Synarel

* Objective.-Reports on the histologic effects of gonadotropin-releasing hormone agonists on uterine leiomyomas provide conflicting results. Most previous studies used qualitative methods of analysis. Using quantitative and semiquantitative stereologic methods of analysis, we assessed volume density of hyalinized areas, cell density, nuclear volume, and cytoplasmic cross-sectional areas of smooth muscle cells in histologic sections and also measured diameters of collagen fibrils in electron micrographs of uterine leiomyomas.

Design.-Thirty leiomyomas from patients treated with gonadotropin-releasing hormone agonists (10 different patient samples), age-matched control patients (10 different patient samples), and postmenopausal women (10 different patient samples) were used. Hyalinization was assessed using a microscope with a projection head and affixed morphometric grid. Cell size and density were evaluated using a video-based, computerized system attached to the microscope, for which morphometric ad hoc programs were

written. Diameters of collagen fibrils were measured from electron micrographs.

Setting-The study was conducted in the Department of Pathology, Mount Sinai Medical Center, New York, NY. Patients.-A total of 30 patient samples were studied, with three groups comprising 10 samples each, including patients treated with gonadotropin-releasing hormone agonists, age-matched control patients, and postmenopausal women.

Results.-Myomas from patients treated with gonadotropin-releasing hormone agonists exhibited more hyalinization, greater cell density, slightly smaller cell sizes, and larger collagen fibrils than those of age-matched control patients and postmenopausal women.

Conclusions.-Shrinkage after treatment with gonadotropin-releasing hormone agonists is attributed to smaller cell size and increased collagenization in myomas. (Arch Pathol Lab Med. 1998;122:442-446)

The use of gonadotropin-releasing hormone agonists (GnRHa) has become increasingly popular over the past several years. These agents are being used in a variety of clinical settings, including treatment of precocious puberty, endometriosis, prostatic cancer, infertility, and in reduction of volume and symptoms related to uterine leiomyomas.l2 Gonadotropin-releasing hormone agonists induce, over the long term, pituitary desensitization through downregulation of pituitary follicle-stimulating hormone and luteinizing hormone receptors, leading to menopausal levels of estrogen and progesterone.l The extent to which GnRHa reduce myoma volume initially increases with the duration of therapy. This reduction in shrinkage varies from 0% to 50% during a 2- to 6-month course of treatment.3 The effect is reversible, because myomas return to their initial size within 4 months after cessation of treatment.l

Studies on the histologic effects of GnRHa on myomas have found conflicting results (Table 1).410 Three of the seven studies cited found an increase in hyalinization in myomas from GnRHa-treated patients. Three of the seven found no change in hyalinization, and one suggested that ischemic injury and cell atrophy might account for the shrinkage. One possible reason for these discrepancies is that studies on the histology of myomas are difficult because of sampling problems. Leiomyomas tend to exhibit enormous heterogeneity in distribution of cells and hyaline. The histologic section examined usually represents a small piece of a large tumor, and, given the nonuniformity of the distribution of cells and acellular areas, a given histologic section may not be representative of the tumor.

The purpose of our study was to further investigate the histologic effects of GnRHa on myomas. We used quantitative stereologic methods of analysis to assess cell density, size of nuclei and cytoplasm, and amount of hyalinization.

MATERIALS AND METHODS

Uterine leiomyomas from 30 women were studied. These were divided into three groups, each composed of 10 patient samples. Group 1 was made up of myomas from patients treated with GnRHa, aged 31 to 47 years, who had received either Synarel, BID, 6 to 8 weeks, or Lupron, 3.75 mg IM, 6 to 8 weeks. In both groups, therapy was started before the time of expected menses. The 10 patients (of 40, see next sentence) selected from the treated group experienced, on average, a 15% to 20% shrinkage of their myomas, as assessed by physical examination in conjunction with ultrasonographic evaluation. These patients were not concurrently receiving other hormonal therapy. Myomas from 40 patients treated with Lupron were screened for possible inclusion in our study. Selection was based on the following criteria: routine histologic appearance of the myoma, high quality of tissue preservation, presence of one slide tissue section per centimeter of tumor, and absence of necrosis. Group 2 consisted of myomas from age-matched control patients who did not receive GnRHa, and group 3 comprised myomas from postmenopausal women, aged 57 to 82 years, who did not receive hormone therapy.

The leiomyomas had been processed according to routine procedures in our histology laboratory. Random sections had been cut from myomas that appeared normal on gross examination for histologic evaluation. Tissue sections were fixed in 10% buffered formalin and subsequently embedded in paraffin wax. Four-micron-thick sections were cut and stained with hematoxylin and eosin. Of these, one section of the several on the histologic slide was chosen for morphometric analysis. Only myomas with the usual histologic appearance, in which all routine sections (from the given myoma) appeared similar on low-power histologic examination, were selected for study. Myomas with extensive areas of necrosis or with nonuniform cellularity on lowpower inspection were omitted. One random section was chosen for study, for each case, from the several available sections of the chosen myomas. For the one random section chosen, the entire tissue section (every field) was assessed by light microscopy. For the electron microscopic studies, several small pieces were removed from the paraffin block of the section chosen for study by light microscopy. Thick sections were initially prepared and, from these, the one showing the best preservation was processed for electron microscopy.

Morphometric Studies

Equipment.-General purpose equipment included a videobased computerized system consisting of the following: a Gateway 2000 486DX2/50E computer augmented with a Sprynt i860 board (image processor, display, and frame grabber) from Synoptics Ltd, Cambridge, UK; a Mitsubishi color monitor outfitted with a touch-sensitive screen (although in this study only the mouse was used as pointing device); and a Nikon microscope with a Sony color camera. The system runs on the XNIX environment and the general graphics program Semper 6 Plus (Synoptics) and has been repeatedly described."'2 Morphometric ad hoc programs were written in C+ + or Fortran.

Hyalinization.-A conventional morphometric light microscopic setup was selected. A ZEISS microscope was outfitted with a projection head, and a photographic coherent test grid of 100 points" was attached to its screen. The total magnification on the screen was x10 the magnification of the objective lens (in this case, also x10), giving a total magnification of x100. Slide sections of the selected areas of the myoma specimens were stained with hematoxylin and eosin and viewed on the projection screen of the ZEISS microscope. Each cross point of the grid was scored as either containing cells (presence of nuclei) or containing hyaline (absence of nuclei). Results were tallied as the number of acellular grid points divided by the total number of grid points counted. The volume density (Vv), which is identical to the percentage of volume occupied by the foci of hyalinization, equals, according to the Delesse theorem, the areal density (ie, the number of points overlying the areas of hyalinization) divided by the total number of points (100) of the grid. Approximately 500 grid points per sample were counted.

Cell Density.-Cell density was defined as the density of profiles of spindle cell nuclei. The histologic image via a X10 lens of a Nikon microscope was projected onto the video screen of the image analysis system. A double-square lattice system grid, with a calibrated total area dimension of 51900 (mu)m2, appeared next over the image. All nuclear profiles lying within the grid were counted, as were nuclei crossing the top and left boundaries defining the grid. Nuclei crossing the bottom or right boundaries of the grid were omitted. Results were tallied as the total number of nuclear profiles per unit of areal density enclosed by the grid. Evaluation of the cellular (nonhyalinized) areas in their entirety amounted to approximately 10 low-power fields per tissue sample.

Cell Size.-The method for determination of the volumeweighted, mean particle volume of the nuclei was based on the new stereologic procedure developed by Cruz-Orive, as described by Gil and Barba.'3 In brief, the cells, under an oil-immersion x 100 objective of the Nikon Optiphot microscope equipped with a Sony model DXC-M2 video camera, are viewed on the video monitor while the computer generates a set of parallel lines at a randomly chosen angle to the edge. The user clicks the mouse indicator on the points of entry and exit, on the parallel lines, by each of the nuclear profiles of the cells of interest in the field (Fig 1). Using the entry and exit points for each nucleus, the computer generates, for each nucleus, a chord length (lo). The final computation is done automatically by the computer. The number-weighted, mean particle volume of nuclei was estimated from the two-dimensional nuclear profiles, according to the method of Cruz-Orive,'3 as follows: v = (rr/3)(lo3). This is an estimator of the true three-dimensional average nuclear size."

Cross-Sectional, Equatorial Cytoplasmic Areas.-Unstained tissue sections (blanks) of the same tissue samples used in the previous studies were prepared and stained with trichrome. Using this stain, the cytoplasm of the smooth muscle cell is red, and the surrounding matrix (hyaline) appears blue. Cytoplasms of individual cells are clearly defined. Areas on the slide section selected for study were those showing only cross-sectional cells in the equatorial position. Cell cytoplasmic cross-sectional areas were calculated using the computerized interactive tracing methodology as follows: the observer, using the mouse, clicks three sequential points on the cytoplasmic outline, thereby delimiting a circumferential arc to be generated by the computer. The procedure is repeated until the segmentation of each profile is complete. From here, the enclosed area is computed following published procedures.'3 This is not a stereologic estimator of nuclear or cell size, but it lends itself well to comparisons.

Collagen Fibril Diameters.-The diameters of collagen fibrils were assessed by measuring the diameters of individual fibrils using a clear plastic ruler and magnifying glass from electron micrographs taken at x20000 magnification. The collagen molecule is a triple-stranded helical structure made up of three polypeptide chains of fibrillar collagens. After the molecule is secreted into the extracellular space, the collagen molecules assemble into ordered polymers called fibrils, which tend to be 10 to 300 nm in diameter. Fibrils, which can be visualized using the electron microscope, often aggregate into larger bundles called fibers, which are visible with a light microscope, and tend to be several microns in diameter.1s In our study, 100 fibrils were counted for each patient sample. Samples had been removed from the paraffin blocks, deparaffinized, and processed according to routine procedures in our electron microscopy laboratory. RESULTS

Table 2 shows the results of our studies. Hyalinization volume density (ie, the percentage of volume occupied by hyalinized areas) was greatest in the myomas from patients treated with GnRHa, followed by myomas from postmenopausal women. Values among the three groups differed significantly, as assessed by analysis of variance (F = 0.95).

Cell density was highest in myomas from patients treated with GnRHa compared with age-matched control patients and postmenopausal women. Values from the GnRHa-treated group differed significantly from those of the other two groups (analysis of variance, F = 0.95). Values from the age-matched control patients and postmenopausal women did not differ significantly from each other. Mean nuclear volumes, as estimated from nuclear profiles, were not significantly different (analysis of variance, F = 0.95). The smallest values were found in the GnRHatreated group.

Mean cross-sectional, equatorial cytoplasmic area did not differ significantly among the three groups (analysis of variance, F = 0.95).

Mean collagen fibril diameter was significantly greater in myomas from patients treated with GnRHa than in agematched control patients (analysis of variance F = 0.95). Figure 2 shows the distribution of collagen fibril diameters in the patients treated with GnRHa and age-matched control patients. Although most fibrils are around 50 run in diameter for both groups, more smaller fibrils (25 and 37.5 nm) were observed in the control group, and more larger fibrils (62.5 and 75 nm) were observed in the GnRHatreated group. Fibrils were not measured in specimens from postmenopausal women. COMMENT

Our results suggest that myomas from women treated with GnRHa have more hyalinization and are more densely cellular in nonhyalinized areas than myomas from agematched control patients and postmenopausal women. Additionally, myomas from GnRHa-treated patients show a trend toward cells with smaller nuclei and cytoplasm. Myomas from GnRHa-treated patients seem to have, in hyalinized areas, collagen fibrils with larger diameters than control patients. The collagen fibril diameters observed correspond to type I collagen.

Shaws suggested that treatment with GnRHa resulted in shrinkage of the cytoplasm in cells making up myomas and that this might account for the reduction in size. Rein et al9 suggested that the number of cells in myomas was unchanged, based on DNA and protein quantitations, and that the amount of hyaline protein was increased. The combined findings suggest that if the number of cells remains the same after GnRHa therapy and their cytoplasm is decreased, then the cells should be more crowded, or densely packed, in nonhyalinized (cellular) areas. Our results, which showed increased cellular density and decreased cytoplasmic area in myomas from patients treated with GnRHa, support these contentions. The appearance of more hyalinization in myomas from GnRHa-treated patients is consistent with the findings of several others46,9 and may be a result of: (1) the production of more hyaline; (2) a change in the structure of hyaline (state of polymerization or depolymerization of collagen fibrils), or (3) as suggested by Rein et al,9 a change in the state of hydration of collagen fibers. Results of our electron microscopic studies, which showed increased collagen fibril diameters in myomas of patients treated with GnRHa, suggest either increased hydration and/ or increased polymerization or decreased depolymerization of the collagen fibrils. The findings reported by Rein et all of a net increase in collagen protein after GnRHa therapy suggest that fibril polymerization, rather than hydration, may be operative. The overall reduction in size of the myomas could be explained by postulating that the hyaline occupies less space than do cells (cytoplasm). Puistola et alle reported antagonistic effects by GnRHa on ovarian estrogen-stimulated collagenase matrix metalloproteinase, which degrades type IV collagen. Our results, which suggest increased polymerization of collagen fibrils after GnRHa therapy, raise the question of whether collagenase activity in myomas might also be antagonized by GnRHa therapy.

Other postulated factors involved in myoma shrinkage, as reported in Table 1, include cell dropout (necrosis) and atrophy. Colgan et al5 reported more necrosis in myomas of patients treated with GnRHa, whereas Shaws found no change in amount of necrosis. In our preliminary survey, it did not appear that myomas of patients treated with GnRHa differed significantly from those of control patients in amounts of necrosis. In fact, the myoma with the most extensive necrosis came from an age-matched control patient. Cell atrophy has been suggested to be a factor in shrinkage.5 We found a trend toward smaller nuclear volume and less cytoplasmic area in myomas of patients treated with GnRHa compared to control patients. Thus, cell atrophy may play a role in size reduction of myomas after GnRHa therapy.

Our results suggest that the physiologic shrinkage observed in myomas with aging (postmenopausal group) might, like the shrinkage after GnRHa therapy, be attributed to an increase in hyalinization. This study showed that for the postmenopausal group, no significant difference in all parameters examined except hyalinization was noted. We would infer from this that increased hyalinization alone or other, additional mechanisms may be operative in myoma reduction in postmenopausal women as compared to women treated with GnRHa.

References

1. Plosker GL, Brogden RN. Leuprorelin: a review of its pharmacology and therapeutic use in prostatic cancer, endometriosis and other sex hormone-related disorders. Drugs. 1994;48:930-967.

2. Charbonnel B. Analogs of gonadoliberin: value in physiology and therapeutics. Presse Med. 1994;23:805-811.

3. Hackenberg R, Gesenhues T, Deichert U, Duda V, Schmidt-Rhode P, Schulz KD. The response of uterine fibroids to GnRH-agonist treatment can be predicted in most cases after one month. Eur J Obstet Gynecol Reprod Biol. 1992;45:125129.

4. Cohen D, Mazur MT, jozefczyk MA, Badawy SZ. Hyalinization and cellular changes in uterine leiomyomata after gonadotropin releasing hormone agonist therapy. J Reprod Med. 1994;39:377-380.

5. Colgan TJ, Pendergast S, LeBlanc M. The histopathology of uterine leiomyomas following treatment with gonadotropin-releasing hormone analogues. Hum Pathol. 1993;24:1073-1077.

6. Uemura T, Mori J, Yoshimura Y, Minaguchi H. Treatment effects of GnRH agonist on the binding of estrogen and progesterone, and the histological findings of uterine leiomyomas. Asia Oceania J Obstet Gynaecol. 1991 ;17:315-320.

7. Upadhyahya NB, Doody MC, Googe PB. Histopathological changes in leiomyomata treated with leuprolide acetate. Fertil Steril. 1990;54:811-814. 8. Shaw RW. Mechanism of LHRH analogue action in uterine fibroids. Horm Res. 1989;32:150-153.

9. Rein MS, Barbieri RL, Welch W, Gleason RE, Caulfield JP, Friedman Al. The concentrations of collagen-associated amino acids are higher in GnRH agonisttreated uterine myomas. Obstet Gynecol. 1993;82:901-905.

10. Sreenan JJ, Prayson RA, Biscotti CV, Thornton MH, Easley KA, Hart WR. Histopathologic findings in 107 uterine leiomyomas treated with leuprolide acetate compared with 126 controls. Am J Surg Pathol. 1996;20:427-432.

11. Gil J, Marchevsky AM, Silage DA. Applications of computerized interactive morphometry in pathology, I: tracing and generation of graphic standards. Lab Invest.1986;54:222-227.

12. Marchevsky AM, Gil). Applications of computerized interactive morphometry in pathology, II: a model for computer generated diagnosis. Lab Invest. 1986; 54:708-716.

13. Gil J, Barba J. Principles of stereology: computerized applications to anatomic pathology. In: Marchevsky AM, Bartels PH, eds. Image Analysis: A Primer for Pathologists. New York, NY: Raven Press; 1994:102-104.

14. Einstein A, Gil J, Wallenstein S, et al. Reproducibility and accuracy of interactive representation of procedures for image analysis in cytology. Microscopy. 1997;188:136-148.

15. Alberts B, Bray D, Lewis), Raff M, Roberts K, Watson J. Molecular Biology of the Cell. New York, NY: Garland Publishing Inc; 1989:809.

16. Puistola U, Westerlund A, Kauppila A, Turpeenniemi-Hujanen T. Regulation of 72-kd type IV collagenase-matrix metalloproteinase-2 by estradiol and gonadotropin-releasing hormone agonist in human granulosa-lutein cells. Fertil Steril.1995;64:81-87.

Accepted for publication January 15, 1998. From the Departments of Pathology (Drs Kalir, Gordon, Deligdisch Wu, and Gil) and Obstetrics, Gynecology, and Reproductive Science (Drs Kalir, Goldstein, Dottino, Brodman, and Deligdisch), Mount Sinai Medical Center, New York, NY.

Reprint requests to Box 1194, Department of Pathology, Mount Sinai Medical Center, One Gustave L. Levy Place, New York, NY 10029 (Dr Kalir).

Copyright College of American Pathologists May 1998
Provided by ProQuest Information and Learning Company. All rights Reserved