Is There Still a Place for Hysterectomy for Treatment of Recurrent Squamous Intraepithelial Lesions?

Introduction

Asquamous intraepithelial lesion (SIL) of the uterine cervix is a premalignant lesion that can be divided further into low-grade SIL (LSIL) and high-grade SIL (HSIL). In cases of HSIL, surgical treatment is indicated because of a higher likelihood of progression to malignant disease. ¹ Excision options for treatment include cold–knife cervical conization, large loop excision of the transformation zone (LLETZ) or hysterectomy. Hysterectomy had been recommended as a primary mode of treatment for cervical intraepithelial lesions (CIN) until 1970. Later, less-invasive treatment options were introduced. ² Today, hysterectomy as a primary mode of treatment still appears to be controversial. Some researchers propose that, with hysterectomy, there is a higher possibility of complete resection of the lesion. Other express concerns about cervical lesions extending to the vagina, where lesions might be hidden in the vaginal cuff scar after the procedure and, thus, could be overlooked, or they believe hysterectomy should only be performed when the existence of invasive cancer is excluded with certainty. ³

Compared to other techniques, hysterectomy requires general anesthesia. It has a higher complication rate and prolonged duration of hospitalization. ² Hysterectomy is limited in younger patients, who desire to preserve their reproductive functions. It is a possible method of choice in postmenopausal women with coexistent uterine pathology or in those who have finished their family planning. Hysterectomy can be performed in cases when other excision methods are not achievable due to shortened cervices. ⁴ A minimally invasive approach via a vaginal (total vaginal hysterectomy; TVH) or a laparoscopic (total laparoscopic hysterectomy; TLH) route is preferable. When this is not feasible, abdominal (total abdominal hysterectomy; TAH) approach can be used. The goal of treatment, regardless of the method, is to excise the abnormal tissue fully in order to prevent progression of the lesion to cancer.^5,6

The objective of the current study was to determine if colposcopy and biopsy could predict the extent of the premalignant lesion in the uterine cervix and if hysterectomy is an appropriate method of treatment in cases, where LLETZ could not be performed due to previous surgical procedures on the uterine cervix.

Materials and Methods

This was a retrospective study that involved patients who had undergone previous excisional techniques (e.g., LLETZ or conization) in which hysterectomy was performed as a final mode of treatment because of recurrent SIL in the authors' department for gynecologic oncology and breast oncology, at the Clinic for Gynecology and Perinatology, at the University Medical Centre Maribor, in Maribor, Slovenia, between 2010 and 2018.

Written informed consent was obtained from the patients for use of their medical records for research purposes. The research was performed in accordance with the ethical standards of the Declaration of Helsinki.

At the University Medical Center Maribor, the method of choice for treatment of SIL is LLETZ. Hysterectomy (minimally invasive or abdominal) is offered in cases when LLETZ cannot be performed due to some patients having shortened uterine cervices. The indications for surgical treatment were in concordance with Slovenian national guidelines for LLETZ: HSIL; recurrent LSIL for 2 years; or pathologic Papanicolaou (Pap) smear with or without a biopsy; and/or when an adequate colposcopy is not feasible due to an incompletely visible transformation zone. Human papilloma virus (HPV) testing was also performed according to Slovenian national guidelines.

All procedures were considered with a curative intent. Preoperative assessment of each patient included a mandatory colposcopy with a Pap smear and biopsy (punch biopsy and/or endocervical curettage in cases of transformation zone types 2 or 3). Transvaginal ultrasound examination was performed to evaluate if a laparoscopic approach was possible.

For each patient, the following data were obtained: age at the diagnosis; histologic and/or cytologic diagnosis before the procedure; surgical margin status after previous LLETZ; histopathologic diagnosis after the procedure; type of hysterectomy; treatment success (defined as negative surgical margins of hysterectomy specimen on histopathologic examination); and possible complications of the procedure. Complications were divided into early (≤ 30 days after the procedure) and late (> 30 days after the procedure). Data were analyzed using IBM's SPSS Statistics Program, version 25.0, and descriptive statistics were calculated.

Results

Fifty patients, all after previous LLETZ, were identified using the current authors' inclusion criteria. One patient was excluded from the analysis because her pre and postoperative data were lost. The median patient age was 56.5 years (Q₁ = 50; Q₃ = 65).

Preoperative cytologic and histopathologic diagnosis were obtained in 43/49 (88%) patients; 7 of whom were diagnosed with LSIL and 36 of whom were diagnosed with HSIL. In 6/49 (12%) women, only cytologic diagnosis were available preoperatively.

In patients with LSIL, preoperative diagnosis matched the postoperative histopathologic diagnosis in 6/7 (86%) cases. In 1/7 (14%) cases, postoperative diagnosis was upgraded to HSIL. In patients with HSIL, preoperative diagnosis matched the postoperative histopathologic diagnosis in 17/36 (47%) cases. The diagnosis was upgraded in 1/36 (3%) patients, in whom microinvasive carcinoma was found, and downgraded in 18/36 (50%) patients. In the group of patients in which the HSIL diagnosis was downgraded, no dysplasia was found in 10/18 (56%).

In 6 patients, the indication for the procedure was pathologic Pap smears, excluding LSIL and HSIL. According to the Bethesda system terminology, the cases included were:

2 cases of atypical squamous cells—cannot exclude HSIL (ASC-H)

Among these 6 cases, only 1 patient was had LSIL postoperatively, whereas, in the other 5 cases, no lesions were found on postoperative histopathologic examinations (Table 1).

Altogether, preoperative histologic diagnoses were obtained for 43 patients. Sensitivity of preoperative biopsy and histopathologic examination to detect HSIL was 97.7%, with a specificity of 25%, a positive predictive value (PPV) of 50%, and a negative predictive value (NPV) of 85.7% (Table 2). As mentioned previously, 1 case of microinvasive carcinoma was detected postoperatively in a patient who was preoperatively diagnosed with an HSIL lesion.

Results of HPV testing were available for 36/49 (74%) patients, 31/36 (86%) of who were positive. All patients with preoperative LSIL were HPV-positive. In 6/36 (14%) women who were HPV-negative, only 1/6 (17%) had HSIL postoperatively. Sensitivity of preoperative HPV testing to predict HSIL was 92.9%, with a specificity of 21.7%, a PPV of 41.9% and an NPV of 83.3%.

Regarding the surgical margins after previous LLETZ, 19/49 (38%) patients had negative surgical margins and 11/49 (22%) had positive surgical margins. In 1/49 (2%) case, the surgical margin was thermally too damaged to allow for an accurate histopathologic evaluation. Margin evaluation was not possible in 5/49 (10%) women and was not available for 14/49 (28%) patients.

As an operative technique, TLH was used in 23/49 (47%) cases, TVH in 16/49 (33%) cases, laparoscopically assisted vaginal hysterectomy (LAVH) in 6/49 (12%) cases, and TAH in 4/49 (8%) cases. Forty-three of 49 (88%) patients had no complications in the postoperative timeperiod. Early complications occurred in 6/49 (12%) patients (3 cases of vaginal wound dehiscence and 3 cases of bladder lesions). A late complication (a postoperative hernia) was found in 1/49 (2%) patient. In 48/49 (98%) patients, the lesions were completely removed, and in 1/49 (2%) case, surgical treatment was not successful. In this latter case, the final pathologic report showed a synchronous cervical cancer (International Federation of Gynecology and Obstetrics [FIGO] IB) and an endometrial (FIGO II) cancer (Table 3).

Discussion

Hysterectomy is the most-radical form of cervical dysplasia treatment. It can be considered in patients who have finished their family planning, are postmenopausal, have coexistent uterine pathology, or have shortened uterine cervices due to previous procedures, such as LLETZ or conization.^3,4

TVH and TLH are considered to be minimally invasive approaches. The advantages of these procedures are lower blood loss, fewer transfusions, faster recovery time, shorter hospital stay, decreased risk of wound infections, and better quality of life. ⁷ Although, according to a study performed in 2017, the rate of performed TLH surpassed the rate of TVH for the first time, ⁸ and different researchers still believe that the latter is the method of choice whenever possible, because it has been shown that it has better outcomes, compared to other techniques. Although recovery time and duration of hospitalization are comparable to the laparoscopic approach, TVH advantages are shorter operative times and lower rates of vaginal cuff dehiscence. The limitations are the shape and size of the vagina and the accessibility of the uterus.^5,6 A vaginal approach might be difficult in a nulliparous woman with a narrow suprapubic arch or in a patient with an enlarged uterus. In addition, uterine mobility might be compromised by leiomyomas or adhesions. ⁷ If TVH cannot be performed, a laparoscopic approach is chosen. ⁶

The most-common contraindication for TLH is abdominal adhesions from previous surgeries or an obstructive uterus. TLH is not advised in cases of suspected malignant disease if the uteri cannot be removed intact. ⁷ If minimally invasive surgery cannot be performed, the procedure of choice is TAH. It is also possible to convert to the abdominal approach if the less-invasive method fails during surgery.^5,6 The risk factors for conversion include obesity, findings of multiple or thick fibrous adhesions, uterine segments >5 cm, and an increased uterine width of >10 cm. ⁷ The current study results were similar to the results of other studies, when implementation of a minimally invasive approach (especially TLH) was associated with a decrease of TAH and even TVH. ⁸ In this current study, the minimally invasive approach was taken in 92% of the cases, with TLH as a preferred method, and the abdominal approach was taken in 8% of patients.

Before treatment, Pap smears were obtained and colposcopy with or without biopsy was performed. As already described in the literature, the Pap smear as a screening method has high specificity (86%–100%) but a low sensitivity (30%–87 %), whereas colposcopy with biopsy is a subjective method and its quality depends on the competence and experience of the colposcopist. Specificity of the latter method is relatively low (45%–90%), but its sensitivity is higher (70%–98%).^9,10 Considering the low sensitivity of the Pap smear and the low specificity of colposcopy, under- or overdiagnosing can occur, although it is important to notice that both cytology and colposcopy sensitivity tends to increase with the severity of the lesion. ¹¹

Although this study had a limitation of a small sample size, the findings were in concordance with data from the literature. In the current series of patients, colposcopy with biopsy had a very high sensitivity (97.7%) but a lower specificity (25%) for detection of an HSIL lesion. HPV testing detects most but not all of the high-risk HPV subtypes. Compared to the Pap smear, HPV testing has a higher sensitivity (64%–97%), but a slightly lower specificity (56%–93%). ¹² In the current study, HPV status was obtained for 76% of the cases. The main reason for this relatively low percentage is that the timespan of the analysis was quite long, and HPV testing was not used nor was it available at the beginning of this research. However, the current authors have to emphasize that obtaining HPV status is crucial, especially because of its high NPV.

It is important to emphasize the significance of precise preoperative diagnostic procedures, including careful colposcopic evaluation, acquisition of adequate biopsy specimens, and a thorough histologic examination of the cervical tissue in order to make the correct initial diagnosis. In their study, Kesic et al. described that finding unexpected invasive cervical cancer in hysterectomy specimens seems significantly higher, compared to conization procedures and that one must be careful to avoid hysterectomy as the primary mode of treatment for women who do not desire further reproduction, postmenopausal women, and women with coincident uterine pathology. ³

It was also stated, that hysterectomy as a mode of treatment for HSIL should be performed only if the existence of invasive cancer was excluded with certainty. ³ To achieve this goal, it is of uttermost importance to adhere to the guidelines for the management of women with cervical cytologic abnormalities. Gynecologic surgeons should be aware that finding invasive cancer during a simple hysterectomy means that the disease was inadequately surgically treated, which carries psychologic, physical, and social consequences that could arise from further treatment. ³ This is especially important in postmenopausal women, women with shortened cervices, and women with severe cervical stenosis, as exclusion of invasive cancer can be difficult in these patients. Therefore, some researchers advise obtaining more specimens from suspicious lesions. ¹³

In the current study, in 1 patient with a preoperative diagnosis of HSIL, unexpected microinvasive cancer was found postoperatively. However, as seen from the study results, colposcopy with adequate inspection of the transformation zone of the uterine cervix can be difficult in menopausal women who have undergone 1 or more excisional procedures. Moreover, cervical stenosis can impede adequate dilatation of the cervical canal to perform endocervical curettage in some cases. Benign diseases, such as atrophy or inflammation, must also be taken into account because they can disrupt the histologic evaluation of the specimen. In the current study, when biopsy could not be performed and indication for the procedure was made solely on the basis of a Pap smear (excluding HSIL), a lesion was found on postoperative histopathologic examination in only 1 of six cases. Even in this 1 case, the indication for the procedure could be debatable given that the lesion was classified as LSIL only.

Regarding margin status, it is still not known if it is the most important prognostic factor for residual disease. For example, a study by Kim et al. discovered that 92 of 172 women who underwent hysterectomy following LLETZ showed residual tumors in subsequent hysterectomy specimens, mostly carcinoma in situ (98%), but 2 cases of microinvasive squamous-cell carcinomas were also identified. ¹⁴ Most residual tumors were present in resection margin–positive samples (98%), but were rarely seen in resection margin–negative samples. Yet, more than one-third of women with resection margin–positive samples had no signs of residual tumors in subsequent hysterectomy specimens.

Kim et al. also found out that, in patients with positive LLETZ margins, negative endocervical biopsy and the absence of a bulbous growth or central necrosis can be used as selection criteria for women who can be monitored closely and potentially avoid additional surgery. ¹⁴ Regarding the surgical margins after previous LLETZ in the current study, 19/49 (38%) patients had negative surgical margins and 11/49 (22%) had positive surgical margins.

Despite its low mortality rate, hysterectomy has higher risks and complication rates compared to LLETZ.^5–7,15,16 The most-common complications after LLETZ are infections and bleeding, and, less frequently, uterine-cervix injury, pelvic pain, and stenosis of the cervical canal. ¹⁵ Complications after hysterectomy vary, depending on the surgical approach and technique. The most-common complications are infections, hemorrhage, venous thromboembolism, ureteral injury, gastrointestinal injury, bladder lesion, nerve injury, and vaginal cuff dehiscence. ¹⁶ In the current study, 6 patients experienced early complications. Vaginal cuff dehiscence occurred in 3 patients; all were after TLH. This is similar to the results of other studies, which have shown that vaginal cuff dehiscence is more common after a laparoscopic approach. ⁶ Lesion of the bladder occurred in 2 cases, both after TLH. One patient had both dehiscence of the vaginal cuff and a lesion of the bladder after TVH. There was 1 case of a late complication, namely, a postoperative hernia after LAVH.

Conclusions

Because of its invasiveness, hysterectomy should be used with care in cases of premalignant diseases of the uterine cervix. If possible, less-invasive procedures, such as excision techniques, should be chosen; yet, this might not be possible in patients who have undergone previous surgical procedures on the uterine cervix. Although colposcopy with biopsy has a very high sensitivity for HSIL diagnosis, the specificity is low. Thus, gynecologic surgeons should aim to improve preoperative diagnostics in combination with HPV testing, and collect adequate biopsy specimens with subsequent careful histologic examination. In addition, hysterectomy should be performed only after the existence of invasive cancer has been excluded with certainty. If a surgeon opts for hysterectomy, a minimally invasive approach is the method of choice.