Cervical Dysplasia and Cancer in Pregnancy - 6
First, there are risks to ANY surgical procedure during pregnancy and though relatively small under these circumstances, must be balanced against the benefits. Because of the normal increases in blood flow to the uterus and cervix that occur during pregnancy, significant hemorrhage may occur even when a simple cervical biopsy is performed. The risk of bleeding is proportional to the area biopsied and when a large ‘biopsy’ such as a ‘cold knife conization’ is performed the amount of bleeding can be considerable both during the procedure and the post-operative period. In our case, we chose to perform a cervical cerclage as a means of both minimizing the risk of bleeding and providing some protection from premature labor under these circumstances and controlled additional bleeding using electrocautery after the specimen was obtained.
Other providers may choose to perform the entire conization, or even simply a wider biopsy of the TZ, during pregnancy using a LEEP (loop electrosurgical excision procedure) as a means of minimizing bleeding. Still others prefer the use of the CO2 laser, but the downside of these latter approaches is the ‘surgical margins’ may be obscured due to heat damage (thermal injury) and that can complicate interpretation of the presence and the depth of invasion. All of these procedures have the potential to increase the risk of premature rupture of membranes, premature delivery, intrauterine infection, and even pregnancy loss during the pregnancy, but these are surprisingly uncommon occurrences, particularly if the surgical procedures are performed prior to 24 weeks.
There are some long-term risks to both fertility and pregnancy success when cervical conization has been performed. We have known for years that scarring after extensive conization, particularly if heavy cauterization of deeper tissues was required, or infectious complications followed the procedure, can lead to stenosis (narrowing) of the cervical canal and even complete occlusion in some cases. This could impair fertility by preventing the transit of sperm into the uterus or increase the risk for obstetrical complications (such as failure to dilate in labor, lacerations, or even cesarean sections).
These procedures also “shorten” and distort the normal configuration of the endocervical canal and destroy endocervical glands that produce the mucous that is involved in the ‘capacitation of sperm’ (important for efficient ‘fertilization’) and the protection of the uterine cavity from ascending infections. Pregnancy consequences of these results are summarized in a metaanalysis of 27 studies last year by Kyrgiou and colleagues (Lancet 2006;367:489-98) in which pregnancy outcomes following cervical conization procedures were evaluated. These investigators found that prior cervical conization was associated with increased risk (generally 2- to 3-fold) in subsequent pregnancies for premature rupture of membranes, preterm delivery, low birth weight, and cesarean delivery. High conizations can disrupt the internal cervical os and lead to cervical incompetence.
Of course, there are also risks of incomplete evaluation of high-grade intraepithelial disease and microinvasive cervical cancer, the greatest being that there might be more extensive disease present than previously thought. If you recall, the patient featured in our case report had “multifocal areas of carcinoma in situ and microinvasion.” Such high-grade disease may encompass a wide surface area and it is not uncommon, even in experienced hands, to find disease in the surgical margins of a conization specimen (indicating incomplete resection) when the final pathology report returns. A recent study by Phongnarisorn (Int J Gynecol Cancer 2006;16:655-9) actually reviewed 129 hysterectomy specimens from women who had cervical conizations with residual high-grade disease or microinvasive disease in the original surgical margins. Of these, 57 (44.2%) had residual high-grade intraepithelial lesions and 20 (15.5%) residual invasive carcinoma (18 were microinvasive and 2 were frankly invasive). Fortunately, with truly ‘microinvasive’ disease, there is only about a 1% chance of having positive lymph node involvement and no more than about 0.2% chance of actually dying from cervical cancer if appropriate therapy is instituted and the patient continues in regular follow-up.
If a high-grade intraepithelial lesion is found in pregnancy with no evidence of invasion based on a thorough evaluation (colposcopy, biopsies, and conization if deemed necessary) by an experienced provider, there is fairly low risk of progression to significant invasive disease during the course of the pregnancy. Robova and colleagues (Eur J Gynaecol Oncol 2005;26:611-4) assessed persistence, progression, and regression of both low-grade and high-grade intraepithelial disease diagnosed during pregnancy and 24 months of follow-up after delivery. Of the sixty-two women with HSIL, they observed complete regression in 14 (22.6%), regression to LSIL in 17 (27.4%), persistence in 25 (40%), and progression to microinvasive disease in 6 (9.7%). Obviously, with the high risk of persistence of HSIL and progression of HSIL to invasive disease, aggressive and ongoing follow-up after delivery is indicated and this should be accompanied by a low threshold for ‘definitive therapy.’
Women will frequently ask if a cesarean delivery is indicated for high-grade intraepithelial lesions. They are usually asking two questions contained within one: Does a vaginal delivery increase the risk of disease progression? And, could I possibly give my disease to my baby? Answering the last question first, I usually tell them that high-grade lesions are the end-result of an HPV infection, but usually by the time HSIL is present, the virus has gained the ability to control cell growth, but that control is at the expense of losing the capacity to produce intact and infectious viral particles that could be passed on to other individuals. With regard to the first question, there has actually been some evidence presented in years past that a vaginal delivery actually increases the chance of disease regression. Recently, Kaneshiro and colleagues (Am J Obstet Gynecol 2005;192:1452-4) reviewed retrospectively data on 201 pregnant women with abnormal Pap tests. Regression rates for vaginal and cesarean section groups for LSIL were 58% vs 42%, respectively, and for HSIL, 53% vs 25%. Although the latter did not reach statistical significance, we can safely say that vaginal delivery certainly does not increase the risk of progression in women with HSIL.