1000 women between 15 and 64 years of age are hospitalized with endometriosis each year.3 The disease is a major cause of infertility and chronic pelvic pain.4 The most common extrauterine sites involved include the ovary, fallopian tube, and adjacent pelvic tissues. Pelvic spread is thought to occur by retrograde menstruation by way of the fallopian tube, during which viable endometrial cells are implanted in dependent portions of the pelvis. Rarely, the disease occurs outside the pelvis in sites such as the lungs, resulting in cyclic hemoptysis, hemothorax, or pneumothorax, and within retroperitoneal lymph nodes and liver parenchyma. These findings suggest lymphatic or vascular dissemination. Finally, the occurrence of endometriosis in men taking estrogen therapy suggests yet a third pathogenetic mechanism: endometriosis arising as a result of coelomic metaplasia.
Endometriotic implants in scars resulting from episiotomy and cesarean section most often occur at the time of vaginal or cesarean delivery and are the result of direct implantation of viable endometrial cells into the subcutaneous or subfascial spaces that are exposed by the surgical incisions on the abdominal wall or perineal body. The classic manifestations of endometriosis in surgical scars are focal cyclic pain and a slow-growing mass.5,6,7,8 The pain can often be timed to the menses in women who have not undergone oophorectomy. Progressive cycles of bleeding and local inflammatory response lead to the development of a complex vascular or fibrotic mass. The overall incidence of endometriomas occurring in cesarean-section scars has been estimated at 0.03 to 0.15 percent.9
Because of the clinical similarity between endometriosis in a surgical scar and an incisional hernia, general surgeons, rather than gynecologists, are often the first to evaluate patients with this disorder. Although the development of endometriosis by way of implanted cells is widely reported in the gynecologic literature, this entity is not well recognized among general surgeons. As a result, the diagnosis is often not considered. In particular, the telltale history of cyclic pain is either not elicited or, as in the case under discussion, not appreciated. As Nirula and Greaney report,10 among 10 cases of incisional endometriosis in caesarean-section scars, in only 2 cases was the diagnosis suspected preoperatively, and these were diagnosed by a single surgeon who was aware of the disease from prior experience.
The most probable diagnosis in this case is endometriosis of the anterior abdominal wall. Although endometrial cells can implant within a ventral hernia sac, it seems far more likely that implantation occurred beneath the upper flap of the Pfannenstiel incision after the patient's primary or repeated cesarean section or during the hysterectomy.
The Hepatic Lesion
There is a second finding that must be explained in this case — the suspicious hepatic lesion noted on CT scanning and subsequent MRI. This may be entirely unrelated to the patient's abdominal-wall mass; although it could not be characterized as benign on the basis of radiologic criteria, its features were thought to be consistent with focal nodular hyperplasia. Malignant degeneration of endometriotic implants has been widely reported in the gynecology and gynecologic-oncology literature.11,12 Clear-cell and endometrioid adenocarcinomas are by far the most common histologic types reported. Endometrioid and clear-cell carcinomas predominate in intraperitoneal and ovarian implants, whereas clear-cell cancers more frequently are associated with extraperitoneal implants.13
In conclusion, the most likely diagnosis for this 45-year-old woman's illness was an implant of endometriosis within her surgical scar. Given her ongoing symptoms, complete excision would be the most appropriate procedure for both diagnosis and therapy. If the mass contained evidence of an endometriosis-associated cancer, a liver metastasis should be suspected.
A Physician: Were all three radiologic studies necessary to make this diagnosis?
Dr. Muto: The ultrasonographic study was a reasonable choice for primary imaging of this ill-defined mass. Once a mass was clearly defined, abdominal and pelvic CT scanning was required to rule out an incarcerated hernia. When the CT scanning revealed the unexpected finding of a suspicious-appearing cystic liver lesion, her physicians were obliged to evaluate it further with an MRI scan.
A Physician: Is MRI useful for the diagnosis of endometriosis?
Dr. O'Neill: Pelvic endometriosis has characteristic increased signal intensity on T1-weighted images; however, when the endometriosis is located within fascial or muscular tissues, it often does not produce the characteristic imaging findings. In this patient, CT scanning would have been the single best imaging test because it can localize the palpable abnormality and allow the clinician to better assess the remainder of the abdomen and pelvis for findings that would affect the differential diagnosis.
Dr. Nancy Lee Harris (Pathology): There was no evidence of endometriosis on the uterus that was removed at hysterectomy. Is the risk of endometriosis in a surgical incision independent of the presence of pelvic endometriosis?
Dr. Muto: Many reported cases had no evidence of pelvic endometriosis.
Dr. Harris: This patient was seen by a general surgeon, whose differential diagnosis included endometriosis in a scar, but the primary diagnosis was a desmoid tumor.
Clinical Diagnosis
Desmoid tumor of the abdominal wall.
Dr. Michael G. Muto's Diagnosis
Endometriosis of the abdominal wall.
Pathological Discussion
Dr. Esther Oliva: A specimen of soft tissue measuring 9.0 cm by 6.0 cm by 1.2 cm was received in the pathology department. Sectioning revealed areas of hemorrhage and cyst formation. On microscopical examination, there was a biphasic growth of glands and stroma in a background of fibrous tissue. Some of the glands were cystically dilated (Figure 3A). The glands were coiled with abundant luminal secretions and were lined by cuboidal and columnar cells with basally located nuclei (Figure 3B). Cytologic atypia and mitotic activity were absent. The glands were surrounded by a cellular stroma that was composed of small cells with scant cytoplasm and oval-to-elongated nuclei, which were whorled around small blood vessels (Figure 3B). Edema, focal elastosis, and areas of recent hemorrhage were evident (Figure 3A).
Figure 3. Histologic Sections of the Mass in the Abdominal Wall (Hematoxylin and Eosin).
An endometrial gland, cystically dilated (Panel A), is embedded in an edematous stroma, adjacent to areas of scarring and elastosis. A higher-power view (Panel B) shows endometrial glands lined by cells with tall cytoplasm and basally located nuclei, with no cytologic atypia.
Endometriosis is defined as the presence of endometrial-type glands and stroma outside the endometrium and myometrium. It most commonly involves the ovaries, the uterine ligaments, the rectovaginal septum, the cul-de-sac and peritoneum of the uterus, the fallopian tubes, the rectosigmoid, the ureter, and the bladder.14 Endometriosis of the skin and soft tissues makes up 3.5 percent of the cases of extrapelvic endometriosis, and most cases occur in surgical scars.15 Because endometriosis frequently occurs after gynecologic or obstetrical procedures, most commonly after delivery by cesarean section, it typically involves the lower abdominal wall and, less frequently, the umbilicus. In one series, the incidence of endometriosis in a scar after hysterotomy of a gravid uterus was approximately 1 percent.16 Although the true incidence of endometriosis in cesarean-section scars is difficult to determine, it has been estimated to be between 0.03 percent and 0.4 percent.17,18 Endometriosis can also involve the lower genital tract in areas of obstetrical or surgical trauma; the most common are episiotomy scars.19 Cases of cutaneous or soft-tissue endometriosis have been described after an appendectomy or repair of inguinal hernia.20 Finally, a small number of cases of spontaneous cutaneous endometriosis have been reported, typically involving the umbilicus and, less commonly, the inguinal and perineal regions.21,22,23 Cutaneous or soft-tissue endometriosis is only rarely associated with pelvic endometriosis, a much more common disease.
On gross examination, soft-tissue endometriotic lesions range in size from microscopic to more than 10 cm in largest dimension, and there may be areas of hemorrhage or even cystification, as there were in this case. Frequently there is prominent fibrosis and variable degrees of recent or old hemorrhage associated with the lesions, as was seen in this case; in some cases prominent myxoid change may be seen.24 Malignant transformation of endometriosis is a well-known phenomenon.25 The frequency of malignant transformation is estimated to be 1 percent or less in ovarian endometriosis; approximately 75 percent of cancers arising from endometriosis originate in the ovary. The incidence of malignant transformation in cutaneous or soft-tissue endometriosis is not known, with only isolated cases reported.26,27,28,29 Areas of cancer typically appear as discrete, firm nodules within the endometriotic tissue; no suspicious foci were seen in this case.
Scar endometriomas are believed to result from spillage of endometrial tissue during surgery with secondary implantation in the abdominal fascia or subcutaneous tissues. However, even though retrograde menstruation occurs in 75 to 90 percent of women and spillage into surgical incisions is probably quite frequent during gynecologic or obstetrical surgery, endometriosis occurs with a much lower frequency, suggesting that additional factors, including environmental and genetic factors (cellular and humoral immune system abnormalities, as well as estrogen abnormalities, among others), may confer susceptibility to the development of endometriosis in a small group of patients.30,31,32,33
Dr. Isaac Schiff (Obstetrics and Gynecology): Would irrigating the wound extensively before closing a Pfannenstiel incision reduce the risk of endometriosis?
Dr. Muto: There is no literature on the subject, but it is reasonable to think that irrigation might help. In addition, it is important to avoid tearing muscle fascicles or fascia, particularly on the second or third entry through the incision. Using electrocautery to define the planes also helps maintain hemostasis. In this particular case, the clinical time course suggests implantation at the time of hysterectomy. This seems counterintuitive, since endometrium is not usually exposed during a total hysterectomy, and suggests a mode of spread other than direct implantation.
Dr. Harris: Dr. O'Neill, was there any follow-up on the liver lesion?
Dr. O'Neill: Repeated CT scanning of the abdomen and pelvis with contrast material, four months after the operation, showed that the liver lesion was unchanged in size. It may well turn out to be a benign lesion that is unrelated to the endometriosis.
Dr. Harris: Dr. Robert Scully, the former editor of the Case Records and former head of gynecologic pathology, is with us today. Do you have any comments?
Dr. Robert E. Scully (Pathology): It is interesting from a historical viewpoint that when Pfannenstiel was practicing in the 19th century, most gynecologic pathology was done by gynecologists instead of surgical pathologists. Pfannenstiel actually wrote a very nice paper on the epithelial tumors of the ovary in which he illustrated the serous borderline tumors. He was the first one to use the term "borderline" for this group of tumors. It is conceivable that someday someone will report a serous borderline tumor of the ovary that presented in a Pfannenstiel scar, so it will be a Pfannenstiel tumor in a Pfannenstiel incision.
Anatomical Diagnosis
Endometriosis associated with a surgical scar.
Source Information
From the Division of Gynecologic Oncology, Gillette Center for Women's Cancers, Dana–Farber Cancer Institute and Brigham and Women's Hospital (M.G.M.); the Departments of Radiology (M.J.O.) and Pathology (E.O.), Massachusetts General Hospital; and the Departments of Obstetrics, Gynecology and Reproductive Biology (M.G.M.), Radiology (M.J.O.), and Pathology (E.O.), Harvard Medical School — all in Boston.
References
Suleiman S, Johnston DE. The abdominal wall: an overlooked source of pain. Am Fam Physician 2001;64:431-438.
Shields CJ, Winter DC, Kirwan WO, Redmond HP. Desmoid tumours. Eur J Surg Oncol 2001;27:701-706.
Speroff L. Endometriosis. In: Speroff L, Glass RH, Kase NG, eds. Clinical gynecologic endocrinology and infertility. 6th ed. Baltimore: Lippincott Williams & Wilkins, 1999:1057-73.
Olive DL, Schwartz LB. Endometriosis. N Engl J Med 1993;328:1759-1759.
Seydel AS, Sickel JZ, Warner ED, Sax HC. Extrapelvic endometriosis: diagnosis and treatment. Am J Surg 1996;171:239-239.
Franklin RR, Navarro C. Extragenital endometriosis. In: Chadha DR, Buttram VC Jr, eds. Current concepts in endometriosis. New York: Liss, 1990:289-308.
Wolf Y, Haddad R, Werbin N, Skornick Y, Kaplan O. Endometriosis in abdominal scars: a diagnostic pitfall. Am Surg 1996;62:1042-1044.
Firilas A, Soi A, Max M. Abdominal incisional endometriomas. Am Surg 1994;60:259-261.
Wolf GC, Singh KB. Cesarean scar endometriosis: a review. Obstet Gynecol Surv 1989;44:89-95.
Nirula R, Greaney GC. Incisional endometriosis: an underappreciated diagnosis in general surgery. J Am Coll Surg 2000;190:404-407.
Ness RB. Endometriosis and ovarian cancer: thoughts on shared pathophysiology. Am J Obstet Gynecol 2003;189:280-294.
Brinton LA, Gridley G, Persson I, Baron J, Bergqvist A. Cancer risk after a hospital discharge diagnosis of endometriosis. Am J Obstet Gynecol 1997;176:572-579.
Stern RC, Dash R, Bentley RC, Snyder MJ, Haney AF, Robboy SJ. Malignancy in endometriosis: frequency and comparison of ovarian and extraovarian types. Int J Gynecol Pathol 2001;20:133-139.
Clement PB. Pathology of endometriosis. Pathol Annu 1990;25:245-295.
Honore GM. Extrapelvic endometriosis. Clin Obstet Gynecol 1999;42:699-711.
Chatterjee SK. Scar endometriosis: a clinicopathologic study of 17 cases. Obstet Gynecol 1980;56:81-84.
Roncoroni L, Costi R, Violi V, Nunziata R. Endometriosis on laparotomy scar: a three-case report. Arch Gynecol Obstet 2001;265:165-167.
Blanco RG, Parithivel VS, Shah AK, Gumbs MA, Schein M, Gerst PH. Abdominal wall endometriomas. Am J Surg 2003;185:596-598.
Isbister WH. Endometriosis in an episiotomy scar preceding pelvic endometriosis. ANZ J Surg 2002;72:314-315.
Steck WD, Helwig EB. Cutaneous endometriosis. Clin Obstet Gynecol 1966;9:373-383.
Singh KK, Lessells AM, Adam DJ, et al. Presentation of endometriosis to general surgeons: a 10-year experience. Br J Surg 1995;82:1349-1351.
Michowitz M, Baratz M, Stavorovsky M. Endometriosis of the umbilicus. Dermatologica 1983;167:326-330.
Ahn GH, Scully RE. Clear cell carcinoma of the inguinal region arising from endometriosis. Cancer 1991;67:116-120.
Ying AL, Copeland LJ, Hameed A. Myxoid change in nondecidualized cutaneous endometriosis resembling malignancy. Gynecol Oncol 1998;68:301-303.
Sampson JA. Endometrial carcinoma of the ovary arising in endometrial tissue at that site. Arch Surg 1925;10:1-72.
Markopoulos C, Gogas H, Eleftheriou G, Floros D. Endometrioid carcinoma arising in a scar of caesarean section: case report. Eur J Gynaecol Oncol 1996;17:520-521.
Gucer F, Pieber D, Arikan MG. Malignancy arising in extraovarian endometriosis during estrogen stimulation. Eur J Gynaecol Oncol 1997;19:39-41.
Park SW, Hong SM, Wu HG, Ha SW. Clear cell carcinoma arising in a Cesarean section scar endometriosis: a case report. J Korean Med Sci 1999;14:217-219.
Todd RW, Kehoe S, Gearty J. A case of clear cell carcinoma arising in extragonadal endometriosis. Int J Gynecol Cancer 2000;10:170-172.
Seli E, Berkkanoglu M, Arici A. Pathogenesis of endometriosis. Obstet Gynecol Clin North Am 2003;30:41-61.
Redwine DB. Was Sampson wrong? Fertil Steril 2002;4:686-693.
Campbell IG, Thomas EJ. Endometriosis: candidate genes. Hum Reprod Update 2001;7:15-20.
Vinatier D, Orazi G, Cosson M, Dufour P. Theories of endometriosis. Eur J Obstet Gynecol Reprod Biol 2001;96:21-34.(Michael G. Muto, M.D., Ma)