In this analysis of consecutively operated patients with cervical carcinoma FIGO IA1 with LVSI to IIA2, preoperative cone biopsy was the strongest factor associated with reduced risk for recurrence. These data support the theory of the influence of intraoperative tumor spread during radical hysterectomy.
After publication of the LACC study, there was a dramatic change in the treatment of early cervical carcinoma [1, 4, 16, 17]. Even before guidelines recommended abdominal radical hysterectomy as the standard procedure, there was a decline in laparoscopic surgery for early cervical cancer [4]. The LACC trial showed significant inferiority of laparoscopic versus open surgery, with a reduction in disease-free survival from 96.5–86.0% after 4.5 years [1]. Similar results were shown in a recent meta-analysis by Nitecki et al, which stated that laparoscopic radical hysterectomy increased the likelihood of recurrence or death by 71% [18]. Unfortunately, they did not evaluate surgical techniques and use of uterine manipulators for their impact on survival. Important results of a retrospective international multicenter study were published by Chiva et al, which showed a disease-free survival at 4.5 years of 79% for laparoscopy and 89% for the abdominal approach [14]. Interestingly, this work demonstrated that the outcomes of laparoscopic surgery were better when no uterine manipulator was used (4.5 years disease-free survival 83% vs. 73%) or a vaginal cuff closure was performed (4.5 years disease-free survival 93% vs. 74%). This correlates with the results of a large patient series published by Köhler et al, in which excellent survival data (4.5 years disease-free survival 95.8%) were achieved for patients undergoing laparoscopic radical hysterectomy with vaginal colpotomy and vaginal cuff closure [11]. A similar approach was suggested by Kanao et al. [19]. Kong et al. hypothesized that intracorporal colpotomy is associated with a 3-fold decrease in disease-free survival [12]. Although patient and tumor characteristics are not completely comparable between those different studies, these results raise the hypothesis that outcomes of laparoscopic radical hysterectomy depend on surgical technique and the possibility of tumor cell spread into the intraperitoneal cavity [11, 12]. For example, tumor cell spread may occur during intracorporal colpotomy, when intravaginal tumor components have contact with the intraperitoneal cavity, as mechanistically demonstrated by our group [13]. In the present study, after laparoscopic surgery patients with tumors < 2 cm showed only one late recurrence after 12 years which is most likely not associated with possible intraoperative tumor cell spillage since all other local recurrences developed during the first 1.5 years after surgery.
In contrast to the results of the LACC study, which indicated impaired survival for all patients after laparoscopic surgery independently from tumor size, several other analyses comparing laparoscopic with open radical hysterectomy detected comparable results between radical laparoscopic hysterectomy and radical abdominal hysterectomy in patients with tumors < 2 cm [10, 20–24]. In a large analysis of patients treated by laparoscopic or open radical hysterectomy, tumor size > 2 cm was the only factor that characterized patients with increased risk of recurrence by laparoscopic surgery [25]. Of particular interest are the results of the SUCCOR study, in which all patients with preoperative cone biopsy were excluded, which leads to a high risk patient collective as shown in our study [14]. This analysis reported a comparably low 4.5 year disease-free survival of 79%. Interestingly, the group of patients in which potential tumor cell contamination was attempted to be reduced by protective measures achieved a significantly better 4.5 year disease-free survival of 93%. However, most of the aforementioned studies that suggested a reduced risk for small tumors did not evaluate preoperative cone biopsy.
Interestingly, in our study, of all patients who had received macroscopic tumor resection by preoperative cone biopsy, only one local recurrence was found. There was no recurrence in patients with completely resected tumors by cone biopsy. Preoperative cone biopsy was the only factor significantly associated with reduced risk for recurrences in multivariate analysis with an OR 5.90 (95% CI 1.11–31.29). These results correlate with the study of Casarin et al. in whose analysis a risk reduction of 71% (HR 0.29, 95%CI 0.13–0.91) was shown for patients who received preoperative cone biopsy [23]. Similar results were reported by Uppal et al [24]. These results raise the question whether there is a causal risk reduction by peroperative cone biopsy. Resection of all macroscopic visual tumor reduces the chances for tumor cell spillage during colpotomy. On the other hand, preoperative cone biopsy, as shown in our study, is associated with smaller tumors which might on the other hand be responsible for improved results. However, multivariate analysis indicates that preoperative cone biopsy is the strongest factor independently associated with risk for recurrence. In our cohort, there were only two patients with residual macroscopic tumor after cone biopsy. Therefore, we cannot conclude whether a macroscopically complete resection is necessary to achieve the optimal results.
In future studies the role of preoperative cone biopsy in order to reduce the visible tumor mass should be evaluated especially in laparoscopic surgery. In our cohort preventive surgical methods to reduce tumor spillage routinely consisted of vaginal colpotomy and non-use of uterine manipulator whereas the formation of a vaginal cuff was more rarely performed. Therefore, it has to be discussed whether in future studies protective surgical methods such as vaginal cuff closure might be omitted in patients with macroscopically resected tumor, since risk for recurrence is extremely low.
This is an exploratory analysis of a retrospective database. Therefore, it can only be used for the generation of hypotheses and the limitations of this analysis must be considered. The aim of this analysis was to assess the recurrences after radical hysterectomy in cervical carcinoma FIGO IA1 with LVSI and above and to evaluate the influence of preoperative cone biopsy believing that selected patients are still be eligible for laparoscopic surgery. We decided not to restrict tumor stage according to LACC inclusion criteria and also include patients with stage up to IIA2. Thereby, we wanted to avoid restricting the cohort too much by retrospective selection. Moreover, additional recurrences provide more data for analysis. More than 90% of all surgeries were performed by two experienced surgeons using standardized surgical technique which allows for a more reliable evaluation on the influence of patient and tumor characteristics compared to multicenter studies. Since there was a selection bias regarding the surgical approach in our cohort, this study was not mainly planned and powered to compare laparoscopic and open surgical approach.
Although the LACC trial did not demonstrate a significant quality-of-life benefit from a laparoscopic procedure, other studies suggest reduced complication rates, shorter hospital stays and lower costs [15, 26–29]. The question arises whether laparoscopic radical hysterectomy is a medically justifiable procedure in certain cases. Our data support the idea that the success of laparoscopic radical hysterectomy depends on the likelihood of intraoperative tumor cell contamination. There appears to be a subgroup of patients (no macroscopic tumor after cone biopsy, tumor diameter < 2 cm) with excellent survival and low risk for recurrence after laparoscopic radical hysterectomy which might still benefit from the advantages of laparoscopic surgery. This analysis supports the initiation of new studies examining laparoscopic radical hysterectomy under conditions that reduce the risk of tumor cell contamination.