Adjuvant Hysterectomy for Cervical Cancer Patients Treated with Chemoradiation Therapy: A Systematic Review on the Pathology-Proven Residual Disease Rate

Simple Summary The treatment for patients with locally advanced cervical cancer generally consists of platinum-based chemotherapy during external beam radiotherapy, followed by brachytherapy. Some medical centers perform adjuvant hysterectomy after chemoradiation therapy, even though the international guideline advises otherwise. Performing adjuvant hysterectomy after chemoradiation therapy is associated with a high complication rate and the percentage residual disease in adjuvant hysterectomy specimen is unknown. Therefore, the aim of our systematic review was to determine the percentage of residual disease in the adjuvant hysterectomy specimen. Furthermore, we want to determine if there is an association between the time of adjuvant hysterectomy and the percentage residual disease in adjuvant hysterectomy specimens. Findings from this research provide insight into potential complications, survival benefits, and an overtreatment rate. Patients need to be well informed before considering an adjuvant hysterectomy. Abstract Objective: To determine the incidence of pathology-proven residual disease in adjuvant hysterectomy specimens in patients with cervical cancer, treated with chemoradiation therapy. Secondly, to assess a possible association for pathology-proven residual disease regarding the time between chemoradiation therapy and adjuvant hysterectomy. Additionally, the survival rate and complication rate were assessed. Methods: PubMed, EMBASE, and the Cochrane database were searched from inception up to 8 March 2021. Results: Of the 4601 screened articles, eleven studies were included. A total of 1205 patients were treated with chemoradiation therapy and adjuvant hysterectomy, ranging from three to twelve weeks after chemoradiation therapy. A total of 411 out of 1205 patients (34%) had pathology-proven residual disease in the adjuvant hysterectomy specimen. There was no association found in the time between chemoradiation therapy and adjuvant hysterectomy. Follow-up ranged from 2.4 to 245 months, during which 270 patients (22%) relapsed, and 298 patients (27%) were deceased. A total of 202 (35%) complications were registered in 578 patients. Conclusion: there is no association found in the time between chemoradiation therapy and residual disease on adjuvant hysterectomy specimens. The survival rates after chemoradiation therapy and adjuvant hysterectomy are suboptimal, while the risk of complications after adjuvant hysterectomy is high.


Introduction
Cervical cancer is the fourth most common malignancy in patients worldwide, within 2018, an estimated number of 570,000 patients were diagnosed with cervical cancer [1,2]. Approximately 30-40% of patients with cervical cancer in developed countries have locally advanced cervical cancer at initial diagnosis [3]. Locally advanced cervical cancer is classified as stage IIB-IVA according to the International Federation of Gynecology and Obstetrics (FIGO) 2018 staging system [2].
According to international guidelines, the standard treatment for patients with locally advanced cervical cancer is concurrent chemoradiation therapy. This treatment generally consists of platinum-based chemotherapy during external beam radiotherapy, followed by brachytherapy [4]. There is still a reported overall local pelvic recurrence rate of 14.5% after treatment with chemoradiation therapy for patients with stage IIB-IVA disease [5,6]. Reported 5-year overall survival rates range from 83% for stage IB to 32% for stage IVA [7]. Even though the international guidelines advise otherwise, some medical centers do perform adjuvant hysterectomy after chemoradiation therapy. Adjuvant hysterectomy after chemoradiation therapy has not been shown to be effective in improving survival outcomes, but it seems to reduce the risk of local recurrence [8,9].
If a patient has received chemoradiation therapy, the morphologic response continues after chemoradiation therapy. Therefore, the total benefit of chemoradiation therapy is achieved for several weeks after the last radiation treatment [10]. Medical centers performing routine adjuvant hysterectomy are inconsistent in the time between chemoradiation therapy and adjuvant surgery. However, the time between chemoradiation therapy and adjuvant hysterectomy has not been assessed before [4,9]. Performing adjuvant hysterectomy after chemoradiation therapy is associated with a high complication risk because of radiation-induced tissue damage and reduced propensity to healing [11,12]. This may negatively influence the quality of life of patients.
The aim of this review is to determine the incidence of pathology-proven residual disease in adjuvant hysterectomy specimens in patients treated with chemoradiation therapy because of cervical cancer. Furthermore, we aimed to assess a possible association with time between chemoradiation therapy and adjuvant hysterectomy. Secondary outcome measures were survival rate and complications after an adjuvant hysterectomy.

Protocol and Registration
A protocol was designed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guideline and registered in PROSPERO (registration number CRD42020196399). In this systematic review, adjuvant hysterectomy was defined as planned surgery after primary treatment with chemoradiation therapy, without determining the presence or absence of residual disease. Adequate chemoradiation therapy is defined as a total dose radiation therapy of at least 70 Gy (external beam radiation therapy and brachytherapy). Preferable, with daily external beam radiotherapy total dose of 45 Gy with concomitant platinum-based chemotherapy (recommended weekly cisplatin 40 mg/m 2 for 5-6 cycles) followed by brachytherapy.

Newcastle-Ottawa Scale
Of the total group of eleven studies, ten studies were assessed as good or fair quality and one study was assessed as poor quality according to the Newcastle-Ottawa Scale for cohort studies (Table 2).

Newcastle-Ottawa Scale
Of the total group of eleven studies, ten studies were assessed as go and one study was assessed as poor quality according to the Newcastle cohort studies (Table 2).

Newcastle-Ottawa Scale
Of the total group of eleven studies, ten studies were as and one study was assessed as poor quality according to th cohort studies (Table 2).

Newcastle-Ottawa Scale
Of the total group of eleven stu and one study was assessed as poor cohort studies (Table 2). Cancers 2021, 13, x 7 of 13

Newcastle-Ottawa Scale
Of the total group of eleven studies, ten studies were assessed as good or fair quality and one study was assessed as poor quality according to the Newcastle-Ottawa Scale for cohort studies (Table 2).

Newcastle-Ottawa Scale
Of the total group of eleven studies, ten studies were assessed as go and one study was assessed as poor quality according to the Newcastle cohort studies (Table 2).

Newcastle-Ottawa Scale
Of the total group of eleven studies, ten studies were as and one study was assessed as poor quality according to th cohort studies (Table 2).

Newcastle-Ottawa Scale
Of the total group of eleven stu and one study was assessed as poor cohort studies (Table 2). Cancers 2021, 13, x

Newcastle-Ottaw
Of the total gro and one study was cohort studies (Tab

Newcastle-Ottawa Scale
Of the total group of eleven studies, ten studies were assessed as good or fair quality and one study was assessed as poor quality according to the Newcastle-Ottawa Scale for cohort studies (Table 2).

Newcastle-Ottawa Scale
Of the total group of eleven studies, ten studies were as and one study was assessed as poor quality according to th cohort studies (Table 2).

Newcastle-Ottawa Scale
Of the total group of eleven studies, ten studies were assessed as good or fair quality and one study was assessed as poor quality according to the Newcastle-Ottawa Scale for cohort studies (Table 2).

Newcastle-Ottawa Scale
Of the total group of eleven studies, ten studies were as and one study was assessed as poor quality according to th cohort studies (Table 2).

Newcastle-Ottawa S
Of the total group and one study was ass cohort studies (

Newcastle-Ottawa Scale
Of the total group of eleven studies, ten studies were assessed as good or fair quality and one study was assessed as poor quality according to the Newcastle-Ottawa Scale for cohort studies ( Table 2).

Newcastle-Ottawa Scale
Of the total group of eleven studies, ten studies were as and one study was assessed as poor quality according to th cohort studies ( Table 2).

Newcastle-Ottawa Scale
Of the total group of eleven studies, ten studies were assessed as good or fair quality and one study was assessed as poor quality according to the Newcastle-Ottawa Scale for cohort studies (Table 2).

Newcastle-Ottawa Scale
Of the total group of eleven studies, ten studies were assessed as go and one study was assessed as poor quality according to the Newcastle cohort studies (Table 2).

Newcastle-Ottawa Scale
Of the total group of eleven studies, ten studies were as and one study was assessed as poor quality according to th cohort studies (Table 2).

Newcastle-Ottawa Scale
Of the total group of eleven stu and one study was assessed as poor cohort studies (Table 2).

Newcastle-Ottaw
Of the total gro and one study was cohort studies (Tab  [19] Cancers 2021, 13, x 7 of 13

Newcastle-Ottawa Scale
Of the total group of eleven studies, ten studies were assessed as good or fair quality and one study was assessed as poor quality according to the Newcastle-Ottawa Scale for cohort studies (Table 2).

Newcastle-Ottawa Scale
Of the total group of eleven studies, ten studies were assessed as go and one study was assessed as poor quality according to the Newcastle cohort studies (Table 2).

Newcastle-Ottawa Scale
Of the total group of eleven studies, ten studies were as and one study was assessed as poor quality according to th cohort studies (Table 2).

Newcastle-Ottawa Scale
Of the total group of eleven stu and one study was assessed as poor cohort studies (Table 2).

Newcastle-Ottawa S
Of the total group and one study was ass cohort studies (

Newcastle-Ottawa Scale
Of the total group of eleven studies, ten studies were assessed as good or fair quality and one study was assessed as poor quality according to the Newcastle-Ottawa Scale for cohort studies (Table 2).

Newcastle-Ottawa Scale
Of the total group of eleven studies, ten studies were assessed as go and one study was assessed as poor quality according to the Newcastle cohort studies (Table 2).

Newcastle-Ottawa Scale
Of the total group of eleven studies, ten studies were as and one study was assessed as poor quality according to th cohort studies (Table 2).

Newcastle-Ottawa Scale
Of the total group of eleven stu and one study was assessed as poor cohort studies (Table 2). Cancers 2021, 13, x 7 of 13

Newcastle-Ottawa Scale
Of the total group of eleven studies, ten studies were assessed as good or fair quality and one study was assessed as poor quality according to the Newcastle-Ottawa Scale for cohort studies (Table 2).

Newcastle-Ottawa Scale
Of the total group of eleven studies, ten studies were assessed as go and one study was assessed as poor quality according to the Newcastle cohort studies (Table 2).

Newcastle-Ottawa Scale
Of the total group of eleven studies, ten studies were as and one study was assessed as poor quality according to th cohort studies (Table 2).

Newcastle-Ottawa Scale
Of the total group of eleven stu and one study was assessed as poor cohort studies (Table 2).  [22] Cancers 2021, 13, x 7 of 13

Newcastle-Ottawa Scale
Of the total group of eleven studies, ten studies were assessed as good or fair quality and one study was assessed as poor quality according to the Newcastle-Ottawa Scale for cohort studies (Table 2).

Newcastle-Ottawa Scale
Of the total group of eleven studies, ten studies were assessed as go and one study was assessed as poor quality according to the Newcastle cohort studies (Table 2).

Newcastle-Ottawa Scale
Of the total group of eleven studies, ten studies were as and one study was assessed as poor quality according to th cohort studies (Table 2).

Newcastle-Ottawa Scale
Of the total group of eleven stu and one study was assessed as poor cohort studies (Table 2). Cancers 2021, 13, x

Newcastle-Ottawa S
Of the total group and one study was ass cohort studies (Table 2  Cancers 2021, 13, x 7 of 13

Newcastle-Ottawa Scale
Of the total group of eleven studies, ten studies were assessed as good or fair quality and one study was assessed as poor quality according to the Newcastle-Ottawa Scale for cohort studies (Table 2).

Newcastle-Ottawa Scale
Of the total group of eleven studies, ten studies were assessed as go and one study was assessed as poor quality according to the Newcastle cohort studies (Table 2).

Newcastle-Ottawa Scale
Of the total group of eleven studies, ten studies were as and one study was assessed as poor quality according to th cohort studies (Table 2).

Newcastle-Ottawa Scale
Of the total group of eleven stu and one study was assessed as poor cohort studies (Table 2).

Newcastle-Ottawa S
Of the total group and one study was ass cohort studies ( * Representativeness of the exposed cohort: all included studies representative for women with residual cervical cancer after chemoradiation therapy treated with salvage surgery. ** Ascertainment of exposure: all with database or medical records. *** Assessment of outcome: all with medical records. **** Follow-up period ≥ 12 months was assessed as long enough for outcomes to occur.

Newcastle-Ottawa Scale
Of the total group of eleven studies, ten studies were a and one study was assessed as poor quality according to th cohort studies (Table 2).

Analysis
Eleven studies were included, representing a total of 2077 patients of which 1205 patients (58%) were treated with chemoradiation therapy and adjuvant hysterectomy. Patients were included with an age range of 19-90 years. Of this group, 955 (79%) patients were diagnosed with squamous cell carcinoma, 158 (13%) with adenocarcinoma, 37 (3%) with adenosquamous carcinoma and 55 (5%) with another histological subtype. Adjuvant hysterectomy was performed between three and twelve weeks after chemoradiation therapy. Residual disease in the pathology specimen was found in 34% of the patients. All studies performed adjuvant hysterectomy at different time points after chemoradiation therapy. No association was found for pathology-proven residual disease regarding the time between chemoradiation therapy and adjuvant hysterectomy ( Figure 2). Sun (2013) [23] NA NA months (range 60-245 months) No statement about lost to follow-up Fair * Representativeness of the exposed cohort: all included studies representative for women with residual cervical cancer after chemoradiation therapy treated with salvage surgery. ** Ascertainment of exposure: all with database or medical records. *** Assessment of outcome: all with medical records. **** Follow-up period ≥12 months was assessed as long enough for outcomes to occur. : fulfilled the criteria of the Newcastle Ottawa Scale for this item.

Analysis
Eleven studies were included, representing a total of 2077 patients of which 1205 patients (58%) were treated with chemoradiation therapy and adjuvant hysterectomy. Patients were included with an age range of 19-90 years. Of this group, 955 (79%) patients were diagnosed with squamous cell carcinoma, 158 (13%) with adenocarcinoma, 37 (3%) with adenosquamous carcinoma and 55 (5%) with another histological subtype. Adjuvant hysterectomy was performed between three and twelve weeks after chemoradiation therapy. Residual disease in the pathology specimen was found in 34% of the patients. All studies performed adjuvant hysterectomy at different time points after chemoradiation therapy. No association was found for pathology-proven residual disease regarding the time between chemoradiation therapy and adjuvant hysterectomy (Figure 2).  Recurrence of disease after treatment with chemoradiation therapy and adjuvant hysterectomy was observed in 270 patients (22%). Recurrence was pathology-proven in the study of Sun et al. In all other studies the way of determining the recurrence of disease is unknown. The survival rate is based on ten studies and 1115 patients, of which 298 patients (27%) deceased because of cervical cancer and 28 patients (3%) deceased because of causes unrelated to cervical cancer. The follow-up period is described in ten studies and the patients were followed during a range of 2.4 and 245 months (Table 1).
A total of 202 complications related to adjuvant hysterectomy were registered in 578 patients treated with chemoradiation therapy and adjuvant hysterectomy (35%). A total of 27 fistulas (5%) was reported, of which sixteen fistulas to the urinary tract and eleven fistulas to the gastro-intestinal tract. A total of 53 complications of the urinary tract was reported (9%) five injuries to the urinary tract and five ureteral stenoses. Six complications to the female genital tract were reported (1%), two patients with vaginal necrosis and two with vaginal stenosis. Eighteen gastro-intestinal complications were reported (3%) of which two intestinal injuries and a total of thirteen infections was reported (2%) of which seven with an abscess, four with peritonitis and two patients with pelvic infections. Other complications were reporter 81 times, of which four deaths related to postoperative morbidity, four ruptures of the iliac vessel during surgery and one pulmonary embolism (Table 3).

Discussion
This systematic review shows a mean pathology-proven residual disease in 34% (range, 14-52%) of all patients undergoing adjuvant hysterectomy after chemoradiation therapy for cervical cancer. No association was found between the time between chemoradiation therapy and for pathology-proven residual disease on adjuvant hysterectomy specimens. The survival rate after chemoradiation therapy and adjuvant hysterectomy seem suboptimal, and adjuvant hysterectomy has a high risk of complications, including death.

Comparison with Existing Literature
In this systematic review, the percentage of residual disease on adjuvant hysterectomy specimens did not diminish in time. In the included studies, adjuvant hysterectomy was performed within twelve weeks after chemoradiation therapy. However, we believe it might be plausible that the incidence of positive pathology could decrease because the morphologic response sometimes continues for several weeks after the last treatment of chemoradiation therapy [10]. In addition to this hypothesis, Eifel et al. reported 22% local recurrences for patients with stage III cervical cancer after treatment with only chemoradiation therapy [5]. Which is lower than the mean 34% residual disease in the surgical specimen in our review. This may indicate that when (salvage) surgery is performed too early after chemoradiation therapy patients are overtreated. Thereby, two previously published studies assessed residual disease detection by biopsy. The study of Boers et al. shows that residual disease detected by biopsy eight to ten weeks after chemoradiation therapy is a poor prognostic factor to identify patients with residual disease who may be salvaged by surgery [24]. The study by Hoeijmakers et al. implies that a biopsy to prove residual disease should be taken twelve until sixteen weeks after completing chemoradiation therapy, to select patients for salvage surgery [25]. These studies indicate that detection of residual disease should not take place early after chemoradiation therapy.
As previously mentioned, two recent meta-analyses determined the survival rates for adjuvant hysterectomy after chemoradiation. They concluded that adjuvant hysterectomy seemed to reduce the risk of local recurrence [8,9]. One limitation of these studies is that the results are based on only two randomized controlled trials [26,27]. The review of literature in our study shows an overall recurrence rate (distant and local recurrence) of 22% after adjuvant hysterectomy and chemoradiation therapy. Previous studies reported an overall recurrence rate after treatment with only chemoradiation therapy of up to 37% [5,6]. We reported an overall survival of 73% after chemoradiation therapy and adjuvant hysterectomy, which seems to be equivalent to treatment with chemoradiation therapy alone. Previous studies reported 5-year overall survival rates for patients with stage IB-IVA ranges between 32% and 83% [7]. A hypothesis for the decreased recurrence rate after adjuvant hysterectomy and chemoradiation therapy, compared to chemoradiation therapy alone, is that patients with residual disease after chemoradiation therapy are directly treated with salvage surgery. Patients with residual disease after chemoradiation therapy are an unfavorable group of patients, which are likely to relapse. However, standard adjuvant hysterectomy after chemoradiation therapy harbors the risk that the majority of patients are unnecessary treated and thus exposed to the (post)operative risks. Besides, more recent studies use a higher total dose of radiation therapy, which results in lower reported local recurrence rates. For example, the retroEMBRACE study used a mean D90 (EQD210) for high and intermediate risk clinical target volumes of 87 ± 15 Gy and 69 ± 8 Gy respectively. They reported 9.6% of local failures, of which 6.4% had true local recurrences [7]. This indicates even less reason to perform an adjuvant hysterectomy.
Additionally, adjuvant hysterectomy after chemoradiation therapy is associated with a high complication risk. The risk of complications after adjuvant hysterectomy is higher in comparison with primary surgery because adjuvant hysterectomy takes place in a previously irradiated area with the unpredictable healing quality of the tissue [28]. In our systematic review adjuvant hysterectomy is associated with a percentage of major complications of 35%. The mortality rate after adjuvant hysterectomy in our systematic review is 0.7%. The study of Magrina et al. reported a mortality rate of 0.5% after primary radical hysterectomy [29]. A major complication with high implications for patients is the development of a fistula. Fistulas can occur due to irradiation, surgery, or a combination of both [30]. The study of Hilton et al. reported an overall fistula rate after a hysterectomy of 0.13% and a fistula rate of 1% after radical hysterectomy for patients with cervical cancer [31]. In this review the studies reported an overall fistula rate of 5%, this implicates that the chance of developing fistulas increases after the combination of chemoradiation therapy and adjuvant hysterectomy. Most of the complications after adjuvant hysterectomy turn out to be unnecessary because 66% of the patients were overtreated with adjuvant hysterectomy. Patients should be clearly informed on the survival benefits and the disadvantages like complications and the over-treatment rate of non-routinely adjuvant hysterectomy. To reduce the number of patients needed to harm by adjuvant hysterectomy, less invasive procedures, such as radiological imaging or biopsies, should be performed first, to determine the presence of residual disease before performing surgery. However, the best period to determine residual disease by radiological imaging, biopsy, as well as performing adjuvant or salvage surgery, should be further investigated.

Strengths and Weaknesses
One of the strengths of this study is that the dose and regimen of chemoradiation therapy that patients received were assessed by a radiation oncologist, this ensures that all patients were treated via a comparable chemoradiation therapy treatment schedule, similar to high income countries. However, the evaluation of data is not without limitations, there is much heterogenicity among the included studies. First, the time to complete chemoradiation therapy was not an inclusion criterion, which could influence survival outcomes. Second, not all studies performed the same type of adjuvant hysterectomy after chemoradiation therapy. Differing between simple hysterectomy and extended hysterectomy with or without lymph node dissection, and between open versus laparoscopic surgery. This could influence the complication rate. Analyzing the different surgery types regarding recurrence, survival and complications was not possible because of missing data. Third, the preoperative doses of chemoradiation therapy given to patients were nearly always less than the planned dose. Fourth, pathological residual disease rate can be underestimated because the viable microscopic residual disease can be missed in surgical specimens. Fifth, it was not possible to calculate the 2-, 3-or 5-years overall survival rate as the data was not representative, which could overestimate or underestimate the survival rate.

Conclusions
The majority of patients treated with adjuvant hysterectomy as a routine procedure are over-treated. No association was found between the time between chemoradiation therapy and pathology-proven residual disease on adjuvant hysterectomy specimens. The survival rate after adjuvant hysterectomy seems suboptimal, and the risk of major complications after adjuvant hysterectomy is high. To prevent high overtreatment rates by adjuvant hysterectomy it should be considered to use less invasive procedures to determine the presence of residual disease before performing surgery after chemoradiation therapy. Patients need to be well informed on potential complications, survival benefits, and a potential overtreatment rate before considering an adjuvant hysterectomy.

Conflicts of Interest:
The authors declare no conflict of interest.