Practical questions in liver metastases of colorectal cancer: general principles of treatment

Abstract

Liver metastases of colorectal cancer are currently treated by multidisciplinary teams using strategies that combine chemotherapy, surgery and ablative techniques. Many patients classically considered non-resectable can now be rescued by neoadjuvant chemotherapy followed by liver resection, with similar results to those obtained in initial resections. While many of those patients will recur, repeat resection is a feasible and safe approach if the recurrence is confined to the liver. Several factors that until recently were considered contraindications are now recognized only as adverse prognostic factors and no longer as contraindications for surgery. The current evaluation process to select patients for surgery is no longer focused on what is to be removed but rather on what will remain. The single most important objective is to achieve a complete (R0) resection within the limits of safety in terms of quantity and quality of the remaining liver. An increasing number of patients with synchronous liver metastases are treated by simultaneous resection of the primary and the liver metastatic tumours. Multilobar disease can also be approached by staged procedures that combine neoadjuvant chemotherapy, limited resections in one lobe, embolization or ligation of the contralateral portal vein and a major resection in a second procedure. Extrahepatic disease is no longer a contraindication for surgery provided that an R0 resection can be achieved. A reverse surgical staged approach (liver metastases first, primary second) is another strategy that has appeared recently. Provided that a careful selection is made, elderly patients can also benefit from surgical treatment of liver metastases.

Introduction

In recent years, advances in surgical techniques and modern chemotherapy have introduced revolutionary changes in the multidisciplinary treatment approach to colorectal liver metastases (CLMs) ¹. The results are highly encouraging and surgeons have the possibility to tailor the therapeutic approach according to each patient's individual characteristics. ‘Which treatment for which patient?’ is no longer a question for the future but the current practice in CLM.

What are the principles of indications/contraindications for surgical treatment?

Until recently, evaluation of the liver focused on the location and number of lesions, and disease-free intervals. Classical contraindications for resection included: bilobar disease, more than four lesions, large tumours (>5 or 10 cm) and extrahepatic disease ²,³,⁴,⁵. In current practice, while the prognostic impact of these factors is acknowledged, none of them are considered absolute contraindications for surgery. Nowadays, the evaluation of the surgical candidate must be conducted to determine if a complete and safe resection can be performed ⁶. Complete (R0) resection is defined as elimination of all metastatic lesions with free margins of resection. If the liver function is normal, and the remaining liver will be at least 25–30%, the resection can be done safely. The San Francisco Consensus Conference ⁶ represented an important shift in the focus of the evaluation before liver resection. Instead of being defined by what is removed, resectability should now be determined by what will remain. This will result in an increasing number of patients being eligible for surgery.

What are the factors that affect the overall prognosis?

The overall prognosis is negatively affected by synchronous presentation, CEA levels >200 ng/ml, positive lymph nodes, serosal invasion, shorter intervals in appearance of metachronous lesions, number of lesions seen on CT scan preoperatively, resection margins <1 cm, metastases >5 cm and bilobar disease. These are adverse prognostic factors, not contraindications for surgery ⁷.

When is the tumour burden too much for surgery?

The tumour burden is too much for surgery only when an R0 resection cannot be expected. In a recently published large prospective single-centre experience, Figueras et al. ⁷ suggested that liver resection combined with preoperative and postoperative chemotherapy offers the possibility of long-term survival to patients with multiple, bilateral, large CLMs, even with resectable extrahepatic disease. Surgery should be offered to patients with expanded indications because hepatic resection may increase longevity and improve quality of life in selected patients with CLM.

What is the current impact of the surgery of CLM?

Surgical resection is presently the only approach that offers patients with CLM a substantial chance of cure. Resection of CLM is safe and effective; it should be considered the treatment of choice for this disease and proposed even for advanced lesions. The survival rate of patients who undergo a complete resection is about 40% at 5 years. Operative mortality within 2 months should be <5%, and postoperative morbidity was 28% in the experience of Figueras et al. ⁷.

What to do with patients with colorectal liver metastases who present with unresectable disease at the time of diagnosis?

Most patients with CLM will present with unresectable disease at the time of diagnosis. The current definition of unresectable includes: >70% of the liver affected or more than six segments. Other causes of unresectability are invasion of both portal branches or the three hepatic veins. Those patients who present with unresectable disease can be rescued by neoadjuvant chemotherapy. The prognosis of unresectable CLM might be improved if a radical surgical resection (rescue surgery) of CLM could be performed after a response to chemotherapy ⁸,⁹.

What is the role of neoadjuvant chemotherapy?

Unresectable disease is the most frequent presentation in patients with CLM. Bismuth et al. ⁸ in 1990 reported patients who had a significant down-sizing of their hepatic lesions with neoadjuvant chemotherapy. Those patients were then brought to surgery for an attempted R0 resection, obtaining a 5-year survival rate of 28%. Mortality and morbidity were similar to those resected de novo. These results represented a revolution in liver surgery ¹⁰.

How to combine neoadjuvant chemotherapy and surgical treatment?

Those patients initially classified as unresectable should be re-evaluated for resectability after neoadjuvant therapy ¹¹,¹²,¹³,¹⁴,¹⁵,¹⁶,¹⁷,¹⁸,¹⁹,²⁰,²¹. One strategy consists of using four to six cycles of one of the newer effective chemotherapeutic combinations such as FOLFOX or FOLFIRI + Bevacizumab or Cetuximab and re-evaluation of resectability. Do not use excessive number of cycles of chemotherapy because important liver steatohepatitis can develop ²²,²³. A recent report suggested that vascular complications manifesting as sinusoidal obstruction are a prominent feature in chemotherapy-induced liver toxicity ²⁴,²⁵. The combination of steatosis and sinusoidal obstruction greatly enhances the risk of serious intraoperative bleeding. Another important notion is that complete radiological response must not be achieved because it produces great difficulty at the time of mapping for surgical resection. Besides that, complete radiological response almost always is associated with recurrence ²⁶. At the time of planning the surgical resection, the CT image that must be taken into account is the first one rather than that following chemotherapy. Hepatic surgery should be performed as soon as the disease becomes resectable. The patient must be operated during the therapeutic window provided by the response to chemotherapy. Otherwise we will be operating a patient during tumour progression and the results will be poor ²⁷. The surgical procedure must be scheduled 6–8 weeks after the discontinuation of Bevacizumab. This period of 6–8 weeks is important so as to avoid the complications produced by the antiangiogenic effects of Bevacizumab ²⁸, although another report did not find any increase in the morbidity following hepatectomy with the perioperative use of Bevacizumab ²⁹. Alternatively one or two cycles of FOLFOX or FOLFIRI without Bevacizumab can be administered during this period before surgery. Modern chemotherapy allows 17–50% of patients with unresectable CLM to be rescued by liver surgery ⁹. Despite a high rate of recurrence, 5-year survival is 33% overall, with a wide use of repeat hepatectomies and extrahepatic resections.

Is surgery of CLM indicated in patients with progression of the tumours while on neoadjuvant chemotherapy?

In this clinical scenario surgery must not be carried out. Liver resection is able to offer long-term survival to patients with multiple CLMs provided that the metastatic disease can be controlled by chemotherapy before surgery. Tumour progression before surgery is associated with a poor outcome, even after potentially curative hepatectomies. Tumour control before surgery is crucial to offer a chance of prolonged remission in patients with multiple metastases ²⁷.

How many patients with colorectal cancer who have undergone ‘curative’ hepatic resection will develop recurrences?

At least 60–80% of patients who undergo a hepatic resection for CLM will develop a local, regional or distant recurrence. Those recurrences must be expected in the 30 months that follow the resection, with 85% of the recurrences occurring in this period. In 30% of those patients that recur the metastases are confined to the liver. In these patients a repeat resection or an ablation procedure must be considered ³⁰.

Are repeat resections a safe approach?

This approach is possible due to the regeneration capacity of the liver ³¹,³²,³³. It takes 3–6 weeks for the liver to regenerate to its presurgical volume. Mortality, morbidity and survival are similar to those following the initial resection. Combined extrahepatic surgery can be required to achieve tumour eradication. Repeat hepatectomies appear worthwhile when potentially curative ³⁴,³⁵,³⁶,³⁷. The definition of ‘potentially curative’ includes: margins of resection free of disease and resection of extrahepatic disease. In repeat resections the general rule also applies that it does not matter how many lesions the patient has provided that an R0 resection can be achieved within limits of safety in terms of liver volume and function. Based on this, the new trends in surgical evaluation now focus on whether a complete resection can be safely performed. The risk factors associated with poorer prognosis in second and third resections are similar to those for primary resections: number of tumours resected, the size of the largest lesion, bilobar involvement, and positive margins. The most important ominous factor is the presence of positive margins.

Can surgical resection be combined with ablative techniques?

At present, radiofrequency ablation (RFA) is the most widely used technique of liver ablation ³⁸,³⁹,⁴⁰,⁴¹,⁴²,⁴³,⁴⁴,⁴⁵,⁴⁶. Recent reports showed that it is especially useful for lesions smaller than 3 cm and located at a certain distance from the important vessels ⁴⁷,⁴⁸. This is a serious limitation of RFA: the dissipation of heat produced by the blood flow in an adjacent vessel greatly diminishes its efficacy. Intraoperative use of RFA is an important tool in achieving an R0 resection in cases otherwise considered unresectable ⁴⁹,⁵⁰. In this sense, RFA increased the number of patients who can be surgically treated. In cases of recurrences, some authors prefer percutaneous RFA to repeat hepatectomy when feasible and safe because it is less invasive, using repeat hepatectomy only when RFA is contraindicated or fails ⁵¹. RFA is safe and as effective as surgical resection in prolonging survival in patients with lesions <3 cm. For lesions >3 cm, surgery must be preferred whenever possible.

Two-stage hepatectomy: how can it be integrated in a multimodal treatment approach?

In cases of multiple, bilateral, unresectable CLMs, the strategy must be multimodal – starting with neoadjuvant chemotherapy. In patients with a normal liver, portal vein embolization (PVE) is indicated when the future liver remnant volume is predicted to be <30% ⁵²,⁵³,⁵⁴,⁵⁵,⁵⁶,⁵⁷,⁵⁸,⁵⁹,⁶⁰,⁶¹,⁶²,⁶³,⁶⁴,⁶⁵,⁶⁶. PVE may also be useful in patients who have evidence of chemotherapy-related liver injury. If there is a positive response to chemotherapy, then surgery could be used ²⁷. In a first stage, limited resection of CLM in one lobe is performed combined with RFA of other lesions and PVE of the contralateral portal vein. Alternatively, ligation and alcoholization of the portal vein can be performed intraoperatively. PVE, ligation and/or alcoholization produces, on the one hand, an important atrophy of the implicated lobe, and on the other hand, a significant hypertrophy of the contralateral lobe. This is a strategy to enhance the volume of the future remnant liver. Studies have shown an approximately 10–25% increase in the size of the liver remnant after PVE ⁵²,⁵³,⁵⁴,⁵⁵,⁵⁶,⁵⁷,⁵⁸,⁵⁹,⁶⁰,⁶¹,⁶²,⁶³,⁶⁴,⁶⁵,⁶⁶. This hypertrophy is complete by the third week. In a second stage, a major resection of the contralateral liver can be carried out in conditions of safety ⁶⁷,⁶⁸,⁶⁹. Usually, those patients will receive postoperative chemotherapy. With this approach, currently there are no limits to resection other than those imposed by the volume and function of the remnant liver.

Combined or staged procedures in the treatment of synchronous CLMs?

The optimal timing of the surgical treatment of synchronous CLM is a matter of controversy ⁷⁰,⁷¹,⁷²,⁷³,⁷⁴,⁷⁵,⁷⁶,⁷⁷,⁷⁸,⁷⁹,⁸⁰,⁸¹,⁸²,⁸³,⁸⁴,⁸⁵,⁸⁶. Studies that compared combined versus staged procedures demonstrated that the overall survival is similar, but the complication rate for the combined procedure was significantly less than for a staged procedure ⁷⁰,⁷¹. However, combined procedures tended to include more right colectomies and more limited hepatic resections. Staged procedures tended to include more difficult low anterior or abdominal perineal resections or major liver resections. In highly specialized centres there is a trend toward combined procedures whenever possible, because simultaneous liver resection is safe, avoids two operations, avoids the possible hepatotoxicity of neoadjuvant chemotherapy, and allows resection of CLM that may ultimately metastasize further ⁷¹. In our experience with 61 patients undergoing simultaneous liver and colorectal resection this approach was safe and had similar short- and long-term results to staged procedures in this carefully selected group of patients ⁸⁷. The most important factor in selection of patients for combined resection in our series was the extent of liver to be resected. It is our policy to avoid combined resection in those patients who need a major liver resection.

Staged surgery: which procedure first? The reverse approach

The classic staged surgical approach consists of treatment of the primary followed by treatment of the CLM. However, in many patients with advanced synchronous CLMs, the metastases progress during treatment of the primary, precluding curative treatment. Based on this observation, Mentha et al. ⁸⁸ designed a management strategy that involves high-impact chemotherapy first, resection of CLM second and finally removal of the primary tumour in those patients with adverse prognostic factors. In the experience of these authors, this new ‘reverse’ approach produced resectability and survival rates better than those expected from the published data on patients with disease of similar severity. The obvious candidate for this novel approach would be a patient with non-obstructive primary colonic tumour. The rationale for the reverse approach is that the lesion that kills the patient is the metastasis.

Is there any role for trans-metastases hepatectomy?

In the setting of multiple, bilateral, unresectable CLMs, sometimes a situation exists in which the only possible future hepatectomy plane passes through a liver metastasis. Elias et al. ⁸⁹,⁹⁰ proposed a new curative technical approach to these patients. Firstly, the ill-sited LM, which is located in the only feasible future resection plane, is ablated using RF, then the hepatectomy is performed through this ablated LM. According to these authors, this is a safe and useful technique.

What is the role of an R0 resection regarding prognosis?

Many of the factors that affect the overall prognosis do that by making R0 resection difficult to achieve ⁹¹,⁹²,⁹³,⁹⁴. But, if an R0 resection can be achieved, they lose their negative impact on survival. Studies showed significant 5-year survival rates in patients with more than four lesions if an adequate R0 surgical resection was performed ⁹¹. If metastases can be resected completely with sufficient remaining liver, it does not matter how many lesions the patient has. Similar considerations can be made regarding the size of the metastatic lesions, the synchronous appearance, a disease-free interval <1 year in the appearance of metachronous lesions, and bilobar involvement ⁷. In summary, at present, the only major predictor of prognosis is an R0 resection.

Free margins. Is the 1 cm rule still valid?

The single most important factor that negatively affects the prognosis after resection is the involvement of the resection margin for the tumour ⁹¹,⁹²,⁹³,⁹⁴. Classically, a 1 cm margin has been considered necessary to avoid liver recurrence. The importance of this factor is enhanced by the fact that it is the only factor that could be directly influenced by the surgeon. This was a matter of controversy for many years. In a large series recently reported by Figueras et al. ⁹⁵, sub-centimetre non-positive resection margin failed to show any influence as an independent factor on liver recurrence. Currently, most authors agree that an expected resection margin <1 cm must not preclude the indication for surgery ⁹⁶,⁹⁷,⁹⁸.

Is the presence of extrahepatic disease an absolute contraindication for liver resection?

For many years extrahepatic disease has been accepted as a contraindication to liver resection. Recently a 5-year survival rate of 12–37% following liver resection was reported in selected patients with extrahepatic disease ⁹⁹. This benefit was independent of the location of the disease (lung, primary colorectal recurrence, retroperitoneal or hepatic pedicle lymph nodes, peritoneal carcinomatosis, miscellaneous). Currently, the presence of extrahepatic disease is no longer considered an absolute contraindication to hepatic resection. A careful selection of patients must be conducted. The minimum requirements are: responsive disease with chemotherapy and the expectation of an R0 resection. However, patients with tumoral invasion of the lymph nodes in the hepatic pedicle, particularly if the celiac nodes are affected, have poorer prognosis than those with solitary pulmonary metastases.

Resection of isolated pulmonary recurrence after resection of colorectal liver metastases. Is it a worthwhile operation?

Prolonged survival can be achieved with resection of isolated pulmonary recurrence after hepatic resection for CLM ¹⁰⁰,¹⁰¹,¹⁰². Studies conducted to identify a subgroup of patients who best benefit from hepatic and pulmonary metastasectomy among those with colorectal carcinoma metastases identified sequential detection of hepatic and pulmonary metastases as the strongest independent favourable prognostic factor. Patients with sequentially detected hepatic and pulmonary metastases from a colorectal primary are good candidates for aggressive metastasectomy. Simultaneous detection of these metastases does not warrant resection ¹⁰³.

Can elderly patients benefit from surgical treatment of liver metastases?

Yes, elderly patients can definitely benefit from surgical treatment of liver metastases. In our experience the elderly have a higher mortality after surgical treatment of CLM compared with the younger group ¹⁰⁴. In recent years, however, that difference was markedly reduced – probably due to improvements in perioperative care. The performance of major liver resections increased the mortality in the elderly. If the postoperative mortality is excluded, the results in terms of survival and disease-free survival are similar in both groups ¹⁰⁴,¹⁰⁵,¹⁰⁶,¹⁰⁷,¹⁰⁸,¹⁰⁹,¹¹⁰,¹¹¹,¹¹²,¹¹³,¹¹⁴,¹¹⁵,¹¹⁶,¹¹⁷. Therefore, the criteria on which the indication for surgery in the elderly is based must be the same as those for the rest of the population ¹⁰⁵.

Conclusions

Treatment of CLM is a dynamic and continuously evolving field. At present, the multimodal treatment approach allows us to individualize the treatment according to each patient's characteristics.