Post-LACC era: critical assessment not “all-or-none” is needed

An all-or-none approach has been adopted by many since the landmark LACC trial, which unexpectedly reported higher recurrence and mortality rates in patients randomized to a minimally invasive surgical (MIS) approach for radical hysterectomy (RH) [1]. We have watched with great interest as subsequent retrospective studies that have reported similar findings have been touted as “confirmatory” and highlighted as plenary presentations at national and international meetings and published in high-impact journals. It is also interesting to see that retrospective studies with contradictory findings have not been highlighted and considered of low quality by many [2].

In the current issue, Li et al. [3] report comparable oncologic outcomes between laparoscopic RH (LRH) and open RH (ORH) among patients with 2009 FIGO stage IB1 cervical cancer and tumor size <2 cm. This multi-institutional retrospective cohort study included 1,484 patients (LRH, n=585 vs. ORH, n=899) -- among the largest to date. It is a sound, high-quality retrospective study, which suffers from the inherent limitations of any retrospective study. There were no observed differences in disease-free survival (DFS) and overall survival (OS) between the two groups, both unmatched and after appropriate propensity score matching. There was not even a “trend” that would lead many to suggest that this study suffered from an underpowering issue. Also, the number of cases included were way more than that of the LACC trial, which further likely negates any concern about power. The authors also performed subgroup analyses according to the visibility of tumor and found that LRH was not associated with recurrence and mortality rates both in the visible tumor and invisible tumor subgroups. Of course, there are limitations, as this study may only apply to the surgeons who contributed cases, the method of follow-up is debatable, and assessment of PFS in a retrospective fashion is quite challenging and heterogeneous. But, whether someone has died and the date of death is not so challenging when determining OS if patients have appropriate follow-up. It is important to remember that randomized controlled trials (RCTs) only include a small fraction of the entire cohort of patients with a particular disease or who are undergoing a procedure. RCTs are the highest level of discovery but are also quite limited by the inability to include all eligible cases, and their external validity is critical. Well-executed retrospective studies have the potential to report outcomes on all cases and not just a sampling. The current one is such an example.

A number of worldwide retrospective studies have been published since we asked for continued critical assessment on this issue [4]. Many studies have reported similar concerning findings to those of the LACC trial, but others have not. Kim et al. [5] reported that LRH did not influence disease recurrence in patients with 2009 FIGO stage IB1 disease with a cervical mass ≤2 cm in their early single-institution retrospective cohort study [5]. They suggested this subgroup as safe candidates for LRH. However, these studies differ in terms of tumor size measurement. While Li et al. [3] (current study) used clinical or pathologic tumor size, Kim et al. [5] used preoperative magnetic resonance imaging (MRI) to determine cervical mass size. Such a difference might originate from the wide utility of MRI in Korea. This brings up an important critique of the current study and others. What is the best way to assess tumor size? Postoperative tumor size does not always reflect preoperative tumor size assessments. Decisions to operate are based on preoperative factors and not final pathologic findings.

In contrast to the above and current study, in their multi-institution retrospective cohort study Uppal et al. [6] reported that patients with tumor ≤2 cm on final pathology also exhibited significantly inferior DFS when RH was performed by MIS. In this analysis, however, cases with no residual tumor were excluded, while 2009 FIGO stage IA1 (lymphovascular space invasion positive), IA2, and IB1 cases were included. Interestingly, Uppal et al. [6] reported that conization before surgery was associated with lower recurrence risk, similar to what was reported by Kim et al. [5].

In the SUCCOR study, an international European retrospective cohort study, patients with 2009 FIGO stage IB1 cervical cancer who had a tumor <4 cm on preoperative MRI were included, whereas those who received conization before surgery were excluded [7]. Inverse probability weighting was performed and showed that MIS RH was associated with higher recurrence and mortality rates compared to those of ORH. However, in a subgroup of patients with a tumor ≤2 cm, similar survival outcomes were observed between MIS RH and ORH.

Why does a cervical tumor size matter? Tumor size is relevant in many cancers, not just cervical cancer. The ability to properly contain and handle tumors as they increase in size is a challenge. This is further complicated when MIS approaches are used. Tumor containment and handling must be approached using basic oncologic principles whenever tumor is still present and there is potential for tumor spillage and unwanted contamination no matter what the surgical approach used. This might be supported by a protective effect of preoperative conization on recurrence, suggested by many retrospective studies. Moreover, the use of uterine manipulators during MIS RH further appears to promote tumor breakdown and spillage. The SUCCOR study noted that patients who underwent MIS RH without using a manipulator seemed to have a similar recurrence rate compared to those who underwent ORH [7]. Additionally, increasing tumor size is correlated with the presence or absence of pathologic high-risk factors, such as parametrial extension, and extracervical and nodal metastases. The LACC trial and many retrospective studies, including the current one, did not utilize routine preoperative MRI. Improper patient selection may also have factored into the results noted by LACC and other studies.

It is important that results of well-executed retrospective large cohort studies continue to be highlighted and not simply negated because they do not support the LACC findings. An all-or-none approach is not the way to make decisions for patients. Not all patients with cervical cancer should undergo surgery, not all should have MIS, and not all surgeons should perform MIS. Similarly, not all patients should be opened. It seems extreme to us to perform an ORH in a patient with cervical cancer after a cone has been performed and has negative margins, while at the same time, if fertility is desired, a cone alone may suffice and MIS safe if performing a radical trachelectomy. This continued critical assessment is needed and provides greater equipoise as we continue to enroll patients on much needed additional RCTs, such as the RACC trial (ClinicalTrials.gov Identifier: NCT03719547). Before we abandon MIS, however, we need to await the pending data.

We are living in a post-LACC era. What are we to do? Because of MIS’s many advantages, although LACC also suggested no advantage to MIS, it is difficult to abandon all MIS RH in the management of early-stage cervical cancer. The LACC trial is a landmark study, and we must learn from it and critically assess why the results were what they were. We need to refine our selection process as well as technique and return to core surgical oncologic principles. Tumor spillage should be avoided and is completely a surgeon- and technique-related event. Köhler et al.'s ‘transvaginal closure technique of vaginal cuff’ [8] and Kanao et al.'s ‘no-look no-touch technique’ [9] are great examples of how to prevent tumor spillage, as are other methods. We also must be more careful when selecting cases for a possible MIS approach. A simple pelvic examination and clinical tumor assessment to decide primary surgical treatment, not just MIS, are not acceptable anymore. MRI, or high-quality ultrasound by experienced sonographers, must be a part of the initial preoperative assessment of these patients.

The potential advantages of MIS, as seen in many other procedures, should not be abandoned because of the unexpected findings of the LACC trial. We hope that MIS RH continues to be critically assessed with thoughtful discussion. Ongoing RCTs and future ones are needed and should be strongly supported. There is enough equipoise to conduct such RCTs and enroll patients.