Determinants of high-grade dysplasia among women with mild dyskaryosis on cervical smear

Determinants of High-Grade Dysplasia Among Women With Mild Dyskaryosis on Cervical Smear FABIO PARAZZINI, MD, MAXI0 SIDERI, MD, SIMONA FRANCESCO SCHETTINO, MD, LILIANE CHATENOUD, PIER GIORGIO CROSIGNANI, MD Objective: women cervical biopsy.

To identify the epidemiologic characteristics of who have mild dyskaryosis on cervical smear but intraepithelial neoplasia (GIN) grade II or III at

Methods: We analyzed information from 291 women (median age 33 years, range 17-69) observed for the first time with a single smear test showing mild dyskaryosis. All subjects underwent colposcopy, and histologic confirmation was obtained by biopsy. We compared the characteristics of women who had CIN I or no evidence of CIN with those of women with CIN II or III at biopsy. Results: Twenty-eight women (10%) had CIN I at biopsy, 46 (15%) CIN II, and 23 (8%) CIN III. The frequency of CIN II or III tended to decrease with increasing education; compared with women reporting 11 or fewer years of education, the multivariate odds ratios (OR) of CIN II or III lesions was 0.5 (95% confidence interval [CI] 0.3-0.9) in those reporting 11 years of education or more. Compared with nulliparas, the OR of CIN II or III was 1.8 (95% CI 1.1-3.5) for parous women. Furthermore, compared with neversmokers, the OR of CIN II or III was 2.3 (95% CI 1.0-5.4) for current smokers. Ex-smokers were at increased risk, too; the estimated multivariate OR was 3.8 (95% CI 1.9-7.6). Compared with women reporting one sexual partner, the multivariate ORs of CIN II or III were 1.4 and 2.3 for women reporting two to three or four or more sexual partners, respectively Cd trend = 6.65, P < .05). Conclusion: Our results show that smoking is a risk indicator of CIN II or III in women with a single smear showing mild dyskaryosis. Parous women, those of low social standing, and those reporting multiple sexual partners also are at increased risk of CIN II or III. (Obstet Gynecol

RESTELLI, MD, ScD, AND

The routine referral for colposcopy and biopsy of women with a single smear test showing mild dyskaryosis is open to debate. ‘G’ Most women with mild dyskaryosis revert to a negative smear without any treatment, although some with mild dyskaryosis have cervical intraepithelial neoplasia (GIN) grade II or III.4-6 Identification of clinical or epidemiologic characteristics of women with a smear test showing mild dyskaryosis but CIN II or III at biopsy may be useful for identifying women who should be referred immediately for colposcopy. The epidemiologic characteristics of women at high risk of CIN II or III among those with mild smear abnormalities may help clarify potential factors associated with the progression of cervical lesions toward severe disease and cervical cancer. Two studies7T8 suggested that the prevalence of CIN was higher among smokers with mild dyskaryosis than among nonsmokers. Less consistent evidence emerged for other factors, such as age and sexual habits. To provide further information on the issue, we report the results of a study conducted on 291 women with a single smear test showing mild dyskaryosis. Specifically, we analyzed the role of suggested risk indicators of CIN II or III in women with mild dyskaryosis at cervical smear in a southern European population with sexual, reproductive, and general lifestyle habits different from the northern European women considered in previous studies on this issue.

1995;86:754-7)

Materials and Methods From the lstituto di Ricerche Fatmw/~gic/~e “Mnrio Nqri,” Mih; 1 Chica Ostetriro Gimmlogic~~ “Ixigi Mm~iaplli,” University of Milan, Milan; and Istituto Europeo di Oncolug~a, Milan, ltnly. We thank Ms. loam Gnrirnoldi, Ms. j&y Bqgott, nrd the G. A. Pfeifer Memorial Lihraty Stnf for editorin/ ossdsrlce.

754

0029.7844/95/$9.50 SSDI 0029.7844(955)00281-2

Included in this analysis are 291 women (median age 33 years, range 17-69) consecutively referred for the first time with one smear test showing mild dyskaryosis during 1990-1992 to the colposcopic unit of the First

Obstetrics & Gynecology

Obstetric and Gynecologic Clinic of the University of Milan. Mild dyskaryosis was defined according to Koss.’ None had had a previous abnormal smear, and the median number of lifelong Papanicolaou smears was three (range zero to 12). Each underwent colposcopy with the application of 5% acetic acid and Lug01 solution and biopsy. Grading of CIN followed the World Health Organization criteria.‘” There were no cases of invasive or microinvasive cervical cancer. Twenty-eight of the 291 women (10%) had CIN I at biopsy, 46 (15%) CIN II, and 23 (8%) CIN III. The physician asked the women about their general characteristics, reproductive history, smoking, sexual habits, and oral contraceptive use before the gynecologic examination. Histologic findings were classified as low-grade dysplasia (GIN I or less) or high-grade dysplasia (CIN II or III). We computed the odds ratios (OR) of high-grade dysplasia together with their 95% approximate confidence intervals (CI) from data stratified for age by the Mantel-Haenszel procedure.” When a factor could be classified in more than two levels, the significance of the linear trend was assessed by the Mantel test.‘* Furthermore, to allow simultaneously for the effects of reciprocal confounding effects of various considered factors, unconditional multiple logistic regression with maximum likelihood fitting was used (GLIM System; Numerical Algorithms Group, Oxford, UK). Included in the regression equation were terms for age, education, and factors found statistically significant in the ageadjusted analysis. Because the age-adjusted and multivariate OR estimates were largely similar, only the latter are presented in the tables.

Results Table 1 shows the distribution of subjects according to age, education, and reproductive factors. The prevalence of high-grade dysplasia tended to decrease with increasing education. Conversely, parity was associated with a higher frequency of high-grade dysplasia; compared with nulliparas, the OR of high-grade dysplasia was almost doubled in parous women. There was no clear relation between age and prevalence of high-grade lesions, but women aged 30-40 years at diagnosis tended to be at higher risk of highgrade lesions (OR 1.7, 95% CI 0.8-3.5) than were women under 30. The relation between induced or spontaneous abortion and high-grade lesions was not statistically significant. Table 2 presents the relation between grade of dysplasia and smoking habits. Compared with neversmokers, the multivariate OR of high-grade dysplasia

VOL.

86, NO.

5, NOVEMBER

1YY.i

Table

1. Age,

Education,

and

Reproductive

Factors

Grade Low Negatiw findings Age (y) C30 30-40 P-40 ,$ trend Education (y) 511 >11 Parity 0 21 Spontaneous abortions 0 21 Induced abortions 0 21

High

OR (95% CI), high us II CIN III low grade”

CIN I CIN

5

I+

67 57 70

18 7 3

20 20 6

14 4

1.7 (0.8-3.5) 0.4 (0.1-1.3) 0.41 (NY

93 101

Y

19

21 23

22 1

1’ 0.5 (0.3-0.9)

139 55

24 4

34 12

14 9

1’ 1.8 (1.1-3.5)

165 29

27 1

44 2

21 2

1’ 0.8 (0.3-2.4)

147 47

22 6

32 14

15 8

1+ 1.3 (0.7-2.4)

OR = odds ratio; CI = confidence interval; NS = not significant. * Multivariate estimates, including terms for age, education, parity, smoking habits, and number of sexual partners. Reference category.

was 2.3 (95% CI 1.0-5.4) for current smokers. Exsmokers were at increased risk, too. Sexual habits are presented in Table 3. Compared with women reporting one sexual partner, the ORs of CIN II or III were 1.4 and 2.3, respectively, for women reporting two to three or four or more sexual partners (d trend 6.65, P < .05). No relation emerged between the prevalence of high-grade lesions and age at first intercourse. We analyzed the potential interaction of smoking and the number of sexual partners on the risk of CIN II or 111. No clear interaction emerged; compared with never smoking or reporting one sexual partner, the ORs of GIN II or III were, respectively, 2.1 in never-smokers reporting three or more sexual partners, 2.3 in ever-

Table

2. Smoking

Habits Grade Low Negative findings

Never-smokers Current smokers Ex-smokers

85 35 74

High CIN 13 3 12

I

CIN

II

CIN

9 Y 28

4 3 16

III

OR (95% CI), high us low grade* 1+ 2.3 (1.0-5.4) 3.8 (1.9-7.6)

OR = odds ratio; CI = confidence intewal. * Multivariate estimates, including terms for age, education, smoking habits, and number of sexual partners. Reference category.

Parazzini

et al

Hi@Grade

Dysplasia

parity,

755

Table

of Sexual Partners and Age at First Intercourse

3. Number

Grade

High

Low Negative finding No. of sexual partners 1 2-3 24 X: trend Age at first intercourse 518 19-20 221 & trend

CIN I

CIN

OR (95% CI),

II CIN III

high us low grade’

84 48 62

5 14 9

14 12 20

4 6 13

1’ 1.4 (0.6-3.2) 2.3 (1.0-3.0) 6.65 (P < .05)

81 53 60

14 7 7

28 6 12

11 10 2

li 0.9 (0.4-1X 0.7 (0.3-l .5) 0.94 (NS)

(y)

OR = odds ratio; CI = confidence interval; NS = not significant. * Multivariate estimates, including terms for age, education, parity, smoking habits, and number of sexual partners. + Reference category.

smokers smokers

reporting reporting

one sexual partner, and 2.6 in everthree or more sexual partners.

Discussion The results of this study suggest that among women with a single smear showing mild dyskaryosis, the prevalence of high-grade lesions (GIN II or III) is increased in less-educated women reporting one or more full-term pregnancies and in ever-smokers; the risk tended to increase with the number of sexual partners. These findings are at least partly in agreement with the scant available data. For example, smokers in the United Kingdom had a higher prevalence of high-grade dysplasia in two studies7,x comprising about 400 women with mild dyskaryosis referred for colposcopy in Aberdeen and Birmingham. In the Aberdeen study, however, the risk was confined to smokers of 20 cigarettes per day or more. Furthermore, in the Aberdeen study, a higher prevalence of CIN II or III was associated with a greater number of sexual partners, but not with age at first intercourse; the crude OR was about 3 for women reporting two or more sexual partners compared with those reporting only one.7 Unlike our results, the study conducted in Birmingham found no association between parity and prevalence of highgrade dysplasia.8 Part of these differences may be attributable to different selection criteria of the study populations. The population included in our analysis was, in fact, a group at lower risk of high-grade lesions than those in

756

Parazzini

et al

High-Gmde

Dysplasin

the other two studies. In the present analysis, the prevalence of CIN II or III among women with a single smear showing mild dyskaryosis was about 25%, an estimate consistent with figures from other series.13 In the two studies conducted in the United Kingdom, the prevalence of CIN II or III was 707 and 42%,* respectively. Potential bias should not explain these findings. In our study, all patients referred during the study period were included in the analysis; most of these women came from a network of collaborating first-level gynecologic outpatient services in Milan. Also, procedures did not change during the study. Although information on the number of sexual partners and the age at first intercourse could be partially misreported, it is unlikely that information bias differs substantially in patients with low- or high-grade dysplasia because information was collected before colposcopy and biopsy. In addition, recall bias can have little effect on variables such as parity or education. Finally, some of our findings are consistent with the recognized risk factors for CIN’4,‘5 and invasive disease, too.” Sexual habits are recognized determinants of risk of cervical disease, and now several studies’4*‘5 have shown that the risk of CIN is higher among smokers, parous women, and women of low social standing. This gives some support to the results of our study.

References 1. Giles JA, Hudson EA, Crow J, Williams D, Walker I’. Colposcopic assessment of the accuracy of cervical cytology screening. BMJ 1988;296:1099-102. 2. Hammond R. Management of women with smears showing mild dyskaryosis. BMJ 1994;308:1383-4. 3. Soutter WP, Wisdom S, Brough AK, Monaghan JM. Should patients with mild atypia in a cervical smear be referred for colposcopy? Br J Obstet Gynaecol 1986;93:70-4. 4. FlanneIIy G, Anderson D, Kitchener HC, et al. Management of women with mild and moderate cervical dyskaryosis. BMJ 1994; 308:1399-403. 5. Soutter WI’, Fletcher A. Invasive cancer of the cervix in women with mild dyskaryosis followed up cytologically. BMJ 1994;308: 1421-3. 6. Walker EM, Dodgson J, Duncan ID. Does mild atypia on a cervical smear warrant further investigation? Lancet 1986;iiz772-3. 7. Anderson DJ, Flannelly GM, Kitchener HC, et al. Mild and moderate dyskaryosis: Can women be selected for colposcopy on the basis of social criteria? BMJ 1992;305:84-7. 8. Luesley D, Blomfield I’, Dunn J, Shafi M, Chenoy R, Buxton J. Cigarette smoking and histological outcome in women with mildly dyskaryotic cervical smears. Br J Obstet Gynaecol 1994;101:49-52. 9. Koss LG. Diagnostic cytology and its histopathologic basis. Philadelphia: Lippincott, 1979. 10. Centers for Disease Control, U.S. Department of Health and Human Services. Classification system for human T-lymphotropic

Obstetrics B Gynecology

virus type III/lymphadenopathy-associated virus infection. Ann Intern Med 1986;105:234-7. 11. Mantel N, Haenszel W. StatistIcal aspects of the analysis of data from retrospective studies of disease. J Nat1 Cancer Inst 1959;22: 719-48. 12. Mantel N. Chi-square tests with one degree of freedom: Extensions of the Mantel-Haenszel procedure. J Am Stat Assoc 1963;58:690-

Address

reprint

requests

to:

Fabio Parazzini, MD lnstituto di Ricerche Farmacologiche via Eritrea, 62 20157 Milano Italy

‘ ‘Mario

Negri”

700. 13. Giles JA, Decry

A, Crow J, Walker I’. The accuracy of repeat cytology in women with mildly dyskaryotic smears. Br J Obstet Gynaecol 1989;96:1067-70. 14. Jones CJ, Brinton LA, Hamman RF, et al. Risk factors for in sifu cervical cancer: Results from a case-control study. Cancer Res 1990;50:3657-62. 15. Parazzini F, La Vecchia C, Negri E, Fed& L, Franceschi S, Gall&a L. Risk factors for cervical intraepithelial neoplasia. Cancer 1992;

Received April 29, 1995. Received in reaised form July 21, 1995. Accepted August 3, 1995.

69~2276-82. 16. Brinton LA, Fraumeni JF Jr. Epidemiology cancer. J Chronic Dis 1986;39:1051-65.

VOL.

86,

NO.

5, NOVEMBER

1995

of uterine

cervical

Copyright 0 1995 by The Gynecologists.

American

Parazzini

et al

College

of Obstetricians

High-Grade Dysplasia

and

757