Colorectal cancer (CRC) is a major health-threatening disease in the United States, with 140,000 new cases in 2012, and CRC is the second leading cause of cancer death worldwide. Due to improvements in screening prevention and treatment, CRC mortality rates have gradually decreased, with 1- and 5-year survival rates of 83.2% and 64.3%, but once distant organ metastases develop, the 5-year survival rate drops to 11.7%.
The most common site of metastasis in CRC is the liver, with approximately 25% of patients initially presenting with liver metastases and another 30% developing liver metastases during the course of the disease. Liver metastases account for 2/3 of deaths from CRC, making multidisciplinary treatment of CRC liver metastases (CLM) very important.
In an article published in J Gastrointest Oncol, Margaret E. Clark, MD, USA, describes how the gold standard of cure for liver metastases from colorectal cancer is surgery, and how several radical surgical standards currently exist alongside each other; and how steps can be taken to prolong survival and provide palliative care for patients who are not candidates for surgery.
Surgical resection
Complications and mortality are less than 30% and 3%, respectively. Multiple risk factors are independent prognostic markers, including age, primary tumor stage, preoperative CEA level, disease-free interval, liver tumor size, number of metastases, margin of resection, and presence of extrahepatic metastases. These factors, when applied together, select 10-20% of CLM patients as suitable for surgical treatment.
In 1990 Fong et al established a clinical prognostic score that identified 7 factors that significantly affected survival after resection of liver metastases, the top 2 being positive margins and extrahepatic metastases, with the risk of death in these patients being 1.7 times higher than the baseline value. The authors concluded that these 2 factors should be considered relative contraindications to hepatic resection.
The remaining 5 factors were disease-free survival <12 months, more than 1 tumor, preoperative CEA >200, positive primary tumor lymph nodes, and tumor >5 cm. 1 point was assigned to each factor. 5-year survival was 60% for patients with a score of 0 and only 14% for patients with a score of 5. A score of 0-2 was indicative of a good prognosis, while a score of 3-4 was relatively poor and should be followed by adjuvant therapy after resection.
In patients with a score of 5, it would be inappropriate to have no effective adjuvant therapy after resection or not to enter an adjuvant therapy trial. This score is still used, but recent analyses have shown that even patients with a score of 5 have a 5-year survival of 31%, and this improvement may be related to a number of factors, such as more effective chemotherapy and expanded treatment indications. Recent studies suggest that only patients with ≥8 metastases and concomitant inflammatory tumor response do not benefit from surgical resection.
While most studies have traditionally focused on clinicopathologic factors to determine which patients benefit from hepatic resection, the focus has now shifted to whether intrahepatic and extrahepatic metastases can be completely resected while ensuring adequate liver function. Complete intrahepatic resection is defined by a minimum of 1 cm of hepatic margin, and the exact definition of more adequate margins is being closely evaluated. Based on the results of a series of studies modern chemotherapy has made the impact of margins on overall survival less important, as long as they are negative.
Safe removal of all lesions from the liver and assurance of negative margins is dependent on residual liver (FLR). If an FLR of ≤40% is expected, this should be calculated for all patients according to standard calculations. There is no standard minimum FLR for safe hepatectomy. Guidelines generally recommend preserving ≥20% of total liver volume in patients without cirrhosis or other underlying liver disease, >30% in patients with severe steatosis or chemotherapy, and >40% in patients with cirrhosis.
Many studies have focused specifically on extended hepatic resection, showing prolonged complication rates, ICU length of stay, and total hospital stay for FLR ≤25%. Another way to assess the safety of resection is the ratio of FLR to body weight, with patients ≤0.5% at significantly increased risk of hepatic insufficiency and death.
Extrahepatic metastases
Long-term survival after hepatic resection is also possible for some patients with extrahepatic metastases (EHD). Most studies have shown that long-term survival with complete resection of EHD is largely dependent on the location of the EHD. survival is best with CLM with pulmonary metastases, worse with paravertebral lymph node and abdominal metastases OS, and poorer with multiple sites and aortic or abdominal lymph node metastases OS is significantly worse in the EHD group than in patients without EHD, with 5-year survival 19-38% compared to less than 5% of patients with chemotherapy only.
In a recent study of 1142 patients with EHD and CLM in 22 studies, complications and mortality were 28% and 1%, respectively, similar to those seen with CLM resection alone. This study found a median 5-year overall survival of 25% for R0 resected patients.
It has been previously noted that survival of EHD patients is related to EHD location, with median 5-year overall survival being 27% for pulmonary EHD, 17% for portal-inferior vena cava nodes, 8% for abdominal metastases, and 7% for multisite metastases. It is important to note that survival is longer in patients with coexisting pulmonary EHD and CLM because liver resection and pneumonectomy are performed in steps, and there may be selection bias, meaning that patients with further progression of intrapulmonary disease may not receive surgical treatment.
Looking also at CLM with abdominal metastases, in a recent multicenter study of 523 patients with CRC abdominal metastases and 77 with CLM, CLM had no effect on OS in the total population, but did have an effect on OS in patients with R0 resection of abdominal metastases. The authors concluded that liver metastases should only be a relative contraindication if the abdominal metastasis score is high. In conclusion, complete resection of CLM and EHD prolongs survival in this patient population after careful patient screening.
Concurrent liver metastases
Liver metastases are both concurrent and heterochronous, and many studies now focus on whether concurrent CLM is worse for survival. The decision to be made for resectable patients is staged or concurrent resection. Complications and mortality are high with concurrent resection, but recent studies have shown that complications and perioperative mortality with concurrent resection are similar to those with staged resection, and no differences in long-term outcomes have been found.
A recent meta-analysis of staged versus concurrent resection showed no difference in the outcome of the cancer itself, and concurrent resection patients had shorter total hospital stays and fewer complications. Retrospective studies have also shown that complications and mortality are similar between concurrent and staged resection even for most hepatic resections.
Concurrent resection is safer in some patients, but most studies have concluded that the selection bias for staged resection is greater and that patients who are expected to have more complications are usually given staged surgery. Simultaneous resection is the preferred option for some patients, thus avoiding secondary surgery, early completion of surgical treatment, and earlier initiation of adjuvant therapy.
According to expert consensus, the decision to perform hepatic resection or resection of the colorectal primary tumor first in staged surgery is based on the comorbidities of the primary tumor, such as the presence of obstruction, perforation, bleeding, and progression of marginal resectable CLM during the course of primary tumor treatment.
The decision of concurrent treatment is based on the complexity of liver and primary tumor resection and patient comorbidities. For rectal cancer, hepatic resection is more appropriate to prevent the liver from not receiving treatment during rectal radiotherapy. It is typical for concurrent treatment to have hepatic resection first so that the central venous pressure is lower. In either order, R0 resection should be achieved at both resection sites.
If liver metastases are unresectable, resection of the primary tumor will not improve survival and should only be performed if local symptoms are not well controlled by other means.
Measures to improve resection
There are several options to improve FLR if marginal FLR is expected, including systemic chemotherapy, portal vein embolization (PVE), second-stage hepatectomy, and liver dissection combined with portal vein ligation (ALPPS).
Systemic chemotherapy
For unresectable patients, systemic chemotherapy remains the standard first-line treatment. For initially unresectable CLM, systemic chemotherapy can reduce the tumor load to some extent and make resection feasible. If the initial start is due to anatomic location, preoperative chemotherapy can result in complete resection rates of 12.5-32.5%. The most commonly used chemotherapy regimens include FOLFOX and FOLFIRI, and more recently monoclonal antibodies such as bevacizumab or cetuximab have been used in combination with chemotherapy to increase treatment response rates.
Hepatic steatosis or steatohepatitis is associated with treatment with fluorouracil and irinotecan, hepatic sinusoidal dilatation and congestion is seen with longer oxaliplatin use, steatohepatitis and hepatic sinus injury are associated with increased perioperative complications, and steatohepatitis is associated with increased mortality.
Scoggins found no increase in mortality or complications at an average of 4.2 months of neoadjuvant chemotherapy, and steatohepatitis was more common in obese patients treated with neoadjuvant chemotherapy. Bevacizumab in combination with chemotherapy does not increase complications, but needs to be discontinued 6-8 weeks before surgery. Data suggest that bevacizumab in combination with oxaliplatin may prevent hepatic sinusoidal injury. There are no published articles suggesting that cetuximab and panitumumab can directly cause liver injury.
Approximately 2/3 of patients experience recurrence after resection of colorectal cancer metastases. Do resectable metastases require preoperative chemotherapy?The EORTC trial of patients with resectable metastases treated with 6 weeks of FOLFOX perioperative chemotherapy + surgery or direct surgery found an improvement in 3-year disease-free survival in the neoadjuvant chemotherapy group. This study was not sufficient to assess the impact on overall survival, and follow-up studies showed no difference in OS between the two groups.
The results of retrospective studies are highly variable; Adam found that neoadjuvant chemotherapy did not improve survival in heterochronous solitary metastases and only increased complications; Zhu found that patients with more than 2 poor prognostic factors benefited from neoadjuvant chemotherapy; Malik retrospectively examined more than 600 patients and did not find differences in DFS and OS between the neoadjuvant and direct surgery groups; Reddy’s large multicenter retrospective study resectable concurrent colorectal cancer with liver metastases, found that chemotherapy after liver resection increased OS over neoadjuvant chemotherapy.
These variations in outcomes have led to differences in expert opinion, such as whether resection should be performed early and whether the duration of neoadjuvant chemotherapy should be chosen carefully for most patients.
Portal vein embolization
PVE is used preoperatively to treat borderline FLR to increase the safety of resection treatment. Physiologic response refers to atrophy-hyperplasia syndrome (AHC), which has the potential to increase portal blood flow to the non-embolized liver lobes PVE is performed under local anesthesia and guided by imaging. It takes at least 3 weeks to achieve a stable state of liver regeneration.
FLR hyperplasia reduces the risk of postoperative hepatic failure and allows for curative extended hepatectomy, especially in patients with small FLRs at the margin of resectability.PVE increases FLR by 7-27%. The indocyanine green secretion assay and 99mTc-GSA scintillation scan assay were used to test functional status after PVE and appeared to improve functional status significantly more than and faster than hyperplasia.
PVE is safe, with a complication rate of less than 10%, and PVE results in resection rates of over 60% and R0 resection in over 70% of patients undergoing resection. liver surgery after PVE is also safe, with complications of 19-55% and perioperative mortality of 1-7%.
However, there is a concern that tumor growth rates may increase after PVE at both embolized and non-embolized sites. This hypothesis is based on the following: by increasing hepatic artery and portal blood flow, local growth factor levels can be increased, leading to tumor growth. Several studies have confirmed that this is indeed the case in liver metastases from colorectal cancer.
The addition of chemotherapy between PVE and surgery can slow down this progression and improve long-term survival. Because of the doctrine of increased growth rate, studies have also examined whether bevacizumab could potentially affect tumor growth after PVE, but did not reach statistical significance. It was initially thought that neoadjuvant chemotherapy could reduce hepatic hyperproliferation if the patient continued, but recent studies have shown this idea to be false and that chemotherapy has no inhibitory effect on hepatic hyperproliferation.
Most of the contraindications to PVE are relative and include tumor invasion of the portal vein, portal thrombosis, severe portal hypertension, uncorrectable coagulation abnormalities, renal insufficiency, and biliary dilatation insufficient for drainage of the FLR.
Imaging is performed 3-6 weeks after PVE to assess the extent of hyperproliferation, determine the patient’s new FLR, and decide whether curative resection is feasible.
Second-stage hepatectomy
Second-stage hepatectomy removes those liver metastases that are initially unresectable and improves survival compared to patients treated with chemotherapy alone. Stage II resection is usually performed after 4-6 cycles of chemotherapy. Patients who have responded to treatment or are stable are first reviewed with images and undergo stage 1 hepatectomy.
Phase I resection is usually the removal of all metastases on the FLR, as small as possible, and avoids hilar resection and damage to the contralateral liver. All metastases on the FLR are removed prior to PVE to avoid increasing the tumor growth rate.
After 4-6 weeks, this period may or may not be treated with chemotherapy, with repeat imaging to assess liver regeneration, followed by stage II hepatectomy. Complications after stage I hepatectomy are approximately 11-17% without death. Post-stage I complications should be as minimal as possible to ensure stage II resection, and stage I resection alone has no benefit for survival.
Seventy-six to 87% of patients who undergo stage I resection can undergo additional stage II surgery. The second-stage R0 resection rate is 58-79%, with a 3-year OS of 50-84%. Survival time is a function of good tumor biology and complete resection of liver metastases.
Liver dissection combined with portal vein ligation / in situ liver dissection
ALPPS is an alternative approach to PVE for increasing FLR, which is still in the developmental phase and has shown good promise. In the first stage, an exploratory procedure is performed, with ligation of the right portal vein and, for further enlargement of the right hepatectomy in the future, separation of the liver parenchyma along the right side of the sickle ligament or along the Cantile line. A CT volume examination was performed one week later and a second surgery to remove the damaged liver was performed soon after.
ALPPS increases FLR in the range of 63-87%, complications 53-71%, and mortality 0-22%. Recently reported mortality rates for ALPPS are lower at 4.7-5.6%, which may be related to refinement of techniques and indications. However, complications and mortality are particularly high in patients with hepatic cholangiocarcinoma with preoperative biliary stasis and bile duct aggregation, and some authors have questioned the appropriateness of these patients as an indication. The nature of this technique is different and no long-term oncologic outcomes have been studied.
The main advantage of ALPPS over PVE is the shorter interval required to complete the procedure. As with PVE, common reasons for failure include disease progression and failure to obtain the expected FLR.
ALPPS takes less than 10 days to achieve FLR hyperplasia compared to PVE, which takes at least 3 weeks. ALPPS takes so little time due to ISS, which completely disconnects the IV segment of the liver from the vessels preventing the formation of collateral circulation between the left and right lobes.
ALPPS can be continued to convert patients to operable when they have not acquired sufficient hyperplasia after PVE, and usually such patients have rapid growth within 3 days after in situ liver transection, with an average volume increase of 63%.
Unresectable liver metastases
Ablation therapy
Ablative treatments include radiofrequency ablation (RFA), microwave ablation (MWA) and cryoablation. Temperature ablation involves immediate cell death by changing the temperature of the metastasis site. The advantages of ablation therapy are that it preserves as much liver parenchyma as possible, uses a percutaneous or endoscopic approach, and does not interfere with future treatment options and has fewer complications.
Ablative techniques are used to kill tumor cells and surrounding normal liver tissue cells by changing the temperature sufficiently to produce irreversible temperature damage, a change called coagulative necrosis.RFA is the most commonly used ablative treatment, primarily for the treatment of CLM.The limitations of these approaches are influenced by the size of the tumor and the probe, and are now used extensively in patients with unresectable or significant complications.
RFA
RFA involves the placement of electrodes into the tumor under imaging guidance and the application of radiofrequency or temperature energy to the damaged cells and surrounding normal liver tissue. Specific, high-frequency alternating current produces hemostasis and protein denaturation by electrocoagulation, and immediate cell death at 60 degrees, producing an ablation zone.
RFA can be performed either percutaneously, endoscopically or open, and is most effective for metastases smaller than 3 cm that can be covered with a single probe. For larger lesions it is necessary to cross multiple probes to obtain adequate ablation, but this is usually technically difficult. Open or laparoscopic placement of probes is preferable to percutaneous placement of probes, along with exploration and intraoperative ultrasound of the liver to detect occult abdominal and hepatic metastases.
RFA has some limitations on probe placement. There is a risk of inadequate ablation when large vessels are approached, mainly due to the flow of blood through the vessels that removes heat from the target site. This thermal ablation can be overcome by temporary vascular obstruction, as in the Pringle strategy, and RFA should not be performed near adjacent bold ductal structures, especially only 1-2 cm from the hepatic hilum, leading to the risk of bile duct stenosis and fistula.
There are only two phase II trials and numerous retrospective studies on cancer outcomes with RFA. median survival after RFA for CLM is 24-45.3 months, with a 5-year OS of 18-33%, and median survival after hepatic resection is 41-80 months, with a 5-year OS of 48-71%. Even the lowest local recurrence rate in RFA is worse than the local recurrence rate after resection.CLM in RFA is unresectable more advanced liver metastases compared to resection, which removes occult liver parenchymal micrometastases.
Three clinical questions exist: is RFA comparable to hepatic resection for resectable liver metastases, does RFA reproduce the curative purpose of hepatic resection, and does RFA combined with chemotherapy benefit unresectable CLM?
The first question is the most difficult to answer. Many authors have retrospectively compared resectable CLM with unresectable CLM treated with RFA, and the results certainly show that RFA is worse than resection in terms of local control rates. There are obvious differences in the comparison process, and it must be flawed to conclude that RFA is worse than resection.
It is certainly true that RFA has a higher rate of local recurrence, which leads to lower survival. These data support the continued use of resection as the gold standard for resectable CLM. Some authors suggest that repeated minimally invasive use of RFA can overcome the pitfalls of local recurrence by selecting the right patient, a model similar to that of breast-conserving treatment for breast cancer.
The ultimate role of RFA should be based on a good identification of the different advantages and disadvantages inherent in RFA and excision, so that the different indications can be selected to better exploit the advantages of both treatments.
Can RFA be used and benefit patients with resectable CLM? 52 unresectable CLM in the phase II trial EORTC40004 were treated with RFA + excision and had a 5-year OS of 43%. Karanicolas treated unresectable, poor-prognosis CLM with ablation + surgery and had a 5-year OS of 56%. These data support the use of RFA for unresectable CLM, the potential use of RFA as a substitute for stage II hepatectomy, faster patient recovery, earlier initiation of adjuvant therapy, and avoidance of disease progression during stage II surgery.
The CLOCC trial randomized 119 patients to either the chemotherapy arm or the chemotherapy + RFA arm. patients in the RFA arm had a PFS of 16.8 months, significantly better than the 9.9 months in the chemotherapy alone arm. However, recruitment to the trial was slow and there was no final assessment of OS and no way to know if the PFS benefit would translate into an OS benefit.
MWA
is a treatment method for rapid delivery of hyperthermia to treat large liver damage. Electrodes are placed into the tumor under ultrasound or CT guidance, microwave coagulator capacity is converted to microwaves, and coagulative necrosis causes cell death and tissue damage, an effect that is less dependent on tissue changes, is in some ways superior to RFA, is safer to apply, and potentially leads to lower local recurrence and complications. The shorter wavelengths allow for rapid heating in tissues of varying density with less energy loss.
However, this technique has 2 disadvantages compared to RFA, thermal fading effects near large vessels and incomplete destruction of large damage after charring. The advantages of this technique have been well demonstrated in animal models, and MWA may be more beneficial for damage larger than 3 cm, because dehydration and charring seem to be less important than in RFA.
Recent multicenter trials have shown a low local recurrence rate of only 6%, yet the greatest effect was seen in recurrence-free survival for damage ≥3 cm, contrary to the results of RFA studies. As with RFA, MWA has not been well studied and the theoretical benefit has not been clearly translated into improved clinical outcomes.
Cryoablation
Cryoablation involves the distribution of liquid nitrogen or argon gas into the liver tumor under ultrasound guidance. Ice crystals can damage cellular structures and kill tumor cells during the rapid freezing process. Cryoablation has fallen out of favor, mainly because of the high complication and recurrence rates compared to RFA. Among the complications is the lethal complication of cold shock, which manifests as hypothermia, coagulation abnormalities, respiratory failure and renal failure.
Hepatic artery infusion
Hepatic arterial infusion (HAI) is performed by a pump connected to a catheter that is then implanted in the hepatoduodenal artery with the catheter tip placed in the hepatoduodenal – hepatic artery junction. This treatment may be used in combination with systemic chemotherapy and may be supplemented with open or laparoscopic hepatectomy or RFA.
Chemotherapy via the hepatic artery reduces toxicity because metastatic liver cancer is almost exclusively supplied by the hepatic artery, whereas normal liver tissue receives its blood supply primarily from the portal vein. The direct action of chemotherapeutic agents increases the amount of cytotoxic drugs without increasing systemic side effects. Because of the high capacity of the liver to take up FUDR, hepatic arterial administration allows for almost equivalent doses of systemic chemotherapy to be used without increased toxicity.
Phase I and II HAI studies have shown response rates of 52-75% in previously treated patients and even higher response rates in patients who have not been treated with chemotherapy. The combination of HAI and systemic chemotherapy showed response rates in excess of 90%, with 24-47% of patients converting to resectable. The conversion rate to resectable is even higher, 53-57%, in patients who have not received chemotherapy, including those with severe liver invasion.
HAI has also been used in adjuvant studies in patients at high risk of recurrence after CLM resection, with significant improvement in DFS but no improvement in OS. comorbidity of HAI pumping is about 20%, and about half can be adjusted for continuation of therapy. Biliary sclerosis is a long-term complication that can be resolved with biliary stenting and does not affect OS.
Radiotherapy
Transarterial radiotherapy (TACE) can be administered in a traditional manner such as using latex or ethidium oil in combination with chemotherapy infusion, or using drug-eluting bead-loaded irinotecan (DEBIRI-TACE). No one compares these two modalities, but usually each institute has its own preference.
DEBIRI was originally reported in 2006 and the toxicology data showed more severe post-embolization complications compared to radioembolization (RE), with 40% of patients having right upper abdominal pain, 80% having fever, 27% having nausea, and 70% having elevated transaminases. Despite these symptoms, 78% of patients achieved a treatment response, with more than 90% reporting an improvement in status beyond 4 months, a median response rate of 6 months, and a median survival of 25 months.
A prospective randomized study of patients with colorectal liver metastases who failed standard chemotherapy were randomized to DEBIRI or FOLFIRI chemotherapy. median survival improved significantly to 22 months in the DEBIRI group and 15 months in the FOLFIRI group.
RE
RE is the most well-studied embolization technique for the treatment of CLM. 90Y can be used in RE, with two commercially available particles, one containing biocompatible resin (SIR-Spheres) and the other containing glass (TheraSphere). Involvement of the portal vein is a contraindication to SIR-Spheres. The most common side effect is gastrointestinal toxicity.
The first step in reducing toxicity is to perform X-rays of the distribution of the abdominal aorta and superior mesenteric artery and to pulsate the hepatic arterial network. Gastrointestinal ulcers are associated with particles from the extrahepatic arteries that nourish the gastrointestinal tract. tc99 m Large particle polymerase protein scans are used to assess the presence of arteriovenous short circuits prior to treatment and to identify non-target organs, such as the gastrointestinal tract and lungs.
The pulmonary shunt fraction (LSF) is calculated based on images and dose reduction. Toxicity is usually mild and resolves spontaneously within 1-4 weeks, with symptoms including fatigue, abdominal pain, nausea, and anorexia. Response rates are 12.9-35.5%, with 24-65% achieving disease stabilization. 90Y median OS is 10.2-12.6 months, which is obtained in patients after chemotherapy failure.
Extracorporeal irradiation
Historically, external radiation therapy has not been used for liver tumors due to the small therapeutic window between benefit and hepatotoxicity. Stereotactic radiation therapy, initially used in neurosurgery, allows for the precise delivery of highly concentrated ionized radiation to the target site, called body stereotactic radiation therapy. 1-year or 2-year local control rates are 67-100% and 55-92%, with a median survival of 20.5-34 months.
Chang reported a dose-dependent local control rate for colorectal cancer liver metastases, with a local control rate of 84% at 18 months for doses ≥42 Gy and only 43% for doses <42 Gy. Accordingly, the authors recommend a total dose of 42 Gy in 3 fractions.
Conclusion
Surgery should still be performed for resectable CLM. There are many options for more progressive patients to undergo extended resection, such as systemic chemotherapy, PVE, phase II hepatectomy, ablation, and HAI, but there are few phase III trials comparing these treatment modalities, and the choice of treatment is largely patient-dependent.
Treating liver metastases requires a multidisciplinary approach, and knowledge about all treatment options for CLM is constantly being updated. If a patient undergoes a treatment modality to increase the likelihood of resection, this is the primary goal of treatment. If the patient is persistently unresectable, then treatment is primarily aimed at prolonging progression-free survival and overall survival.