Fax: (416) 340-4579
Hormonal effects of Depo-Provera in cervical smears
A comparison with Triphasil and postmenopausal effects
Article first published online: 15 MAR 2002
Copyright © 2002 American Cancer Society
Volume 96, Issue 2, pages 74–82, 25 April 2002
How to Cite
Kaptain, S., Bloom, L. I. and Weir, M. M. (2002), Hormonal effects of Depo-Provera in cervical smears. Cancer, 96: 74–82. doi: 10.1002/cncr.10476
- Issue published online: 8 APR 2002
- Article first published online: 15 MAR 2002
- Manuscript Accepted: 27 JUL 2001
- Manuscript Revised: 26 JUN 2001
- Manuscript Received: 15 FEB 2001
- depot-medroxyprogesterone acetate (Depo-Provera [DP]);
- cervical smears;
- hormonal effects;
- levonorgestrel and ethinyl estradiol (Triphasil [TP]);
Long-acting injectable contraceptive agents may cause changes in cervical smears that could impair the detection of epithelial abnormalities. The objectives of the current study were to 1) compare the hormonal effects of depot-medroxyprogesterone acetate (Depo-Provera®) (DP) in cervical smears with those of levonorgestrel and ethinyl estradiol (Triphasil®) (TP) and postmenopausal (PM) changes; and 2) determine whether the duration of DP use affects squamous maturation.
Satisfactory cervical smears from 50 DP users, 55 TP users, and 51 PM patients were evaluated blindly for: 1) squamous cell curling, crowding, cytolysis, and navicular cell formation; 2) pseudoparakeratosis, blue blobs, and histiocytes; 3) endometrial cells and blood; 4) single or enlarged endocervical nuclei and mucin-depleted endocervical cells; 5) lactobacilli and coccobacilli amounts; and 6) squamous maturation (ratio of parabasal:intermediate:superficial cells).
No statistically significant differences were observed for blue blobs, histiocytes, blood, endometrial cells, or single or enlarged endocervical nuclei among smears from the three groups. More smears from DP and TP users demonstrated squamous cell curling, crowding, and cytolysis as well as navicular cells and abundant lactobacilli compared with smears from PM patients. There were more PM smears with pseudoparakeratosis and mucin-depleted endocervical cells compared with the other groups. The majority of PM smears (98%) demonstrated predominantly parabasal cells with some intermediate cells. The majority of DP (86%) and TP (93%) smears demonstrated mostly intermediate and some superficial cells, regardless of the duration of DP use.
Certain progestational-dependent effects (i.e., curling, crowding, navicular cells, and abundant lactobacilli) were identified more often in TP users compared with DP users and less often in PM patients. The mostly parabasal pattern observed in smears from PM patients contrasted with the predominantly intermediate pattern found in smears from DP and TP users. The duration of DP use did not appear to have any effect on squamous maturation. Cancer (Cancer Cytopathol) 2002;96:000–000. © 2002 American Cancer Society.
The hormonal cellular changes in cervicovaginal smears from pregnant and postmenopausal patients are well-known. However, to our knowledge the hormonal effects of oral and injectable contraceptive agents on human cervical and vaginal epithelium are underreported in the literature. This appears to be especially true for the new long-acting injectable or implantable progestin-only contraceptive agents depot-medroxyprogesterone acetate (Depo-Provera®; UpJohn Co., Kalamazoo, MI) (DP) and levonorgestrel (Norplant®; Wyeth-Ayerst Laboratories, Philadelphia, PA). It is interesting to note that studies that examine the risk of cervical dysplasia and/or carcinoma in DP users report contradictory results.1–7 Similarly, recent studies have reported conflicting data regarding the increased rate of cervical smear epithelial abnormalities among DP users.8–11 Some groups conclude that the hormonal effects of DP (“postpartum-like or atrophic pattern”) in cervical smears results in difficulty in recognizing high-grade squamous intraepithelial lesions or in overcalling reactive cellular changes as epithelial abnormalities in cervical smears.12, 13
Given the recent interest in the hormonal effects of DP on cervical smears, we undertook the current study to 1) examine the hormonal effects of DP in cervical smears; 2) compare the effects of DP with those of levonorgestrel and ethinyl estradiol (Triphasil®; Wyeth-Ayerst Laboratories) (TP) and postmenopausal (PM) changes; and 3) determine the effect of the duration of DP use on the squamous maturation pattern. We chose not to examine whether hormonal changes impaired the detection of epithelial abnormalities, but instead attempted to determine only differences in hormonal changes between the three groups (DP users, TP users, and postmenopausal patients).
The study was reviewed and approved by the Massachusetts General Hospital Human Research Ethics Board.
MATERIALS AND METHODS
We examined 157 satisfactory cervical smears from 50 DP users (50 smears), 55 TP users (56 smears), and 51 PM patients (51 smears) taken from the cytology files at Massachusetts General Hospital in a blinded retrospective fashion. A satisfactory smear was defined using the Bethesda criteria.14 The 50 DP users were selected randomly from 290 DP users who were identified through a computer search from 1996, without knowledge of their menstrual status or the duration of DP use. The 55 TP users were selected from a review of 300 cytology requisitions from 1996. The 51 PM patients were selected randomly from 1835 patients with atrophic smears who were identified through a computer search from 1996.
Cases were excluded for the following: pregnancy, postpartum or postabortion status, long-term steroid use, and untreated endocrine disease or the presence of significant inflammation (> 50% obscuring), Trichomonas vaginalis, Candida species, or herpes simplex viral changes. DP and TP users age > 45 years were excluded to avoid a perimenopausal hormone state.
For the 157 cervical smears examined, the original diagnoses were 142 within normal limits or benign cellular changes; 6 atypical squamous cells of undetermined significance (ASCUS) in smears taken from 1 DP user, 1 PM patient, and 4 TP users; and 9 squamous intraepithelial lesions (SIL) (6 low-grade [LSIL] and 3 high-grade [HSIL] lesions) from 5 DP users and 4 TP users.
We examined the 157 cervical smears for the following parameters: the presence of blue blobs (degenerated cells without nuclear detail), pseudoparakeratosis (small squamous cells with intense orange cytoplasm and pyknotic nuclei) (Fig. 1), blood (none or partially obscuring), multinucleated histiocytes, and endometrial cells. In addition, the squamous cells were examined for crowding (greater than four cells in a tightly cohesive group), curling (rounded-up cell edges), cytolysis (Fig. 2), and navicular cell formation (boat-shaped intermediate cell with curled edges and central glycogen) (Fig. 3). Endocervical cells were examined for the presence of single bare nuclei, enlarged nuclei, and a mucin-depleted cuboidal appearance. The numbers of lactobacilli (Fig. 4) were scored semiquantitatively as follows: 0: rare to absent, 1+: few or proliferative phase amounts, and 2+: abundant or luteal phase amounts. Coccobacilli numbers also were scored as 0: rare to absent, 1+: few, and 2+: abundant (Fig. 5).
Univariate analysis of the categoric variables was performed using the chi-square test and, when appropriate, the Fisher exact test. A P value < 0.05 was considered to be statistically significant. Significance tests compared DP users with PM patients, and DP users with TP users.
Squamous maturation on the 157 cervical smears then was assessed. The degree of maturation was assigned to one of nine numeric categories based on the proportion of parabasal (PB), intermediate (INT), and superficial (SUP) cells. These numeric categories were assigned in ascending order with 1.00 indicating 100% PB cells and 3.00 indicating 100% SUP cells. (Table 1) (Figs. 6 and 7). The categories do not represent a maturation index but rather indicate a maturation perception scale that we devised.
|Proportion of cell types||Superficial (%)|
|Parabasal (%)||Intermediate (%)|
The duration of DP hormone use was confirmed using outpatient charts or by direct communication with clinicians. TP use was confirmed based on the cervical smear requisition or from outpatient charts. The date of the patient's last menstrual period (LMP) was obtained only for DP users from outpatient charts or by direct communication with clinicians.
DP users ranged in age from 15–44 years (average, 27 years) compared with TP users, who ranged in age from 20–41 years (average, 28 years) (Table 2). PM patients ranged in age from 54–90 years (average, 68 years). The distribution of clinic sites for the three groups was not comparable (Table 2). The PM group had a nearly equal distribution between hospital-based and community-based clinics. In contrast, the majority of DP users were from community-based clinics, whereas many of the TP users were from hospital-based clinics (Table 2).
|DP (n = 50)||TP (n = 56)||PM (n = 51)|
|Age range (yrs)||15–44||20–41||54–90|
|Average age (yrs)||27||28||68|
|Community-based clinic||37 (74%)||22 (39%)||28 (55%)|
|Hospital-based clinic||13 (26%)||34 (61%)||23 (45%)|
There were no statistically significant differences using univariate analysis with regard to the numbers of cervical smears demonstrating endometrial cells, blood, single or enlarged endocervical cells, or blue blobs among the three groups (Table 3).
|DP (n = 50)||TP (n = 56)||PM (n = 51)||P valuea|
|DP vs. TP||DP vs. PM|
|Endometrial cells||0||1 (2%)||0||1.00||1.00|
|Blood||22 (44%)||25 (45%)||16 (31%)||0.78||0.82|
|Single EC nuclei||1 (2%)||1 (2%)||1 (2%)||1.00||1.00|
|Enlarged EC nuclei||2 (4%)||0||0||0.22||0.24|
|Blue blobs||1 (2%)||0||5 (10%)||0.47||0.21|
On univariate analysis, there were statistically significantly higher numbers of cervical smears that demonstrated squamous cell curling and crowding from TP users compared with DP users and from DP users compared with PM patients (Table 4). There was no statistically significant difference between the number of cervical smears demonstrating cytolysis from DP users (30%) and TP users (25%) (P = 0.56) (Table 4), a finding that is in contrast with the 0% cytolysis rate found for PM patients. Navicular cells were identified in more TP users compared with DP users (P = 0.01) and PM patients (Table 4).
|DP (n = 50)||TP (n = 56)||PM (n = 51)||P value|
|DP vs. TP||DP vs. PM|
|Crowding||34 (68%)||49 (88%)||8 (16%)||0.015||0.001|
|Curling||39 (78%)||55 (98%)||5 (10%)||0.001||0.001|
|Cytolysis||15 (30%)||14 (25%)||0||0.56||0.001|
|Navicular cells||8 (16%)||21 (38%)||6 (12%)||0.01||0.54|
Pseudoparakeratosis was apparent in more PM patients (49%) compared with the other two groups (10% in DP users and 7% in TP users; P = 0.001) (Table 5) Multinucleated histiocytes were observed more commonly in the smears from the DP users (14%) and PM patients (23%) compared with the TP users (2%) (Table 5). Mucin-depleted cuboidal endocervical cells were identified in more cervical smears from PM patients compared with DP and TP users (53% vs. 4% vs. 0%, respectively) (Table 5).
|DP (n = 50)||TP (n = 56)||PM (n = 51)||P value|
|DP vs. TP||DP vs. PM|
|Pseudopk||5 (10%)||4 (7%)||25 (49%)||0.73a||0.001|
|MNH||7 (14%)||1 (2%)||12 (24%)||0.03a||0.22|
|EC||2 (4%)||0||27 (53%)||0.22a||0.001|
|Lactobacilli 2+||26 (52%)||44 (79%)||0||0.006||0.001|
|Coccobacilli 2+||15 (30%)||6 (11%)||0||0.018a||0.001|
Lactobacilli 2+ quantities were identified in more cervical smears from TP users compared with DP users (79% vs. 52%) (P = 0.006), a finding that is in contrast with the 0% rate of lactobacilli 2+ among PM patients (Table 5). Coccobacilli 2+ quantities were present in statistically significantly higher numbers in the cervical smears from DP users (30%) compared with TP users (11%) (P = 0.018) and PM patients (0%) (P = 0.001) (Table 5).
Comparison of lactobacilli and coccobacilli quantities among the DP and TP users (Table 6) shows that when coccobacilli 2+ quantities were present, the majority of the smears from DP users (13 of 15 smears; 87%) and TP users (5 of 6 smears; 83%) demonstrated absent to rare lactobacilli, and none demonstrated 2+ lactobacilli quantities. In addition, when coccobacilli were absent to rare, the majority of the smears from DP users (30 of 35 smears; 86%) and TP users (47 of 48 smears; 98%) demonstrated 1+ or 2+ lactobacilli (Table 6). Because only “satisfactory” cervical smears were selected for review, a significant inflammatory component was not present in the smears with 1+ or 2+ coccobacilli.
|DP coccobacilli (n = 50)||0||5||4||26||35|
|TP coccobacilli (n = 56)||0||1||4||43||48|
The majority of smears from PM patients (50 of 51 smears; 98%) demonstrated maturation categories of 1.25 (25 smears), 1.00 (12 smears) (Fig. 6), 1.50 (11 smears), or 1.75 (2 smears). These categories included predominantly PB cells (category 1.00), or PB cells in combination with INT cells (Fig. 8). One cervical smear from a PM patient had a 2.25 maturation category; this patient apparently was not receiving estrogen therapy. The majority of the smears from DP users (43 of 50 smears; 86%) and TP users (52 of 56 smears; 93%) had cervical smears with maturation categories of 2.25 (Fig. 4), 2.50 (Fig. 5), or 2.75. These categories corresponded to a combination of INT and SUP cells (Fig. 8). There were cervical smears from 3 DP users and 4 TP users with a maturation category of 2.00 (predominantly INT cells) (Figs. 7 and 8). Among the four remaining DP users, the cervical smears had maturation categories of 1.75 (in 2), 1.50 (in 1), and 1.25 (in 1).
The average duration of DP use was 18 months (range, 1–84 months) among 49 DP users (data unavailable for 1 DP user). The standard dosage of DP was 150 mg intramuscularly every 3 months. One patient also received topical vaginal estrogen cream. The comparison between the duration of DP use and the squamous maturation category (Table 7) demonstrates a duration of use varying from 17 months, 11 months, 13 months, and 22 months for maturation categories of < 2.00, 2.25, 2.50, and 2.75, respectively (Table 7). The longest DP user had a 2.75 maturation category, whereas the topical estrogen user had a 2.00 maturation category. There were 8 DP users with a duration of DP use > 2 years and maturation categories of 2.00 (2 DP users), 2.25 (4 DP users), or 2.75 (2 DP users).
|DP use (mos)||LMP (days)a|
|Squamous maturation category (n = 50)||≤2.00 (n = 7)||17||5–34b||N/A||98, 952|
|2.25 (n = 28)||11||1–36||34||1–90|
|2.50 (n = 7)||13||3–19||31||3–90|
|2.75 (n = 8)||22||3–84||14||7–20|
The LMP dates were available for only 23 of 50 DP users, and were on average 73 days (range, 1–952 days) or 33 days (if the “952 day” is excluded). There were 6 patients with LMP dates > 40 days (60 days, > 60 days, 90 days, 90 days, 98 days, and 952 days, respectively) with squamous maturation categories (Table 7) of 1.75 (for 1 DP user with an LMP of 98 days with the duration of DP use unknown), 2.00 (for 1 DP user with an LMP of 952 days and 34 months of DP use), 2.25 (for 3 DP users with an LMP of 60 days, > 60 days, and 90 days, respectively, and 2 months, 13 months, and 3 months, respectively, of DP use), or 2.50 (for 1 DP user with an LMP of 90 days and 3 months of DP use).
There were nine patients with a SIL diagnosis (six cases of LSIL and three cases of HSIL). The follow-up information is as follows. For four DP users with LSIL, there was no follow up in one case, LSIL was detected on biopsy in one case, HSIL was detected on biopsy and loop excision in one case, and follow-up was negative (curettage and cervical smear) in one case. In two TP users diagnosed with LSIL, LSIL was detected on biopsy in both patients. In one DP user diagnosed with HSIL, there was negative follow-up (curettages and loop excision of transformation zone [LETZ] excision) and for two TP users diagnosed with HSIL, HSIL was detected on biopsy and LETZ excision in one patient and LSIL was detected on LETZ excision in the other patient.
For the six patients with a diagnosis of ASCUS (favoring LSIL or SIL), follow-up information included repeat ASCUS findings in two TP users with negative follow-up cervical smear(s); LSIL was detected on cervical smear in one TP user with biopsy results suggestive of LSIL; negative cervical smears were reported in one TP user; LSIL was detected on cervical smear in one DP user, with negative follow-up cervical smear and biopsy. No follow-up information was available for one PM patient.
To our knowledge the hormonal effects of oral and injectable contraceptive agents on the cervicovaginal epithelium are underreported in the literature. This finding is in contrast to the hormonal cellular changes reported in cervicovaginal smears from pregnant and PM women, which are well known. To our knowledge even fewer studies have been published to date examining the effects of the new long-acting injectable progestin-only contraceptive agent DP on the cervicovaginal epithelium. Some authors have suggested an increase in the rate of cervical smear epithelial abnormalities among DP users, whereas others have refuted this finding.8–11 In addition, some groups have reported that the hormonal effects of DP (“postpartum or atrophic pattern”) in cervical smears results in difficulty in recognizing HSIL or in overcalling reactive cellular changes as epithelial abnormalities in cervical smears.12, 13
Given the recent interest in the hormonal effects of DP on cervical smears, the current study was performed to 1) examine the hormonal effects of DP in cervical smears taken from a population of DP users who were not selected on the basis of the duration of DP use or amenorrheal status; 2) compare the effects of DP with TP and PM changes in cervical smears; and 3) determine whether the duration of DP use affects the squamous maturation pattern in cervical smears. We chose not to examine whether hormonal changes impaired the detection of epithelial abnormalities, but wanted to determine only differences in hormonal changes between the three groups of patients (DP users, TP users, and PM patients).
The age distributions were similar for the DP and TP users but not for the PM patients, a finding that was expected given that an “older” PM status was the selection criteria for this latter group. The distributions with regard to clinic origin were different between the three groups, with disproportionately higher numbers of DP users coming from community-based clinics and TP users coming from hospital-based clinics. This finding may be related to the different patient populations at each site; injectable long-acting contraceptive agents are the preferred method of contraception among the patient population at the community-based clinics. It is not clear whether this unequal clinic site distribution contributed to a selection bias. It is interesting to note that there were 8–12 different locations comprising each of the community-based and hospital-based sites, eliminating a neighborhood selection bias.
The current study did not demonstrate any statistically significant differences between DP and TP users or PM patients for a variety of cervical smear parameters (the presence of endometrial cells, blood, single or enlarged endocervical cell nuclei, or blue blobs). Valente et al. also reported no statistically significant difference between 29 DP users and 25 nonusers with regard to enlarged reactive endocervical nuclei (62.5% and 33%, respectively).13
With regard to “progestational effects,” there were some interesting findings among the three study groups. There were statistically significant differences identified for the presence of squamous cell curling and crowding in the cervical smears, with these two parameters identified in the majority of TP users but reported less commonly in the DP users and in only a minority of the PM patients. Navicular cells were identified more often as well among the TP users compared with the other two groups. Cytolysis rates were comparable between the DP and TP users, which contrasted with the rate of 0% reported in PM patients. Of note, Volk et al. reported increased rates of glycogenated intermediate squamous cells among the cervical smears from 74 DP users compared with smears from 38 non-DP users, and showed that the time since the LMP was inversely proportional to glycogenation among the DP users.15
With regard to the cervical smear parameters examined, the DP users had significantly more smears that demonstrated progestational-dependent changes (squamous cell curling, crowding, and cytolysis) but not navicular cell formation, and fewer smears demonstrating pseudoparakeratosis and mucin-depleted endocervical cells compared with the PM patients, findings that are expected given the different hormonal status of the two groups. It is interesting to note that the TP users had significantly more cervical smears than the DP users demonstrating some progestational-dependent changes (squamous cell curling, crowding, and navicular cell formation), but not cytolysis.
One interesting finding in the current study was the difference in lactobacilli and coccobacilli rates, especially among the DP and TP users. Luteal phase amounts (2+) of lactobacilli were identified in more cervical smears from TP users compared with those from DP users, whereas abundant (2+) coccobacilli were observed in more cervical smears from DP users. The lactobacilli and coccobacilli amounts were inversely proportional, which most likely reflects the shift in vaginal flora when lactobacilli numbers are decreased or absent. The reason for the higher numbers of cervical smears with abundant coccobacilli among the DP users compared with the TP users was not apparent.
The squamous maturation pattern for the PM patients typically was “atrophic,” with a predominance of PB cells (75–100%) or with PB cells in combination with INT cells (25–50%) (Fig. 8). The maturation patterns of the PM patients were strikingly different from the those of the DP and TP users, with the majority of DP and TP users demonstrating a predominance of INT cells (75–100%) or INT cells in combination with SUP cells (25–50%) (Fig. 8). The maturation patterns of the DP and TP users could not be classified as “atrophic” for the majority of the smears. However, there were rare exceptions, where the maturation patterns among four DP users overlapped with the PM patients' patterns.
Comparison of the duration of DP use and squamous maturation did not demonstrate a well defined relation (Table 7). The results of the current study did not demonstrate a left shift in the maturation pattern (i.e., more “atrophic”) among patients with longer DP use. Although the LMP data are limited (available in only 23 of 50 DP users) and the group numbers are low, the 2 patients with LMP dates > 3 months (98 days and 952 days, respectively) did appear to have more left-shifted maturation patterns (categories 1.75 or 2.00), in contrast to the 4 patients with LMP dates between 40–90 days (categories 2.25 or 2.50).
Davidson and Wu reported a predominantly INT cell pattern with limited maturation in the majority of 77 DP and Norplant® users, with a predominance of SUP cells in some cases, although no control group was reported.12 Other authors have reported a “pregnancy,” “post-partum,” or “atrophic pattern” for DP users.8, 13, 15 One study compared the cervical smear maturation values for DP users with that for non-DP users (38.45 vs. 64.60; P < 0.001), in which the DP users were selected specifically to be amenorrheal for > 5 months (average, 22 months).13 A recent study reported a significant decrease in the vaginal epithelial thickness, number of epithelial cell layers, and number of keratinized epithelial cells in Rhesus monkeys receiving DP or Norplant, compared with the pretreatment control cycle.16 These effects were reversible after removal of the progestin agent.16 Liu et al. reported a predominant INT cell pattern with SUP cells for patients receiving oral contraceptive agents, with decreased numbers of SUP cells found compared with the control group (who received no contraceptive agent).17 Based on the literature, it would appear that oral and injectable contraceptive agents have comparable squamous maturation patterns in cervical smears that are more left-shifted compared with nonhormonal users.
Although the current study was blinded, had comparable numbers in each study group, and attempted to examine the cervical smear hormonal pattern for DP users without selection of patients based on duration of DP use and amenorrheal status, there were some biases. By reviewing only satisfactory cervical smears, there was a selection bias against smears with significant inflammation, obscuring blood, or noticeable air-drying artifact, which may be typical of smears from DP users. We chose to include only satisfactory smears to allow more accurate assessment of the hormonal effects of DP without the compounding influence of significant inflammation or artifactual processes.
By comparing DP users with TP users and PM patients only, the results of the current study did not demonstrate how the hormonal effects of DP in cervical smears compared with age-matched controls who were not using hormonal contraception. Unfortunately, we were unable to identify an age-matched control group of patients in the study laboratory who we were absolutely certain were not using hormonal contraception based on 1996 cervical smear computer lists.
The results of the current study demonstrated that many of the progestational hormonal effects in the cervical smears from DP users were not comparable to those of the PM patients. Only a small number of DP users demonstrated cervical smear squamous maturation patterns that were similar to those of PM patients. The reasons for these findings are not known. Because DP is a low-dose progestational agent, there still may be some endogenous estrogen influence on the cervical epithelium, causing a more right-shifted maturation pattern compared with that in PM patients. This study also demonstrated similar squamous maturation patterns between the cervical smears of DP and TP users, and more cervical smears from TP users were found to demonstrate progestational effects compared with smears from DP users. It is interesting to note that more cervical smears from DP users demonstrated abundant coccobacilli compared with the smears from the other groups, a factor that possibly could contribute to difficult screening of the cervical smears. The results of the current study were unable to demonstrate a relation between the duration of DP use and the squamous maturation pattern, although there were rare DP users whose cervical smears demonstrated left-shifted squamous maturation patterns when the LMP was > 95 days.
The authors would like to thank Ms. Merna Cardoso and Ms. Melissa Skarban for typing the article.
- 8Cervical cytology associated with medroxyprogesterone acetate contraception. Acta Cytol. 1995; 39: 981A., , , , , .
- 9Epithelial cell abnormality in the Depo-Provera patient: a diagnostic pitfall. Acta Cytol. 1998; 42: 1235A., , .
- 10Cytologic findings on (Papanicolaou) smears of patients on Depot-Provera. Acta Cytol. 1998; 42: 1235A., .
- 11Squamous intraepithelial lesions following diagnoses of atypical squamous cells of undetermined significance (ASCUS) in progesterone-treated women: comparison with an untreated control group. Acta Cytol. 1998; 42: 1224A., .
- 12Underestimation of cervical lesion severity in cytology specimens from patients receiving long-term antifertility treatment. Acta Cytol. 1997; 41: 1566A., .
- 14The Bethesda System for reporting cervical vaginal cytological diagnoses. New York: Springer-Verlag, 1994: 6, 8., .