J Korean Radiol Soc 199?; 36: 1 08?-1 092

Mammographic Changes in Postmenopausal W omen : Comparative Effects between Continuous Combined Hormone and

Single Estrogen Replacement Therapy 1

Sug Oh, M.D. , Jong-Tae Choi, M.D., Kyoon Soon Jung, M.D.2, Seung Hye JungM.D. 2

Purpose : As the use of hormone replacement therapy for the menopausal women increases, some caution is advised, since there is an increased risk ofbreast cancer. Ac­cordingly, the importance of regular mammography has been addressed. This cross-sectional study analyzed the effects of different hormone therapies on mammographic density.

Materials and Methods: Sixty-seven postemenopausal women who had completed one year ofhormone therapy and had undergone follow-up mammography, were divided into two groups: Group 1 : continuous conjugated equine estrogen, O. 625mg, plus continuous medroxyprogesterone acetate, 2.5mg (n=48), Group II : con­tinuous conjugated equine estrogen 0.625 mg (n = 19). The mammograms were read by two radiologists.

Res비ts : With regard to the radiologists involved , interobserver reliability (kappa) was 0.70 and intrao bserver reliability (kappa) was 0.51 and 0.67 . Before hormone ther­apy, factors related to decreased mammographic density were age and number of full term pregnancies (p < 0.05) . After one year ofhormone therapy, body fat showed a sig­nificant increase (p < 0.05), but in spite of this, increased mammographic density induced by hormone therapy remained significantly high (p < 0.05). Compared with Group II , Group 1 showed a significant increase in mammographic density (p < 0 .05). In Group 1 , mammographic density increased from P2 to DY pattern in two cases, but there was no such change in Group II.

Conclusion : The increase of mammographic density seen in Group II was much more significant statistically than that seen in Group 1. The mammograms ofwomen who have undergone continuous combined hormone therapy should therefore be interpreted very cautiously.

Index Words : Breast, parenchymal pattern Breast radiography

Increasing numbers of postmenopausal women are undergoing hormone replacement therapy (HRT) for a variety ofreasons : to relieve menopausal symptoms, to prevent menopause- induced osteoporosis , and for the beneficial effects on the prevention and prognosis of cardiovascular disease (1, 2). Even though the risk of breast cancer is a matter of controversy, some prospec-

IDepartmentofFamily Medicine, Jeil Women’s Hospita l, Seoul, Korea ' Department of Radiology, Jeil Women’s Hospita l, Seoul, Korea Received December 23, 1996; Accepted April3, 1997

Address reprint requests to : Sug Oh, M.D., Department of Family Medicine, Yongdong Jeil Women ’ s Hospital, # 1021-4, Daechi-dong, Kangnam-Ku, Seoul,

135-280 Korea. Tel. 82-2-561-6 100 Fax.82-2-552-7964

tive studies have indicated a gradually-increasing risk of this after long term estrogen use (3, 4); in a signifi­cant percentage ofwomen undergoing HRT, increased breast density on mammography has also been reported (5 , 6). There are many factors affecting breast density; we analyzed the differing effects of different HRT regimens in postmenopausal women.

Subjects and Methods

Subjects Among 1934 women who visited Yongdong Jeil

1087 -

Sug Oh. et' a/: Mammographic Changes in Postmenopausal Women

Women ’ s hospital for their first general examination, 67 postmenopausal women who had fulfilled the fol­lowing criteria were selected : 1) last menstrual period at least one year previously; 2) FSH) 4O m1U and last menstrual period at least 6 months previously; 3) no previous history of hormone replacement therapy ; 4) successful follow-up of second mammography one year after baseline mammography. These 67 women were treated for 12 months with one of two regimens : Group 1 : 48 women who received 0.625mg/day con­jugated equine estrogen (CEE) (Premarin, Wyeth-Ayer­st Laboratories, Radnor, PA) plus 2.5mg/day medrox­yprogesterone acetate (Provera, Upjohn Co. , Kalamaz-00, MI) for 30 consecutive days, Group II : 19 women who received CEE, 0.625mg/day , for 30 consecutive days.

Mammography

Baseline craniocaudal and mediolateral oblique mammograms were obtained with a Senographe 600T (G.E. , Milwaukee, U.S. A.), using the maximum tolerated compression. All subjects underwent a follow-up mammogram, using the same equipment, one year after their first visit. All mammograms were obtained by the same radiologic technologist, who maintained strict quality control, and were evaluated by two board certified radiologists . Mammographic density was assessed according to the Wolfe classification (7 , 8) : Nl , breasts composed primarily of fat; Pl , prominent ducts in the subareolar area involving approximately one-third of the breast; P2, prominent ducts involving the major portion of the breast; DY, a considerable amount of collagen or dysplasia, with or without identified ducts .

Data Analysis

An estimation of the reliability ofthe two radiologis­ts used kappa statistics (9) to compare interobserver consistency, and 3-month interval readings ofthe same 40 films to determine intraobserver reproducibility. To assess the factors that have been shown to influence the appearance of the breast on mammograms before hormone replacement treatment (10 - 12), simple cor­relation and multiple regression analysis were perfor med. These factors include interval since last men strual period, age at menarche, number of children, weight, body fat, and lactation. Comparisons of those factors before and after hormone therapy were perfor­med using student’s t-test for paired quantitative data. After adjustment for those factors which significantly changed after treatment, differences in mammographic changes between the two regimens were compared by

ANOV A. We used the Statistical Package for Social Sciences for Windows TM(SPSS Inc, Chicago, 1L, U.S.A.) for calculation; the minimum level of statistical significance for all analyses was p (0.05.

Results

Reliability of Mammographic Readings between two Radio/ogists

1nterobserver reliability (kappa) was 0.70 while intraobserver reliability (kappa) was 0.51 and 0.67

General Characteristics of Study Population (Tab/e

1)

Mean age of the study population was 58 :t 7.8 years, and median ages of menarche and menopause were 16 and 49 years, respectively. As older women disliked bleeding during hormone therapy, we had, for them, preferred to prescribe a continuous combined regimen. Table 1 clearly shows that those women who underwent this regimen were older and had a much higher body fat component than those whose therapy involved only estrogen (p (0.05).

Factors influencing Mammographic Density before Hormone Replacement Therapy (HRT) (Tab/e2)

As women aged, they had less dense breast (p = O. 002) and the more children they had, the less dense were their breasts (p=O.OOI); tall women had denser breasts (p=O.027). Age of menarche and menopause, smoking history , body weight, body fat , body mass index (BM1), and lactation were not statistically significant factors influencing mammographic change

Table 1. Comparison of Demographic & Anthropometric Characteristics of Two Groups

Group I Group n n mean (SD) n mean (SD)

Age (years) 48 60.44 (7.1)* 19 52.00 (6.2) Pregnancy No. 47 3.72 (1.7)* 19 2.정 (0.8) Menarche (years) 46 16.15 (1.9) 19 15.89 (1. 8) Menopause (years) 48 48.77 (4.3) 19 47.50 (4.9) Weight (kg) 46 55.80 (7 .9) 19 56.49 (5.7) Height (cm) 46 154.45 (5 .1) 19 155.93 (4.3) Body fat (%) 42 32.09 (4.7)* 18 29.42 (4.4) BMI (kg/cm2

) 46 23.41 (3 .2) 19 23.22 (2.0)

* p < 0.05 vs. Group II Group 1 : continuous conjugated equine estrogen

0.625mg plus continuous medroxyprogesterone acetate 2.5mg

Group II : continuous conjugated estrogen 0.625mg

- 1088 -

J Körean Radi이 S∞ 1997; 36: 1087-1092

When age was controlled by using multiple regression analysis, these factors lost their statistical significan­ces .

Mammographic Change after Hormone Replace­

ment Therapy (HR T)

After hormone therapy, increased mammographic density was noted in both groups; it was seen in 30 of 67 postmenopausal women (44.8 %) who had under­gone HRT. The mammographs of the 48 women who had undergone continuous combined HRT showed sig­nificantly increased dens띠1 in 27 cases (56 .3 %) ; this was seen in only 3 of 19 (15 .8 %) who underwent single estrogen replacement therapy (ERT) (p < 0.05) (Fig 1, 2) (Tabel 3). Because hormone therapy changes some anthropometric components, we analyzed factors in f1uencing mammographic change after HRT. Only hormone therapy itself had a significant impact on mammographic change (p = O.OOI) (Table 4). In women who had undergone continuous combined HRT, two cases showed the DY pattern, which changed from P2, while there was no such change in the single ERT group.

Discussion

Previously little kaown effects of HRT, such as

came to light, and this has increased demand upon this treatment. With increasing numbers ofwomen under­going HRT, physicians are very concerned about the possible risk of breast cancer and advocate regular mammography. Mammographic change associated with HRT has been reported in many studies (6,

13 - 15), including ours, in which 44.8 % (30/67) of postmenopausal women showed increased density, a figure much higher than that ofStomper et al. (14), who reported that 24 % (12/50) of postmenopausal women undergoing HRT showed mammographic change, or of Berkowitz et al. (15) whose data showed that change

Table 2. Correlation of Mammographic Density with Demographic & Anthropometric Variables before Hormone Replacement Therapy (HRT)

Significant (p < 0.05)

Age (yeras) r = 0.3756 p = 0.002 n=67

Height(cm) r = 0.2744 p = 0.027 n= 65

Pregnancy number r = -0.4205 p = 0.001 n=65

Not Significant

Menarche (years) Menopause (years) Smoking Weight(kg)

Body fat(%) BMI (kg /cm2

)

Lactation

dementia and cardiovascular protection have recently BMI = Body Mass Index

A B c D

Fig. 1. Significant mammographic change of 54 year old woman (A, 8) and 67 year old woman (C, D) before (A, C) and after (8, D) continuous combined hormone replacement therapy.

- 1089

Sug Oh. et a/: Mammographic Changes in Postmenopausal Women

decreased or has no effect on mitotic activity and pro­liferation of breast epithelial cells, continues to be very controversial. Progesterone acts in synergism with estrogen on the distal portion of the ducts , favoring differentiation into acini, and promotes the growth of lobuloalveolar structures. It can reduce estrogen induced proliferation and converts the proliferative effect of estrogen on the ductal cells into cellular differ entiation (16; 17). As ducta l, periductal connective, and lobular tissue proliferates in response to a combination of estrogen and progesterone, the greatest increase in mammographic breast density would be expected in women being treated with combination

had occurred in 17 % of women. Both these studies showed that treatment with both estrogen and pro­gesterone more often led to mammographic change than treatment with estrogen alone. This latter increases the mitogenic activity of epithelial cells, which leads to duct growth; connective tissue sur­rounding the ducts is also very sensitive to the action of estrogen. Whether progesterone causes increased or

Table 3. Mammographic Density before & after Hormone Re­placement Therapy (HRT) between Two Gro때S

*Group 1 (n=48) Group n (n=19)

Table 4. Correlation of Mammographic Changes with Anthropometric Changes after Hormone Replacement Therapy (HRT)

r=-0.3876 p=O.OOl n=67 NS

NS

NS

1ι 。。

。。1i

1i

。。n?

1i

M

R

R

N

m R

R

N

mη 깅 7

before HRT

5

애 깅 2

after HRT

CJ

때

G 뼈 빼

안 R

B

T

m

삐% 써 -빡

t

서매 oe-s

뺑 파싸 -m

nu

o h -=

v B

B -M

* p < 0.05 vs. Group n Group 1 : continuous conjugated equine estrogen

0.625mg plus continuous medroxyprogesterone acetate 2. 5mg

Group n : continuous conjugated estrogen 0.625mg BMI = Body Mass lndex

A 8 C D Fig. 2. No significant mammographic change of 50 year old woman (A, 8), and 55 year old woman (C , D) before (A , C) and after (8, D) singe estrogen replacement therapy.

- 1090 -

J Korean Radiol Soc 1997; 36: 1087-1092

therapy (5). There has, however, been no report

describing the likely effect of different drug combi­nation on mammographic change. Our study showed that the continuous combined regimen we used

resulted in significantly increased mammographic density after one year ’s treatment; we observed that

mammographic density after 50 years of age, or after menopause decreased, and breast parenchyma became less dense (7, 8, 12). Boyd et a1. (18) reported the re­

lationship between mammographic density and hist이ogic risk factor for breast cancer, and Wolfe (7, 8) and Saftla (19) reported an increased risk of breast can­

cer in women with mammographically dense breasts. According to Saftla’s retrospective review of cases detected during the fifth year of the Breast Cancer De­tection Demonstration 안이ects ' (19), high risk patterns were particularly significant in women with a positive family history of breast cancer ; in those with P2 or DY patterns, the likelihood ofbreast cancer was 7.37 times higher than in those with an N1 pattern. In women with no family history, the corresponding odds ratio was 2.42. In our study, two cases who had undergone

the continuous combined HRT showed a DY pattern after one year' s treatment, while there was no such change in the single ERT group. This implies that in the

continuous combined HRT group, we should be alert to the risk of increased risk of breast cancer. Most postmenopausal women who underwent continuous combined therapy complained of more frequent mastalgia (data not shown) than those whose regimen was different. Plu-Bureau et a1. (20) asserted that mastalgia is an easy and early marker of breast suscep­tibility to estrogen, and reported that it was associated with the occurrence of breast cancer. Beca use of breast tenderness, adequate breast compression could not be performed, resulting in an apparent increase in density and a decreased sensitivity ofmammograms. In women under 50, breast density had a greater in f1uence on mammographic sensitivity than was the case in women aged 50 or over (12); in women with primarily fatty breasts, sensitivity was higher. Irrespective of the cause of increased density, improved quality and increased sensitivity are very important for reliable mammographic interpretation. Our study showed that mammographic density increased significantly as a re­sult of continuous combined HRT; this treatment could thus lead to decreased m

know whether increased density would disappear

after a couple ofyears under the combined regimen. To be aware of further mammographic change, which

might occur after the first year of treatment, a longer follow-up period is therefore needed; longitudinal ob­servation to assess the relative risk of eventual breast

cancer with the use of the continuous combined regi­men is also required. In conclusion, this regimen, with estrogen and progesterone, led to a significant increase in mammographic density, compared to that seen after

single estrogen therapy. To determine whether this in­crease remains or disappears - in other words, to de­

termine its clinical significance - a longer follow-up period is required; after resolving such factors as mastalgia and mammographic sensitivity, further studies are therefore needed. Dense mammograms

should also be interpreted more carefully.

References

1. European Menopause Society. European consensus develop­ment conference on menopause. Human Reproduction 1996; 11 975-979

2. Harlap S. The benefits and risks of hormone replacement ther apy: an epidemiologic overview. Am J Obstet Gynecol 1992; 166 1986-1992

3. Bergkvist I, Adami H, Persson I, Hoover R, Scharer C. The risk of breast cancer after estrogen and estrogen-progestin replace ment. N Engl J Med 1989; 321 : 293-7

4. Colditz GA , Hankinson SE , Hunter DJ, et al. The use of estrogens and progestins and the risk of breast cancer in postmenopausal women. N Engl J Med 1995; 332: 1589-93

5. Cylark D, Wong CH. Mammographic changes in postmenopausal women undergoing hormonal replacemcnt therapy. AJR 1993; 161: 111 7-1183

6. Mcnicholas MMJ , Heneghan JP, Milner MH, et al. Pain and increased mammographic density in women receiving hormone replacement therapy. ARJ 1994; 163:311-3 15

7. Wolfe JN. Risk for breast cancer development determined by mammographic parenchymal pattern. Cancer 1976; 37: 2486-2492

8. Wolfe JN. Breast patterns as an index of risk for developing breast cancer. AJR 1976 ; 126 ‘ 1130-1139

9. Rosner B. Fundamentals of Biostatistics. 2nd ed. Boston ’

Duxbury Press, 1990; 424-429 10. Buchanan JB , Weisberg BF, Sandoz JP, Gray LA, Bland KI.

Selected prognostic variables for mammographic parenchymal variables. Cancer 1981; 47: 2135-2137

11. Leinster SJ , Whitehouse GH. The mammographic breast pattern and oral contraception. Br J Radiol 1986; 59: 237-239

12. Kerlikowske K. Grady D, et al. Effect of age, breast density,

and family history on the sensitivity of first screening mammography. JAMA 1996; 276: 33-38

13. Peck DR, Lowman RM. Estrogen and the postmenopausal breast: mammographic considerations. JAMA 1978; 240: 1733-1735

14. Stomper Pc. VonVoorhis BJ, Ravnikar VA. Meyer JE. Mammograhic changes associated with postmenopausal hor-

- 1091 -

Sug Oh. et a/: Mammographic Changes in PostmenopausalWomen

mone replacement therapy : a longitudinal study. Radiology 1990 ; gesterone on human breast epitheli꾀 cell cycJe in vivo. Fertil and

174: 487-490 Steri11995; 63: 785-791

15. Berkowitz JE, Gatewood OMB, Goldblum LE, Gayler BW. Hor- 18. Boyd NF, Jensen HM, Cooke G, Han HL. Relationship between

monal replacement therapy: mammographic manifestations. mammographic and hisologic risk factor for breast cancer. J Natl

Radiology 1990; 174: 199-201 Cancer Inst 1992 ; 84: 1170-9

16. Kuttenn F, Fournier S, Sitruk-W are R, Martin P, Mauvais-Jarvis P. 19. Saftla AF, Szklo M. Mammographic parenchymal patterns and

Progesterone deficiency in benign breast disease. In: Angeli A, breast cancer risk . Epidemiol Rev 1987; 9: 146-1174

Bradlow ML, Dogliotti L, eds. Endocrinology of cystic breast disease . 20. Plu-Bureau G, Thalabard JC, Sitruk-Ware R, Asselain B, New York: Raven, 1983; 231-252 Mauvais-Jarvis P. Cyclical mastalgia as a marker of breast cancer

17. Chang KJ, Lee TTY, Linares-Cruz G, Fournier S, Lignieres B. susceptibility: result of a case-control study among French

Influences of percutaneous administration of estradiol and pro- women . Br J Câncer 1992 ; 65 : 945-94

대한빙사선의학회지 1997; 36: 1087-1092

폐경 여성에서 지속적 병합요법과 에스트로젠 단독요법간의 유방음영 변화의 비교1

l 영동제일병원 가정의학과

2영동제일병원 방사선과

오 석 · 최종태 · 정균순2 • 정숭혜2

목 적 : 폐경 여성에서 심혈관계 질환， 골다공증의 예방 및 갱년기에 나타나는 여러 증상을 예 방하고 치료하

기 위해 호르몬 제제들을 사용하고 있다. 호르몬제의 사용에 따른 유방암의 위험도에 대한 논란이 많은바 호르

몬 사용시 유방촬영은 필수적인 검사가 되었다. 따라서 저자들은 지속적인 에스트로젠요볍과 지속적인 에스트

로젠과프로제스테론의 병합요법간의 유방실질음영의 변화를살펴보고이에 대한 임상적 의미를고찰하였다.

대상 및 방법 : 결합 에스트로젠 (CEE 0.625mg, Premarin)과 MPA 2.5mg을 연속적으로 투여한 48명과 CEE

0.625mg을 연속적으로 단독 투여한 19명， 총 67명을 대상으로 치료전 유방촬영 소견과 치료 l년후 소견을 비교

분석하였다.

결 과 : 유방촬영의 판독은 2명의 방사선과 전문의가 하였으며， interobserver reliabili ty는 0.70 (kappa) ,

intraobserver reliabili ty는 kappa값이 각각 0.51 ， 0.67이였다. 치료 시작전 유방음영의 감소소견과 관련된 요인

들은 임신 회수 · 연령의 증가였다 (p<0.05 ). 호르몬투여 1년후 투여전에 비해 체지방의 유의한증가를보였다

(p<0.05). 체지방의 증가를 조절한 후에도 호르몬 투여에 따른 유방음영의 증가된 소견은 변하지 않았다 (p<O.

05) . 에스트로젠 단독투여군에 비해 병합투여군에서는 유방음영의 유의한 증가를 보였다(p<0.05). 에스트로젠

단독투여군에서는 P2에서 DY로의 음영증가는 없었으나 병합투여군의 경우， 2명에서 DY로 증가된 소견을 보

였다.

결 론 : 호르몬 제제에 따른 유방음영의 증가는 지속적 병합투여군에서 통계학적으로 유의한 증가를 보였다.

따라서 지속적인 호르몬투여여성에서 유방촬영 판독시 이러한사항을고려해야 하겠으며 병합투여군에서 DY

로 음영증가를 보인 여성의 경우 유방암 위험도에 대해 재평가하고， 호르몬 요볍이 지속적으로 필요한 경우 철

저한유방검사와자가진찰이 필요하다고생각된다.

1092 -