A leiomyoma is a benign tumor composed mainly of smooth muscle cells but containing varying amounts of fibrous connective tissue. The tumor is well circumscribed but not encapsulated. Various terms are used to refer to the tumor, such as fibromyoma, myofibroma, leiomyofibroma, fibroleiomyoma, myoma, fibroma, and fibroid. The latter designation is the one most commonly used, but it is the least accurate and acceptable. The term leiomyoma is a reasonably accurate one that emphasizes the origin of this tumor from smooth muscle cells and the predominance of the smooth muscle component. The tissue culture work of Miller and Ludovici suggested an origin from smooth muscle cells, and the studies of Townsend and associates affirm a unicellular origin for leiomyomata.

Leiomyomata are the most common tumors of the uterus and female pelvis. It is impossible to determine their true incidence accurately, although the frequently quoted incidence of 50% found at postmortem examinations seems reasonable. Leiomyomata are responsible for about one third of all hospital admissions to gynecology services. It is well recognized that the incidence is much higher in African American than in Caucasian women. In a careful study of leiomyomata among women in Augusta, Georgia, Torpin and associates found the incidence among African American women to be three and one third times that among Caucasian women. There is no known explanation for this racial difference. Leiomyomata also are larger and occur at a younger age in African Americans. In fact, many African American women develop leiomyomata before 30 years of age. However, development prior to age 20 is extremely rare, regardless of race. Patients with uterine leiomyomata often have a positive family history of uterine leiomyomata. This suggests the presence of a gene, which has yet to be discovered, encoding for their development.

About 40% to 50% of leiomyomas show karyotypically detectable chromosomal abnormalities that are both nonrandom and tumor specific. Identified chromosomal abnormalities include t(12;14) (q15;q23-24), del(7) (q22q32), rearrangements involving 6p21, 10q, trisomy 12, and deletions of 3q. Interestingly, a recent study of 217 myomas found a positive correlation between the presence of a cytogenetic abnormality and the anatomic location of the myoma. In this study by Brosens and colleagues, submucous myomas were consistently shown to have fewer cytogenetic abnormalities when compared with intramural and subserous lesions (12% vs. 35% and 29%, respectively). An increased prevalence in certain races, twin studies indicating higher correlation with hysterectomy in monozygotic twins, and increased incidence in first-degree relatives all seem to support an inherited predisposition. The true genetic contribution to the development of uterine leiomyoma remains to be defined.

Most of the data concerning the incidence of uterine leiomyomata are based on gross examination of the uterus, routine pathology reports, or the clinical diagnosis of uterine leiomyomata. Cramer and Patel subjected 100 uteri to gross serial sectioning at 2-mm intervals. They found 649 leiomyomata, roughly threefold the number identified by routine pathologic examination. Admittedly, some were only a few millimeters in diameter, but all were grossly visible. In 48 uteri with no mention of leiomyomata in the routine report, 27 were found to have small tumors. The incidence of leiomyomata was the same in premenopausal and postmenopausal uteri, although the average number of leiomyomata and the average size of the largest leiomyoma were greater in the premenopausal women. This work has important implications for future epidemiologic studies. It also suggests that it is almost never possible to surgically remove all leiomyomata when a myomectomy is performed.

The growth of leiomyomata is dependent on estrogen production. The tumors thrive during the years of greatest ovarian activity. Continuous estrogen secretion, especially when uninterrupted by pregnancy and lactation, is thought to be the most important underlying risk factor in the development of myomata. After menopause, with regression of ovarian estrogen secretion, growth of leiomyomata usually ceases. Actual regression in the tumor size may occur. There are rare instances, however, of postmenopausal growth of benign leiomyomata, suggesting the possibility of postmenopausal estrogen production either in the ovary or elsewhere. Postmenopausal ovarian cortical stromal hyperplasia may be associated with an increase in estrogen secretion by the ovary. The postmenopausal ovarian stroma in a variety of presumably inactive ovarian tumors, including mucinous cysts and Brenner tumors, can also produce estrogen. When a central pelvic tumor presumed to be uterine leiomyomata enlarges after menopause, one should think of the possibility of malignant change in the leiomyoma itself or in the adjacent myometrium, or of the growth of a new pelvic tumor of extrauterine origin.

Older nulliparous women have an increased risk of developing leiomyomata. However, in multiparous women, the relative risk decreases with each pregnancy. A woman who has had five term pregnancies has only one fifth the risk of a nulliparous woman of developing myomata. The risk is reduced in women who smoke and is increased in obese women; this is possibly related to the conversion of androgens to estrogen by fat aromatase.

The observation that leiomyomata may show significant enlargement during pregnancy provides further clinical evidence of the relation of estrogen and progesterone to the growth of these tumors. However, a better blood supply during pregnancy might also encourage their growth. In a prospective ultrasonographic study of 29 pregnant patients with uterine leiomyomata, Aharoni and associates found no evidence of enlargement of the myomata in 78%. Lev-Toaff and colleagues also confirmed that some but not all leiomyomata enlarge during pregnancy in response to estrogen and progesterone.

In the initial two decades following the introduction of oral contraceptives containing high-dose estrogen, there was a striking increase in the occurrence of large leiomyomata among young women of all racial backgrounds who took these pills. Although the growth of uterine leiomyomata is not invariably stimulated, oral contraceptives containing highdose estrogen should not be prescribed for women with these tumors. Oral contraceptives with low-dose estrogen are less likely to stimulate growth. According to Parazzini and associates, there is no significant relation between the occurrence or growth of leiomyomata and the newer oral contraceptives that contain much smaller amounts of estrogens and progestins, and some believe that the risk of developing myomata is reduced with these low-dose pills.

Scientific investigators have been intrigued by the observation that leiomyomata develop during the reproductive years, sometimes grow during pregnancy, and regress after menopause. Nelson, Lipschutz, and others have produced multiple leiomyomata artificially on the serosal surface of the uterus and other peritoneal surfaces in guinea pigs given prolonged estrogen injections. Spellacy and coworkers found that levels of plasma estradiol were the same in patients with and without leiomyomata. However, Wilson and associates found a significantly higher concentration of estrogen receptors in leiomyomata than in myometrium. Farber and colleagues reported that these tumors bind about 20% more estradiol per milligram of cytoplasmic protein than does the normal myometrium of the same organ. This observation was not uniformly true for all leiomyomata, suggesting that different cellular components with a leiomyoma may be associated with different biologic activity. Otubu and coworkers found the concentration of estradiol to be significantly higher in leiomyomata than in normal myometrium, especially in the proliferative phase of the menstrual cycle. Soules and McCarty reported that leiomyomata had more estrogen receptors than did normal uterine tissues in the first phase (days 1 through 9) and in the second phase (days 10 through 18) of the menstrual cycle. Gabb and Stone found that the ability to convert estradiol to estrone was similar in leiomyomata and myometrium. However, Pollow and associates found the conversion of estradiol into estrone to be significantly lower in leiomyomata than in myometrium. This difference in conversion rate could result in a relative accumulation of estrogen in a leiomyoma, causing a hyperestrogenic state within the tumor and surrounding tissues. The enzyme 17β-hydroxy dehydrogenase accelerates the conversion of estradiol to estrone. Leiomyomata have a low concentration of 17β-hydroxy dehydrogenase, which results in a relative accumulation of estradiol in leiomyomatous tissue. These findings may explain the myometrial hypertrophy that is invariably present with leiomyomata.

Other abnormalities in endocrine function have also been suggested. Ylikorkala and colleagues found that pituitary function may be abnormal in women with leiomyomata. Patients with leiomyomata had a low follicle-stimulating hormone level and a diminished follicle-stimulating hormone response to pituitary GnRH. There was an excessive prolactin response to thyrotropin-releasing hormone. Spellacy and colleagues found that the peak levels of human growth hormone reached during a hypoglycemic test were twice as high in patients with leiomyomata as in the control group. Reddy and Rose suggested the possibility that 5α-reduced androgens may play a role in the pathophysiology of uterine leiomyomata, because a significant increase in 5α-reductase has been found in leiomyoma tissue as compared with the myometrium and endometrium. Influenced by the experimental investigations of Lipschutz and associates, Goodman in 1946 treated patients with uterine leiomyomata with progesterone and noted a decrease in tumor size in all patients. However, Segaloff and colleagues reported no effect in their study. Goldzieher and coworkers produced histologic evidence of extensive degenerative changes in leiomyomata by administering high-dose progestin therapy (medrogestone in high doses for 21 days). Filiceri and associates have reported the regression of a uterine leiomyoma after longterm administration of a long-acting luteinizing hormone-releasing hormone agonist given to suppress ovarian estrogen secretion. Coutinho successfully used a potent 19-norsteroid antiestrogen-antiprogesterone to treat excessive uterine bleeding in 16 patients with uterine leiomyomata. A reduction in the size of the tumors was noted.

Although the exact etiology of uterine leiomyomata is not known, the puzzle may be solved bit by bit by the research of Kornyei and colleagues, Wilson and coworkers, Tamaya and associates, Buchi and Keller, Sadan and colleagues, and others who continue to investigate estrogen and progesterone as possible growth factors. Although some data are conflicting, evidence suggests that both estrogen and progesterone are involved in the growth of uterine leiomyomata. The possibility that progesterone may play a role in the growth of leiomyomata is suggested by the work of Kawaguchi and coworkers, who found a higher mitotic count in leiomyomata obtained in the proliferative phase of the menstrual cycle.

Anderson and associates have shown that medroxyprogesterone acetate, a progestin, causes a decrease in connexin-43 messenger ribonucleic acid levels in primary cultures of human myometrium and leiomyoma. Connexin-43 is a gap junction protein whose formation is stimulated by 17β-estradiol.

According to the research data of Brandon and colleagues, progesterone receptor messenger ribonucleic acid is overexpressed in uterine leiomyomata, compared with normal adjacent myometrium, suggesting that amplified progesteronemediated signaling is instrumental in the abnormal growth of these tumors. It is possible that the increased amount of progesterone receptor is caused by an alteration of estrogen or estrogen receptors in leiomyomata. The work of Kastner and coworkers and Nardulli and associates has demonstrated that progesterone receptor expression is regulated by estrogen.

Research in recent years has also focused on polypeptide growth factors in the stimulation of growth of leiomyomata. Polypeptide growth factors that have been investigated include epidermal growth factor, transforming growth factor alpha and beta, insulin-like growth factor (IGF), platelet-derived growth factor, vascular endothelial growth factor, and basic fibroblast growth factor. Other growth factors may also be involved. Polypeptide growth factor research has been performed by Goustin and colleagues, by Hoffmann and coworkers, by Lumsden and associates, and by others. A brief review of this research has been written by Vollenhoven and associates, who have been involved in the study of IGFs in uterine leiomyomata.

Results from a study by Strawn and colleagues demonstrate that IGF-I stimulates leiomyoma growth in a dose-related manner over that of normal myometrial tissue in monolayer culture. This stimulatory effect, in the absence of sex steroid hormones or other growth factors, provides additional support that IGF-I may play an important direct role in the pathogenesis of these tumors, possibly by modulating the response of these tumors to various levels of sex steroids. Dawood and Kahn-Dawood were unable to find any significant elevation in peripheral levels of serum IGF-I in nonpregnant premenopausal women with uterine leiomyomata of 14 weeks gestational size. The authors state, “Nevertheless, the finding does not detract from the potential paracrine or autocrine role that
IGF-I produced by leiomyoma cells may have either on the growth of its own or adjacent myomas or on the vascular supply and blood flow of the uterus and myomas.”

Rein and coworkers have proposed a hypothesis to explain the pathogenesis of myomata. This hypothesis suggests a critical role for progesterone in the growth of myomata. They state:

To treat patients with uterine leiomyomata properly, the gynecologic surgeon must be familiar with their pathology, growth characteristics, and clinical features. Leiomyomata may be single, but most are multiple. They develop most commonly in the uterine corpus and much less often in the cervix. They may develop in the round ligaments, but this is rare. Because they arise in the myometrium, they are all interstitial or intramural in the beginning. As they enlarge, they can remain intramural, but growth often extends in an internal or external direction. Thus, the tumor can eventually become subserous or submucous in location. A subserous tumor can become pedunculated and occasionally parasitic, receiving its blood supply from another source, usually the omentum. A submucous tumor can also become pedunculated and may gradually dilate the endocervical canal and protrude through the cervical os. Indeed, a submucous myoma may descend through the vagina. Rarely, chronic uterine inversion results if the prolapsing submucous leiomyoma is attached to the top of the endometrial cavity and pulls the uterine fundus downward through the cervix.

In general, subserous leiomyomata contain more fibrous tissue than submucous leiomyomata. However, submucous leiomyomata contain more smooth muscle tissue than subserous leiomyomata. Sarcomatous change is more common in submucous tumors.

The typical uterine leiomyoma is a firm multinodular structure of variable size. The largest tumor, reported by Hunt in 1888, weighed more than 65 kg. Tumors of 4 to 5 kg are not rare, but most are smaller. In the operating room, leiomyomata appear as nodular tumors of different sizes that distort the uterus in various ways, depending on their size, location, and direction of growth. Growth between the leaves of the broad ligament and origin from the cervix may make surgical removal difficult. Subserous and subserous pedunculated tumors, as well as intraligamentous tumors, may create problems in diagnosis because they are difficult to distinguish from tumors arising from the adnexal organs (Fig. 31.2). When tumors cause symmetric enlargement of the uterus, they may be mistaken for a pregnant uterus on bimanual examination.

The “normal” intramural leiomyoma on section protrudes from the surrounding compressed myometrium. Ordinarily, there is a clear distinction between the myoma and the myometrium so that dissection between the two is easy to accomplish. Myomata usually can be removed from surrounding myometrium with ease. Although these tumors are not encapsulated, a clear distinction can usually be made between a myoma and the myometrium that surrounds it. The cut surface appears as glistening pinkish white and gray. It is firm, and there is a whorllike arrangement of the muscle and the fibrous tissue. In contrast to this typical appearance, the myometrium may be thickened by a diffuse, ill-defined nodularity of smooth muscle. This so-called diffuse leiomyomatosis usually involves all parts of the myometrium and causes symmetric enlargement of the uterine corpus. The nodules of smooth muscle are not distinct, contain little collagen, and merge with one another and the surrounding hypertrophied myometrium.

The extracellular matrix of leiomyomata is composed mostly of collagen but also contains proteoglycans and fibronectin. According to Fujita, myomata contain 50% more collagen than does normal myometrium, and the ratio of collagen type I to collagen type III is increased in myomata. Proteoglycans provide hydrated spaces between myoma cells. Fibronectin is a glycoprotein that mediates adhesion between myoma cells and extracellular matrix.

The most common change in leiomyomata is hyaline degeneration. The cut surface of a hyalinized area is smooth and homogeneous and does not show the whorllike arrangement of the rest of the leiomyoma. Almost all leiomyomata, except the smallest, have scattered areas of hyaline degeneration. Eventually, these may become liquefied and form cystic cavities filled with clear liquid or gelatinous material
(Fig. 31.3). Sometimes, the cystic change is so great that the leiomyoma becomes a mere shell and is truly a cystic tumor. Softness of a tumor does not necessarily indicate cystic degeneration. Fleshy leiomyomata may be equally soft.

Over time, with continued diminished blood supply and ischemic necrosis of tissue, calcium phosphates and carbonates are deposited in myomata. Their presence is evidence of a continuum of degenerative changes. The calcium may be deposited in varying amounts. If it is deposited at the periphery of the tumor, the leiomyoma may resemble a calcified cyst. Other calcified leiomyomata may show an irregular or diffuse distribution throughout with a honeycomb or mulberry appearance. When the degenerative change is advanced, the leiomyoma may become solidly calcified. Such calcified tumors have been called “womb stones.” Calcified leiomyomata are seen most often in elderly women, in African American women, and in women who have pedunculated subserous tumors. They are easily seen radiographically (Fig. 31.4).

Leiomyomata may undergo changes as a result of infection. Submucous leiomyomata are most commonly infected when they protrude into the uterine cavity, or especially into the vagina (Fig. 31.5). The pedunculated submucous leiomyoma thins out the endometrium as it grows inward, and eventually, the surface becomes ulcerated and infected (Fig. 31.6). An intramural leiomyoma in an involuting puerperal uterus can also become infected when endometritis is present. Microscopic abscesses can be found, and gross abscesses occasionally occur, particularly if the leiomyoma descends as low as the cervical canal. Such infections are usually streptococcal and may be virulent. Bacteroides fragilis infections also occur. Parametritis, peritonitis, and even septicemia may result.

Necrosis of a leiomyoma is caused by interference with its blood supply. Occasionally, a pedunculated subserous leiomyoma twists, and if an operation is not done immediately, infarction results. Necrosis sometimes occurs in the center of
a large tumor simply as a result of poor circulation. Necrotic leiomyomata are dark and hemorrhagic in the interior. Eventually, the tissue breaks down completely. So-called red or carneous degeneration is seen occasionally, especially in association with pregnancy. This condition is thought to result from poor circulation of blood through a rapidly growing tumor. Thrombosis and extravasation of blood into the myoma tissue are responsible for the reddish discoloration (Fig. 31.7).

A subserous and especially a subserous pedunculated myoma may gradually outgrow its blood supply (Fig. 31.8). To keep the myoma tissue from undergoing complete ischemic necrosis, the omentum becomes adherent to the peritoneal surface of a pedunculated subserous myoma and provides whatever blood supply is needed. Eventually, the pedicle may disappear or twist, and the myoma will become completely free from the uterus, wander in the upper abdomen, and receive its “parasitic” blood supply from the omentum and other sources.

On occasion, fat occurs in leiomyomata as true fatty degeneration. The cut surface may have a yellowish discoloration. Infrequently, a deposit of true fat may form a fibrolipoma; however, the presence of fat in a leiomyoma is rare. Indeed, if fat is seen grossly or microscopically in a curettage specimen, one should not assume that it represents fatty degeneration of a leiomyoma. One should assume that the uterus has been perforated and that fragments of fat have been curetted from the mesentery or omentum.

The most important, but rare, change in a leiomyoma is sarcomatous degeneration. There is much variation in the reported incidence of sarcoma in leiomyomata. The incidence given by
Novak is 0.7%. However, a review of 13,000 myomata by Montague and associates at Johns Hopkins Hospital revealed 38 cases of malignant change, the incidence of sarcoma thus being 0.29%. Corscaden and Singh indicated by their study that the true incidence of sarcoma developing within uterine leiomyomata is no higher than 0.13% and is probably as low as 0.04%. It should be remembered that because most women with uterine leiomyomata do not undergo surgical removal, the true incidence of sarcoma in leiomyomata is probably much lower than 1 per 1,000 (0.1%).

After hysterectomy in 1,429 patients with presumed benign leiomyomata, the histologic diagnosis of leiomyosarcoma was made in seven (0.49%), according to a study by Leibsohn and coworkers. There was no evidence of malignancy in the endometrial sampling of any of these seven patients, and the diagnosis was suspected intraoperatively in only three. Uterine weights ranged from 120 to 1,100 g. In a woman between 41 and 50 years of age with presumed symptomatic leiomyomata, there is a 1 in 112 chance of a leiomyosarcoma being present, according to these authors. This information has important implications in the consideration of conservative or delayed treatment for these women. Parker and associates found that the total incidence of uterine sarcomas (leiomyosarcoma, endometrial stromal sarcoma, and mixed mesodermal tumor) among patients operated on for presumed benign uterine leiomyoma is lower (0.23%) than the 0.49% reported by Leibsohn and coworkers.

The difficulty in defining the true incidence of sarcomatous change is understandable if one is familiar with the histology of leiomyomata. Abundantly, cellular leiomyomata are relatively common, and at first glance, they suggest sarcoma; however, they lack a significant number of mitotic figures, and patients from whom such tumors are removed all remain well. Misinterpretation of the histologic picture of this type of cellular leiomyoma undoubtedly accounts for the increased incidence of leiomyosarcoma reported by some. When cutting into leiomyomata in the operating room, the surgeon finds that sarcomatous areas have a somewhat characteristic appearance, although the histologic diagnosis certainly cannot be made by gross examination. A sarcoma is likely to occur in a rather large leiomyoma and toward the center of the tumor, where the blood supply is poorest. Instead of being firm fibrous tissue that grates when scraped with a knife blade, the tissue is soft and homogeneous and is described as resembling raw pork. Later, as necrosis of the malignant tissue occurs, it becomes more friable and hemorrhagic.

It has been difficult to understand uterine leiomyosarcoma, because pathologists do not agree on the criteria necessary for diagnosis. Some pathologists rely on the mitotic count. All tumors with less than 5 mitotic figures per 10 high-power fields are considered benign. All tumors with more than 10 mitotic figures per 10 high-power fields are called malignant. Those in between can be called smooth muscle tumors of uncertain malignant potential.

Other pathologists believe the mitotic count may have some significance but choose to rely instead on the presence of nuclear hyperchromatism, nuclear pleomorphism, or giant cells and other bizarre cell forms to make the diagnosis. Corscaden and Singh believe that no combination of histologic features is reliable and that only smooth muscle tumors that metastasize or recur are definitely malignant. We believe that all of these features should be taken into consideration for diagnosis and prognosis. When the tumor is confined to the uterus, both mitotic grade and histologic grade are important in the diagnosis and prognosis. A poor prognosis is associated with high mitotic counts and extremely atypical and anaplastic cytologic features. Bell and colleagues at Stanford University Medical Center assessed a variety of histopathologic features of 213 problematic smooth muscle neoplasms for which there were at least 2 years of clinical follow-up data. From the wide variety of light microscopic features assessed, the important predictors that emerged were mitotic index, the degree of cytologic atypia, and the presence or absence of coagulative tumor cell necrosis. Previously, the mitotic index was relied on exclusively to determine whether a uterine smooth muscle tumor was benign or malignant, but currently, an approach is used that incorporates additional histopathologic features.

A normal chromosome complement (46,XX) was observed by Meloni and coworkers in about 50% of leiomyoma cases. About 50% showed clonal abnormalities, such as those of chromosomes 1, 7, and 13, and t(12;14). Interstitial deletions of chromosome 7 were the ones most often involved, suggesting that this abnormality may be of primary importance in the cellular proliferation of leiomyomata. A relation between more aggressive histology and chromosomal abnormalities was also suggested.

Tumors that show obvious evidence of blood vessel invasion or spread to contiguous organs are rarely cured. The extent of the disease at the time of initial diagnosis is of even greater significance. In other words, when the diagnosis is suspected for the first time by the pathologist when he or she examines routine sections from a uterine leiomyoma, the patient almost always survives. However, if the diagnosis is made preoperatively by the gynecologist or is suspected during the operative procedure because of invasion of surrounding organs, the prognosis is grave.

An unusual atypical smooth muscle tumor was first described in the stomach by Martin and associates in 1960. Variously called bizarre leiomyoma, leiomyoblastoma, clear cell leiomyoma, and plexiform tumorlet, these atypical smooth tumors probably all belong together. The term epithelioid leiomyoma was adopted by the World Health Organization. Kurman and Norris have proposed that this term be used for all atypical leiomyomata. Histologically, the characteristic feature is the mixture of rounded polygonal cells and multinucleated giant cells present in epithelioid clear cell and plexiform patterns. Clinically, in the uterus most of these tumors are benign. They may rarely exhibit malignant potential. Malignancy is difficult to predict from histologic criteria because some metastases occurred from tumors that demonstrated very few mitoses. Kurman and Norris have suggested, however, that epithelioid neoplasms having more than 5 mitotic figures per 10 highpower fields should be called epithelioid leiomyosarcomas and that the term epithelioid leiomyoma should be applied when there is a lower level of mitotic activity. Although combination therapy (surgery plus radiation therapy or chemotherapy) may not be indicated for a patient with an epithelioid leiomyoma, follow-up should be considered essential, as emphasized by Klunder and colleagues.

An unusual benign form of leiomyomata uteri, intravenous leiomyomatosis, was first recognized at the turn of the 20th century and has been reported sporadically since then. Before 1982, about 50 cases had been reported, according to Bahary and coworkers. Probably, at least that many have been reported since. Marshall and Morris presented the first detailed report of this entity in the American literature in 1959. The characteristic feature of this peculiar smooth muscle tumor is the extension of the polypoid intravascular projections into the veins of the parametrium and broad ligaments. Although there may be some difficulty in distinguishing such lesions from lowgrade sarcoma, they are distinctly different histologically from stromatosis uteri because the intravenous plugs are mainly smooth muscle in origin. In 1966, Edwards and Peacock collected 32 cases of intravenous leiomyomatosis, including two cases of their own, and reviewed the clinical experience with this condition. In approximately 50% of the cases, the intravenous tumor was confined to the parametrium; in 75%, it extended no further than the veins of the broad ligament. The observations of Edwards and Peacock suggest that the severed intravenous extensions are probably incapable of independent parasitic existence and remain dormant after removal of the uterus. However, the cases presented by Bahary and associates tend to refute this idea. Total surgical excision of the tumor should be attempted for successful therapy. Some patients have survived for many years after incomplete resection of the tumor. A review of 14 cases of this rare uterine tumor from the file of the Armed Forces Institute of Pathology has been reported by Norris and Parmley. In this series, two of three patients with incomplete resection had a recurrence; the recurrent tumor was excised surgically, and the patients were alive and free of disease 5 and 11 years after operation. The authors concluded that this tumor behaves clinically like a benign neoplasm, although its wormlike extensions may involve uterine, vaginal, ovarian, and iliac veins. The uterine veins in the broad ligaments are the most common sites of extension. The mitotic index is quite low, with the most active lesions showing only one mitosis per 15 high-power fields. The material from the Armed Forces Institute of Pathology provides histologic evidence consistent with both theories of origin of intravenous leiomyomatosis, namely, that it may be the result of unusual vascular invasion from a leiomyoma or may arise de novo from the wall of veins within the myometrium.

Extension of benign leiomyomatosis up the vena cava and into the right atrium has been reported in several cases, with a fatal outcome in some. Before 1994, approximately 27 cases of intravenous leiomyomatosis extending to the heart were reported. Several recent cases requiring open heart surgery to remove the intracardiac tumor thrombosis have been successful and without recurrence. All reported cases occurred in women. Tierney and colleagues reported that substantial quantities of cytoplasmic estradiol and progesterone receptors were found in the right atrial tumor removed from a patient with intravenous leiomyomatosis. Their patient was treated with the antiestrogen tamoxifen because of residual tumor in the vena cava that could be estrogen dependent. Irey and Norris have presented evidence that female reproductive steroids can produce intimal proliferation of veins in predisposed persons. Interestingly, of the 30 patients with leiomyomata and leiomyosarcomas of the vena cava reviewed by Wray and Dawkins, 80% were female. Both intravenous leiomyomatosis and benign metastasizing leiomyoma have been reported to metastasize to the lung. As suggested by Banner and coworkers, by Horstmann and associates, and by Evans and colleagues, oophorectomy may be indicated in patients with these conditions, again because of the possibility that these tumors may be estrogen dependent or that estrogens may have the ability to stimulate their development, whether in a uterine or extrauterine location and whether they appear to be endothelial or mesenchymal in origin.

The possibility of metastases from a histologically benign uterine leiomyoma has been discussed by Idelson and Davids and by Clark and Weed. When such a case occurs, it is usually settled by finding a sarcomatous component in the leiomyoma or by finding evidence of intravenous leiomyomatosis. However, multiple cases have now been reported in which a benign uterine leiomyoma metastasized. Idelson and Davids’ case showed metastases to the aortic lymph nodes. The patient reported by Cramer and associates had metastatic tumor to the omentum, ovary, periaortic lymph node, and lung. In each location, the histology and estrogen receptor content of the tumor resembled those of a benign leiomyoma. The recommended treatment consists of surgical removal with castration and little or no estrogen replacement.

Leiomyomatosis peritonealis disseminata is sometimes confused with intravenous leiomyomatosis. However, only subperitoneal surfaces of the uterus and other pelvic and abdominal viscera are involved with leiomyomatosis peritonealis disseminata, and invasion of the lumen of blood vessels does not occur. Only about 15 cases have been reported, according to Pearce. All occurred in patients in the reproductive years who often had large uterine leiomyomata and were usually pregnant or taking oral contraceptives. The condition is likely to be confused with a disseminated intraabdominal malignancy, but it is entirely benign histologically and clinically. Parmley and colleagues have demonstrated the histologic similarities between this peritoneal lesion and the decidual change of the mesothelium in the pelvis, and they propose that the condition represents a benign reparative process in which fibroblasts replace soft peritoneal decidua. They suggest that this fibrotic reaction occurs during pregnancy and especially in the postpartum period, resulting in nodules with a pseudoleiomyomatous pattern. Similar findings have been noted in patients with endometriosis treated with prolonged Enovid therapy. These findings indicate that prolonged and continuous stimulation of subperitoneal decidua by either endogenous or exogenous estrogen and progesterone is important in the pathogenesis of this condition. Parmley and coworkers suggest that the condition is more appropriately called disseminated fibrosing deciduosis. Goldberg and associates, on the other
hand, on the basis of electron microscopy studies, believe that the tumors arise from smooth muscles of small blood vessels. This has been confirmed by Ceccacci and colleagues. It has been possible to show a continuum from fibroblastic cells through myofibroblasts to leiomyocytes. Although the cell of origin of this tumor is still controversial, the tumor is benign, and the acceptable treatment to date is total abdominal hysterectomy and bilateral salpingo-oophorectomy. If this tumor occurs in the omentum, an omentectomy should also be performed to define more clearly the histologic nature of the lesion.

In attempting to distinguish between benign and malignant disease in a patient with uterine leiomyomata who also has unusual clinical findings, it is appropriate to keep the entities mentioned earlier (intravenous leiomyomatosis, atypical bizarre leiomyoma, benign metastasizing leiomyoma, and disseminated intraperitoneal leiomyomatosis) in mind. Although they all have features similar to those of malignant disease, they are almost always benign and amenable to treatment. One should also remember that benign uterine leiomyomata have been associated with pseudo-Meigs syndrome in a few cases. Meigs reported five cases in 1954. In these cases, the ascites did not reappear after removal of the uterine leiomyomata.

There is a high frequency of endometrial hyperplasia when the uterus contains leiomyomata. Degligdish and Loewenthal reported that cystic glandular hyperplasia is often found in the endometrium at the margin of the leiomyoma. Yamamoto and coworkers have reported high concentrations of estrone and estrone sulfatase activity in the endometrium overlying a myoma. They suggest that the local hyperestrogenism in the endometrium overlying a leiomyoma may assist in the genesis or enlargement of these tumors.

Gynecologic surgeons are especially concerned about the vascularity of individual leiomyomata and about the blood flow to the uterus in the presence of multiple and sometimes very large leiomyomata. These considerations are pertinent when surgery, especially myomectomy, is contemplated.

According to Vollenhoven and associates, the vascularization of leiomyomata was studied by Vasserman and colleagues, and the findings were presented to the World Congress of Gynecology and Obstetrics in 1988. Using femoral arteriography, selective intraoperative angiography, radiography, and injection of surgical specimens, these investigators showed that leiomyomata have a rich vascular supply, including blood lakes within tumors. They found more than one nutrient vessel per myoma. Venous channels were predominantly peripheral, whereas the arterial supply was both internal and peripheral. Farrer-Brown and coworkers, using radiologic methods, demonstrated that myomata in various locations within the myometrium can cause congestion and dilatation of endometrial venous plexuses by obstructing venous return. These obstructions can result in ectasia of endometrial and myometrial venules (Fig. 31.9). The degree of vascularity of leiomyomata was also studied by Karlsson and Persson. Vascularity varied from many, to few, to no intrinsic vessels demonstrable. Generally, the sum of the width of the uterine arteries increases with the size of the uterus, but the diameter of the two sides sometimes differs markedly. A rich vascularity was found in 22 of 34 uteri with leiomyomata, but with increasing size, there is a tendency to less vascularity. In none of five cases with very large (20 cm or more) leiomyomata uteri was the vascularity rich. The intrinsic vessels were few in two cases and absent in three cases.