Minimally invasive right colectomy with transrectal natural orifice extraction: could this be the next step forward?

Minimally invasive surgery for colorectal disease is a global trend. Over the past 30 years, minimally invasive colorectal surgery has been shown to cause less postoperative pain, earlier return of bowel function, shorter hospital stays, and fewer wound complications than open surgery [1, 2]. In addition to providing more favorable short-term outcomes than conventional surgery, the long-term outcomes of minimally invasive colectomy and those of open surgery are comparable [2‐6]. However, in colorectal surgery, the specimen is extracted through a mini-laparotomy, which entails a 3–8-cm incision depending on the size of specimen. Abdominal incision wounds may negate many of the benefits of minimally invasive surgery [7]. Thanks to the advances made in minimally invasive colorectal surgery, natural orifice specimen extraction (NOSE) surgery can prevent the mini-laparotomy wound.

NOSE surgery was first published in early 1990s. In 1991 and 1992, reports by Stewart et al. [8] and Nezhat [9], respectively, described the extraction of a colectomy specimen through the vagina. Franklin et al. first described colectomy with specimen extraction through the anus in 1993 [10]. Left-side colectomy by the transrectal NOSE method is safe and feasible in some patients; the short-term outcomes include less postoperative pain and a shorter hospital stay than the conventional method. Furthermore, the long-term oncological outcomes are comparable to those of conventional mini-laparotomy specimen extraction [11‐13]. Extracting a specimen of the right-sided colon from the resection wound of the colon using a colonoscope is challenging because of the anatomically narrow and tortuous shape of the sigmoid colon. Eshuis’ case series reported specimen extraction through colotomy; however, the extraction failed in two of ten patients because of the bulk of the specimen [14]. Karagul reported that only approximately two-thirds of the unselected laparoscopic colectomy patients were suitable for NOSE. The success rate of the NOSE method was lower in male than in female patients, and also lower for large than for small tumors [15]. Because of technical difficulty, use of NOSE is limited. Transvaginal NOSE has remained the most commonly used path to extract specimens of the right colon [16‐18]. However, transvaginal NOSE is limited to female patients and sexual dysfunction after vaginal incision may cause concern. There is currently no scientific literature on the use of transrectal NOSE for extracting specimens of the right colon.

In this study, the short-term outcomes in patients who had undergone minimally invasive right hemicolectomy by the transrectal NOSE approach and those in patients who underwent the surgery by the conventional mini-laparotomy specimen extraction were compared.

We enrolled 297 patients (151 males, mean age 64.9 ± 12.8 years) who underwent minimally invasive right hemicolectomy. In these 297 patients, 272 patients had conventional surgery with specimen extraction through mini-laparotomy and 25 patients had NOSE surgery (23 transrectal, 2 transvaginal). The demographic data of these patients are presented in Table 1. The two groups did not differ significantly in terms of age, sex, BMI, presence of medical illness (including hypertension, cardiac disease, cardiovascular accident, asthma, diabetes mellitus, peptic ulcer, hepatitis, liver cirrhosis, gallstone disease, and thyroid problems), and preoperative laboratory data (hemoglobin levels, white blood cell [WBC] counts, percentages of segmented WBC, serum albumin levels, blood urea nitrogen levels, creatinine levels, aspartate aminotransferase levels, and total bilirubin levels). The NOSE group had a higher rate of robotic surgery (12.0% vs. 3.3%, p = 0.035) and a higher rate of intracorporeal anastomosis (100.0% vs. 52.6%, p < 0.001) than the conventional group. The diagnosis of malignant colon disease did not differ significantly between the two groups (malignancy rate in conventional vs. NOSE, 89.3% vs. 88.0%, p = 0.836). Among the patients with malignancy (conventional vs. NOSE, 243 vs. 22), the tumors were significantly larger in the conventional than in the NOSE group (conventional vs. NOSE, 4.4 ± 2.2 cm vs. 3.4 ± 1.6 cm, p = 0.007).

The postoperative short-term outcomes are listed in Table 2. The operating time did not differ significantly between the two groups (conventional vs. NOSE, 248.0 ± 78.3 min vs. 247.8 ± 84.4 min, p = 0.988). Furthermore, the blood loss during surgery did not differ significantly between the two groups. The rates of surgery combined with resection of other involved organs were similar in these two groups (conventional vs. NOSE, 9.6% vs. 8.0%, p = 0.799). The overall postoperative morbidity rate did not differ significantly between the two groups although the conventional group had a higher morbidity rate than the NOSE group (conventional vs. NOSE, 12.9% vs. 4.0%, p = 0.194). In the subgroup of postoperative complications, the two groups did not differ significantly in any of the postoperative variables (wound, pulmonary, cardiovascular, urinary, gastrointestinal, abdominal, and anastomosis). There was no deep or organ space surgical site infection in the NOSE group. None of the patients complained about anal bleeding, anal pain, and fecal or gas incontinence after the NOSE procedure. The reoperation rate and readmission rate did not differ significantly between the two groups. Postoperative mortality rates were comparable (p = 0.761) in the conventional group (1 patient, 0.4%) and NOSE group (0 patient). The mean follow-up time of the NOSE group was 13.36 months (range 1–25) months. None of the patients in the NOSE group had rectal tumor seeding during the follow-up period.

Postoperative clinical information is presented in Table 3. Laboratory data checked on postoperative day (POD) 3 did not exhibit significant differences in WBC counts, percentages of segmented WBC, and C-reactive protein (CRP) levels between the two groups. The time of first flatus was comparable in the two groups (POD 1.8 ± 0.7 vs. POD 2.4 ± 1.4, p = 0.066). The NOSE group had earlier bowel movements (POD 3.0 ± 1.2 vs. POD 4.2 ± 2.0, p < 0.001), tolerance to liquid diet (POD 2.6 ± 1.1 vs. POD 4.3 ± 2.9, p = 0.004) and tolerance to soft diet (POD 4.5 ± 2.5 vs. POD 6.1 ± 3.3, p = 0.020) than the conventional group. The postoperative hospital stay was significantly shorter in the NOSE group (mean 5.2 ± 2.8 days, median 5 days, range 3–17 days) than in the conventional group (mean 8.3 ± 5.1 days, median 7 days, range 3–45 days) (p = 0.004).

Figure 2 shows the difference in scores on the NRS on 3 consecutive days after operation in the two groups. The NRS scores of patients without PCA (conventional 263, NOSE 24) are shown in Fig. 2a and the scores of the patients with PCA (conventional 9, NOSE 1) are shown in Fig. 2b.The conventional group exhibited significantly higher NRS scores on POD1 and POD2 than the NOSE group (POD1, conventional vs. NOSE: 4.5 ± 1.8 vs. 3.6 ± 2.0, p = 0.026; POD2, conventional vs. NOSE: 3.3 ± 1.5 vs. 2.6 ± 1.0, p = 0.002) (Fig. 2a). The NRS did not differ in POD1-3 between the two groups (Fig. 2b).

Minimally invasive surgery for colorectal disease has been widely accepted; this surgery causes smaller abdominal wounds than open surgery. However, a mini-laparotomy wound is necessary to remove specimen. To reduce wound-related complications and achieve superior cosmetic results, the NOSE procedure was developed. The advantages and disadvantages of conventional and NOSE surgery for right colectomy are summarized in Table 4. Some studies have reported that laparoscopic right hemicolectomy with transvaginal specimen retrieval is feasible with favorable short-term surgical outcomes [16‐18]. Transanal extraction of the sigmoid colon and rectum specimens also has acceptable short-term and long-term outcomes [11, 13]. In this study, we mainly performed minimally invasive right hemicolectomy using the NOSE method to extract the specimen transrectally. The transvaginal method can also be used in the NOSE procedure; however, colpotomy incision repair is more difficult than rectal repair. An incision was made longitudinally over the upper rectum, and suture repair was quite simple. In addition, transrectal specimen removal can be performed in male and female patients and can avoid adverse effects on sexual function. Some researchers have expressed concern about the bacteriological safety of the transrectal method because the rectum is opened for specimen retrieval. We observed that no significant postoperative morbidity and changes in laboratory data are observed if the rectum is cleaned properly. To our knowledge, our study is the first evaluating short-term surgical outcomes of minimally invasive right hemicolectomy performed using transrectal NOSE.

Avoidance of the mini-laparotomy wound for specimen extraction is one of the most crucial features of NOSE. Compared with conventional minimally invasive surgery, NOSE causes less postoperative pain, results in faster recovery, and provides superior cosmetic results. A previous study demonstrated favorable short-term surgical outcomes after performing laparoscopic right hemicolectomy using the transvaginal NOSE [18]. Patients who had undergone NOSE experienced less pain and required shorter hospital stays than those who had undergone conventional surgery, without significant differences in surgical morbidity. Two randomized clinical trials that compared the short-term operative outcomes in patients with left-sided colonic disease [19, 20]. They demonstrated that NOSE group experienced less wound pain and had a lower wound infection rate than the conventional group. In our study, no significant differences were observed in the operation time, blood loss, and postoperative morbidity between the two groups. The postoperative pain scores did not differ significantly in the two groups among patients with PCA. As regards patients without PCA, the NOSE group experienced less wound pain than the conventional group. The benefit of NOSE was observed on POD 1 and POD2. Previous studies have reported time to flatus passage of approximately 2.7–3 days and time to resumption of a regular diet of approximately 4–4.6 days for patients who had undergone transvaginal NOSE [17, 18]. Faster bowel recovery and earlier food intake were observed in the NOSE group than in the conventional group. Early ambulation and less use of analgesic agents because of less postoperative pain may be the reasons. This can help to shorten the hospital stay. Surgical site infection occurs after conventional laparoscopic right hemicolectomy in 5–7% of cases [21, 22]. None of our patients who had NOSE experienced wound-related complications.

Bacterial contamination is always a concern during the NOSE procedure. We strongly suggest that mechanical bowel preparation, intraoperative transanal lavage with povidone iodine solution, transluminal wound protector, and prophylactic antibiotics are applied to reduce the bacterial load [23]. Recently, a study showed that the risk of bacterial contamination with NOSE was not significantly higher than that in conventional laparoscopic surgery [24]. In our study, patients who had NOSE did not experience significant postoperative morbidity or laboratory data changes, such as leukocytosis, or CRP level elevation, than the conventional group. None of our patients have had rectal wound-related complications or leakage thus far.

Tumor size is considered before applying the NOSE procedure. Many authors limit indications to tumors smaller than 3 [25, 26], 4 [27, 28], 5 [29, 30], 6 [31], or 6.5 [32] cm. The average tumor size in the NOSE group in our study was 3.4 cm, and it was significantly smaller than that in the conventional group. Some authors have stated that obese patients are not suitable for transrectal specimen extraction and set the BMI cutoff at > 28 kg/m² [32], > 30 kg/m² [29], or > 35 kg/m² [33]. In our study, although no significant difference in patients’ BMI was seen in the two groups, the highest BMI was 32 kg/m² in the NOSE group and 40 kg/m² in the conventional group. Patients with a bulky mesocolon, a narrow pelvis, and previous pelvic surgery with severe adhesions were not eligible for NOSE.

This study has some limitations. First, the retrospective analysis of prospectively collected data might have caused some selection bias. Notably, the NOSE group is a highly selective patient group, which is not comparable to the conventional group. Second, the sample size is relatively small in the NOSE group, which might have resulted in a lack of statistical power. Third, this study only reports on short-term outcomes and lacks long-term oncologic outcome follow-up; however, the use of the NOSE procedure in the left colon is well established.

Transrectal NOSE can be performed in some patients who require minimally invasive right hemicolectomy with postoperative short-term outcomes that are comparable to those of conventional laparoscopic surgery. NOSE is associated with less postoperative wound pain, faster bowel recovery, and shorter hospital stay than the conventional method. Additional prospective studies with larger patient populations and longer follow-up are warranted.