Article about fibroid biology in detail - technical
The Single-Cell Origin of Fibroids
Fibroids are benign tumors of the myometrium of the uterus composed of a widely variable amount of smooth muscle and extracellular fibrous material. The vast majority arise randomly in the population, while a small number of fibroids develop as part of the hereditary leiomyomatosis and renal cell cancer syndrome. Fibroids occur in the body of the uterus more frequently than in the cervix.
Each fibroid arises from 1 cell in the myometrium. This sweeping generalization is based on studies of women with multiple fibroids who were also heterozygous for the glucose-6-phosphate dehydrogenase (G6PD) gene on the X chromosome. In females, each somatic cell carries 2 X chromosomes, while each male carries only 1. One female X chromosome is silenced by packaging it in repressive heterochromatin. Which X chromosome will be inactivated is randomly determined, and once a cell’s X chromosome is inactivated, it remains inactive throughout that cell’s life. In these women, all cells in any given fibroid were either G6PD variant A or variant B, and not a mixture of the two. From this observation it was concluded by Linder and Gartler and by Townsend et al. that each fibroid arises from 1 cell in the myometrium. X-chromosome inactivation patterns in DNA methylation, in monoclonal androgen receptor DNA, and in enzyme phosphoglycerokinase have also demonstrated that each fibroid arises from a single cell within the myometrium.
Thus, multiple fibroids in the same uterus are not clonally related; each fibroid arises independently. Since many types of fibroids contain considerable quantities of fibrous tissue, presumably generated by fibroblasts, this evidence suggests that in addition to myometrial cells within fibroids being derived from a single cell, fibroblast within fibroids are derived from the same cell line. This further implies that the progenitor cell from which fibroids are derived are pluripotential and able to generate at least myometrial and fibroblast cell types. Furthermore, the high incidence of uterine fibroids in the general population suggests that their origin is connected with a frequently occurring cellular event. Menstruation is a very frequently occurring event in modern women, and the positive correlation of the risk of fibroids with age (which is a surrogate for number of menstrual cycles) suggests that excessive menstrual cycles may be causally related.
Genes and Fibroids
The familial predisposition to fibroids, a higher incidence of fibroids in black women versus white, and a higher risk of fibroids in monozygous twins than dizygotic all indicate that susceptibility to fibroids has a genetic component. It has been widely assumed, but never demonstrated, that fibroids arise from myometrial cell mutations. Documenting this assumption will be very difficult. Single nucleotide polymorphisms exist widely in the human genome, generating many alleles for each gene, some of which have no effect at all upon protein expression and others which produce 2 different but normally functioning proteins. In diseases in which hundreds of genes might be mutated to cause a specific disease, sorting out which single nucleotide substitutions are associated with a specific disease and which are not is a daunting task. For example, only recently has the genome of breast cancer been defined. In an analysis of 13,023 genes, a little more than half of the human genome, Sjoblom et al. identified 105 mutant genes thought to be causally related to breast cancer. A project of this magnitude will be needed to identify the genetic basis of fibroids.
Furthermore, cell proliferation and differentiation can occur without gene mutation. The fertilized egg differentiates and proliferates into all of the diverse tissue types of a baby’s body through differential activation of the transcriptome. If normal genes that regulate growth are upregulated and normal genes that promote apoptosis are downregulated and the state of regulation is stable, cell growth will occur as the result of a transcriptome difference, not a genome difference. Epigenetic phenomena may be at work in the development, maintenance, and degeneration of fi broids.
The transcriptome is commonly defined by complementary DNA gene-array techniques that define the state of mRNA at a given point in time of a cell’s life. However, gene-array technology, though widely used, is in its infancy and is not yet mature and stable. For example, when given mRNA samples from the same cohort of cells, different microarray brands produced different gene-array results. Despite such disturbing validity problems, many microarray studies comparing fibroids with adjacent myometrium have been published. Catherino and Segars compared 2 microarray studies in an editorial. In one study, 106 fibroid-related genes were identified; in the other, 68 separate fibroid-related genes were found. The vast majority of identified fibroid genes were unique to each study. Only 11 genes were differentially expressed in fibroids and myometrium in both studies. Catherino et al. have attempted to defi ne strategies for minimizing this type of disparity in fibroids gene-array studies. At the present time, comparing upregulated and downregulated genes in matched pairs of fibroid and myometrial tissue is often confusing. However, one of the themes that has emerged from this nascent science is that many extracellular matrix genes are deranged in fibroids when compared to myometrium.
Abnormal Chromosomes and Fibroids
Chromosomal abnormalities in fibroid cells are not rare. Nonetheless, the cytogenic abnormalities seen in fibroids are most likely secondary phenomena, developing after a fibroid is established and growing rapidly, and not causal.
Fibroid Malignant Potential
Conventional wisdom holds that malignant transformation of fibroids leads to leiomyosarcoma. However, fibroids are very benign tumors and they are very, very common. Leiomyosarcomas, on the other hand, are very, very rare. If leiomyosarcomas arise from fibroids, it would be expected that their incidence would be much higher. Some question whether fibroids ever progress to malignant sarcomas. From chromosomal analysis by genomic hybridization of 14 cases of uterine fibroids and 8 cases of uterine leiomyosarcomas, Packenham et al. concluded, “Our results do not provide evidence for the progression from benign leiomyoma [fibroids] to malignant leiomyosarcoma.” To further confuse the issue, fibroids can, microscopically, appear to be sarcomas.
a. Are Common
Multiple fibroids are common in women who have fibroids. In a study of 100 consecutive hysterectomy specimens, serially sectioned at 2-mm intervals, multiple fibroids were identified in 84% of the specimens from women with at least 1 fibroid. Multiple fibroids imply that multiple genome or transcriptome changes have occurred at different times during a woman’s life. Furthermore, fibroids developing at different times in a woman’s life imply that at any point in time, fibroids at different stages of development (growth, stability, or degeneration) and, consequently, of different sizes will be present within the same uterus. For example, in Sampson’s classic study of the blood supply of fibroids, published in 1912, he identified 1008 fibroids in 100 women, for an average count of 10 fibroids per uterus. Furthermore, these fibroids differed widely in size.
Multiple fibroids in the same uterus, with some being small, undetectable, and asymptomatic and others being larger, detectable, and symptomatic, pose a significant problem for treatments aimed at selectively or focally removing or killing 1 fibroid at a time. A wide range of focal, nonexcisional methods of killing fibroids, in situ, have been tested clinically. Radiofrequency energy delivered by bipolar needle (heat), laser (heat), focused ultrasound (heat), cryoablation probe (cold), and even the direct injection of 98% ethanol have all been demonstrated to kill fibroids. If a women has many fibroids, which ones should be treated—big ones, small ones, young ones, old ones?
b. Continue to Appear throughout Some Women’s Reproductive Lives
When fibroids are treated with myomectomy or myolysis, in approximately half of the cases, fibroids are again seen in the uterus. The appearance of fibroids growing on a uterus following either surgical removal or in situ ablation has two explanations. A third explanation, which is a mixture of the two, is certainly possible but, for brevity, will not be explored.
On the one hand, some of the original fibroid could have survived ablation or surgery and regrown. In that case, the appearance of fibroids following myomectomy or myolysis would truly be a “recurrence.” Some favor this interpretation. Further growth of ablated fi broids has been described. On the other hand, if all (or even nearly all) of the original fibroids were removed or killed during surgery, the appearance of fibroids at some time interval after myomectomy or myolysis could simply represent continued growth of small, untreated tumors or the interval appearance of new fibroids not present at the time of the intervention. In this instance, referring to the appearance of fibroids as “recurrences” seems inappropriate. A fibroid seen at some time interval after surgery would not be a recurrence of the original fibroid. It would be either the clinical appearance of a small, overlooked fibroid or the clinical appearance of an altogether new fibroid. I believe that in the vast majority of cases, fibroids seen after myomectomy or myolysis do not represent regrowth of the treated fibroid. Fibroids that appear after focal treatment are either untreated fibroids that have simply grown larger during the interval since surgery or de novo fibroids that have arisen in the follow-up interval. By the time follow-up imaging is performed at 1 to 5 years after surgery, these fibroids are large enough to be clinically appreciated. Consequently, I prefer the term “appearance of new fibroids” or “new fibroid appearance rate” instead of “recurrence” or “recurrence rate,” even though in some instances, a fibroid seen in a patient previously treated may actually be a recurrence.
Clinical data implies that women can be divided into 2 groups: those that develop a few fibroids during their life and those that develop many. Malone clinically distinguished women who were treated with myomectomy for a single fibroid and women who were treated with myomectomy for multiple fibroids in a series of 125 myomectomies. Imaging methods (US or MRI) were not used to detect fibroids preoperatively or postoperatively. The entire study is based upon clinical findings. The postoperative appearance of new fibroids was observed in 26.6% of women with solitary myomectomies and in 58.8% of women with multiple fibroids. Malone gives the colorful description of the “appearance of new crops of fibroids” to characterize the appearance of fibroids after myomectomy. The appearance of new fibroids following surgery is more common in women who have many fibroids in their uterus than in women who have few. Friedman et al. used ultrasound scanning to detect fibroids after myomectomy. They followed 18 women 27 to 38 months after myomectomy. They observed that in women who had 3 or fewer fibroids removed, new fibroid appearance rate was 25%. In women who had more than 3 fibroids removed, new fibroid appearance rate was 90%. Nishiyama et al. have demonstrated high new fibroids appearance rates in Japanese women. Attempts have been made to decrease the new fibroid appearance rate following myomectomy by giving a gonadotropinreleasing hormone analogue either pre- or postoperatively without success.
Fibroid Size Is a Balance between Cell Growth and Cell Death
Fibroid growth is a balance between mitosis, which produces more fibroid cells, and necrosis or apoptosis, which kill fibroid cells. If mitosis increases, and necrosis or apoptosis remain static, fibroids will grow. If mitosis remains constant, and necrosis or apoptosis decrease, fibroids will grow. If mitosis increases, and necrosis or apoptosis decrease, fibroids will grow even more. An increase in mitosis or a decrease in necrosis or apoptosis or a combination of the processes can lead to fibroid growth. Spontaneous necrosis of fibroids occurs, leading to fibroid degeneration and shrinkage.
a. Fibroids Grow from Their Periphery
Fibroids grow primarily from their periphery, somewhat like the trunk of a tree. Bourlev et al. demonstrated this process by examining proliferative and apoptotic activity and the expression of sex steroid receptors in central and peripheral parts of fibroids in different phases of the menstrual cycle. Paired biopsy specimens from myometrium and from peripheral and central parts of fibroids were obtained from 15 women in the proliferative phase and 8 women in the secretory phase of their cycles. During the secretory phase, mitosis was significantly higher in the peripheral than in the central parts of the fibroids. During the proliferative phase, apoptosis was significantly higher in the peripheral compared with the central parts. Taken together, this data indicates that fibroids grow from their periphery, and they grow during the secretory phase of the menstrual cycle.
Wei et al. obtained tissue from large (>10 cm in diameter) and small (<2 cm in diameter) fibroids from the same uterus and compared these fibroid tissues to adjacent normal myometrial tissues from 7 hysterectomy specimens. Each large fibroid was spatially divided into 6 zones, with zone 1 starting at the outer margin of the fibroid and proceeding to zone 6 in the fibroid’s center. Within each woman, differential microarray expression of selected genes in fibroids was compared with that of adjacent normal myometrium. Hypoxia, as measured by HIF-1 expression, was higher throughout large fibroids than in normal myometrium or in small fibroids. Furthermore, hypoxia increased toward the center of large fibroids. Cellular proliferation was higher in large and small fibroids than in normal myometrium and decreased from very high levels in the periphery of large fibroids to lower but still elevated levels, centrally. Conversely, hyaline degeneration was absent in normal myometrium and in small fibroids and present throughout large fibroids, increasingly so toward the center of large fibroids. The authors concluded that small fibroids and the periphery of large fibroids are more “biologically active” than myometrium and that large fibroids grow from their periphery.
b. Inhibition of Apoptosis: How Important?
i. Not Important
To determine the role of apoptosis in fibroid growth, Wu et al. obtained paired myometrial and fibroid tissue from 10 women in the proliferative phase of the menstrual cycle and from 8 women in the secretory phase and measured an apoptosis index in both. No difference in their apoptosis scores was observed between fibroid and adjacent myometrium, and no difference was seen between the proliferative and the secretory phases of the menstrual cycle. These authors concluded that cell proliferation, and not inhibition of apoptosis, is responsible for fibroid growth. Similarly, Dixon et al. measured the protein expression of 2 apoptosis genes, Bcl-2 (an antiapoptosis protein) and Bax protein (a proapoptosis protein) in women with fibroids. Cell proliferation in fibroids was measured in 3 ways: by measuring the concentration of proliferating cell nuclear antigen (PCNA) and Ki-67, and by mitotic counts. All 5 variables were measured in fibroids and matched myometrium from premenopausal women. Apoptosis was not a prominent feature of uterine fibroids or myometrium. On the other hand, PCNA, Ki-67, and mitotic counts were significantly higher in fibroids than in matched myometrial samples (p<0.05). The authors concluded that increased cell proliferation was the most significant contributor to fibroid growth
Several investigators have published studies that indicate that the suppression of apoptosis is an important determinant in fibroid growth. Soini and Paakko demonstrated that Bcl-2 (an antiapoptosis protein) was expressed in 6 of 14 (42.9%) fibroids examined. Matsuo et al. examined paired myometrial and fibroid tissues from hysterectomy specimens in premenopausal women and observed that Bcl-2 (an antiapoptosis protein) was “abundantly present in the cytoplasm of leiomyoma cell [while] scarcely present in normal myometrial smooth muscle cells.” Furthermore, in fibroids, Bcl-2 protein increased from the secretory phase to the proliferative phase. In contrast, in adjacent myometrium, Bcl-2 protein concentration did not change during the menstrual cycle. Studying fibroid cell cultures, Maruo et al. demonstrated that progesterone stimulates the production of Bcl-2 (antiapoptosis protein), whereas estrogen diminishes the production of Bcl-2. Wu et al. obtained paired myometrial and fibroid tissue samples from 18 premenopausal and 6 postmenopausal women. A significant difference in Bax protein (a proapoptosis protein) was observed between fibroids and myometrium in tissues obtained from women in the secretory phase of the menstrual cycle. The Bcl-2 protein (an antiapoptosis protein) was more abundant in fibroids than in myometrium only in tissues obtained in the proliferative phase of the cycle. Further, Bcl-2 protein was more abundant in fibroids from fertile women than in fibroids from menopausal women. No significant differences were observed for the Bax proteins between menopausal and premenopausal women. Maruo et al. studied the effect of progesterone on Bcl-2 production in fibroids and demonstrated that progesterone upregulates Bcl-2 production in fibroids. In a review article, Martel demonstrated that Bcl-2 protein is expressed cyclically in fibroids, with more expression in the secretory than in the proliferative phase.
Hormonal Influence on Fibroids
a. Clinical Observations
i. Nonpregnant Women
Because fibroids have not been described before menarche and because they regress after a woman reaches menopause, it has been clinically deduced that estrogens and progesterone strongly infl uence fibroid maintenance and growth. Furthermore, when estrogens and progesterone are administered to postmenopausal women, quiescent fibroids begin to grow again. For example, in a study of 44 premenopausal and 12 postmenopausal women, Lamminen et al. demonstrated that fibroids in postmenopausal women receiving no hormones had low proliferative activity. In contrast, women receiving estrogen and progesterone replacement had fibroid proliferative activity equal to premenopausal women. However, the effects of estrogens and progesterone on fibroids are not simple. Blood levels of estrogens and progesterone do not predict fibroid development. In a study by Buttram, no difference was observed in peripheral blood levels of estrogens or progesterone in women who develop fibroids and those that don’t. Furthermore, quantitative estrogen and progesterone receptor concentration in fibroids changed in parallel with changes observed in myometrium in 17 women studied by Soules and McCarty. On its surface, the most convincing clinical demonstration that estrogens and progesterone profoundly affect the maintenance of growth of fibroids is the dramatic decrease in fibroid volume following treatment with a gonadotropin-releasing hormone (GnRH) analogue. A decade or so ago, it was believed that GnRH analogues first decrease the production of these 2 ovarian sex steroids and then, secondarily, that low estrogen and progesterone levels cause fibroids to shrink. As we shall see, the GnRH analogue story is no longer that simple. Fibroids have nuclear receptors for GnRH. GnRH analogues may act directly upon fibroids.
ii. Pregnant Women
Even though by 1943 it had been documented to the contrary, it has been stated time and again in the world’s gynecology literature that fibroids grow during pregnancy because it is a time when progesterone and estrogens concentrations are elevated. For example, Wallach and Vhahos state, “Rapid growth of myomas is common during pregnancy,” and Haney states, “leiomyomata … often dramatically enlarge during pregnancy.”
The fact is, most fibroids do not grow during pregnancy. Farquhar et al. estimated that 80% of fibroids either become smaller during pregnancy or show no change in size during pregnancy, and when they do enlarge, volume increase is rarely greater than 25%. Longitudinal US studies have shown that, in general, fibroids are stable in size during pregnancy. These and other studies have also shown that following delivery, many fibroids actually shrink in size. Childbirth appears to kill fibroids. In one study, 81 (92%) of 89 women showed no change in fibroid diameter from the first trimester to delivery. Thirty-one of these women were available for rescanning 6 weeks postpartum. In these 31 women, all fibroids showed a “marked decrease” in size, in the 50% diameter range. In a second study of 113 pregnant women, Lev-Toaff et al. observed that most fibroids decreased in size during the third trimester. In a third study, individual fibroid volumes were followed serially with US in 29 pregnant women. In 25 (78.1%) of 32 fibroids, no change in fibroid size was noted during the course of pregnancy. At 6 weeks postpartum, 13 women were available for rescanning. In 5 (38.4%) of these 13 women, fibroid volume was unchanged. In the remaining 8 (61.6%) women, fibroids decreased so much in volume that they could no longer be identified. In a fourth study, fibroids did not change in volume during pregnancy in 25 (69.4%) of 36 women. In 34 women re-examined at 4 weeks postpartum, average fibroid volume reduction was 12.9%. Finally, in a fifth study, Strobelt et al. examined 134 pregnant women who had fibroids. Sonograms were obtained at 2-week intervals until 20 weeks of gestation, and monthly thereafter. They observed that the majority (85.1%) of fibroids remained stable or decreased in size during the second and third trimesters. Consistent with these US observations, estrogen receptor expression in fibroids was observed throughout the menstrual cycle in nonpregnant women but was suppressed during pregnancy. Finally, Sidorova et al. directly measured regional blood flow in 35 new mothers with fibroids and compared their measurements with 25 women without fibroids. Using an electrical current method to estimate blood flow, they observed decreased blood flow within and around fibroids in these new mothers.
b. Transcription and Growth Factors
Estrogens and progesterone affect all tissues indirectly, through nuclear receptors, and compared to myometrium, fibroids overexpress these receptors. To cause an effect, estrogens and progesterone first bind with their corresponding soluble intracellular receptors. Once activated, these receptors move to the nucleus, where they function as a ligand-dependent transcription factor. These transcription factors then up- or down-regulate the expression of genes. Gene transcription then leads to the production of local growth factors, which, in turn, promote fibroid maintenance or growth. The growth factors exert their effects upon the cells that produced them (an autocrine effect) or upon neighboring cells (a paracrine effect). Growth factors include epidermal growth factor, insulin-like growth factors, heparin-binding growth factor, transforming growth factor β, platelet-derived growth factor, interferon-α growth factor, basic fibroblast growth factor, and various angiogenic growth factors. The number and type of intracellular and intercellular growth factors is quite complex and, as yet, incompletely understood. Recent reviews of growth factors and fibroids have been published. In addition, “cross-talk” exists between estrogen receptors and progesterone receptors, which makes the study of the fibroid effects of estrogens and progesterone even more complex.
i. Effects on Fibroids
Progesterone must be present for fibroids to grow. As early as 1949, Seagaloff et al. demonstrated that exogenous progesterone caused increased mitotic activity and cellularity in fibroids. Tiltman reported that the mitotic activity in fibroids increased in women taking progestin-containing birth control preparations. Increased mitotic activity would be presumed to be linked with fibroid growth. However, in a comparison of 910 Thailand women who had used quarterly injections of depo-medroxyprogesterone acetate for birth control with 2709 controls who had not, Lumbiganon et al. observed that the odds ratio for developing fibroids was 0.44 for the progestin users compared with 1.00 for nonusers. Continuous progesterone exposure appears to have a protective effect on the risk of developing fibroids. Cyclical progesterone exposure may not.
The effect of progesterone on fibroids is not uniform throughout the menstrual cycle. Fibroids grow primarily during the secretory phase of each menstrual cycle. In their 1989 study, Kawaguchi et al. performed high-power field mitotic counts in hysterectomy specimens containing fibroids from women in various points of the menstrual cycle. The mean mitotic count in the secretory phase, 12.7 per 100 high-power fields, was significantly higher than that of the proliferative phase or menses (3.8 per 100 high-power fields, [p<0.01] and 8.3 per 100 high-power fields [p<0.05], respectively). The highest mitotic count in fibroids, 54 per 100 high-power fields, was observed during the early secretory phase. Fibroids from younger women exhibited significantly higher mitotic counts in the secretory phase than fibroids from the older women. In their 1991 study, Kawaguchi et al. investigated the immunohistochemical distribution of estrogen receptors (ER), progesterone receptors (PR), and the cell proliferation-associated antigen Ki-67 in fibroids during the menstrual cycle and in pregnancy. In the myometrium, ER expression was observed in the proliferative phase but was suppressed in the secretory phase and during pregnancy. In fibroids, ER expression was observed throughout the menstrual cycle but was suppressed during pregnancy. However, PR was expressed both in the myometrium and fibroids throughout the menstrual cycle and pregnancy. In both the myometrium and fibroids, a higher number of Ki-67-positive cells was observed during pregnancy than in the secretory phase, and Ki-67 was negative during menopause. The Ki-67-positive cell count in fibroids was significantly higher than that in the myometrium throughout the menstrual cycle and pregnancy. Thus both myometrium and fibroids have high-growth activity under the hormonal milieu of high progesterone levels. The growth potential of fibroids is apparently higher than that of myometrium throughout the menstrual cycle and during pregnancy.
Supporting these menstrual phase observations is a study by Harrison-Woolrych et al. demonstrating that epidermal growth factor in fibroids is only increased during the secretory phase of the menstrual cycle. Weighing the relative effect of estrogens and progesterone on fibroid shrinkage following GnRH analogue therapy, Wang et al. demonstrated that fibroid shrinkage is inversely related to the concentration of unbound progesterone receptors within fibroids. As the concentration of unbound PR decreased, fibroid shrinkage increased. No relationship, either direct or inverse, was observed between fibroid shrinkage and unbound estrogen receptors.
Progesterone can negate the GnRH agonist-induced shrinkage of fibroids. For example, Carr et al. compared the effectiveness of administering a progestin in conjunction with a 6-month course of a GnRH analogue When the progestin was not administered, total uterine volume decreased to 73% of the baseline following 12 weeks of GnRH analogue therapy. Uterine volume did not drop when the progestin was administered in conjunction with the GnRH analogue. The role of progesterone and fibroid growth has been carefully reviewed.
ii. Progesterone Blockers, Antiprogesterones, and Selective Progesterone Receptor Modulators
Antiprogesterones can reduce fibroid volume and menstrual symptoms to the same general degree as GnRH analogues can. RU- 486, or mifepristone, has a direct antiprogesterone effect. In a review of 6 mifepristone-fibroid treatment clinical trials, average uterine volume decreased 27%–49% and average fibroid volume decreased 26%– 74%. When mifepristone was stopped, regrowth to pretreatment levels occurred. The Selective Progesterone Receptor Modulators (SPRM) asoprisnil can also reduce fibroid volume and menorrhagia.
i. Effects on Fibroids
Estrogens are also needed for the well-being of fibroids, but their role may be more supportive than stimulatory. Estrogens exert their physiological effects on target cells by binding to 2 subtypes of nuclear receptor, estrogen receptor alpha (ERα) and estrogen receptor beta (ERβ). Both ERα and ERβ are expressed in myometrium and fibroids and fluctuate during the menstrual cycle. The expression of estrogen regulated genes is elevated in fibroids compared to adjacent myometrium. Kitawaki et al. measured the frequency of 3 genotypes of the ERα gene in 67 women with fibroids and in 206 controls. The frequencies of these genotypes were significantly different between the fibroid group and controls. The authors concluded that polymorphism of the ERα gene is associated with increased fibroid risk.
Aromatase synthetase converts testosterone to estrogens. Fibroids overexpress the gene for aromatase synthetase and produce estrogens from testosterone, in situ, acting in an autocrine/paracrine mechanism to stimulate fibroids. Estrogen concentration has been shown to be higher in fibroids than in adjacent myometrium.
ii. Estrogen Blockers, Antiestrogens, and Selective Estrogen Receptor Modulators
Because the effects of estrogens are tissue specific, the word antiestrogen has proven to be inadequate. The mouthful “selective estrogen receptor modulator,” or SERM, has been chosen instead to reflect the complexity of estrogen interactions with different tissue types: ovary, uterus, bone, and the cardiovascular system. Different SERMs have been studied to determine their effect on fibroids.
In a 6-month study in which tamoxifen was administered to 6 premenopausal women with uterine fibroids, no change in the size of fibroids was observed. Women experienced variability only in menstrual cycle duration, with a significant lengthening of the luteal phase (p<0.02).
In a study of raloxifene and fibroids, 90 women with fibroids were treated with 2 different dosages of raloxifene and a placebo. At entry into the study and after 3 and 6 menstrual cycles, uterine and fibroid sizes were measured by transvaginal ultrasonography. No significant changes were observed in uterine and fibroid size between treated groups and controls. Similarly, the length and severity of uterine bleedings were not significantly different between treated and control groups. In another study of raloxifene, 25 women with fibroids were randomized to treatment or no treatment. Raloxifene treatment prevented the progression of fibroid growth.
Compared to antiprogesterones and SPRMs, SERM effects upon fibroids seem minimal.
e. GnRH analogues
i. Indirect Action
GnRH is produced in and released from the hypothalamus of the brain and binds to its receptors in anterior pituitary cell causing de-novo synthesis and release of gonadotropins from the anterior pituitary gland, which, in turn, stimulate the production of ovarian sex steroids. GnRH analogues block the production of gonadotropins, which decrease the ovarian production of estrogens and progesterone. However, this is not the only effect GnRH analogues have on fibroids, endometrium, and myometrium.
ii. Direct Action
A variety of direct effects of GnRH analogues have been reported. Fibroids, endometrium, and myometrium cells express GnRH and GnRH receptors and contain GnRH binding sites. During the 12-week administration of a GnRH agonist to 18 women with symptomatic fibroids, blood levels of growth hormone and insulin-like growth factors were serially measured. Concentrations of growth hormone and insulinlike growth factors dropped in the circulation during the 12-week study. Rein et al. compared fibroid tissue culture explants from 17 women treated with a GnRH agonist and from controls. Tissue was obtained during myomectomy. Secretions from the fibroid tissue cultures from women treated with the GnRH analogue were lower in insulin-like growth factors. Downregulation of transforming growth factor, upregulation of collagens, and upregulation of vascular endothelial growth factor have been reported following GnRH analogue treatment. Shozu et al. reported that GnRH agonist therapy inhibits the expression of fibroid aromatase gene expression, suppressing the in situ production of estrogen. In women receiving no GnRH agonist medication, fibroids express aromatase mRNA at levels 20 times higher than that in the surrounding myometrium. GnRH agonist treatment for 12 to 24 weeks reduced the expression of aromatase mRNA in fibroid tissue as well as in the myometrium, to approximately one-tenth of that in the myometrium of untreated women. In a side study of fibroid cell cultures, the addition of estradiol did not affect the aromatase activity of fibroid cells, suggesting that deprivation of circulating (ovarian) estrogen is not a cause of decreased expression of aromatase during GnRH agonist therapy.
iii. Apoptosis or Not?
Huang et al. measured an apoptosis marker in fibroid tissue from 15 women receiving GnRH analogue therapy and compared those results with tissue from 44 controls. Apoptosis was observed at the same rate in treated and control specimens. A follow-up study of 20 women examining a wide range of apoptosis markers confirmed these results. The authors concluded that GnRH analogues do not induce apoptosis in fibroids. Higashijima et al. compared apoptosis levels in fibroids among 26 premenopausal women treated with a GnRH analogue, 20 premenopausal women not treated, and 15 postmenopausal women. Apoptotic cells were observed in only 14 (53.8%) of the 26 treated fibroids, 10 (50.0%) of the untreated, but 12 (80.0%) of the postmenopausal women. Even though the authors stated that their findings supported apoptosis as a mechanism of action for GnRH analogues, their data did not.
The opposite conclusion has been reached by others. Mizutani et al. compared a GnRH analogue treatment group of 43 women treated for 16 weeks to a control group of 18. Tissue was obtained following GnRH analogue therapy at myomectomy or hysterectomy. Apoptosis was transiently increased in treated fibroids during week 4. In a cell culture study, Hatsuo et al. observed an increase in apoptosis following treatment of a fibroid cell culture with a GnRH analogue.
On balance, experts believe that apoptosis plays little role in the shrinkage of fibroids following GnRH analogue treatment.
iv. GnRH Analogue Types
Two types of GnRH analogs have been synthesized: agonists and antagonists. Agonists bind to GnRH receptors, decrease the number of available GnRH receptors, and suppress anterior pituitary gonadotropin synthesis. While bound to GnRH receptors, agonists mimic the action of GnRH and then interfere with action of GnRH. During the time that they mimic the action of GnRH, a “flare” is said to occur because gonadotropin synthesis is increased. Antagonists bind to GnRH receptors, have no physiological effect on the receptors, and consequently do not mimic the action of GnRH. However, antagonists do interfere with the action of GnRH. Antagonists do not cause a flare.
(a) GnRH Agonists
More is known about gonadotropin-releasing hormone agonists. Agonists reduce fibroid size by approximately 50% after 3 months of therapy. The degree of shrinkage has been shown to be inversely related to circulating estradiol levels. As estrogen levels fall, fi broid shrinkage increases. However, as indicated above, this relationship may not be causal. GnRH agonists may directly cause both a drop of circulating estrogen levels and fibroid shrinkage.
The effect of GnRH agonists on fibroid size is dependent on continued administration of the agonist. Three months following the cessation of GnRH agonists therapy, uterine volume returns to pretreatment levels. Rein and coworkers compared estrogen receptor status in fibroids from 10 women on GnRH agonist therapy with 10 control women. They demonstrated a significant increase in estrogen receptor content in fibroids from agonist-treated women. They concluded that “the significant increase in fibroid ER may be an explanation for the rapid regrowth of fibroids observed after the cessation of GNRH-a therapy.” Shortly after starting GnRH analogue therapy, patients become amenorrheic. Four to 10 weeks following therapy, menses return. Wang et al. reported the regrowth of fibroids in 5 women. GnRH agonist therapy is known to shrink fi roids, but the molecular mechanisms responsible for this effect remain poorly understood.
GnRH agonist therapy does not induce significant apoptosis in fibroids. Huang et al. investigated the effects of a GnRH agonist on uterine fibroids by profiling the expression levels of apoptosis-related molecules such as Fas/Fas ligand (FasL); caspases 3, 6, 7, 8, 9, and 10; and Bcl-2 from specimens of 20 patients receiving a GnRH agonist before myomectomy and 24 controls. The authors concluded that GnRH agonist therapy fails to increase apoptosis in uterine fibroids.
(b) GnRH Antagonists
GnRH antagonist effects on fibroids have not been as extensively studied as agonists. The GnRH antagonist Cetrorelix was administered to 20 premenopausal women with symptomatic fibroids prior to either hysterectomy or myomectomy. Pretreatment and weekly MR images were obtained to measure uterine and fibroid volumes. Volume reductions were comparable to those reported in GnRH agonist studies without patient experience of initial flare-up. Others have reported the use of this antagonist preoperatively.
v. GnRH Analogue Side Effects
GnRH analogue therapy is medically induced menopause. Long-term gonadotropin-releasing hormone analogue therapy is, consequently, poorly tolerated. Patients experience hot flushes, headache, insomnia, decreased libido, irritability, depression, vaginal dryness, and amenorrhea. In addition, prolonged therapy has led to bone loss in some women. In addition to its effect on fibroids, agonists may harm myometrium. Following GnRH analogue therapy, Uemura et al. described hyaline degeneration in the myometrium, in addition to hyaline degeneration in fibroids.
(a) Estrogen Add-back Therapy
Because long-term GnRH analogue therapy is poorly tolerated and may be damaging to bone mass and normal uterine tissue, analogue therapy without add-back hormone therapy is now limited to use prior to myomectomy or hysterectomy to decrease fibroid size and vascularity at the time of surgery. Furthermore, analogue therapy for even as short as 2 months prior to surgery may correct anemia. Some women whose fibroid size might exclude them from vaginal hysterectomy may become suitable for vaginal hysterectomy following analogue therapy. However, others question the value of preoperative analogue therapy. Continuous GnRH analogue treatment in conjunction with estrogen and progesterone add-back therapy to counteract the menopausal symptoms of estrogen deficiency has been demonstrated a safe and effective treatment for fibroids for up to 2 years.
(b) Progestin Add-back Therapy
Friedman et al. compared the effects of an add-back of progestin in women treated with a GnRH analogue. Total uterine volume, fibroid volume, and myometrial volume were measured from MR images before the study started and at weeks 12 and 24 during the study. In a cross-over design, when the progestin was administered during the first 12 weeks of GnRH analogue therapy, uterine volume did not decrease on 12-week MR images. When the progestin was withdrawn during the last 12 weeks of GnRH analogue therapy, uterine volume decreased significantly. Similarly, when the progestin was added back to the second treatment group, uterine volume increased significantly. In 1994, the same group published a report describing the effect of progestin on 51 women after receiving a GnRH analogue as primary therapy for 2 years. They observed that average uterine volume in women receiving progestin add-back increased 87% by 12 months and 95% by 24. Clearly, if uterine and fibroid volume reduction is the treatment goal in women with symptomatic fibroids, add-back of a progestin makes no sense.
vi. What GnRH Analogues Studies Teach
Even though GnRH analogue therapy may have limited clinical use for the treatment of fibroids, many carefully performed and interesting GnRH analogue studies exist that shed light on fibroid biology.
(a) Histology and Fibroid Volume Change
Microscopic studies that compare fibroid and myometrial tissue from women treated with GnRH analogues and controls have been performed to determine which microscopic elements change in treated women. In general, these studies teach that the cellular fibroid elements diminish and that extracellular fibroid elements increase in treated women. It seems that extracellular fibroid elements take up less space overall than cellular fibroid elements, which results in overall fibroid shrinkage. Kalir et al. examined fibroids from 10 women treated with GnRH agonist and compared the treated fibroids to untreated fibroids from control women. They also reviewed 7 similar published studies that were similar in design to their own. They observed that “myomas from patients treated with GnRH agonists exhibited more hyalinization, greater cell density, slightly smaller cell sizes, and larger collagen fibrils.” In their review of the 7 similar investigations, 1 study found no histologic explanation for fibroid shrinkage, 1 study had findings contrary to the reviewers’, and 5 studies made similar observations. Similar conclusions have also been published by others. Higashijima et al. and Deligdisch et al. observed hyaline degeneration in fibroids treated with agonists. Wang et al. studied 5 women treated with an agonist and observed “marked cellular shrinkage” in the fibroid cells.
(b) Blood Flow and GnRH Analogues
Some believe that GnRH analogue therapy leads to fibroid shrinkage by first decreasing blood flow to the uterus. They argue that as fibroids become inadequately perfused, they are starved of needed nutrients and shrink. This line of thinking is partly based on the work of Matta et al. These authors stated, “Our results would suggest that the reduction in size of the uterine fibroids is related to hypoestrogenism, possibly via hypoestrogenic mediated reduction in blood flow.” They base their conclusion on hormone and Doppler velocity measurements of 8 women undergoing GnRH agonist therapy. Doppler velocitometry in uterine and fibroid arteries was recorded prior to GnRH analogue treatment and then at 2 months and 4 months into treatment. Their data showed that as estrogen levels fell, uterine and fibroid volumes decreased (41.2% and 67.2%, respectively), and resistance in the uterine and fibroid arteries increased. Similar results have been observed by Aleem and Predanic. They studied arterial resistance by Doppler velocitometry serially in 23 women receiving a GnRH agonist over 12 weeks. They noted that arterial resistive index increased first in fibroid arteries and that the increase in resistance correlated in time with the decrease in blood levels of estradiol. The increase in arterial resistance and the decrease in estradiol level occurred before fibroid volume decreased. The authors concluded that “the first signifi cant effect of GnRH analogue therapy in the process of uterine and leiomyoma volume shrinkage is the reduction of leiomyometrial rather than uterine blood flow.” It should be noted that other authors have noted no change in Doppler-measured arterial resistance following GnRH agonist therapy.
Did agonist therapy cause an increase in fibroid arterial resistance, which led to reduced arterial blood flow to fibroids, starving them of needed nutrients? Or did agonist therapy first reduce metabolic activity in fibroid cells, which secondarily led to reduction in blood flow? Did the chicken come first or the egg?
Throughout the body, organ blood flow is autoregulated to meet the metabolic need of that organ. Circulating hormones do not cause blood flow changes by organ. The regulation of blood flow to the uterus through the uterine arteries is particularly complex and varies during pregnancy. Furthermore, the diameter of a uterine artery can grow or shrink to accommodate changes in blood flow. However, a structure as simple as an arteriovenous fistula in the myometrium can cause an increase in uterine artery blood flow, and obliteration of the fistula can cause the increased flow to decrease. Specialists in this area do not believe that GnRH analogues directly regulate the arterial blood flow to the uterus. A study by Alatas et al. supports this view. These gynecologists studied Doppler blood flow in the uterine arteries in healthy women and in women with fibroids. The resistance in the uterine arteries correlated most closely with uterine volume. They observed that when the uterus was larger, blood flow was higher. Looked at in reverse, they could not differentiate women with fibroids from those without by Doppler velocitometry. Furthermore, high arterial resistance within fibroids with high cell proliferation rates has been reported. Arguably the most convincing observation that metabolic activity within the uterus infl uences blood flow to the uterus comes from 3 case reports of Mengert et al. They describe successful full-term pregnancies in 3 women who had previously had surgical ligation of both internal iliac arteries and both ovarian arteries. By inference, blood fl ow from collateral arteries (arteries with no known estrogenic sensitivity) increased enough during these 3 pregnancies to support the huge blood fl ow requirements of each gestation.
(c) Which Fibroids Respond to GnRH Analogues?
Microscopic studies have also been performed to determine histologic features that predict which fibroids will respond best to GnRH analogue therapy. In general, these studies teach that the degree of fibroid shrinkage is a function of the proportion of cellular and extracellular tissue within a given fibroid. The more cellular tissue present, the more a fibroid will decrease in volume during GnRH analogue therapy. The more extracellular fibroid tissue present, the less a fibroid will decrease in volume during GnRH analogue therapy. Kawamura et al. correlated pre-GnRH analogue therapy microscopic anatomy from transcervical fibroid biopsies (cellularity, hyaline content, and collagen content) with fibroid shrinkage. They observed 3 major findings: As fibroid cellularity increased, fibroid volume decreased more with agonist therapy; as hyaline content increased, fibroid volume decreased less with therapy; as collagen increased, fibroid volume decreased less with therapy. Similarly, Oguchi et al. showed that fibroid volume dropped more following GnRH analogue treatment in fibroids that had higher smooth muscle counts and higher proliferative cell indices. They concluded that the response to GnRH analogue correlated with “cellularity and proliferative activity of fibroids.” Finally, Yamashita et al. compared MRI contrast medium enhancement and fibroid shrinkage for various types of fibroids during agonist therapy. They found that cellular fibroids, which are made up almost exclusively of cellular fi broid tissue, enhanced more and decreased more in volume than fi broids composed of extracellular fi broid tissue.
Fibroid Tissue Is Extremely Heterogeneous
a. Fibroids Spontaneously “Degenerate”
As fibroids grow, they are said to undergo “degeneration.” It is a commonly held, but unproven, belief that a fibroid degenerates because it outgrows its blood supply. In support of this notion is the observation that large fibroids are more hypoxic than small fibroids and are more frequently associated with degenerative changes. But even in very small fibroids, fibrosis is more frequently seen as size increases. In a study of fibroids less than 1 cm in diameter, which are very, very small fibroids, Cramer et al. noted “substantial fibrosis” in 23 of 101 (23%) fibroids 2–4 mm in diameter versus 45 of 112 (40%) of fibroids 5–9 mm in diameter, a difference significant at p<0.05. In a study of 298 fibroids in women from Jamaica, West Indies, Persaud et al. observed that “some form of degenerative change was encountered in 195 (65.4%)” of the fibroids studied. Hyaline degeneration was most common (63%); myxomatous, next most common (13%). Hyaline degeneration (or hyaline fibrosis) is characterized by the presence of extracellular material that has a glass-like or hyaline appearance with routine H&E staining under light microscopy. Hyalinization can progress to the point where no identifiable smooth muscle cells are seen microscopically. Myxomatous degeneration is characterized by the presence of an extracellular gelatinous substance between smooth muscle cells. Like hyaline degeneration, myxoid degeneration can be so extensive that it is difficult microscopically to identify smooth muscle cells. Mucoid degeneration was also defined but was very similar in appearance to myxomatous degeneration. Cystic degeneration of fibroids occurs. Hemorrhagic degeneration of fibroids, referred to as “red degeneration,” involves acute hemorrhage into a previously hyalinized fibroid. And finally, degenerated fibroids can calcify, appearing on plain film radiographs as “popcorn.” In all forms of degeneration, as the process progresses, living smooth muscle cells become less plentiful, and extracellular material (hyaline, myxoid, fat, cyst fluid, or calcium) becomes more plentiful.
An alternative explanation for these observations sees different types of fibroids as different phenotypic expressions of each fibroid’s underlying genome or transcriptome. The production of hyaline, myxomatous material, and cystic fluid may be genetically programmed in the nucleus of each fibroid. Instead of undergoing “degeneration,” the production of extracellular material by each fibroid may be that fibroid following its genetic plan. For example, a human cell culture study demonstrated that fibroids cells produce more collagen than myometrium. When Tranilast, an agent used to treat scar formation, was added to fibroid cell cultures, fibroid growth stopped. Palmer et al. examined levels of 4 matrix metalloproteinases (MMPs) in fibroids and myometrium from 22 women, during the proliferative phase of the menstrual cycle in 6 women and the secretory in 16. Two MMPs were elevated in concentration in fibroid tissues compared to myometrium, suggesting that fibroids form more extracellular matrix than is formed in myometrium. Stewart et al. observed elevated concentrations of collagen mRNA in fibroids compared to adjacent myometrium.
b. “Alive” Elements = Cells
Microscopic fibroids, which are generally younger and healthier fibroids, are sometimes referred to as “seedling” fibroids and are composed of smooth muscle cells, fibroblasts, and collagen. Of these 3 elements, only smooth muscle cells and fibroblasts are metabolically active or “alive.”
c. “Dead” Elements = Extracellular Matrix
Collagen, the third major component, is a protein secreted from fibroid cells, is extracellular in location, and is inert. Though collagen is not alive, in all fibroid subtypes but cellular fibroids, it is produced in abundance. In fact, the production of collagen may be a defining genetic feature of fibroids. Stewart et al. characterized the localization of 3 extracellular matrix (ECM) proteins in fibroids and adjacent normal myometrium: collagen type I, collagen type III, and fibronectin. The authors studied the expression of messenger ribonucleic acid (mRNA) levels for these proteins from women who were in various stages of the menstrual cycle and from multiple fibroids from the same patient. Fibroids showed more intense staining for collagen types I and III than corresponding normal myometrium; mRNA was consistently elevated in fibroids relative to the adjacent myometrium in patients who were in the proliferative phase of the menstrual cycle (p<0.02 for both), a difference not seen in the secretory phase. Catherino et al. have demonstrated that in comparison to adjacent myometrial cells, the expression of ECM genes in fibroids are deranged. Shime et al. demonstrated that Tranilast, a drug that interferes with collagen accumulation, inhibited the proliferation of cultured human fibroid cells in a dose-dependent manner without inducing apoptosis.
The dead fibroid elements cannot, of themselves, grow. They are not alive. The living elements in fibroids, however, can grow quite rapidly. Schlaff and colleagues observed a 43.5% increase in uterine volume and a 56.2% increase in fibroid volume over a 6-month period in 6 untreated women with fibroids who were the control arm in a hormone drug study. In a remarkable case report, Szczurowicz and Wegrzycki reported 6 kg of fibroid growth over 6 months.
d. The Spectrum of Fibroids’ “Aliveness”
It is possible then to imagine a spectrum characterizing how “alive” a fibroid is. Cellular fibroids, an uncommon fibroid subtype, are at one end of the spectrum and contain more smooth muscle cells per unit volume than the surrounding normal myometrium and little or no ECM. Next along the spectrum are young, undegenerated seedling fibroids, which are only slightly more cellular than adjacent myometrium. Finally, as degeneration occurs, the concentration of living cells within a fibroid decreases until it is difficult in some fibroids to identify cellular elements at all. Because fibroids develop at different times from separate cells, in a single uterus one fibroid could be composed of all living fibroid tissue, another could be composed of 50% living fibroid tissue and 50% dead fibroid tissue, and a third fibroid could be composed of all dead fibroid tissue. Any therapy that is directed at living fibroid tissue will have little effect on fibroids composed of primarily dead fibroid tissue. Conversely, a fibroid composed primarily of living fibroid tissue can potentially respond dramatically to a therapy that affects living tissues. Therapies aimed at living fibroid tissue would be predicted to have measurably different effects in each of these 3 hypothetical fibroids.
Fibroids Vary Greatly in Size
As we have seen, the normal, non-fibroid uterus is a small organ that varies mildly in size from woman to woman by age and parity. In contrast, variation in size of the fibroid uterus is enormous. Fibroids vary in size from microscopic (1–2 mm in diameter) to quite large (10–20 cm in diameter), at times reaching “the size of a man’s head.” In similar everyday terms, Borell and Fernstrom observed that fibroids “varied between the size of a hazel-nut and that of a croquet-ball.” The size of a fibroid uterus is a function of the numbers of fibroids present in that uterus, their respective sizes, and the increase in myometrial mass that occurs in the fibroid uterus.
Because myometrium grows when fibroids are present and because fibroids are often multiple, overall volume of the fibroid uterus is a good measure of fibroid burden in a given patient. In some fibroids, size may be genetically predetermined. Rein studied 114 fibroid karyotypes from myomectomy and hysterectomy specimens from 92 women. The average fibroid diameter for karyotypes that were normal was 5.9 cm; for nonmosaic abnormal karyotypes, the average was 10.2 cm, a difference significant at p≤0.001. Other variables that directly influence fibroid uterine weight, in descending order of influence, are menopausal state (fibroid uterine weight in postmenopausal women is lower than in premenopausal), race (fibroid uterine weight in white women is lower than in black women), and body mass (fibroid uterine weight in thin women is lower than in obese women).
Fibroids Are Generally Hypovascular Compared with Myometrium
Despite overwhelming evidence to the contrary, some still believe that “leiomoymas are well-vascularized tumors.” They are not. In general, in comparison to adjacent myometrial tissue, fibroids are less vascular. Fibroid vascularity relative to myometrium can be characterized by gross and microscopic tissue examination, radioactive isotope tracer studies, and contrast-medium– enhanced MRI examinations.
a. Anatomy and Histology Studies
Fibroids receive their blood supply from the intrinsic arteries of the uterus, primarily from branches of the arcuate arteries. Sampson used a radiopaque gelatin vascular injection technique and observed that “the degree of arterial vascularity of the tumors themselves varies greatly in individual cases. … Small tumors are usually (but not always) less vascular (arterial) than the neighboring myometrium, while larger tumors are usually more vascular.” Sampson got a lot right in his research, but here he was wrong. Faulkner corrected Sampson’s misconception; Larger fibroids have no more vessels than small fibroids when both are adjusted for tissue volume.
Holmgren also studied uterine vessels with injected radiopaque gelatin and concluded that some fibroids are “relatively poor in vessels” when compared to adjacent normal myometrium and that others are “frequently richer in vessels than the surrounding tissue.” Holmgren also pointed out that fibroids have “a double arterial supply, both through a number of smaller arteries passing direct from the capsule into the myoma and supplying its peripheral portions, and through one or two large vessels, penetrating deeper into the myoma and supplying its central portions.” In this paper,Holmgren demonstrated that the corkscrew shape (or WWWW shape) of the uterine arteries is an inherent trait that is present even before birth and not one that develops only after pregnancy.
More recent studies of the anatomy of fibroid blood supply have demonstrated that the arterial network of a fibroid is usually less dense than the surrounding myometrium. Farrer-Brown et al. have shown that even though the arteries that supply a fibroid may be quite large (larger than the arteries supplying the adjacent myometrium), the relative vascularity (the number of arteries per volume of tissue) is generally greater for normal myometrium than for a fibroid. They have observed that “the arterial density is less than the surrounding myometrium, an observation confirmed by histological section.”
Walocha et al. obtained uteri from 22 women age 22 to 71 at autopsy who had died from causes not related to the uterus. They injected a synthetic resin into the uterine arteries, polymerized the resin, and then dissolved all tissue within each specimen. The resultant vascular cast was then examined by scanning electron microscope to the level of capillaries. They observed that 1–3 mm diameter fibroids were “avascular.” Fibroids 5–10 mm in diameter had “a few small vessels” invading from their periphery. Fibroids greater than 10 mm in diameter contained irregular networks of blood vessels at their periphery with vascular density similar to or lower than that of the adjacent myometrium. The authors concluded that as fibroids grow, preexisting blood vessels undergo regression, and new blood vessels invade the fibroid from the periphery, forming a “vascular capsule.”
Immunohistochemistry techniques have demonstrated that vascular area, microvascular density, and vascular luminal diameter are all greater in myometrium than in fibroids. Pollard et al. compared microvascular density in paired fibroid and myometrial samples from 11 women. Vessel density in fibroids was approximately half that in adjacent myometrium, a difference significant at the p<0.0001 level.
b. Perfusion Studies
Radioactive xenon distribution demonstrates that blood flow is lower in fibroids than in myometrium.
The relative vascularity of fibroids compared to myometrium has been studied with contrast-medium–enhanced MR imaging. In a study of 46 patients who had 115 fibroids, Hricak et al. characterized the vascularity of fibroids. Images were enhanced with gadolinium injected as an intravenous bolus immediately before the MRI sequence was acquired. Dr. Hricak observed that “on contrast-enhanced T1-weighted images, 65% of fibroids had a signal intensity lower than that of adjacent myometrium. … 23% had a signal intensity similar to that of adjacent myometrium … and 12% had a signal intensity predominately higher than that of adjacent myometrium.” In approximately 88% of fibroids, the fibroid enhanced less than or equal to the enhancement of adjacent myometrium.
Burn et al. studied gadolinium- enhanced MRI images from 32 fibroids in 18 women prior to and following uterine artery embolization. They reported the MRI enhancement for 2 groups: 4 out of 18 women (22.2%) had fibroids whose enhancement was equal to or greater than the enhancement of the adjacent myometrium (iso- and hypervascular fibroids, respectively); 13 (77.8%) women had fibroids that were hypovascular. A third MR study also showed that most fibroids enhanced less than myometrium. Yamashita et al. demonstrated that cellular fi broids, which are high in living fibroid tissue, “had marked contrast enhancement in the early phase, while degenerated fibroids [low in living fibroid tissue] showed slight or irregular enhancement.” In the extreme, 9% of fibroids examined with contrast-medium–enhanced MRI prior to uterine artery occlusion were avascular. Myometrium has never been described as avascular anywhere in the world’s MR literature.
c. Gene Array Studies
Weston et al. extracted total mRNA from fibroid and adjacent myometrium tissue from 12 hysterectomy specimens and screened 10,500 genes for differential expression. Two angiogenesis promoter genes were reduced in expression in fi broids compared to the adjacent myometrium. An angiogenesis inhibitor was increased in fibroids relative to myometrium. The authors postulate that these genes help to explain the reduced microvascular density seen in fibroids relative to myometrium. Pollard et al. compared fibroid tissue from women with hereditary leiomyomatosis and renal cell cancer syndrome with run of- the-mill, “sporadic” fibroids. Sporadic fibroids were less vascular than adjacent myometrium.
Fibroids Stimulate Myometrial Growth
As we have seen, when the fetus and placenta grow within the lumen of uterus, the musculature of the uterus grows dramatically as well. Similarly, when fibroids grow within the myometrium of a uterus, myometrium increases in mass. That is, as the fibroid burden in a uterus increases, myometrial mass increases too. Stated another way, the myometrial mass of a uterus supporting fibroids is larger than the myometrial mass of a normal uterus.
How much does myometrium grow when it supports fibroids?
A quantitative estimate of myometrial growth in the fibroid uterus can be made by following uterine volume change after fibroids have been removed by myomectomy. The amount of decrease in myometrial mass following myomectomy is assumed to roughly equal the increase in myometrial mass that occurred as fibroids grew inside the uterus. Two investigations have been published. In a longitudinal study of uterine volume following myomectomy, Tsuji et al. evaluated MR images before and after myomectomy and showed a 37.7% decrease in uterine volume between the 6th postoperative week to the end of the study at 52 weeks. Because the first postoperative look at the uterus was 6 weeks following myomectomy, rapid uterine volume change following myomectomy was missed. In an US study that examined the postmyomectomy uterus beginning at week 1 following myomectomy, Beyth et al. measured uterine volume serially. Their data demonstrated a decelerating drop in myometrial volume following myomectomy from the first postoperative week to the end of the study at 24 weeks. Myomectomy caused an immediate 21.8% drop in average uterine volume from 464 cc to 363 cc. Thereafter, average uterine volume dropped from 363 cc to 146 cc, a 46.8% decrease from the starting volume. Myometrium shrinkage after childbirth follows a similar time course. In the Beyth series, the removal of fibroids themselves by myomectomy caused less overall decrease in uterine volume than subsequent myometrial atrophy. Seen in reverse, myometrial growth contributed more to uterine volume than fibroids contributed.
Because the MR signal intensity of small fibroids is essentially identical to the MR signal intensity of adjacent normal myometrium, small fibroids may not be identified on an MR image. Failure to identify small fibroids does not mean that no small fibroid is present. Small fibroids can be overlooked on MR studies. Not withstanding this limitation, investigators have attempted to estimate how much myometrium increases in volume in a fibroid uterus by measuring overall uterine volume and subtracting fibroid volume. In a hormone study, Carr and et al. calculated “non-myoma volume,” which can be equated to myometrial volume. Their average pretreatment myometrium volume was 808 cc, nearly twice their average fibroid volume of 467 cc. Similarly, in a 6-month hormone study, Schlaff et al. followed fibroid and “non-myoma” or myometrial volume in treated and untreated women with fibroids. In the untreated group, average myometrial volume started at 199 cc and increased to 239 cc over the 6 months of observation, a 20.1% increase.
Endometrial surface area and uterine volume are correlated. Sehgal and Haskins demonstrated that the endometrial surface area of the fibroid uterus is markedly enlarged compared to normal. They observed surface areas as large as 225 cm2 in women with fibroids compared to 15 cm2 in their controls.
Whether one follows the decrease in myometrial volume following myomectomy or the increase in myometrial volume in women with fibroids, it is clear that like the pregnant uterus, normal myometrium increases in mass when a uterus supports fibroids. The mechanism by which myometrium increases in mass is suggested by an immunohistochemistry study by Hague et al. These authors measured vascular density and the activity of 5 angiogenic growth factors in hysterectomy specimens from 91 women. Fifty-two of the specimens contained fibroids; the remainder were used as controls. Angiogenic growth factors were upregulated and vascular density increased in myometrium from the uteri supporting fibroids compared to uteri without fibroids. Fibroids appear to induce increased vascularity in the myometrium of the fibroid uterus. Consistent with this view is the finding that vascular resistance, as measured by velocitometry in the uterine arteries of women with fibroids, is lower than in women without fibroids and that vascular resistance increases as fibroid size increases.