Simultaneous pure laparoscopic resection of primary colorectal cancer and synchronous liver metastases: a single institution experience with propensity score matching analysis

Colorectal cancer is one of the most common causes of cancer related death in the Western world.1 At the time of diagnosis of the primary tumor, up to 25% of patients present with synchronous colorectal liver metastases (SCLM).2 Complete surgical resection of both primary tumor and SCLM remains the only treatment option providing long-term survival and chance for cure.3

Different oncological strategies have been described including traditional two-stage colon first approach,4 liver first procedure5 or a one-step surgical resection of both the primary tumor and SCLM.6 Several reports have shown the benefit of a simultaneous open resection of primary tumor and SCLM compared with a staged approach.7-8

Minimally invasive colorectal and liver surgery are both accepted worldwide.9, 10, 11 Technical refinements and the development of advanced laparoscopic techniques has made simultaneous resection an attractive option. However, despite the increasing use of laparoscopy in colorectal and liver resections, simultaneous pure laparoscopic resection (SPLR) of the primary colorectal cancer and SCLM is still rarely performed.

Sporadic case reports and single institution series have shown the feasibility and safety of simultaneous laparoscopic resection of both primary CRC and SCLM.12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28 Recently, two large international multicenter retrospective studies, which were published from the same group of surgeons, confirmed that in selected patient’s laparoscopic surgery allowed similar outcomes compared with the traditional open approach.29-30

The aim of this study was to compare the surgical and oncological outcomes of patients undergoing simultaneous resection of primary colorectal cancer and SCLM by laparoscopic or open surgery using propensity score matching.

The optimal strategy for resectable SCLM has not been established yet. In selected cases, the simultaneous surgery approach gives the advantages to avoid two surgical procedures thus reducing risk for patients and provides for economic savings while keeping acceptable morbidity and adequate oncologic results.6, 7, 8 However, open resection of both primary tumor and SCLM often requires an extensive incision, especially if the location of the liver metastasis is opposite to the primary tumor location. Using laparoscopic approach exposure can be improved thus avoiding extensive laparotomies.20 Consequently, surgical stress and pain associated with large incisions can be reduced and patient’s recovery enhanced. There is an ongoing effort to show the potential benefit of totally laparoscopic strategies for the radical treatment of stage IV colorectal cancer.12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 35

The present study was designed to investigate the perioperative results, oncologic outcomes and survival of patients undergoing pure laparoscopic simultaneous resection of both the primary colorectal cancer and SCLM. During this study period, the choice of intervention was subject to selection bias since simultaneous laparoscopic liver procedures were reserved for the most ideal candidates. Moreover, the laparoscopic treatment of patients with an extraperitoneal rectal cancer remains clinically challenging compared to upper rectal or colon cancer. To minimize these biases, the LAP group was compared to OPEN group by propensity score matching on the factors known to influence the choice of the approach. Importantly, there were no differences in patient demographics, tumor characteristics, or the extent of the operation, so the present characteristics well identify the most ideal candidates to a combined laparoscopic liver and colorectal surgery. Patients selected for this approach were without severe comorbidities (ASA I–II in majority of cases), with a solitary (median number: 1), small (median diameter: 2 cm), unilateral SCLM, located in accessible anterolateral segments and were mostly resectable by atypical wedge resections or left lateral sectionectomy. Consequently, only minor liver resections were performed. These results are consistent with previous studies showing well selected patients and demonstrating a high rate of minor (89%) and nonanatomical resections (60%).17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 35 In this study colorectal cancer characteristics did not preclude the possibility to perform a laparoscopic surgery as well with majority of patients presenting a T3 primary tumor. There were no differences in N stage and importantly, the number of harvested lymph nodes was even higher in the LAP group. Half of the patients in the LAP group presented with rectal cancer. Surprisingly, previous studies have demonstrated that despite the high rate of rectal resections in some series, the number of the temporary ileostomies was low.25, 27, 35 The results of the present study contradict this: all of the rectal cancer patients had either temporary ileostomy covering low anastomosis or terminal colostomy. This finding might have been due to a high rate of preoperative radiotherapy, which highlights that most rectal resections were performed for cancer in the lower rectum. Moreover, according to some reports simultaneous resection generally is considered unsuitable for rectal cancer due to a high rate of anastomotic leakage.36 Another possible explanation is a concern regarding prolonged vascular clamping which is responsible for transient portal hypertension with edema of the intestinal mucosa that might be leading to colorectal anastomotic failure.21,35

The feasibility of simultaneous laparoscopic operations was clearly demonstrated in the present study: no conversion to open surgery has been performed. This result reflects a findings of a recent review, where a conversion rate of only 0.7% has been reported.35 Similarly, Tranchart et al. reported a low conversion rate of 7% in their large multicenter study.30 It has been highlighted, that simultaneous resections executed by two different specialized teams allowing good results in terms of conversion rates and perioperative outcomes.35 Of note, the present study was performed at high-volume tertiary referral center, where all procedures were performed by an expert surgical team, composed of experienced colorectal and hepatobiliary surgeons. Controversy still exist, which procedure should be carried out first and many authors reported colorectal resection was the first step of the simultaneous surgery.15-16, 20, 22, 29-30 Instead, in the present study liver resection precede colorectal surgery and the same strategy has been reported by others.18-19, 21, 35 The underlying rationality is that the resection of liver metastases requires a low central venous pressure to minimize the blood loss and the preceding liver resection will not interfere with the subsequent fluid resuscitation in the process of colorectal resection. In addition, the choice of carrying out the liver resection as first step of treatment gives to the surgeon the opportunity to change surgical strategy from a simultaneous procedure to a “liver first” resection which has been showed to be another effective treatment of SCLM.5 Moreover, liver metastases are the main determinant of patient prognosis and the present authors believe, that SCLM are leading the decision making process: if simultaneous strategies are not feasible, then the metastases should be managed first. However, no decisions were modified and all procedures were finished simultaneously.

In terms of intraoperative measurements of outcome, the median operating time of 261 min after LAP is comparable with the time reported by Jung et al.25, and is shorter than the time reported by others, where it was more than 300 min.13, 14, 15, 16, 19, 22, 23, 24, 30 Nevertheless, operative time was not different between OPEN and LAP group. This result is inconsistent with some previous studies showing that the duration of operation was significantly longer in the laparoscopic groups.15, 22, 25 A possible explanation can be, that in contrast to present study, these reports included major liver resections as well, which are known to prolong operative time. However, several authors reported equivalent operative times for both groups.16, 23, 30

Intraoperative blood loss is a major concern especially in hepatic resection and perioperative transfusion has been associated with a poor prognosis.37 The findings of decreased blood loss with the laparoscopic approach have been reported in several studies, explained in part by the laparoscopic magnification, and decreased venous oozing from the cut surface under pneumoperitoneum.15-16, 22-23 Results of present study contradicts this as laparoscopic approach did not allow diminishing blood loss and transfusion rate. Contrast to prior studies, in this analysis blood loss was not based only on estimation but it has been objectivized by measuring the volume of blood in the suction canister and precisely weighing the absorbed blood in the sponges. Moreover, the pre- and postoperative hemoglobin levels were demonstrated exactly as well. The reason for this was to exclude the possibility of differences in preoperative hemoglobin levels which can contribute to the difference in blood loss and transfusion rate between two groups since it is well known that patients with SCLM can develop anemia. Moreover, it has been suggested that the blood loss can be underestimated especially in the open surgery.38 Despite analyzing additional objective variable in this study, there were no differences between groups. It should be noted that outcome observed after open surgery was excellent as well (median estimated blood loss 170 ml). Similarly, several recent studies have suggested that there is no decreased blood loss and transfusion rate in the laparoscopic group.19, 25, 30

In respect to postoperative oncological results, a high CRS has been found to be an independent negative prognostic predictor in previous studies.32, 39 The impacts of CRS and colorectal tumor differentiation on survival analysis were eliminated, since the OPEN and LAP groups were comparable in that terms as well. The risk of inadequate oncologic resection is the major concern for the use of laparoscopic resection for malignancy. A positive surgical margin has been shown to predict worse DFS after resection.40 Castaing et al. emphasized that the increase in R1 resections did not affect OS and suggest it might reflect complex anatomic locations of metastases adjacent to major vascular structures.41 To avoid an inherent selection bias with more difficult cases potentially being relegated to open approach, it is important to point out that the two groups of the present study were perfectly matched in term of SCLM proximity to major vessels (hilar or hepatic confluence) as well; there were no tumors with such characteristics in any of groups. Importantly, complete R0 resections of both the primary and SCLM were achieved in all patients confirming the reports of others, which reported a high rate of R0 resections of 100% and 90% as well.25, 30 Surprisingly, the rate of R0 resections has been not reported by many authors.14, 15, 16, 22 The width of the surgical margin was not different between two groups and interestingly, the width was even larger in the LAP group. Moreover, a recent study failed to demonstrate that the width of negative margin correlated with recurrence or survival.42 Importantly, in this study neither peritoneal carcinomatosis nor port site metastases after simultaneous resections of SCLM by laparoscopic means were found. These results suggest that laparoscopic procedure does not compromise oncologic principles.

The major findings in this study were lesser length of hospital stay and faster solid food oral intake for LAP group. The median hospital stay of 8 days in the LAP group is similar to or shorter than the time reported by others, where it has been in a range from 7.4 to 16 days. The benefit of decreased hospital stay was recognized in these analyses as well.15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30 Interestingly, substantial geographical differences exist which depends on national health care systems. Takasu et al. confirmed the benefit of lesser length of hospital stay which were 16 days for the laparoscopy and 36 days for the open group.23 However, Tranchart et al. found no differences in the length of stay between laparoscopic and open groups, and noted that simultaneous resections still appear as difficult procedures for surgeons, which could have influence their decisions concerning patient’s hospital discharge.30 Patients undergoing laparoscopic procedure resumed solid food oral intake earlier, however it has no impact on earlier stool passing in this study, thus the benefit of this parameter remains questionable. Similarly, several authors reported a shorter time to resume a bowel movement and starting an oral intake.19, 25

In terms of other postoperative measurements of outcome none of the patients required intensive care unit stay and the high dependency unit stay was comparable between groups. Ferretti et al. reported a median one-day intensive care unit stay.29 However, these relevant parameters were not reported in several previous analyses.15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 35

Pain control is one of the potential advantage of laparoscopy and in this study, it was evaluated by intravenous narcotics requirement. Laparoscopy offered a benefit of less narcotic requirements, patients in the LAP group needed intravenous narcotics for 4.5 days compared with the OPEN group where these were necessary for 6.5 days. Notwithstanding, statistical significance was reached only marginally (P = 0.08). Hu et al. reported that patients having undergone laparoscopic resection had less severe postoperative pain, but it is not clear how they asses this parameter.19 Surprisingly, in several other reports on this topic pain control was not investigated.13, 14, 15, 16, 17, 18, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30 Some surgeons combined a laparoscopic colorectal resection with an open procedure for the liver, as reported by Hatwell et al.20 However, this technique also left a large incision in the upper abdomen, and the advantage of laparoscopy was not realized fully.

The present study found that LAP and OPEN were not significantly different in terms of postoperative morbidity and mortality. Of note, no liver specific morbidity was observed in the LAP group. Nevertheless, some bile collections which occurred in the OPEN group were easily managed by percutaneous drainage. Importantly, there were no organ failures and postoperative deaths in two groups either. Similarly, no differences in morbidity and zero mortality were reported in several other analyses.15-16, 19, 22-23 Polignano et al. reported that the surgical morbidity was mainly related to the colonic surgery.18 Contrary, in the present series only one anastomotic leak occurred in OPEN and one in the LAP group, respectively. Jung et al. reported a higher rate of minor complications in the open surgery group (superficial surgical site infections and adhesive ileus), which seem to be strongly related to the presence of the bigger wound.25 However, the number of minor complications were equal in the present analysis. Interestingly, despite laparoscopy has the potential to reduce the incisional hernia rates, none of the studies on the present topic investigated this parameter which is responsible for long-term morbidity.12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 35 In the present study none of the LAP whereas three patients from the OPEN group developed incisional hernia, but the difference was not significant.

In the present study, neither the 3-year OS nor the DFS was found to be different for LAP compared with the OPEN group. Similarly, no differences were found between groups in terms of OS reported by others 15-16,19,22-23, 30 After long and adequate median follow-up of 78 months this analysis revealed excellent long-term results among patients in OPEN with 5-year OS/DFS of 80 and 60%, respectively. However, due to a shorter follow-up period, only mid-term results were available for the LAP group with 3-year OS of 75% and DFS of 57%, respectively. Similarly, in the current literature some of the authors reported the short- and midterm results only and the 3-year OS rates ranged from 52 to 78%.15, 30 Unfortunately, data regarding recurrence were inconsistently reported in several case-matched series on this topic.15-16, 19, 23, 25

The present study is a subject to a number of limitations. First and foremost, only a small cohort of patients from a single center were analyzed which were collected across a long interval of time. However, the majority of case-matched studies in the literature derived from a small single center series, which included from 7 to 24 patients.15-16, 19, 22-23, 25 Even in the largest multicenter analysis, which included 142 patients from 14 centers the average number of patients per hospital was ten.29 It demonstrates clearly the limited indications of simultaneous laparoscopic procedures even in the most experienced centers worldwide. Second, given the retrospective nature of this study, selection bias may have been present. Although propensity score matching was performed to overcome potential bias and to make the two groups similar it is less effective than a prospective randomized trial. Notwithstanding, to conduct a randomized study on this topic is still an unresolved issue. Third, since laparoscopic simultaneous procedures started in 2012 the median follow-up data between groups differs and the excellent long-term survival results achieved with OPEN are still waiting to be proved by LAP. However, except for introduction of laparoscopy, the management of SCLM was homogeneous over the study period and the same surgical team made the treatment decisions and performed the procedures.

Indeed, the power of the present study is limited, but importantly the two surgical approaches brought about similar outcomes in terms of postoperative morbidity, mortality, survival and recurrence. Nevertheless, some surgeons may insist on traditional open approach and not to accept widespread changes in clinical practice given the lack of several investigated variables to support the superiority of laparoscopic approach. However, the results of a recently published studies investigating inflammatory markers are promising and have a potential to prove that the reduced inflammatory response by laparoscopy might have positive impact on oncogenesis.43

Based on limited evidence, LAP patients experienced some clinically relevant perioperative benefits without compromise of oncologic outcomes. In high-volume centers experienced in both laparoscopic colorectal and liver surgery, simultaneous minimally invasive approach is appropriate in well selected patients presenting with limited SCLM.