- Email: [email protected]
The effect of route of delivery on regression of abnormal cervical cytologic findings in the postpartum period
The effect of route of delivery on regression of abnormal cervical cytologic findings in the postpartum period David Ahdoot, MD,a Kristi M. Van Nostrand, MD,b Nicole J. Nguyen, BA,a Devansu S. Tewari, MD,a Tom Kurasaki, MS,a, c Philip J. DiSaia, MD,a and G. Scott Rose, MDa Orange and Irvine, California, and Stony Brook, New York OBJECTIVES: We sought to determine whether pregnant women with abnormal antepartum cervical cytologic findings differ in their postpartum rates of regression with respect to mode of delivery. STUDY DESIGN: Between 1990 and 1997, 446 pregnant women with antepartum abnormal cervical cytologic findings were identified. Complete demographic, clinical, and cytologic reports were available for 138 women. Papanicolaou smear data were collected and separated into three groups by use of the Bethesda classification system (atypical squamous cells of undetermined significance, low-grade squamous intraepithelial cells, and high-grade intraepithelial cells). Postpartum regression rates of antepartum Papanicolaou smears, with respect to degree of squamous epithelial cell abnormality and mode of delivery, were analyzed by Fisher’s exact and Wilcoxon rank sum tests. RESULTS: Of the 138 women, 109 (79%) were delivered vaginally and 29 (21%) by cesarean section. No statistically significant difference was found between women delivered vaginally and those delivered by cesarean section with respect to age, parity, and smoking history within the three groups (atypical squamous cells of undetermined significance, low-grade squamous intraepithelial cells, and high-grade squamous intraepithelial cells). The overall postpartum regression rate for the 59 women with antepartum high-grade squamous intraepithelial cells was 48%. Of the 47 women with high-grade squamous intraepithelial cells who were delivered vaginally, 28 showed regression in the postpartum period versus none of the 12 women delivered by cesarean section (60% vs 0%, p < 0.0002). CONCLUSION: Postpartum spontaneous regression of Papanicolaou smears consistent with high-grade squamous intraepithelial cells occurs with increased frequency among women who are delivered vaginally versus by cesarean section. (Am J Obstet Gynecol 1998;178:1116-20.)
Key words: Cervical dysplasia, cervical intraepithelial neoplasia, pregnancy, vaginal delivery, cesarean section Although cervical cancer rates have decreased dramatically with the widespread use of the Papanicolaou smear, the incidence of cervical intraepithelial neoplasia (CIN) has gradually increased, especially in younger women.1 In addition, with improved prenatal care and more diligent screening practices, an increased diagnosis of CIN has been seen during pregnancy.2 Abnormal Papanicolaou smears are common among gravid women (0.5% to 3%) because the peak incidence for both CIN and childbearing
From the Division of Gynecologic Oncology, University of California, Irvine, Medical Center,a the Department of Obstetrics and Gynecology, State University Hospital at Stony Brook,b and the Department of Epidemiology, University of California at Irvine.c Supported in part by grants from Memorial Health Services, Long Beach Memorial Medical Center (Long Beach, California), and the United States Army. Frank Lynch Memorial Prize Essay, presented at the Sixty-fourth Annual Meeting of the Pacific Coast Obstetrical and Gynecological Society, Coeur d’Alene, Idaho, September 17-21, 1997. This article represents the opinions of the authors and is not necessarily the opinions of the United States Government. Reprint requests: G. Scott Rose, MD, Division of Gynecologic Oncology, University of California, Irvine, Medical Center, Orange, CA 92868. Copyright © 1998 by Mosby, Inc. 0002-9378/98 $5.00 + 0 6/6/89325
1116
occurs in women during their third decade of life.3 Therefore screening for detection of carcinoma of the uterine cervix has become a standard part of prenatal care.4-6 The major objective of an antepartum cytologic and colposcopic evaluation is to rule out the presence of an invasive lesion while allowing the pregnancy to proceed to term.7, 8 The natural history of CIN during pregnancy and the postpartum period has been examined in several studies. Some authors describe pregnancy as having no effect on CIN,3, 9, 10 whereas others have reported higher regression rates of cervical dysplasia in the postpartum period compared with spontaneous regression rates of dysplasia for nonpregnant women.8-12 The effect of mode of delivery, however, on regression of cervical dysplasia in the postpartum period has not been examined. The objective of this study is to determine whether abnormal antepartum cervical cytologic findings result in differing postpartum regression rates with respect to mode of delivery. Material and methods This study was approved by the institutional review boards of the University of California, Irvine, Medical
Ahdoot et al. 1117
Volume 178, Number 6 Am J Obstet Gynecol
Table I. Degree of cervical dysplasia and mode of delivery Mode of delivery
ASCUS LGSIL HGSIL
*The
No. of subjects
Vaginal delivery (No.)
26 53 59
20 42 47
Cesarean section* No.
%
6 11 12
23.1 20.8 20.3
average cesarean section rate for the two institutions during the study period was 20.5%.
Table II. Population demographics Mode of delivery Vaginal delivery ASCUS Age (yr, mean) Parity (median) Smokers LGSIL Age (yr, mean) Parity (median) Smokers HGSIL Age (yr, mean) Parity (median) Smokers
Cesarean section
Statistical significance
28 1 33%
29 3 17%
NS NS NS
29 1 12%
28 0 30%
NS NS NS
28 1 18%
29 1 18%
NS NS NS
The obstetric population at the study institutions is predominantly Hispanic. NS, Not significant.
Center and the State University Hospital at Stony Brook. Women with abnormal cervical cytologic findings were referred to either the resident colposcopy clinic or the gynecologic oncology group. Between 1990 and 1997, 483 women with abnormal cervical cytologic findings at their initial antepartum visit were identified through clinic and pathologic data logs. We retrospectively reviewed inpatient and outpatient medical records, which were available for 446 women. Data were collected for age, gravidity, parity, history of sexually transmitted diseases, smoking history, colposcopic evaluation with antepartum and postpartum biopsies, cytologic and histologic reports of pathologic cervical conditions, and postpartum treatment and follow-up of abnormal cervical cytologic findings. Complete demographic, clinical, and cytologic reports in the antepartum and postpartum period were available for 138 patients. By use of the Bethesda classification system, the initial antepartum cytologic data were separated into three groups; atypical squamous cells of undetermined significance (ASCUS), low-grade squamous intraepithelial lesions (LGSIL), and high-grade squamous intraepithelial lesions (HGSIL). Antepartum and postpartum cervical cytologic data were compared between women who delivered vaginally versus by cesarean section. The highestgrade abnormal antepartum Papanicolaou smear was
compared with the highest-grade Papanicolaou smear within 12 weeks from the date of delivery. Regression was defined as either complete normalization of Papanicolaou smear findings or regression of HGSIL to LGSIL. Statistical analysis was performed by means of Fisher’s exact test and the Wilcoxon rank sum test. Results Complete demographic, clinical, and cytologic reports in the antepartum and postpartum period were available for 138 women. Table I depicts the number of patients with respect to degree of squamous intraepithelial abnormality and mode of delivery. At their initial antepartum visit, 26 women had ASCUS, 53 had LGSIL, and 59 had HGSIL. The overall cesarean section rate was 21%. This rate was equally distributed among the three groups and was comparable to the prevailing cesarean section rate at the two institutions. Table II depicts the demographics of the study population. No statistically significant differences were found between women who were delivered vaginally versus those delivered by cesarean section with respect to age, parity, and smoking among women with ASCUS, LGSIL, or HGSIL. The regression and persistence rates of ASCUS, LGSIL, and HGSIL in the postpartum period are depicted in Table III. Of the 26 women with cytologic find-
1118 Ahdoot et al.
June 1998 Am J Obstet Gynecol
Table III. Postpartum regression rates of cervical dysplasia Mode of delivery
Vaginal ASCUS Persistence Regression LGSIL Persistence Regression HGSIL Persistence Regression
20 (77%) 6 (30%) 14 (70%) 42 (79%) 15 (36%) 27 (64%) 47 (80%) 19 (40%) 28 (60%)
Cesarean Statistical section significance* 6 (23%) 3 (50%) 3 (50%) 11 (21%) 4 (36%) 7 (64%) 12 (20%) 12 (100%) 0 (0%)
p = 0.628 p = 1.000 p = 0.0002
*The p values compare the regression rates of cervical dysplasia with respect to mode of delivery.
ings of ASCUS, 20 (77%) were delivered vaginally and 6 (23%) by cesarean section. Normalization of Papanicolaou smears in the postpartum period were observed in 14 women delivered vaginally versus 3 women delivered by cesarean section (70% vs 50%, p = 0.628). With respect to the 53 women with antepartum cytologic findings of LGSIL, the overall regression rate was 64%; 42 (79%) were delivered vaginally and 11 (21%) by cesarean section. Regression was noted in 27 women delivered vaginally versus 7 women delivered by cesarean section (64.3% vs 63.7%, p = l.0). Of the 59 women with HGSIL, 47 (80%) were delivered vaginally and 12 (20%) by cesarean section, with an overall regression rate of 48%. Cytologic regression was noted in 28 of 47 women delivered vaginally versus none of the women delivered by cesarean section (60% vs 0%, p < 0.0002). Of the 28 women who were delivered vaginally and exhibited cytologic regression of HGSIL post partum, 20 had continued follow-up for cytologic evaluation and only 2 of these women had recurrence of HGSIL 9 months after the date of delivery. In the 12 women with HGSIL delivered by cesarean section, none entered the second stage of labor; 9 had a cesarean section electively, 2 of which were for fetal intolerance before active labor and 1 for failed induction. Antepartum colposcopically directed biopsies were performed for histologic diagnosis in 27 of 47 women with HGSIL as part of their antepartum evaluation. The rate of spontaneous postpartum regression was higher for women who had not undergone an antepartum biopsy compared with those who had (70% vs 52%). This difference in regression was not statistically significant, and no difference was noted with respect to mode of delivery. Comment Epidemiologic studies of CIN during pregnancy have exhibited varying outcomes.8, 10, 12 Although it is a sub-
ject of controversy, many authors believe CIN is not accelerated by pregnancy. In fact, cervical dysplasia often regresses completely in the postpartum period.3, 9, 10 In a study by Hall and Walton,13 spontaneous regression rates for cervical dysplasia in nonpregnant women were reported as 62%, 33%, and 19% for mild, moderate, and severe dysplasia, respectively. Studies on regression rates of dysplasia during pregnancy have found higher regression rates in the postpartum period compared with rates in nonpregnant women.8, 10, 12 In this study, 483 women were originally identified with abnormal antepartum cervical cytologic findings, of which records for 446 were obtained for review. Records for all women who had postpartum follow-up for abnormal cytologic findings were reviewed. Complete demographic, clinical, and cytologic reports in the antepartum and postpartum periods were available for 138 women. Postpartum records were not available for the remaining 308 women because of noncompliance with postpartum appointments. Of the women with antepartum abnormal cervical cytologic findings 138 underwent a postpartum follow-up examination to show regression or normalization of the cytologic findings. Of these 138 women, 9 were initially referred to the private practice in the gynecologic oncology department for evaluation. No significant difference was found with respect to severity of abnormal cervical cytologic findings, route of delivery, or cytologic regression between women referred to the private gynecologic oncology practice and those referred to the resident clinic. Postpartum evaluation consisted of a Papanicolaou smear for all women, with biopsy and colposcopy when indicated. No significant difference in regression rates was seen among women with cytologic findings of ASCUS or LGSIL. This is not surprising when one considers the epidemiologic characteristics of low-grade lesions. When left untreated, as many as 62% of low-grade lesions will resolve spontaneously.13 Although these data were established in nonpregnant patients, they have been confirmed in pregnancy by Kiguchi et al.,8 who found high spontaneous regression rates of minimally abnormal Papanicolaou smears in the postpartum period. In this study, on comparison with nonpregnant women, no significant difference in the postpartum regression was noted with respect to low-grade lesions. Because of the high spontaneous regression pattern associated with these low-grade lesions, we did not have an adequate number of patients to demonstrate a significant difference in regression patterns with respect to mode of delivery. In our study the overall postpartum regression rate for women with HGSIL was 48%. Kiguchi et al.8 examined 78 women with high-grade dysplasia during pregnancy and found that regression was demonstrated in 54% within 12 weeks of delivery. The reported spontaneous
Ahdoot et al. 1119
Volume 178, Number 6 Am J Obstet Gynecol
regression rate of HGSIL ranges from 6% to 31% in nonpregnant women and 30% to 54% in pregnancy.8, 10, 12 Therefore our study supports these previous studies and the clinical observation that the spontaneous regression of cervical dysplasia occurs with increased frequency in the postpartum period. Although this study demonstrated increased regression for all women with HGSIL, regression was most significant for women who were delivered vaginally. Cytologic regression was noted in 28 of 47 women delivered vaginally versus none of the women delivered by cesarean section (60% vs 0%, p < 0.0002). In addition, of the 20 women with HGSIL who were delivered vaginally and exhibited cytologic regression in the postpartum period, only 2 women (10%) demonstrated recurrence of HGSIL within 9 months of delivery. The 12 women delivered by cesarean section who demonstrated no cytologic regression were treated with an excision procedure. Interestingly, the recurrence rate for women who had spontaneous regression of HGSIL after vaginal delivery was similar to the spontaneous recurrence rate in women with dysplasia who were treated post partum with an excision procedure (10% vs 8%, respectively). Explanations for the observed discrepancy of postpartum cytologic regression with respect to mode of delivery may include problems inherent to Papanicolaou smear studies. Attaining an adequate sampling of the transformation zone is enhanced by the physiologic cervical eversion, which occurs in the gravid state. Postpartum falsenegative Papanicolaou smears resulting from inadequate cytologic sampling from the physiologic involution of the transformation zone might account for the increased observed regression. This regression would be expected to be evenly distributed between the two modes of delivery. However, as noted in this study, the women delivered by cesarean section exhibited no regression. In addition, the rate of regression could theoretically be increased by the performance of cervical biopsies in the antepartum evaluation. In effect, multiple biopsies could mechanically remove the lesions, thereby giving the appearance of spontaneous regression. Interestingly, in this study the rate of regression was higher for women who had no histologic cervical biopsy as part of the antenatal evaluation compared with women who underwent a biopsy (70% vs 52%). Because the aforementioned study biases cannot explain the observed regression discrepancy, perhaps factors associated with parturition could be implicated. Desquamation of the cervical epithelium or enhancement of a localized reparative immunologic response after vaginal delivery are two theories that might explain this selective regression. In support of cervical desquamation, it is interesting to note that the women delivered by cesarean section not only did not exhibit cytologic regression postpartum but also did not enter the second
stage of labor. Of the 12 deliveries by cesarean section, 9 were performed electively, 2 because of fetal intolerance to labor, and 1 because of failed induction of labor. Thus none of these women entered the second or third stage of labor, during which cervical desquamation could occur, thereby resulting in cytologic regression. These data support the theory that traumatic cervical desquamation during the second and third stages of labor might partially account for the higher regression rates seen among women who had a vaginal delivery. Although not examined in this study, enhancement of a localized reparative immunologic response after vaginal delivery, resulting in cytologic regression, is an interesting theory that should be considered. Cervical biopsy findings suggestive of impaired immune responses have been reported in women with persistent cervical dysplasia. Fukuda et al.14 examined cervical biopsy specimens of nonpregnant women with dysplasia that had spontaneously regressed and from women with persistent dysplasia; they found that women with persistent dysplasia exhibited significantly lower numbers of Langerhans cells and T-helper cells in the biopsy specimens. These cells are intrinsic to the inflammatory wound healing process. Therefore it follows that an enhanced localized immune response after mechanical cervical dilatation and parturition might account for the increased regression of dysplasia that we see associated with vaginal delivery.10, 14 All women with cervical dysplasia diagnosed antepartum should undergo postpartum colposcopy with cytologic or histologic evaluation before proceeding with therapy. This is especially important among women with HGSIL who are delivered vaginally because of the high spontaneous regression rate of cervical dysplasia after vaginal delivery. Optimally, colposcopic evaluation should be performed 8 to 12 weeks post partum when most of the local inflammatory reactions and reparatory processes have resolved. Traumatic cervical desquamation and stimulation of local immune factors associated with vaginal delivery could play an important role in the spontaneous regression of cervical dysplasia in the postpartum period. Future prospective clinical and immunologic studies, such as immunohistochemical analysis of inflammatory infiltrates, are needed to confirm these theories. REFERENCES
1. Pfenninger JL. Colposcopy in a family practice residency: the first 200 cases. J Fam Pract 1992;34:67-71. 2. Kohan S, Beckman EM, Bigelow B, Klein SA, Douglas GW. The role of colposcopy in the management of cervical intraepithelial neoplasia during pregnancy and postpartum. J Reprod Med 1980;25:279-84. 3. Benedet JL, Selke PA, Nickerson KG. Colposcopic evaluation of abnormal Papanicolaou smears in pregnancy. Am J Obstet Gynecol 1987;157:932-7. 4. Ostergard DR, Nieberg RK. Evaluation of abnormal cervical cy-
1120 Ahdoot et al.
5.
6.
7.
8.
9.
tology during pregnancy with colposcopy. Am J Obstet Gynecol 1979;134:756-8. Bertini-Oliveira AM, Keppler MM, Luisi A, Tarico S, Lopes VLD, Delascio D, et al. Comparative evaluation of abnormal cytology, colposcopy and histopathology in preclinical cervical malignancy during pregnancy. Acta Cytol 1982;26:636-44. Economos K, Perez Veridiano N, Delke I, Collado ML, Tancer ML. Abnormal cervical cytology in pregnancy: a 17 year experience. Obstet Gynecol 1993;81:915-8. DePetrillo AD, Townsend DE, Morrow CP, Lickrish GM, DiSaia PJ, Roy M. Colposcopic evaluation of the abnormal Papanicolaou test in pregnancy. Am J Obstet Gynecol 1975;121:441-5. Kiguchi K, Bibbo M, Ilasegawa T, Kurihara S, Tsutsui F, Wied GIN. Dysplasia during pregnancy: a cytologic follow-up study. J Reprod Med 1981;26:66-72. Ueki M, Ueda M, Kumagai K, Ikamoto Y, Noda S, Matsuoka M.
June 1998 Am J Obstet Gynecol
10.
11. 12.
13. 14.
Cervical cytology and conservative management of cervical neoplasias during pregnancy. Int J Gynecol Pathol 1995;14:63-9. Yoonessi M, Wieckowska W, Mariniello D, Antkowiak J. Cervical intra-epithelial neoplasia during pregnancy. Int J Gynaecol Obstet 1982;20:111-8. Patsner B. Management of low-grade cervical dysplasia during pregnancy. South Med J 1990;83:1405-12. Kaplan AL, Kaufman RH. Diagnosis and management of dysplasia and carcinoma in situ of the cervix in pregnancy. Clin Obstet Gynecol 1967;10:871-8. Hall JE, Walton L. Dysplasia of the cervix: a prospective study of 206 cases. Am J Obstet Gynecol 1968; 100:662-71. Fukuda K, Hachisuga T, Nakamura S, Matsuo N, Twasaka T, Sugimori H. Local immune response in persistent cervical dysplasia. Obstet Gynecol 1993;82:941-5.
Bound volumes available to subscribers Bound volumes of the American Journal of Obstetrics and Gynecology are available to subscribers (only) for the 1998 issues from the publisher, at a cost of $116.00 for domestic, $148.73 for Canada, and $139.00 for international for Vol. 178 (January-June) and Vol. 179 (July-December). Shipping charges are included. Each bound volume contains a subject and author index, and all advertising is removed. Copies are shipped within 60 days after publication of the last issue in the volume. The binding is durable buckram with the Journal name, volume number, and year stamped in gold on the spine. Payment must accompany all orders. Contact Mosby, Inc., Subscription Services, 11830 Westline Industrial Drive, St. Louis, MO 63146-3318, USA; phone (800)453-4351 or (314)453-4351. Subscriptions must be in force to qualify. Bound volumes are not available in place of a regular Journal subscription.
Key words: Cervical dysplasia, cervical intraepithelial neoplasia, pregnancy, vaginal delivery, cesarean section Although cervical cancer rates have decreased dramatically with the widespread use of the Papanicolaou smear, the incidence of cervical intraepithelial neoplasia (CIN) has gradually increased, especially in younger women.1 In addition, with improved prenatal care and more diligent screening practices, an increased diagnosis of CIN has been seen during pregnancy.2 Abnormal Papanicolaou smears are common among gravid women (0.5% to 3%) because the peak incidence for both CIN and childbearing
From the Division of Gynecologic Oncology, University of California, Irvine, Medical Center,a the Department of Obstetrics and Gynecology, State University Hospital at Stony Brook,b and the Department of Epidemiology, University of California at Irvine.c Supported in part by grants from Memorial Health Services, Long Beach Memorial Medical Center (Long Beach, California), and the United States Army. Frank Lynch Memorial Prize Essay, presented at the Sixty-fourth Annual Meeting of the Pacific Coast Obstetrical and Gynecological Society, Coeur d’Alene, Idaho, September 17-21, 1997. This article represents the opinions of the authors and is not necessarily the opinions of the United States Government. Reprint requests: G. Scott Rose, MD, Division of Gynecologic Oncology, University of California, Irvine, Medical Center, Orange, CA 92868. Copyright © 1998 by Mosby, Inc. 0002-9378/98 $5.00 + 0 6/6/89325
1116
occurs in women during their third decade of life.3 Therefore screening for detection of carcinoma of the uterine cervix has become a standard part of prenatal care.4-6 The major objective of an antepartum cytologic and colposcopic evaluation is to rule out the presence of an invasive lesion while allowing the pregnancy to proceed to term.7, 8 The natural history of CIN during pregnancy and the postpartum period has been examined in several studies. Some authors describe pregnancy as having no effect on CIN,3, 9, 10 whereas others have reported higher regression rates of cervical dysplasia in the postpartum period compared with spontaneous regression rates of dysplasia for nonpregnant women.8-12 The effect of mode of delivery, however, on regression of cervical dysplasia in the postpartum period has not been examined. The objective of this study is to determine whether abnormal antepartum cervical cytologic findings result in differing postpartum regression rates with respect to mode of delivery. Material and methods This study was approved by the institutional review boards of the University of California, Irvine, Medical
Ahdoot et al. 1117
Volume 178, Number 6 Am J Obstet Gynecol
Table I. Degree of cervical dysplasia and mode of delivery Mode of delivery
ASCUS LGSIL HGSIL
*The
No. of subjects
Vaginal delivery (No.)
26 53 59
20 42 47
Cesarean section* No.
%
6 11 12
23.1 20.8 20.3
average cesarean section rate for the two institutions during the study period was 20.5%.
Table II. Population demographics Mode of delivery Vaginal delivery ASCUS Age (yr, mean) Parity (median) Smokers LGSIL Age (yr, mean) Parity (median) Smokers HGSIL Age (yr, mean) Parity (median) Smokers
Cesarean section
Statistical significance
28 1 33%
29 3 17%
NS NS NS
29 1 12%
28 0 30%
NS NS NS
28 1 18%
29 1 18%
NS NS NS
The obstetric population at the study institutions is predominantly Hispanic. NS, Not significant.
Center and the State University Hospital at Stony Brook. Women with abnormal cervical cytologic findings were referred to either the resident colposcopy clinic or the gynecologic oncology group. Between 1990 and 1997, 483 women with abnormal cervical cytologic findings at their initial antepartum visit were identified through clinic and pathologic data logs. We retrospectively reviewed inpatient and outpatient medical records, which were available for 446 women. Data were collected for age, gravidity, parity, history of sexually transmitted diseases, smoking history, colposcopic evaluation with antepartum and postpartum biopsies, cytologic and histologic reports of pathologic cervical conditions, and postpartum treatment and follow-up of abnormal cervical cytologic findings. Complete demographic, clinical, and cytologic reports in the antepartum and postpartum period were available for 138 patients. By use of the Bethesda classification system, the initial antepartum cytologic data were separated into three groups; atypical squamous cells of undetermined significance (ASCUS), low-grade squamous intraepithelial lesions (LGSIL), and high-grade squamous intraepithelial lesions (HGSIL). Antepartum and postpartum cervical cytologic data were compared between women who delivered vaginally versus by cesarean section. The highestgrade abnormal antepartum Papanicolaou smear was
compared with the highest-grade Papanicolaou smear within 12 weeks from the date of delivery. Regression was defined as either complete normalization of Papanicolaou smear findings or regression of HGSIL to LGSIL. Statistical analysis was performed by means of Fisher’s exact test and the Wilcoxon rank sum test. Results Complete demographic, clinical, and cytologic reports in the antepartum and postpartum period were available for 138 women. Table I depicts the number of patients with respect to degree of squamous intraepithelial abnormality and mode of delivery. At their initial antepartum visit, 26 women had ASCUS, 53 had LGSIL, and 59 had HGSIL. The overall cesarean section rate was 21%. This rate was equally distributed among the three groups and was comparable to the prevailing cesarean section rate at the two institutions. Table II depicts the demographics of the study population. No statistically significant differences were found between women who were delivered vaginally versus those delivered by cesarean section with respect to age, parity, and smoking among women with ASCUS, LGSIL, or HGSIL. The regression and persistence rates of ASCUS, LGSIL, and HGSIL in the postpartum period are depicted in Table III. Of the 26 women with cytologic find-
1118 Ahdoot et al.
June 1998 Am J Obstet Gynecol
Table III. Postpartum regression rates of cervical dysplasia Mode of delivery
Vaginal ASCUS Persistence Regression LGSIL Persistence Regression HGSIL Persistence Regression
20 (77%) 6 (30%) 14 (70%) 42 (79%) 15 (36%) 27 (64%) 47 (80%) 19 (40%) 28 (60%)
Cesarean Statistical section significance* 6 (23%) 3 (50%) 3 (50%) 11 (21%) 4 (36%) 7 (64%) 12 (20%) 12 (100%) 0 (0%)
p = 0.628 p = 1.000 p = 0.0002
*The p values compare the regression rates of cervical dysplasia with respect to mode of delivery.
ings of ASCUS, 20 (77%) were delivered vaginally and 6 (23%) by cesarean section. Normalization of Papanicolaou smears in the postpartum period were observed in 14 women delivered vaginally versus 3 women delivered by cesarean section (70% vs 50%, p = 0.628). With respect to the 53 women with antepartum cytologic findings of LGSIL, the overall regression rate was 64%; 42 (79%) were delivered vaginally and 11 (21%) by cesarean section. Regression was noted in 27 women delivered vaginally versus 7 women delivered by cesarean section (64.3% vs 63.7%, p = l.0). Of the 59 women with HGSIL, 47 (80%) were delivered vaginally and 12 (20%) by cesarean section, with an overall regression rate of 48%. Cytologic regression was noted in 28 of 47 women delivered vaginally versus none of the women delivered by cesarean section (60% vs 0%, p < 0.0002). Of the 28 women who were delivered vaginally and exhibited cytologic regression of HGSIL post partum, 20 had continued follow-up for cytologic evaluation and only 2 of these women had recurrence of HGSIL 9 months after the date of delivery. In the 12 women with HGSIL delivered by cesarean section, none entered the second stage of labor; 9 had a cesarean section electively, 2 of which were for fetal intolerance before active labor and 1 for failed induction. Antepartum colposcopically directed biopsies were performed for histologic diagnosis in 27 of 47 women with HGSIL as part of their antepartum evaluation. The rate of spontaneous postpartum regression was higher for women who had not undergone an antepartum biopsy compared with those who had (70% vs 52%). This difference in regression was not statistically significant, and no difference was noted with respect to mode of delivery. Comment Epidemiologic studies of CIN during pregnancy have exhibited varying outcomes.8, 10, 12 Although it is a sub-
ject of controversy, many authors believe CIN is not accelerated by pregnancy. In fact, cervical dysplasia often regresses completely in the postpartum period.3, 9, 10 In a study by Hall and Walton,13 spontaneous regression rates for cervical dysplasia in nonpregnant women were reported as 62%, 33%, and 19% for mild, moderate, and severe dysplasia, respectively. Studies on regression rates of dysplasia during pregnancy have found higher regression rates in the postpartum period compared with rates in nonpregnant women.8, 10, 12 In this study, 483 women were originally identified with abnormal antepartum cervical cytologic findings, of which records for 446 were obtained for review. Records for all women who had postpartum follow-up for abnormal cytologic findings were reviewed. Complete demographic, clinical, and cytologic reports in the antepartum and postpartum periods were available for 138 women. Postpartum records were not available for the remaining 308 women because of noncompliance with postpartum appointments. Of the women with antepartum abnormal cervical cytologic findings 138 underwent a postpartum follow-up examination to show regression or normalization of the cytologic findings. Of these 138 women, 9 were initially referred to the private practice in the gynecologic oncology department for evaluation. No significant difference was found with respect to severity of abnormal cervical cytologic findings, route of delivery, or cytologic regression between women referred to the private gynecologic oncology practice and those referred to the resident clinic. Postpartum evaluation consisted of a Papanicolaou smear for all women, with biopsy and colposcopy when indicated. No significant difference in regression rates was seen among women with cytologic findings of ASCUS or LGSIL. This is not surprising when one considers the epidemiologic characteristics of low-grade lesions. When left untreated, as many as 62% of low-grade lesions will resolve spontaneously.13 Although these data were established in nonpregnant patients, they have been confirmed in pregnancy by Kiguchi et al.,8 who found high spontaneous regression rates of minimally abnormal Papanicolaou smears in the postpartum period. In this study, on comparison with nonpregnant women, no significant difference in the postpartum regression was noted with respect to low-grade lesions. Because of the high spontaneous regression pattern associated with these low-grade lesions, we did not have an adequate number of patients to demonstrate a significant difference in regression patterns with respect to mode of delivery. In our study the overall postpartum regression rate for women with HGSIL was 48%. Kiguchi et al.8 examined 78 women with high-grade dysplasia during pregnancy and found that regression was demonstrated in 54% within 12 weeks of delivery. The reported spontaneous
Ahdoot et al. 1119
Volume 178, Number 6 Am J Obstet Gynecol
regression rate of HGSIL ranges from 6% to 31% in nonpregnant women and 30% to 54% in pregnancy.8, 10, 12 Therefore our study supports these previous studies and the clinical observation that the spontaneous regression of cervical dysplasia occurs with increased frequency in the postpartum period. Although this study demonstrated increased regression for all women with HGSIL, regression was most significant for women who were delivered vaginally. Cytologic regression was noted in 28 of 47 women delivered vaginally versus none of the women delivered by cesarean section (60% vs 0%, p < 0.0002). In addition, of the 20 women with HGSIL who were delivered vaginally and exhibited cytologic regression in the postpartum period, only 2 women (10%) demonstrated recurrence of HGSIL within 9 months of delivery. The 12 women delivered by cesarean section who demonstrated no cytologic regression were treated with an excision procedure. Interestingly, the recurrence rate for women who had spontaneous regression of HGSIL after vaginal delivery was similar to the spontaneous recurrence rate in women with dysplasia who were treated post partum with an excision procedure (10% vs 8%, respectively). Explanations for the observed discrepancy of postpartum cytologic regression with respect to mode of delivery may include problems inherent to Papanicolaou smear studies. Attaining an adequate sampling of the transformation zone is enhanced by the physiologic cervical eversion, which occurs in the gravid state. Postpartum falsenegative Papanicolaou smears resulting from inadequate cytologic sampling from the physiologic involution of the transformation zone might account for the increased observed regression. This regression would be expected to be evenly distributed between the two modes of delivery. However, as noted in this study, the women delivered by cesarean section exhibited no regression. In addition, the rate of regression could theoretically be increased by the performance of cervical biopsies in the antepartum evaluation. In effect, multiple biopsies could mechanically remove the lesions, thereby giving the appearance of spontaneous regression. Interestingly, in this study the rate of regression was higher for women who had no histologic cervical biopsy as part of the antenatal evaluation compared with women who underwent a biopsy (70% vs 52%). Because the aforementioned study biases cannot explain the observed regression discrepancy, perhaps factors associated with parturition could be implicated. Desquamation of the cervical epithelium or enhancement of a localized reparative immunologic response after vaginal delivery are two theories that might explain this selective regression. In support of cervical desquamation, it is interesting to note that the women delivered by cesarean section not only did not exhibit cytologic regression postpartum but also did not enter the second
stage of labor. Of the 12 deliveries by cesarean section, 9 were performed electively, 2 because of fetal intolerance to labor, and 1 because of failed induction of labor. Thus none of these women entered the second or third stage of labor, during which cervical desquamation could occur, thereby resulting in cytologic regression. These data support the theory that traumatic cervical desquamation during the second and third stages of labor might partially account for the higher regression rates seen among women who had a vaginal delivery. Although not examined in this study, enhancement of a localized reparative immunologic response after vaginal delivery, resulting in cytologic regression, is an interesting theory that should be considered. Cervical biopsy findings suggestive of impaired immune responses have been reported in women with persistent cervical dysplasia. Fukuda et al.14 examined cervical biopsy specimens of nonpregnant women with dysplasia that had spontaneously regressed and from women with persistent dysplasia; they found that women with persistent dysplasia exhibited significantly lower numbers of Langerhans cells and T-helper cells in the biopsy specimens. These cells are intrinsic to the inflammatory wound healing process. Therefore it follows that an enhanced localized immune response after mechanical cervical dilatation and parturition might account for the increased regression of dysplasia that we see associated with vaginal delivery.10, 14 All women with cervical dysplasia diagnosed antepartum should undergo postpartum colposcopy with cytologic or histologic evaluation before proceeding with therapy. This is especially important among women with HGSIL who are delivered vaginally because of the high spontaneous regression rate of cervical dysplasia after vaginal delivery. Optimally, colposcopic evaluation should be performed 8 to 12 weeks post partum when most of the local inflammatory reactions and reparatory processes have resolved. Traumatic cervical desquamation and stimulation of local immune factors associated with vaginal delivery could play an important role in the spontaneous regression of cervical dysplasia in the postpartum period. Future prospective clinical and immunologic studies, such as immunohistochemical analysis of inflammatory infiltrates, are needed to confirm these theories. REFERENCES
1. Pfenninger JL. Colposcopy in a family practice residency: the first 200 cases. J Fam Pract 1992;34:67-71. 2. Kohan S, Beckman EM, Bigelow B, Klein SA, Douglas GW. The role of colposcopy in the management of cervical intraepithelial neoplasia during pregnancy and postpartum. J Reprod Med 1980;25:279-84. 3. Benedet JL, Selke PA, Nickerson KG. Colposcopic evaluation of abnormal Papanicolaou smears in pregnancy. Am J Obstet Gynecol 1987;157:932-7. 4. Ostergard DR, Nieberg RK. Evaluation of abnormal cervical cy-
1120 Ahdoot et al.
5.
6.
7.
8.
9.
tology during pregnancy with colposcopy. Am J Obstet Gynecol 1979;134:756-8. Bertini-Oliveira AM, Keppler MM, Luisi A, Tarico S, Lopes VLD, Delascio D, et al. Comparative evaluation of abnormal cytology, colposcopy and histopathology in preclinical cervical malignancy during pregnancy. Acta Cytol 1982;26:636-44. Economos K, Perez Veridiano N, Delke I, Collado ML, Tancer ML. Abnormal cervical cytology in pregnancy: a 17 year experience. Obstet Gynecol 1993;81:915-8. DePetrillo AD, Townsend DE, Morrow CP, Lickrish GM, DiSaia PJ, Roy M. Colposcopic evaluation of the abnormal Papanicolaou test in pregnancy. Am J Obstet Gynecol 1975;121:441-5. Kiguchi K, Bibbo M, Ilasegawa T, Kurihara S, Tsutsui F, Wied GIN. Dysplasia during pregnancy: a cytologic follow-up study. J Reprod Med 1981;26:66-72. Ueki M, Ueda M, Kumagai K, Ikamoto Y, Noda S, Matsuoka M.
June 1998 Am J Obstet Gynecol
10.
11. 12.
13. 14.
Cervical cytology and conservative management of cervical neoplasias during pregnancy. Int J Gynecol Pathol 1995;14:63-9. Yoonessi M, Wieckowska W, Mariniello D, Antkowiak J. Cervical intra-epithelial neoplasia during pregnancy. Int J Gynaecol Obstet 1982;20:111-8. Patsner B. Management of low-grade cervical dysplasia during pregnancy. South Med J 1990;83:1405-12. Kaplan AL, Kaufman RH. Diagnosis and management of dysplasia and carcinoma in situ of the cervix in pregnancy. Clin Obstet Gynecol 1967;10:871-8. Hall JE, Walton L. Dysplasia of the cervix: a prospective study of 206 cases. Am J Obstet Gynecol 1968; 100:662-71. Fukuda K, Hachisuga T, Nakamura S, Matsuo N, Twasaka T, Sugimori H. Local immune response in persistent cervical dysplasia. Obstet Gynecol 1993;82:941-5.
Bound volumes available to subscribers Bound volumes of the American Journal of Obstetrics and Gynecology are available to subscribers (only) for the 1998 issues from the publisher, at a cost of $116.00 for domestic, $148.73 for Canada, and $139.00 for international for Vol. 178 (January-June) and Vol. 179 (July-December). Shipping charges are included. Each bound volume contains a subject and author index, and all advertising is removed. Copies are shipped within 60 days after publication of the last issue in the volume. The binding is durable buckram with the Journal name, volume number, and year stamped in gold on the spine. Payment must accompany all orders. Contact Mosby, Inc., Subscription Services, 11830 Westline Industrial Drive, St. Louis, MO 63146-3318, USA; phone (800)453-4351 or (314)453-4351. Subscriptions must be in force to qualify. Bound volumes are not available in place of a regular Journal subscription.