Three-dimensional versus conventional two-dimensional laparoscopic colectomy for colon cancer: A 3-year follow-up study.

Background: Three-dimensional (3D) laparoscopy was developed to overcome the drawbacks of two-dimensional (2D) laparoscopy, namely lack of depth perception. However, the benefit of 3D laparoscopy in colorectal surgery is inconclusive. Here, we compare the 3-year follow-up outcomes of 3D and 2D laparoscopic colectomy. Patients and
Methods: A total of 91 consecutive patients who underwent either 3D or 2D laparoscopy colectomy from October 2015 to November 2017 by a single surgical team for colon cancer were enrolled. Data were collected from a prospectively constructed database, including clinico-pathological features and operative parameters. The pathological results, recurrence, survival and systemic treatment were collected from the Taiwan Cancer Database.
Results: There were 47 patients in the 3D group and 44 in the 2D group. There were no significant differences in characteristics of patients, operation data, pathological results, complications, operative time, blood loss or the number of lymph node harvested between the two groups. In addition, disease-free survival and overall survival were equal between the two groups. Conclusions: This is the first long-term result of a 3D laparoscopic colectomy. In our 3-year follow-up, there was no difference in long-term outcomes between 2D and 3D laparoscopy for colorectal surgery in an experienced centre.

INTRODUCTION

While laparoscopic surgery was developed >30 years ago,[1] laparoscopic-assisted right hemi-colectomy and sigmoid resection were first documented in 1991.[2] Today, the benefit of laparoscopic surgery in terms of blood loss and post-operative recovery has been demonstrated by meta-analyses. Furthermore, studies have shown that the oncological outcome of laparoscopic surgery was not inferior to that of open surgery.[3456] With continuous improvement of the camera and monitor system, laparoscopy provides a more precise view of the operative field; however, depth perception remained a drawback of the conventional system. Eyes measure depth in five ways: size difference on the retina from previous experience, depth from the defocus, overlapping of objects, object movement while moving parallax and stereo vision merging from two eyes. The first four methods can be used in conventional two-dimensional (2D) laparoscopic images. The last one comprises depth perception according to two different images through two eyes, called binocular disparity, which cannot be achieved with a 2D system. The three-dimensional (3D) laparoscopic system demonstrates depth perception by polarised 3D glasses, monitors and 3D cameras and is a significant advancement in laparoscopic surgery. Analysis of the short-term outcomes of colorectal cancer by 3D laparoscopy revealed a non-significant decrease in operation time and an increase in the number of lymph nodes (LNs) harvested compared with that of 2D laparoscopy.[789] However, the long-term oncological effect of the 3D laparoscopic system for colon cancer has not been discussed. To the best of our knowledge, this is the first long-term study to compare the oncological outcomes in patients who underwent elective colon surgery using 2D and 3D imaging systems.

PATIENTS AND METHODS

Patients and data collection

Patients who underwent elective laparoscopic surgery for colorectal cancer consecutively at Taipei Veterans General Hospital (TPEVGH) from October 2015 to November 2017 were reviewed retrospectively. Clinical information was prospectively obtained and recorded in a computerised database, including patient demographics (age, sex, comorbidities and history of previous abdominal surgery), operative parameters (operative method, operative time, blood loss and conversion rate) and tumour characteristics (tumour-node-metastasis stage, differentiation, LN harvest number and prognostic features). The history of chemotherapy, recurrence and survival status was based on data from the Taiwan Cancer Database (TCDB) from the Big Data Center (BDC). Patients with distant metastasis; adenoma or tumour in situ only; direct invasion of other organs (T4b); histology revealing no adenocarcinoma or mucinous adenocarcinoma; multiple primary tumours of the colon, tumours located in the rectum, transverse colon, splenic flexure, synchronous intra-abdominal surgery, previous colon surgery and other malignancies were excluded.

This study is based in part on data from the BDC, TPEVGH. The interpretations and conclusions contained herein do not represent the position of the TPEVGA. The clinical course and complications were collected by reviewing medical notes. Informed consent was not required because only medical files and databases were reviewed. The study design was approved by the Institutional Review Board of the (IRB-TPEVGH No.: 2020-03-004CC).

Surgical technique

All the patients were under the service of one single colorectal surgical team with two visiting staff members and up to three fellows. TPEVGH is a medical centre in Taiwan that performed 633 colorectal cancer surgeries in 2018, 81.5% of which were laparoscopic surgeries performed in large part by the surgical team involved in this study. The surgical procedure was the same in both the 2D and 3D groups. The operation was performed with four trocars: one 12-mm trocar at the para-umbilicus, one 5-mm trocar at the tumour side and a 5-mm and 12-mm trocar on the other side. In left-sided colon cancer, high ligation of inferior mesenteric artery (IMA) was routinely performed. The left-sided anastomosis was conducted with hand-sewn or circular stapler end-to-end anastomosis. In right-sided colon cancer, the root of the ileocolic vein was exposed. Part of the D3 LN was sampled, but D3 dissection was not routinely performed. The ileocolic anastomosis was performed with hand-sewn end-to-end anastomosis or intra-corporeal side-to-side anastomosis with endo-GIA in rare cases.

The 2D system was an Olympus 2D HD Imaging System (EVIS EXERA III Universal Imaging Platform and ENDOEYE FLEX 2D laparoscope, Olympus America, Center Valley, PA, USA), while the 3D system was an Olympus 3D HD Imaging System (EVIS EXERA III Universal Imaging Platform and ENDOEYE FLEX 3D laparoscope, Olympus America, Center Valley, PA, USA).

Statistical analysis

Differences between quantitative variables were analysed using Student's t-test. Differences in the distribution of nominal parameters were assessed using the Chi-square test or Fisher's exact test as appropriate. Disease-free survival and overall survival after surgery were assessed using Kaplan–Meier curves. Multivariate analysis was performed using Cox regression analysis. All P values were two-sided, and P < 0.05 was considered statistically significant. Data were edited with Microsoft Excel (Microsoft Corp., Redmond, USA), and the analyses were performed with RStudio (Boston, USA).

RESULTS

There were 259 patients who underwent laparoscopic operations for colorectal cancer from October 2015 to November 2017. A total of 168 patients were excluded for stage 0 and 4 (n = 81), pathological T4b (n = 1), not adenocarcinoma and not mucinous adenocarcinoma (n = 11), multiple tumours (n = 6), another malignancy (n = 11) and tumours located in the transverse colon, splenic flexure and rectum (n = 58).

A total of 91 patients were enrolled, of which 47 patients underwent 3D laparoscopic colectomy and 44 patients underwent 2D laparoscopic colectomy.

There were no significant differences in demographics, tumour characteristics and pathology and operation data between the two groups [Tables 1 and 2]. The median age was 65 years. The ratio of men and women was equal (50.0%:49.5%). One patient had American Society of Anesthesiologists Class 4 due to severe heart disease, 56% of patients had carcinoembryonic antigen levels >5 ng/mL and 15.4% of patients had lumen narrowing with partial obstruction. Similarly, the operation time, post-operative hospital stay, blood loss and complications did not differ between the groups [Table 3].

Table 1

Demographics of the patients

	2D (n=44)	3D (n=47)	Total (n=91)	P
Age	65.0 (16.5)	64.0 (20.5)	65.0 (17.5)	0.462
Sex
Women	20 (45.5)	26 (55.3)	46 (50.5)	0.465
Men	24 (54.5)	21 (44.7)	45 (49.5)
DM	9 (20.5)	9 (19.1)	18 (19.8)	1.000
Smoke	8 (18.2)	5 (10.6)	13 (14.3)	0.467
Alcohol	4 (9.1)	3 (6.4)	7 (7.7)	0.708
BMI >25	16 (36.4)	11 (23.4)	27 (29.7)	0.262
Previous abdominal surgery	5 (11.4)	7 (14.9)	12 (13.2)	0.851
ASA
1	3 (6.8)	6 (12.8)	9 (9.9)	0.288
2	27 (61.4)	33 (70.2)	60 (65.9)
3	13 (29.5)	8 (17.0)	21 (23.1)
4	1 (2.3)	0	1 (1.1)
CEA ≥5	22 (50.0)	29 (61.7)	51 (56.0)	0.361
Adjuvant chemotherapy	14 (31.8)	21 (44.7)	35 (38.5)	0.296
Tumour obstruction	7 (15.9)	7 (14.9)	14 (15.4)	1.000
Tumour perforation	0	1 (2.1)	1 (1.1)	1.000

Data are number of patients with percentages in parentheses or median with IQR in parentheses. 2D: Two-dimensional, 3D: Three-dimensional, DM: Diabetes mellitus, BMI: Body mass index, ASA: American society of Anesthesiologists, CEA: Carcinoembryonic antigen, IQR: Interquartile range

Table 2

Pathological findings

	2D (n=44)	3D (n=47)	Total (n=91)	P
Histology
Adenocarcinoma	42 (95.5)	45 (95.7)	87 (95.6)	1.000
Mucinous adenocarcinoma	2 (4.5)	2 (4.3)	4 (4.4)
Tumour size (mm)	37.5 (26.2)	40.0 (30.0)	40.0 (25.0)	0.553
Lymphovascular invasion	11 (25.0)	11 (23.4)	22 (24.2)	1.000
Peri-neural invasion	0	1 (2.1)	1 (1.1)	1.000
MSS
MSS	26 (59.1)	35 (74.5)	61 (67.0)	0.093
MLH1 and PMS2 loss	2 (4.5)	4 (8.5)	6 (6.6)
MSH2 and MSH6 loss	0	1 (2.1)	1 (1.1)
NA	16 (36.4)	7 (14.9)	23 (25.3)
n
0	28 (63.6)	29 (61.7)	57 (62.6)	0.502
1A	8 (18.2)	7 (14.9)	15 (16.5)
1B	4 (9.1)	2 (4.3)	6 (6.6)
1C	1 (2.3)	0	1 (1.1)
2A	2 (4.5)	6 (12.8)	8 (8.8)
2B	1 (2.3)	3 (6.4)	4 (4.4)
T
1	11 (25.0)	11 (23.4)	22 (24.2)	0.909
2	3 (6.8)	4 (8.5)	7 (7.7)
3	23 (52.3)	22 (46.8)	45 (49.5)
4A	7 (15.9)	10 (21.3)	17 (18.7)
Stage
I	14 (31.8)	10 (21.3)	24 (26.4)	0.486
II	14 (31.8)	19 (40.4)	33 (36.3)
III	16 (36.4)	18 (38.3)	34 (37.4)
Lymph node harvested number	20.0 (9.2)	18.0 (12.5)	19.0 (11.0)	0.408
Root lymph node harvested number	2.0 (3.2)	3.0 (4.0)	2.0 (4.0)	0.248
Root lymph node dissection	36 (81.8)	39 (83.0)	75 (82.4)	1.000
Margin not free	0	0	0
Differentiation
Well	1 (2.3)	1 (2.1)	2 (2.2)	0.757
Moderate	34 (77.3)	41 (87.2)	75 (82.4)
Poor	4 (9.1)	3 (6.4)	7 (7.7)
Undifferentiated	1 (2.3)	0	1 (1.1)
NA	4 (9.1)	2 (4.3)	6 (6.6)

Data are number of patients with percentages in parentheses or median with IQR in parentheses. 2D: Two-dimensional, 3D: Three-dimensional, MSS: Microsatellite stable, NA: Not applicable, IQR: Interquartile range, MLH1: MutL homolog 1, PMS2: Postmeiotic segregation increased 2, MSH2: MutS Homolog 2, MSH6: MutS Homolog 6

Table 3

Operative data

	2D (n=44)	3D (n=47)	Total (n=91)	P
Operation
Anterior resection	21 (47.7)	26 (55.3)	47 (51.6)	0.801
Left hemi-colectomy	3 (6.8)	3 (6.4)	6 (6.6)
Right hemi-colectomy	20 (45.5)	18 (38.3)	38 (41.8)
Operation time (min)	212.5 (76.2)	210.0 (75.0)	210.0 (70.5)	0.100
Blood loss >50 ml	3 (6.8)	3 (6.4)	6 (6.6)	1.000
Anastomosis
Staples	23 (52.3)	27 (57.4)	50 (54.9)	0.776
Hand-sewn	21 (47.7)	20 (42.6)	41 (45.1)
Post-operative hospital stays (days)	8.0 (2.2)	8.0 (2.0)	8.0 (2.0)	0.859
Complications
Anastomosis leakage	0	1 (2.1)	1 (1.1)	1.000
Ileus	2 (4.5)	4 (8.5)	6 (6.6)	0.678
Adhesion obstruction	0	2 (4.3)	2 (2.2)	0.495
Chyle leak	1 (2.3)	2 (4.3)	3 (3.3)	1.000
Wound infection	1 (2.3)	0	1 (1.1)	0.484
Mortality (within 28 days after surgery)	0	0	0

Data are the number of patients with percentages in parentheses or median with IQR in parentheses. IQR: Interquartile range

There were no differences in histology, tumour size, prognostic features (lymphovascular invasion, peri-neural invasion, microsatellite stable [MSS]), differentiation, LN harvested number and root LN sampling number. Only one patient who underwent anterior resection with 3D had 11 LN harvested. The others had qualified LN harvested numbers. All specimens were margin-free. Some specimens had no differentiation or MSS data because the tumour was completely removed by previous colonoscopy (polypectomy), and no detailed pathological reports could be reviewed.

There was no difference in disease-free and overall survival between the two groups [Figure 1]. The 3-year overall survival was 97.6% in 2D and 94.5% in 3D (P = 0.480). The 3-year disease-free survival was 90.7% in 2D and 84.9% in 3D (P = 0.400). The median follow-up time was 44 months (interquartile range: 36–52 months). Three patients developed lung metastasis (stage III), one patient developed metastases in the lung, liver and mediastinum (stage III; expired) and another patient with MSH2 and MSH6 loss developed extra-luminal local recurrence and metastasis in the bone and liver (stage II) in the 3D group. In the 2D group, one developed lung metastasis (stage II), one regional recurrence at the pancreas tail (stage II), one peritoneum seeding (stage III), one local recurrence at anastomosis and lung metastasis (stage II) and one metastasis at the liver, lung and bone (stage III; expired). Another 87-year-old stage II patient died due to hypovolemic shock according to the TCDB. Multivariable analysis (Cox regression) of disease-free and overall survival was not feasible because of the small number of events in the subgroup. None of the factors revealed a significant hazard ratio.

Figure 1

Kaplan–Meier curves for comparing 3-year overall survival and disease-free survival of patients with stage I–III colon cancer

DISCUSSION

We are the first to compare the 3-year follow-up outcomes of 2D and 3D laparoscopic surgery for colon cancer in stages I–III retrospectively. We found that 3D laparoscopy may not shorten operation time, decrease blood loss, facilitate root LN dissection or improve the survival rate.

Tao et al. found that 3D laparoscopy decreased the performance time of laparoscopic right hemi-colectomy.[10] In addition, Zeng and Su found that it improved the operation time for rectal cancer.[11] However, our results showed no difference in the operation time of the entire procedure, which was in consistent with the findings of Currò et al. and some other studies.[91213] Pantalos et al. concluded there were small differences in favour of the 3D group (mean difference = 7.5 min; P = 0.06).[7] Currò et al. found that 3D imaging seemed not to influence the performance time of laparoscopic right colon cancer surgery when performed by a surgeon experienced in 2D laparoscopy, although the 3D system seems to offer better depth perception and to subjectively determine less physical strain compared to 2D vision.[9] Thus, it is reasonable to conclude that the lack of difference observed in the procedures of this study was because our team has >15 years of experience in laparoscopic colectomy. Similarly, Yoon et al., whose study included right hemi-colectomy, left hemi-colectomy and anterior resection also showed no difference in operative time.[8]

Singh et al.[14] suggested better survival for patients with LN metastasis receiving high ligation of the IMA, while Zeng and Su suggested that there was no difference in 5-year survival.[11] Fujii et al. reported that they achieved similar IMA LN harvest numbers in both low and high ligation groups and that there was no difference in 5-year survival.[15] Our surgical team routinely intended to perform D3 dissection for the left side colon, and D3 sampling for patients revealed more than clinical T1 disease. We attempted to confirm the conclusions of Yoon et al. and Pantalos et al., which suggest that better survival may be associated with the greater number of harvested LNs in 3D laparoscopy. Yoon et al. harvested 41 and 47 LNs in 2D and 3D laparoscopy, respectively. Moreover, Pantalos et al. showed a non-significant difference of 0.25 more LNs harvested in 3D versus 2D laparoscopy after excluding the study by Yoon et al. and 1.39 more LNs harvested in 3D compared with that in 2D laparoscopy before its exclusion (P = 0.26).[78] However, the number of harvested LNs and the number of root LN did not differ between the two groups in our study. Previous studies have revealed that stage II colorectal cancer patients have better survival when more LNs are harvested.[1617] However, these studies had minimal cut-off LN numbers or categorised number groups.[1819] More evidence is needed regarding the impact of the small difference in the number of LNs harvested (6 in Yoon et al.) on the survival rate in all stages.

To obtain comparable survival data, our enrolment criteria were based on the COLOR trial.[20] Therefore, many patients were excluded from this study. The 3-year overall survival and disease-free survival after laparoscopic colectomy were 81.8% and 74.2%, respectively, in the COLOR trial. Our study had a wonderful outcome; however, 5-year results in a larger patient number should be collected to verify our results. Through our 3-year follow-up study, we observed that surgical resectable local recurrence, peritoneal seeding or/and lung metastasis could still have a good prognosis, although rapid multi-organ metastases after surgery had a worse prognosis.

Our results were obtained by an experienced surgical team in a medical centre with a high volume of colon cancer. Studies such as Curtis et al., Currò et al. and Yoon et al. were also conducted by surgeons experienced in 2D laparoscopy.[8912] These studies also considered surgeon experience as a confounding factor and a limitation. Tao et al. thought that the difference between 2D and 3D is smaller for experienced surgeons.[10] Results from some studies revealed that the 3D system benefited both experienced and novices surgeons.[2122]

Realistically, not everyone likes the 3D system due to dizziness, discomfort or cost.[2324] Most manufacturers have stopped manufacturing home-used 3D televisions, which indicate that the 3D monitor technique has room to improve. On the other hand, some movies are still shot and played in 3D because people are willing to pay more money to see them. Conceptually, stereoscopy is the right way to develop in laparoscopic surgery, although the technology may not satisfy every surgeon. Just like 3D movies require suitable film categories, display methods and customers, the 3D imaging system must explore appropriate indications and better display devices.

Limitations

These data were from a single surgical team that was highly experienced in laparoscopic surgery for colon cancer and thus may not represent the entire surgical society; meanwhile, the case number might not be large enough for generalisability. Moreover, this was a retrospective analysis, and the two groups were not randomly distributed. Most of the 2D cases were before August 2016 because the 3D system was available only after August 2016 in our institute.

CONCLUSIONS

This is the first retrospective study to compare the 3-year follow-up outcomes between 2D and 3D laparoscopic surgery for colon cancer in stages I–III. We found that 3D laparoscopy may not shorten the operation time or decrease the blood loss as compared to 2D laparoscopy when performed by an experienced surgical team. Furthermore, our 3-year follow-up study revealed no difference in the survival rate following 2D or 3D laparoscopic colectomy.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.