How to increase the resectability of initially unresectable colorectal liver metastases: A surgical perspective

Abstract Although surgical resection is the only treatment of choice that can offer prolonged survival and a chance of cure in patients with colorectal liver metastases (CRLM), nearly 80% of patients are deemed to be unresectable at the time of diagnosis. Considerable efforts have been made to overcome this initial unresectability, including expanding the indication of surgery, the advent of conversion chemotherapy, and development and modification of specific surgical techniques, regulated under multidisciplinary approaches. In terms of specific surgical techniques, portal vein ligation/embolization can increase the volume of future liver remnant and thereby reduce the risk of hepatic insufficiency and death after major hepatectomy. For multiple bilobar CRLM that were traditionally considered unresectable even with preoperative chemotherapy and portal vein embolization, two‐stage hepatectomy was introduced and has been adopted worldwide with acceptable short‐ and long‐term outcomes. Recently, ALPPS (associating liver partition and portal vein ligation for staged hepatectomy) was reported as a novel variant of two‐stage hepatectomy. Although issues regarding safety remain unresolved, rapid future liver remnant hypertrophy and subsequent shorter intervals between the two stages lead to a higher feasibility rate, reaching 98%. In addition, adding radiofrequency ablation and vascular resection and reconstruction techniques can allow expansion of the pool of patients with CRLM who are candidates for liver resection and thus a cure. In this review, we discuss specific techniques that may expand the criteria for resectability in patients with initially unresectable CRLM.

| 477 IMAI et Al. and liver metastases remain the predominant cause of death for colorectal cancer patients. 15,16 In patients with metastatic colorectal cancer treated with chemotherapy alone, survival beyond 5 years is unusual. 8,[17][18][19][20][21] The development of surgical techniques, the increasing efficacy of modern chemotherapy with or without biological agents, and the emergence of multidisciplinary approaches have allowed patients with conventionally unresectable CRLM to undergo surgery. Given that surgical resection remains the only form of treatment that offers the possibility of prolonged survival, expanding the potentially resectable pool of patients is crucial.
In confronting unresectability, one of the most important issues is that the definition of unresectability differs among institutions. The definition of unresectability is also evolving with the development of surgical techniques and chemotherapy. Nowadays, unresectability should be considered based on both technical and oncological criteria. [22][23][24] Historically, several tumor factors that represent massive tumor loads such as multinodular tumors, larger tumor size, bilobar distribution, and the presence of extrahepatic disease were used to classify liver metastases as unresectable, because these variables were prognostic factors for survival after hepatectomy for CRLM.
However, recent advances in surgical techniques, in combination with modern effective chemotherapy, enable surgeons to overcome these issues, and these factors are no longer an absolute contraindication for surgery. According to European Society of Medical Oncology (ESMO) consensus guideline 2016, patients with CRLM can be categorized into groups on technological and oncological criteria (Table 1) Among the three, this review discusses the specific surgical techniques that may expand the criteria for resectability in patients with initially unresectable CRLM.

| PVE
PVE was first described by Makuuchi et al 25 in the 1980s. This procedure induces atrophy of the embolized liver lobe with compensatory hypertrophy of the non-embolized contralateral liver lobe, and can thereby reduce the risk of hepatic insufficiency after major hepatectomy in patients with an insufficient future liver remnant (FLR). [26][27][28] Several studies have reported that PVE has no adverse effect on survival in patients with CRLM who have undergone major hepatectomy. [28][29][30][31][32] PVE approaches include transileocolic and transhepatic techniques. The transileocolic approach requires a mini-laparotomy.
Although a previous meta-analysis demonstrated that the major complication rates of transileocolic and transhepatic PVE were comparable, 27 the transhepatic approach has become standard due to the recent development of radiological intervention techniques.
Transhepatic PVE is performed percutaneously, by an ipsilateral or contralateral approach. The advantage of the ipsilateral approach is that it does not require puncture of the FLR; that is, it can reduce the risk of injury of vessels in the FLR. Given that the vessels of the FLR should be carefully preserved for the subsequent planned surgery, the transhepatic ipsilateral approach is recommended if possible, despite its relative complexity.
One of the concerns about preoperative PVE is that stimulation of liver hypertrophy can also accelerate tumor growth in the embolized and non-embolized liver lobe. [32][33][34][35][36] Portal flow reduction in the embolized liver leads to an increase of arterial blood flow and, subsequently, growth of tumors including micrometastases could be induced, because liver tumors are mostly supplied by arterial blood. Although tumors in the embolized liver lobe will be removed by surgery, those in the non-embolized liver lobe could be also stimulated, leading to a cause of unresectability or a risk of early tumor recurrence. One possible way to reduce the risk of unresectability would be chemotherapy after PVE, although, to our knowledge, there are no studies providing evidence of the efficacy of chemotherapy between PVE and hepatectomy. Meanwhile, preoperative PVE should probably be indicated for unilobar disease, and in the case of multiple bilobar diseases, which requires hypertrophy of the FLR, two-stage surgery may be indicated.   37 We previously reported that RFA in combination with hepatectomy achieved outcomes comparable to hepatectomy alone in a propensity scorematched setting. 38 Another study reported that RFA in combination with hepatectomy gave poorer survival than hepatectomy alone;

| Local ablation therapy in combination with hepatectomy
however, in the setting of tumor number ≥ 4, long-term survival rates were similar between the two groups. 39 These results suggest that in selected patients and selected tumors, adding RFA to hepatectomy may be justified in the treatment of multiple CRLM.
In essence, we should not support the unconsidered use of RFA in the treatment of CRLM. Surgeons should always strive for resection with a clear margin. However, most patients with CRLM are not candidates for hepatectomy at the time of diagnosis. In these circumstances, adding RFA to hepatectomy would expand the pool of patients who can receive radical treatment. 23 With this view, RFA should be strictly indicated for appropriate patients and appropriate tumors. We consider appropriate indications to be as follows: (a) unresectable or deeply located small tumors (≤2 cm) within the FLR, which require extended parenchymal resection; and (b) tumors that have responded to preoperative chemotherapy. Tumors that demonstrate no response to chemotherapy have potentially more malignant behavior and a higher local recurrence rate, and RFA should therefore be contraindicated for such tumors. 40 In addition to its lower invasiveness, the benefit of adding RFA to hepatectomy is that it allows the uninvolved functional liver parenchyma to be preserved. Several recent studies have reported the usefulness of parenchyma-preserving hepatectomy for CRLM, and Mise and colleagues found that parenchyma-preserving hepatectomy improved survival in the event of intrahepatic recurrence, possibly due to an increase in the likelihood that the patients would be able to undergo salvage repeat hepatectomy. 41 In our previous study, salvage hepatectomy for intrahepatic recurrence was more frequently performed in patients who underwent RFA in combination with hepatectomy than in those who underwent hepatectomy alone. 38 Endeavoring to preserve liver parenchyma should lead to an increase in the number of patients who can undergo salvage hepatectomy for recurrence and in subsequent prolonged survival.
During the past several years MWA has gained acceptance alternative to RFA. Theoretical benefits of MWA over RFA include F I G U R E 1 Three material factors for increasing resectability. EHD, extrahepatic disease; ALPPS, associating liver partition and portal vein ligation for staged hepatectomy larger ablation zone, shorter duration, and no heat-sink effect. 42 There are some studies reporting the outcomes of MWA for colorectal liver metastases as a safe and effective modality for use in the treatment of CRLM patients. 43

| Vascular resection and reconstruction
Major vascular invasion is one of the common reasons for unresectability in patients with CRLM. When the tumor is invading either the portal vein bifurcation, all three hepatic veins, or the inferior vena cava (IVC), vascular resection and reconstruction would be required.
Removal of a tumor with involvement of these major vessels has been thought to be technically demanding. However, recently, considerable advances in vascular surgery and liver transplantation have made these procedures possible.
For portal vein resection and reconstruction, primary end-toend anastomosis is preferred over the use of grafts if possible. 51

| Two-stage hepatectomy
From 1992, a Paul Brousse team introduced the concept of TSH, based on two sequential procedures to remove multiple bilateral tumors that were impossible to remove by a single hepatectomy and then using the liver regeneration obtained after the first procedure; the work was published in 2000. 57  The main drawback of the TSH strategy is the failure to complete both sequential procedures; that is, dropout from the second-stage hepatectomy. A systematic review demonstrated that the failure rate of TSH ranged 0%-36% (median, 23%), the main reason for failure being disease progression after first-stage surgery. 59 We previously reported a failure rate of 35.2%, and main reason was disease progression after first-stage surgery (88.6%). 60 In this study, four independent predictive factors for the failure of

| ALPPS
ALPPS is a novel form of TSH. The first report of 25 cases who underwent the ALPPS procedure in five German centers was published in 2012. 63 In this paper, rapid growth of the FLR with a median hypertrophy rate of 74% after 9 days and a complete resection rate of 100% were reported. Despite initial concerns about high mortality and morbidity rates, this innovative concept has been adopted by many specialized centers around the world.
The main criticism of ALPPS was its high morbidity and mortality rates. Indeed, the first paper reported morbidity and mortality rates of 68% and 12%, respectively. 63   TA B L E 2 (Continued) short-and long-term outcomes, including anterior-approach ALPPS (using anterior approach with or without the liver hanging technique), 66,67 hybrid ALPPS (anterior approach + PVE after the first stage), 68 partial ALPPS (partial liver partition), 69 mini-ALPPS (partial liver partition to a depth not exceeding 3-5 cm + intraoperative PVE) 70 , ALTPS (tourniquet instead of liver partition), 71 RALLP (radiofrequency ablation instead of liver partition), 72 ALPTIPS (partial liver partition until the middle hepatic vein + intraoperative transileocecal PVE), 73 and modified ALPPS with preservation of portal pedicles (preserving portal pedicles at the transection plane) ( Table 2). 74 In addition, ALPPS modifications using minimally invasive approaches, such as laparoscopic 75 or robotic ALPPS 76 have been described. However, the reports of these modifications consist of small case series, and larger-scale or prospective controlled studies will be needed to establish the real roles of these modifications.
Compared to conventional TSH, ALPPS has the following ad- recurrence with disease progression after ALPPS for CRLM. 78 We previously published a study comparing ALPPS and conventional TSH for CRLM in an early phase of introduction of ALPPS, demonstrating that the OS rate of ALPPS was significantly worse than that of TSH, even in intention-to-treat analysis. 79 However, two other studies reported no difference in OS between ALPPS and TSH. 80,81 Recently, multicenter randomized controlled trial (LINGO Trial) including 97 patients with CRLM and FLR < 30% (ALPPS, n = 48; TSH, n = 49) reported that resection rate was significantly higher in the ALPPS arm than in the TSH arm (92% vs 57%, P < .0001), and there was no difference in terms of complication and 90-day mortality rates. 82

| Minimally invasive liver surgery
Laparoscopic liver resection is nowadays adopted worldwide for the treatment of liver tumors including CRLM. Numerous studies have reported the benefits of laparoscopic liver resection compared with the standard open liver resection, such as reduced intraoperative bleeding, a lower morbidity rate, cost-effectiveness, and shorter in-hospital stay. [85][86][87][88][89][90] In addition, laparoscopic liver resection has been reported to provide comparable long-term oncological outcomes with open liver resection. 88,89,91 Another possible benefit is that less invasiveness of laparoscopic liver resection may lead to rapid initiation of postoperative chemotherapy. [92][93][94] In light of this reduced invasiveness without compromising short-and long-term outcomes, laparoscopic liver resection will probably continue to expand in the treatment of CRLM.
Robotic liver surgery is another possible minimally invasive treatment for CRLM. Robotic liver surgery has the potential to overcome some of the limits of laparoscopic surgery, such as limited degrees of motion of the instruments, unstable camera platforms, and twodimensional vision. 95 However, in the consensus opinion during the second international laparoscopic liver forum held in 2014, there was insufficient data to comment on the applicability of robotic liver resection. 96 Further data accumulation and prospective large-scale studies will be required.

| Liver transplantation
The indication for liver transplantation for CRLM has been restrictive over time because of high recurrence rates and poor outcomes.
Given the systemic nature of metastatic disease and need for immunosuppressive therapy after surgery, liver transplantation for CRLM has been controversial. However, recent advances in surgical techniques, advent of more effective systemic chemotherapy, development of imaging modality, and improvements of perioperative management including novel immune-modulating agents have led to the re-evaluation of liver transplantation as a treatment option for unresectable CRLM. A recent systematic review reported that estimated cumulative survival rate after liver transplantation was 85.2% (at 1 year), 48% (at 3 years), and 34.6% (at 5 years), respectively, and disease-free survival rate at 1 year was 38.9%. 97 Due to the small sample size and lack of studies comparing liver transplantation with current standard of care, it is difficult to evaluate its applicability as a treatment of choice in patients with unresectable CRLM. Now some clinical trials are ongoing and will provide high-quality evidence regarding the role of liver transplantation in the management of CRLM.

| CON CLUS IONS
For patients with CRLM, surgical resection is the only treatment of choice for prolonged survival and a chance of cure. Given this, how to increase resectability is one of the most crucial outstanding issues. Because we currently have various surgical options to improve resectability, we should use all available techniques to explore the possibilities of surgical resection, working in a multidisciplinary manner with specialists in hepatobiliary, colorectal, and thoracic surgery, and across hepatology, oncology, and radiology.

D I SCLOS U R E
Funding: There was no funding for the present study.
Conflict of interest: The authors declare they have no conflict of interest.