Challenges and perspectives on less invasive surgery for early-stage cervical cancer: a critical analysis of the SHAPE trial and its implications

Cervical cancer represents a global public health issue, particularly burdening many low-income and middle-income countries. Effective interventions, such as vaccination against the human papillomavirus and screening, render cervical cancer a preventable disease [1]. By implementing these interventions, the incidence of cervical cancer can be reduced, simultaneously enhancing the quality of life for cancer survivors, particularly those in the early stages. Currently, the standard treatment for cervical cancer is surgery with radical hysterectomy (RH), which can lead to complications such as urinary incontinence.

The SHAPE trial results have been published in the New England Journal of Medicine [2], generating enthusiasm among patients with early-stage cervical cancer and gynecologic oncologists. The trial demonstrated that less invasive surgery is non-inferior to standard surgery for low-risk patients. The trial results demonstrate that a lot of patients with early-stage cervical cancer can survive with a good quality of life in clinical practice. However, I am composing this article due to certain concerns regarding this trial.

First, there are concerns related to the quality assurance of the surgical procedures.

This trial excels in terms of external validity, specifically in terms of study generalization. However, doubts arise when considering whether the trial was conducted with adequate internal validity.

A crucial issue related to this trial is the quality assurance of the trial intervention, i.e., surgical therapy. To what extent was quality assurance implemented for RH or surgical therapy? Upon reviewing the trial results, in the RH group, there were 9 cases (2.7%) with positive vaginal margins, which is higher than the expected occurrence rate. The study included low-risk patients alone based on preoperative assessment. Considering the frequency of positive vaginal margins when performing RH on the eligible trial participants, questions arise regarding the quality assurance of surgical therapy. For quality assurance, an audit of surgical therapy is crucial, and central review through photographs (or videos) should be considered mandatory. In particular, did the trial manager ensure that the components deemed essential for RH, as outlined in the trial protocol, were appropriately performed in the standard treatment group? According to the protocol, the following procedures are specified for RH: the en block removal of the uterus, cervix, medial one-third of parametria, 2 cm of the uterosacral ligaments, and upper 1–2 cm of the vagina, the ligation of uterine arteries lateral to the ureters, and the unroofing of ureters from the ureterovesical junction. If these techniques were not adequately conducted, the trial could not evaluate its original research question, which means the trial may inadvertently compare “hysterectomy-mimicked RH” to simple hysterectomy. In the SHAPE trial, only photographic images of the surgical specimen were requested for quality assurance [2], which may not be sufficient. Considering these factors, although the external validity is somewhat lacking, it might have been prudent to conduct the trial by limiting participating facilities and surgeons to enhance the internal validity.

Secondly, there are concerns about the primary endpoint and non-inferiority margin.

The primary endpoint was changed from pelvic relapse-free survival to pelvic recurrence rate at 3 years (PRR3) due to the very low event rate. However, the choice of the primary endpoint as PRR3 raises concerns. Can we assert the establishment of surrogacy for the essential ultimate goal of cancer treatment, i.e., overall survival, based on PRR3? It is questionable whether sufficient observation has been conducted for cases with an observation period <3 years. Furthermore, the outcomes may vary depending on the number of patients lost within the 3-year period and how they were labeled in the process of analysis. Additionally, extra-pelvic recurrence was reported in only 2 cases (0.6%) in the RH group compared to 7 cases (2.0%) in the simple hysterectomy group. It is unclear whether it is clinically acceptable for extra-pelvic recurrence to be occur in only 0.6% of patients with standard treatment (RH), but 7 cases undergoing the less invasive approach of simple hysterectomy. This warrants discussion even though the primary point (PRR3) was reached.

Moreover, there are concerns regarding the non-interiority margin in this trial. The non-inferiority margin in this non-inferiority trial is set at 4%. On the other hand, the PRR3 in each group is 2.52% for the simple hysterectomy group and 2.17% for the RH group. With only approximately 2% event occurrence in the population, a non-inferiority margin of 4% seems excessively large. Is it possible that both the setting of the primary end point and the non-inferiority margin were too favorable for the researchers?

Third, there are concerns regarding the application of the minimally invasive surgery (MIS: laparoscopic and robotic surgery) in this trial.

In the LACC trial, non-inferiority of MIS compared to open surgery for stage IA2–IB1 cervical cancer (according to International Federation of Gynecology and Obstetrics [FIGO] 2008) could not be demonstrated [3]. Therefore, globally, when performing surgery for early-stage cervical cancer, open surgery is considered the gold standard [4,5]. However, in the SHAPE trial, a majority of cases underwent surgery using MIS (79.9% in the simple hysterectomy group and 69.5% in the RH group), leading to confusion in interpreting the results.

In conclusion, the patient population eligible for simple hysterectomy may be even more limited than the subjects of the SHAPE trial.

At the 2023 American Society of Clinical Oncology meeting, another confirmatory trial aimed at evaluating the effectiveness of less invasive surgery for stage IB1 cervical cancer was reported—the Japan Clinical Oncology Group (JCOG) 1101 trial [6]. The JCOG1101 trial, conducted in Japan from 2012 to 2017, focused on cases with a preoperative diagnosis of cervical cancer with a size of <2 cm and examined the non-inferiority of modified RH (mRH) compared to RH. The results demonstrated the non-inferiority of mRH with a 5-year overall survival rate of 98.2% (90% confidence interval [CI]=96.0–99.2) as the primary endpoint and a 5-year relapse-free survival rate of 94.7% (95% CI=90.8–96.9) as the secondary endpoint. Additionally, the occurrence rate of postoperative late complications, such as urinary incontinence and urinary retention, were reported as 1.3% and 8.9%, respectively, which were lower than those reported for traditional RH.

A comparison table between the SHAPE and JCOG1101 trials is presented. The main difference between the 2 trials lies in the eligibility criteria and surgical procedures (Table 1).

Table 1. Comparison table between the SHAPE and JCOG1101 trials.

Characteristics		SHAPE trial	JCOG1101 trial
Design		Phase III	Phase III (single arm)
Sample size		700	225
MIS		Permitted	Not permitted
Eligibility criteria
	Stage (FIGO 2008)	IA2–IB1	IB1
	Tumor diameter	Less than 2 cm	Less than 2 cm
	Stromal invasion	Less than 50% on MRI	NA

FIGO, International Federation of Gynecology and Obstetrics; MIS, minimum invasive surgery (laparoscopic/robotic surgery); NA, not applicable; JCOG, Japan Clinical Oncology Group.

Significant differences in eligibility criteria between the 2 trials include the disease stage and depth of stromal infiltration. With regard to the disease stage, the SHAPE trial allowed the inclusion of stage IA2, whereas the JCOG1101 trial did not. With regard to the depth of stromal infiltration, the SHAPE trial considers stromal infiltration as <50%, whereas the JCOG1101 trial allows inclusion even if infiltration >50%. In other words, the JCOG1101 trial involved a patient population with a higher risk of recurrence. For such patient populations, the JCOG1101 trial adopted mRH as the surgical approach for less invasive surgery. Indeed, in the postoperative pathological findings of excised organs, the SHAPE trial reported 25 cases (3.7%) with lymph node metastases (11 cases in the simple hysterectomy group and 14 cases in the RH group), whereas the JCOG1101 trial reported 16 cases (7.1%).

MRH is considered equivalent to class II (Piver classification) [7] or type B (Querleu classification) RH [8] and is identical to the original Westheim operation method [9,10]. The mRH is a hysterectomy method considered as intermediate between simple hysterectomy and RH, allowing for almost complete preservation of the pelvic splanchnic nerves and hypogastric plexus. This surgical procedure involves cutting the anterior leaf, rather than the posterior one, of the vesico-uterine ligament of the uterus. The ureter is mobilized laterally, a portion of the parametrial tissue is transected at the level of the ureteral tunnel, and at least 10–15 mm of the vaginal wall from uterine cervix is removed. The favorable efficacy of mRH for FIGO IB1 patients in the low-risk group mentioned above has also been reported retrospectively [11,12,13,14], and the surgical procedure was considered the most appropriate for a clinical trial in Japan that tested non-inferiority by comparing it with RH (JCOG11101 trial).

Less invasive surgery for stage IB1 cervical cancer (according to FIGO 2018) is non-inferior to RH, as reported by multiple confirmatory trials. This should be offered to patients as a new treatment option in routine clinical practice. However, the selection of the ideal patient population for less invasive surgery and the choice of the appropriate surgical approach warrant further investigations.