This Article Abstract Full Text (PDF) eLetters: Submit a response to this article Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services E-mail this article link to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints/Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Kemeny, N. Search for Related Content PubMed PubMed Citation Articles by Kemeny, N.

Hepatobiliary

Presurgical Chemotherapy in Patients Being Considered for Liver Resection

Memorial Sloan-Kettering Cancer Center, New York, New York, USA

Key Words. Neoadjuvant chemotherapy • Liver resection • Hepatotoxicity from chemotherapy

Correspondence: Nancy Kemeny, M.D., Memorial Sloan-Kettering Cancer Center, Gastrointestinal Oncology Service, 1275 York Avenue, Suite H916, New York, New York 10021, USA. Telephone: 212-639-8068; Fax: 212-794-7186

Received February 2, 2007; accepted for publication April 13, 2007.

	Learning Objectives

Top Learning Objectives Abstract Introduction How to Define Resectable... Initial Chemotherapy in... Synchronous Liver Metastases Neoadjuvant Therapy for Patients... Conclusion Disclosure of Potential... Acknowledgments References

 

1. Describe the approach to patients with unresectable liver metastases from colorectal cancer.
   
2. Identify the hepatotoxicities associated with chemotherapy.
   
3. Explain what happens to metastatic sites when there is a response.
   

	ABSTRACT

Top Learning Objectives Abstract Introduction How to Define Resectable... Initial Chemotherapy in... Synchronous Liver Metastases Neoadjuvant Therapy for Patients... Conclusion Disclosure of Potential... Acknowledgments References

 
The liver is a frequent site of metastatic disease for colorectal cancer patients. Approximately 15% of patients have liver metastases at diagnosis and another 50% develop metastatic disease to the liver over the course of their disease. Advances in systemic chemotherapy and surgical techniques for hepatic resection have led to longer survival times for these patients. There is no doubt that unresectable patients benefit from systemic chemotherapy. For patients who have resectable disease, the timing of chemotherapy is still not clear. This review addresses the pros and cons of presurgical chemotherapy. The benefits of preoperative chemotherapy include decreasing tumor size, controlling micrometastatic disease, assessing activity of chemotherapy, improving chemotherapy tolerance, and perhaps allowing some prediction of the success of liver resection. The risks for presurgical chemotherapy include liver toxicity, the risk for progression or growth of new sites, secondary splenomegaly, selection of resistant clones, and the possibility of leaving behind active tumor that is no longer seen because of a complete radiographic response. The challenge for the future is to develop a multidisciplinary team approach that can design the best treatment plan for patients with liver metastases.

Disclosure of potential conflicts of interest is found at the end of this article.

	INTRODUCTION

Top Learning Objectives Abstract Introduction How to Define Resectable... Initial Chemotherapy in... Synchronous Liver Metastases Neoadjuvant Therapy for Patients... Conclusion Disclosure of Potential... Acknowledgments References

 
There will be approximately 130,000 new cases of colon cancer in the U.S. in 2007. Fifteen percent of these patients will have liver metastases at diagnosis, and another 50% will develop liver metastases during the course of their disease [1]. With advances in adjuvant therapy after primary colon resection, there is hope that the number of cases with metastatic disease will decrease [2].

For those who develop metastatic disease, there are new chemotherapy agents (irinotecan [3] and oxaliplatin [4, 5]) and new targeted agents (cetuximab [6] and bevacizumab [7]) that have improved response rates and survival times. Unfortunately, even with these new agents, the 2-year survival rate remains about 40% for patients with metastatic disease. Therefore, the development of other treatment modalities is important.

Surgeons have demonstrated that resection of liver metastases increases survival [1, 8]. Several reports document a 30% 5-year survival rate for patients who undergo curative resection of their liver metastases [1, 8]. Although there are no randomized studies of surgery versus no surgery, there are natural history studies of patients who had resectable disease and did not undergo resection. In a review by Scheele et al. [9], of 902 patients with unresected liver metastases, only 21 patients survived 3 years, and none were alive at 5 years. Sixty-two patients had resectable disease but were not resected, and none were alive at 5 years. Those who were resected had a 5-year survival rate of 40%. Unfortunately, only 10%–25% of patients with liver metastases are candidates for liver resection.

New effective chemotherapeutic agents have increased response rates and also the possibility of resection. In 1996, Bismuth et al. [10] reported that patients who were initially unresectable could be treated with systemic chemotherapy, and if a good response was obtained, resection was possible in 16% (53 of 330 patients). These patients had a survival duration similar to those who were initially resectable. Thus, in patients who are clearly unresectable, few would dispute the utility of chemotherapy. However, in patients who are initially resectable, the benefit of preoperative chemotherapy is uncertain. This review covers some of the advantages and disadvantages of neoadjuvant therapy.

	HOW TO DEFINE RESECTABLE DISEASE

Top Learning Objectives Abstract Introduction How to Define Resectable... Initial Chemotherapy in... Synchronous Liver Metastases Neoadjuvant Therapy for Patients... Conclusion Disclosure of Potential... Acknowledgments References

 
In the early history of surgery for liver metastases, it was felt that patients with more than three metastatic liver lesions or bilobar disease were not appropriate for liver resection. However, newer data have demonstrated that even patients with poor prognostic characteristics can survive 5 years after curative liver resection [11–13]. Trials looking for prognostic indicators have devised a variety of clinical risk scores. A French review [12] of 1,568 patients found that age, number of metastases, size of the largest metastasis, carcino-embryonic antigen (CEA) level, primary tumor stage, disease-free interval, and positive resection margins were all significant prognostic factors. Giving one point to each of these factors, they divided patients into three risk groups, with corresponding 2-year survival rates of 79%, 60%, and 43% for groups with scores of 0–2, 3–4, and 5–7, respectively. At Memorial Sloan-Kettering Cancer Center (MSKCC), a multivariate analysis of many factors found that the size of the tumor (>5 cm), disease-free interval (<12 months), number of tumors (>1), positive lymph nodes from the primary, and preoperative CEA level >200 ng/ml were the most significant indicators of poor prognosis [11]. If one point is assigned for each of these factors, a score can be obtained that correlates with survival. Thus, a clinical risk score (CRS) of 0, 1–2, 3, or 4–5 predicted 5-year survival rates of 60%, 42%, 20%, and 18%, respectively. While poor characteristics definitely decrease survival, survival was still better than in those without resection, who rarely survive 5 years. An international group (LiverMetSurvey) looking at 2,122 patients undergoing liver resection, found overall 5- and 10-year survival rates of 42% and 26%, respectively, and showed that patients with greater or less than three metastases had 5-year survival rates of 48% and 24%, respectively (p = .001) [13]. In none of these series do the authors suggest that surgery should not be done in the patients with poor prognostic characteristics.

A computer program was developed to aid in selecting patients for resection (OncoSurge model) [14]. A panel of 16 experts decided that absolute contraindications for resection were extrahepatic disease, >70% liver involvement, liver failure, and/or being surgically unfit. Other factors such as age, primary tumor stage, and time to the development of metastases were not contraindications for resection. Today, even extrahepatic disease is not a clear indicator of unresectable disease [15]. There are now reports of resecting liver metastases in patients with positive lymph nodes [16] and small lung or ovarian metastases [15]. In a recent report, the 5-year survival rate was 28% for patients who had extrahepatic disease resected as well, with the number of metastases being a stronger indication of progression than location of disease [15].

With recent advances in surgery, more patients are being considered for resection [17]. One preoperative technique involves portal vein embolization, to remove the blood supply to the affected liver and thus induce hypertrophy of the nondiseased portion of the liver, allowing for more radical liver resections [18]. Better vascular clamping techniques [19], controlled anatomic resection [20], and the use of radiofrequency ablation [21] to small lesions of the remaining liver also expand the options for liver resection. The definition of resectable disease varies greatly. A larger consensus is needed to clearly define what is unresectable disease.

	INITIAL CHEMOTHERAPY IN UNRESECTABLE DISEASE

Top Learning Objectives Abstract Introduction How to Define Resectable... Initial Chemotherapy in... Synchronous Liver Metastases Neoadjuvant Therapy for Patients... Conclusion Disclosure of Potential... Acknowledgments References

 
For patients who do not have resectable disease, initial therapy, often called neoadjuvant therapy, may be a way to reduce tumor size and facilitate resection. Early retrospective studies from France [10, 22] were updated in a report of 1,439 patients with colorectal cancer managed from 1988–1999, where 1,104 patients were considered to have initially unresectable disease. After receiving chemotherapy (70% received 5-fluorouracil, leucovorin, and oxaliplatin [FOLFOX]), 12.5% became resectable, with 5-year survival and disease-free survival rates in the resected patients of 33% and 22%, respectively, with a median follow-up time of 48.7 months [23]. Eighty percent had had a tumor recurrence, of which 71% recurred in the liver. Four preoperative factors were associated with shorter survival times: (a) rectal primary, (b) three or more metastases, (c) tumor size >10 cm, and (d) cancer antigen (CA)19–9 level >100 u/l [23]. The mean adjusted 5-year survival rate with zero, one, two, three, or four of the above factors was 59%, 30%, 7%, 0%, and 0%, respectively.

Recent trials have prospectively evaluated neoadjuvant chemotherapy. Comparisons are somewhat difficult because of diverse reasons for patient unresectability. Some include the number of lesions or bilobar disease, while others look at technical reasons for why patients are unresectable, such as involvement of all three hepatic veins, both portal veins, or the retrohepatic vena cava, or that a resection would leave less than two hepatic segments or leave inadequate hepatic reserve. Many of these trials are listed in Table 1. In the trial by Pozzo et al. [24], patients were considered unresectable if they had six metastases with three per lobe; one lesion >5 cm if six metastases were present; or continuity with two hepatic veins, the inferior vena cava, or the liver hilum. The response rate to neoadjuvant chemotherapy was 47.5%, and 13 patients (32.5%) went on to resection. Although further systemic chemotherapy was given, the median disease-free interval was 14.3 months. In the Mayo Clinic trial [25], unresectability was defined as (a) involvement of three major hepatic veins, the portal vein bifurcation, or the retrohepatic vena cava; (b) involvement of the main right or left portal vein and the main hepatic vein of the opposite lobe; (c) disease requiring more than a right or left trisegmentectomy; and (d) six or more metastatic lesions distributed diffusely to both lobes of the liver. Twenty-five patients (60%) had a tumor reduction, and 17 (40%) underwent surgery. Retrospective review of these data demonstrated that 10% were actually resectable prior to neoadjuvant chemotherapy. There is no description of which patients became resectable, that is, patients with six or more lesions or patients with central lesions. With a median follow-up of 22 months, 11 patients have had recurrence, mostly in the liver. In the study by Ho et al. [26] of liver resection after 5-fluorouracil, leucovorin, and irinotecan (FOLFIRI), 40 patients were treated, 55% had a response, and four patients underwent liver resection. However, of these four patients, one had stable disease, again suggesting that resection could have been possible without the preoperative chemotherapy in some patients.

Chronomodulation of FOLFOX versus standard FOLFOX did not increase resectability [28]. In the European Organization for Research and Treatment of Cancer (EORTC) 05963 study, including 554 patients, 50 were resectable in each arm. However, we do not know yet whether the chronomodulation may have decreased hepatic toxicity.

The use of targeted agents may increase response rates and resectability. Rougier et al. [29] treated 23 patients with cetuximab and FOLFIRI, and seven (30%) became resectable. Another approach to maximize response in the liver is to use hepatic arterial infusion (HAI), because hepatic tumors are perfused by the hepatic artery [30]. Because floxuridine (FUDR) has an extraction rate of 94%–99% during the first pass [31], it is an ideal drug to use in treating liver metastases, and HAI may be combined with almost full doses of systemic therapy. Even in previously treated patients, high response rates can be achieved with the use of HAI therapy coupled with systemic therapy. A trial using HAI FUDR and dexamethasone with systemic irinotecan produced a 73% response rate in 38 patients (all previously treated), with responses in 13 of 15 patients who had received previous irinotecan [32]. Another study using HAI FUDR/dexamethasone with systemic oxaliplatin produced an 88% response rate in 36 patients (89% of whom were previously treated with systemic chemotherapy) [33]. In an updated analysis, 44 patients with clearly unresectable liver metastases were treated with combined HAI FUDR/dexamethasone and systemic oxaliplatin/irinotecan, and 35% were able to undergo liver resection after treatment (personal observation). Clavien et al. [34] treated 23 previously treated patients with HAI FUDR plus bolus cisplatin and doxorubicin and achieved a 23% resection rate.

The main toxicity from HAI FUDR is to the bile ducts rather than the liver parenchyma. Bile ducts, as well as liver metastases, are perfused by the hepatic artery, thus the main toxicity of HAI is biliary, which, when severe, is known as sclerosing cholangitis [35]. The toxicity usually occurs after multiple treatments. Therefore, if a response occurs rapidly, biliary toxicity should be less of an issue. In a study comparing the time needed to achieve a maximum response with HAI plus systemic therapy versus systemic therapy alone, the mean tumor shrinkage after 2 months of treatment was 60% with HAI plus systemic therapy, and 20% with systemic therapy alone. Thus, conversion to resectability after therapy may occur more quickly with HAI therapy [36]. More studies are needed to define how best to decrease liver metastases and enable resection.

	SYNCHRONOUS LIVER METASTASES

Top Learning Objectives Abstract Introduction How to Define Resectable... Initial Chemotherapy in... Synchronous Liver Metastases Neoadjuvant Therapy for Patients... Conclusion Disclosure of Potential... Acknowledgments References

 
The question of whether liver metastases can safely be resected at the time of resection of the primary colorectal carcinoma is still uncertain. Small metastases seen at the time of laparotomy for the primary lesion can easily be resected. When patients are evaluated for their primary and are found to have a large burden of metastatic disease, a question may be whether a simultaneous resection of the primary tumor and the metastases should be done, or whether preoperative chemotherapy should be offered. Bolton and Fuhrman [37] reported a 12% operative mortality rate with simultaneous surgeries that increased to 24% if the surgery included a major liver resection. Nordlinger et al. [12] reported a 7% mortality rate for simultaneous resections compared with 2% for staged resections (p = .01). Tanaka et al. [38], Martin et al. [39], and Weber et al. [40], on the other hand, reported that operative mortality and morbidity were comparable whether staged or simultaneous procedures were done. Tanaka et al. [38] reported that the resected liver volume was the only factor associated with postoperative complications of simultaneous colorectal and liver resection. It is the practice in many institutions to do a simultaneous resection with right colon primaries or when single synchronous metastases are found in the liver, and staged resections for rectal primaries or for patients with multiple liver metastases [41].

	NEOADJUVANT THERAPY FOR PATIENTS WHO ARE RESECTABLE

Top Learning Objectives Abstract Introduction How to Define Resectable... Initial Chemotherapy in... Synchronous Liver Metastases Neoadjuvant Therapy for Patients... Conclusion Disclosure of Potential... Acknowledgments References

 
The following sections discuss the pros and cons of neoadjuvant therapy for patients who are resectable (Table 2).

Decrease Tumor Size
It is known that large liver tumor size (>5 cm) indicates a worse prognosis [11, 12]. Although one might reason that decreasing the tumor size with chemotherapy would improve survival, this has not yet been demonstrated. Decreasing the size may make the surgery easier, but the question is whether residual microscopic disease remains. Resection of a smaller volume of disease may not improve prognosis if residual disease remains.

Control Micrometastatic Disease
When patients present with liver metastases, there is the possibility that there is undetectable disease in other parts of the body, and therefore it would be appropriate to control this disease before attempting liver resection. Also, if the patient is on chemotherapy for a period of time before resection, one can then eliminate patients who develop extrahepatic disease while on treatment and avoid the risks associated with major hepatic resection. Supporting this concept, the LiverMetSurvey group found that patients with more than five metastases survived longer if they were given neoadjuvant chemotherapy, with 5-year survival rates of 22% and 12% (p = .07) for the preoperatively and nonpreoperatively treated groups, respectively. However, for all patients, the group receiving preoperative chemotherapy had a 5-year survival rate of 38%, versus 50% for those without preoperative chemotherapy.

In another retrospective review at MSKCC, preoperative chemotherapy did not improve survival. In a review of 230 patients who received HAI plus systemic therapy after liver resection [42], the median survival times in patients who did and did not receive preoperative (neoadjuvant) therapy were 63 and 115 months, respectively (p = .26). The group who received neoadjuvant therapy had a higher CRS and a greater number of metastases, which may explain the longer survival in those without preoperative chemotherapy. In order to determine if neoadjuvant therapy benefits patients with poor prognostic indicators (CRS, 3–5), these patients were separated from those with a good CRS (0–2). Among patients with a poor CRS (3–5), the median survival times were 60 and 55 months, respectively, for those who received preoperative chemotherapy and those who did not (p = .79). Among patients with a CRS of 0–2, in those who received neoadjuvant therapy versus those who did not, the median survival times were 67 versus 114 months, respectively (p = .41). The results in this report may have been influenced by the addition of postoperative therapy after the liver resection. It is not clear that preoperative chemotherapy is helpful in resectable patients, but this concept requires further study.

Assessment of Chemotherapy Activity
Another advantage of preoperative chemotherapy is that it can be used to assess response to chemotherapeutic agents prior to resection. By knowing if certain drugs produce a response, postresection therapy can be designed more efficiently.

Better Chemotherapy Tolerance
After liver resection, patients may not be able to tolerate full doses of chemotherapy. Two adjuvant studies using systemic irinotecan [32] or oxaliplatin/5-FU/LV [43] demonstrated that full doses of systemic chemotherapy could not be administered postresection with HAI therapy. Higher doses of preoperative chemotherapy may impact the ability to treat microscopic disease. In trials evaluating preoperative chemotherapy and radiation for rectal cancer, toxicity was less when treatment was given in the preoperative period [44]. However, at this time, whether preoperative chemotherapy for colorectal liver metastases is better tolerated than postoperative therapy is unknown.

Surrogate Markers for Success of Liver Surgery
Some reports suggest that tumor progression while on chemotherapy predicts worse survival after liver resection. Adam et al. [45] reviewed 131 consecutive patients who underwent liver resection for multiple lesions (>4 lesions) after systemic chemotherapy. This represented approximately 30% of the patients undergoing resection at their institution. They divided the patients into three groups: those who had an objective response (group 1), those who had tumor stabilization (group 2), and those with tumor progression (group 3). The 5-year survival rates for groups 1, 2, and 3 were 37%, 30%, and 8%, respectively (p = .0001). However, there were imbalances among the groups. For example, group 1 (the response group) had a lower preoperative and postoperative CEA level and there were fewer patients with rectal carcinoma, characteristics that would lead to a higher survival rate. The disease-free survival rates were 21%, 20%, and 3%, respectively (p = .02). In the entire group of 131 patients, 78% had liver recurrence after resection. In a multivariate analysis, tumor progression on chemotherapy, elevated preoperative serum CA19–9, number of resected metastases, and number of regimens of chemotherapy were independently associated with a shorter survival duration.

Allen and colleagues looked at patients with synchronous liver metastases from January 1995–2000 at MSKCC. Neoadjuvant chemotherapy was given to 52 of the 106 patients. The 5-year survival rates were similar in those who did and did not receive neoadjuvant chemotherapy, 43% and 35%, respectively (p = .49). Patients whose tumors did not progress while receiving chemotherapy had an increased survival compared to those who did not receive preoperative therapy (p = .03) [46]. An updated analysis showed no significant difference, with median survival times of 48 and 39 months (p = NS) for patients whose tumor did not progress on preoperative chemotherapy and those who did not receive neoadjuvant therapy, respectively (personal observation).

When analyzing patients who are considered unresectable, Folprecht et al. [47] found a strong relationship (r = 0.96, p = .002) between tumor response to neoadjuvant chemotherapy and resection in an analysis of five retrospective studies of patients with unresectable hepatic metastases. Considering only phase III studies, the correlation was not as significant (r = 0.67, p = .024).

Cons

Liver Toxicity
In treating patients with resectable hepatic disease, one must consider the clinical impact of hepatotoxicity from systemic chemotherapy. Hepatotoxicity has been reported with 5-FU, and this was increased when levamisole was added [48]. Additional reports of liver toxicity have been published, which may reflect the use of new drugs such as irinotecan and oxaliplatin, or new ways of administering 5-FU. Types of liver toxicity include steatosis [49], sinusoidal changes [50], steatohepatitis [51], and hemorrhagic central lobular necrosis (HCN) [52].

Steatosis represents fatty changes in the liver, as demonstrated by the presence of fat droplets within the hepatocytes (Fig. 1). Why would steatosis interfere with hepatic resection? It has been shown that steatosis may interfere with circulation through sinusoids and impair regeneration. The liver's protective mechanism against oxidative stress appear to be impaired by steatosis [53, 54], observed in liver transplantation. Figure 2 shows how normal liver parenchyma after preoperative chemotherapy–induced steatosis becomes less dense and therefore darker on computed tomography scans.

View larger version (106K): [in this window] [in a new window]   Figure 1. Steatosis. Fatty change, represented by the presence of fat droplets (arrow) within hepatocytes.

View larger version (59K): [in this window] [in a new window]   Figure 2. Computed tomography (CT) scans of the liver demonstrating preoperative chemotherapy–induced steatosis. (A): Normal liver with tumor. (B): The tumor has decreased in size, but the liver now has fatty changes that make it less dense, and therefore darker on CT and darker than the spleen.

Benoist et al. [60] examined the results of 60 patients who underwent large liver resections, which required portal triad clamping or hepatic vascular exclusion, after receiving six cycles of preoperative chemotherapy. The morbidity rate was 18% with triad clamping and 43% after vascular exclusion (p = .044). Transfusion rates were high in both groups: 50% and 40%, respectively, which is higher than those reported by others after major liver resections [55]. Their conclusion, however, was that vascular occlusion can be used with acceptable morbidity in patients who need a major liver resection after prolonged chemotherapy.

Preoperative chemotherapy can also lead to vascular changes, including sinusoidal dilatation, which involves congestion of the liver tissue surrounding the central venules (Fig. 3). Rubbia-Brandt et al. [50] reviewed pathologic slides from 153 patients who underwent liver resection in 1994–2002 at their institutions. They noted that 51% of those who had preoperative chemotherapy developed sinusoidal dilatation, which was not seen in those who did not receive preoperative chemotherapy. For those who had received previous oxaliplatin, 78% had striking sinusoidal changes. The authors hypothesized that the chemotherapy may have caused an injury to the endothelial cells that resulted in sinusoidal disruption. The observed sinusoidal lesions were similar to those seen in veno-occlusive disease. Sebagh et al. [61] reported on a series of 52 patients treated with oxaliplatin-based chemotherapy and found that only 38% had vascular lesions; they did not observe the veno-occlusive disease described by Rubbia-Brandt et al. [50]. In their series, the patients were given a chronomodulated infusion of chemotherapy, which they postulate could possibly have explained the difference in toxicity.

View larger version (112K): [in this window] [in a new window]   Figure 3. Sinusoidal dilatation. Liver tissue around the central venule (black arrows) shows congestion and dilatation of sinusoidal spaces (red arrows outline a dilated sinusoid). The sinusoids are filled with red blood cells. Source: Courtesy of Dr. Jinru Shia, Memorial Sloan-Kettering Cancer Center, New York.

A potentially more severe form of liver toxicity is steatohepatitis, depicted in Figure 4, with balloon degeneration and Mallory body formation. In a retrospective review of 406 patients who underwent resection in 1992–2005, Vauthey et al. [51] divided the patients into groups dependent on the type of preoperative chemotherapy or no chemotherapy and described the nontumor liver pathology in these groups. Four pathologists examined the liver and scored the degree of steatosis, steatohepatitis, and sinusoidal injury. One hundred fifty-eight patients had no prior chemotherapy, 63 had 5-FU alone, 94 had 5-FU plus irinotecan, and 79 had 5-FU plus oxaliplatin. Table 6 lists the type of hepatic injury stratified by the type of chemotherapy. Steatohepatitis was seen most often in patients who had received preoperative irinotecan (20%). The incidence of steatohepatitis was greater in patients with a BMI >25 kg/m². In those treated with preoperative irinotecan, the incidences were 12.1% and 24.6%, and in those treated with preoperative oxaliplatin, the incidences were 0% and 11.9%, for patients with a BMI <25 kg/m² versus those with a BMI >25 kg/m², respectively. In patients who had received preoperative oxaliplatin, 18.9% had sinusoidal dilatation, and only 6.3% had steatohepatitis. The 90-day mortality rates were 14.7% and 1.6% for patients with or without steatohepatitis (p = .001), respectively, and the postoperative liver failure rates were also higher, 5.8% and 0.8%, respectively (p = .01). Death within 90 days was seen in 6.5% of patients with steatohepatitis and in 1.6% without steatohepatitis (p = .01). No patient died from sinusoidal injury.

View larger version (122K): [in this window] [in a new window]   Figure 4. Steatohepatitis. Black arrows indicate hepatocyte injury in the form of ballooning degeneration and yellow arrows indicate Mallory body formation (meaning degenerated and condensed cytoskeleton filaments). Red arrows point to inflammatory cells. The liver also has pericellular fibrosis, which is best seen using trichrome stain. Source: Courtesy of Dr. Jinru Shia, Memorial Sloan-Kettering Cancer Center, New York.

Does Perioperative Bevacizumab Increase Morbidity?
The question of whether bevacizumab increases toxicity if given prior to or after liver resection is being investigated. There have been reports of elevated factor VIII and von Willebrand factor with bevacizumab and chemotherapy treatment [67]. Anti–vascular endothelial growth factor (VEGF) antagonism could also decrease wound healing [68] and, in mice, it prevented liver regeneration [69]. There are small studies suggesting that it does not increase toxicity. In an article by Scappaticci et al. [70], data were pooled from two studies in which patients were randomized to receive either bevacizumab and chemotherapy or chemotherapy alone. One of the problems with the study is that they eliminated all patients with significant atherosclerotic vascular disease, as well as those on cyclo-oxygenase-2 inhibitors, aspirin, and low-dose warfarin. When bevacizumab was given after liver resection, the complication rates for chemotherapy with or without bevacizumab were 1.5% and 0.5%, respectively. In the second study, in which bevacizumab was given before resection, the complication rates were 3.4% with no bevacizumab and 13% for those who received bevacizumab. The total number of surgical procedures was higher in the bevacizumab-treated patients, 75/616 (12%) versus 29/516 (5.6%) in the group not receiving bevacizumab [70].

In a review from MSKCC [71], 32 patients underwent hepatic resection receiving perioperative bevacizumab, eight before, eight before and after, and 16 after surgery. Of the 16 patients who received preoperative bevacizumab, seven (43%) had complications, including pulmonary embolus, deep vein thrombosis, subphrenic abscess, three wound infections, and one urinary tract infection. In the 24 patients who received postoperative bevacizumab, there was one ventricular arrhythmia, three wound infections, one groin abscess, and one superficial thrombophlebitis. Thus, there were four (12.5%) cardiovascular events in these 32 patients. In the 32 matched controls, there were no cardiovascular events, although the matched controls seemed to have undergone more extensive surgery, with more extrahepatic disease resected. In another retrospective database of 1,186 patients who underwent liver resection, two subgroups were identified: those who had preoperative chemotherapy and bevacizumab (group 1) and those who had preoperative chemotherapy without bevacizumab (group 2). Forty-five patients from group 1 were compared with 82 patients from group 2. For patients who had concomitant surgical procedures, wound complications were more frequent in group 1 (p = .05) as were greater blood loss and hepatobiliary complications. Overall complications were not greater with the use of bevacizumab perioperatively [72].

To evaluate whether preoperative bevacizumab affects patients going for liver resection, the Bevacizumab Expanded Access Trial (BEAT) study was designed, in which bevacizumab (5 mg/kg) is given every 2 weeks with systemic chemotherapy (FOLFIRI or FOLFOX). Of the 43 patients who have undergone resection thus far, the median time from first bevacizumab treatment to surgery was 183 days, and the median time to resection after stopping bevacizumab was 67 days. Twenty-eight percent of patients have had complications, including operative site infection, gastric perforation, right pleural effusion, thrombosis of the portal vein, and myocardial infarct, as well as other toxicities, but the conclusion was that metastasectomy was feasible after bevacizumab treatment [73]. The results of larger studies are needed to ascertain whether bevacizumab is safe before resection. Ellis et al. [74] suggested that surgery 6–8 weeks after the last bevacizumab treatment is safe. However, waiting two half-lives or 6 weeks would still leave about 1.25 mg/kg of bevacizumab in the circulation.

Secondary Splenomegaly
As some patients develop liver toxicity, they get secondary hypertrophy of the spleen and may develop a clinical syndrome similar to portal hypertension [52] with varices and thrombocytopenia (Fig. 5). Hepatotoxicity increases with the duration of treatment [75]. However, even short courses of chemotherapy can cause hepatotoxicity. Figure 6 illustrates that these problems can occur early in treatment, although a longer time on chemotherapy generally produces more toxicity [75].

View larger version (56K): [in this window] [in a new window]   Figure 5. Enlarged spleen after systemic oxaliplatin/5-fluorouracil/leucovorin (FOLFOX) chemotherapy. (A): Computed tomography (CT) scan prior to chemotherapy. (B): CT scan after 1 year of FOLFOX showing enlarged spleen.

View larger version (53K): [in this window] [in a new window]   Figure 6. Hepatic toxicity after 6 weeks of chemotherapy. (A): 62-year-old male patient with colorectal cancer metastatic to the liver, received three treatments of oxaliplatin/5-fluorouracil/leucovorin (6 weeks of therapy) as neoadjuvant treatment prior to resection. (B): After 6 weeks of treatment, the liver appeared fatty, and there was new ascites. The patient's bilirubin level increased from 0.8 to 2.0 mg/dl, and his alkaline phosphatase increased from 101 to 261 u/l. Ascites indicated by arrow.

Risk for Progression
During preoperative therapy, there can be progression in existing sites, development in new sites, and/or spread to extrahepatic sites if the tumor fails to respond to preoperative chemotherapy. Certainly, surgeons have seen hepatic lymph node involvement at the time of liver resection, suggesting that liver metastases may spread to these lymph nodes. In one study of 29 resectable patients, Bathe demonstrated that five of the patients (17%) became unresectable after preoperative chemotherapy, three because of the development of extrahepatic disease, one as a result of clinical deterioration, and one refusal [76].

The EORTC has completed a study (EORTC 40983) to evaluate 3 months of preoperative FOLFOX and 3 months of postoperative FOLFOX versus immediate surgery alone. The eligibility criteria included no previous oxaliplatin chemotherapy and no extrahepatic disease. Of the 365 patients randomized to date, 94% have one to three metastases, and 42% have stage II disease. Of the 182 patients randomized to preoperative chemotherapy, 21 (11.5%) could not undergo surgery: seven because of disease progression, seven because of refusal or toxicity, and another seven for unstated reasons. Of the 182 patients without preoperative chemotherapy, nine (4.9%) were not operated on: five because of progression and four for unstated reasons [77]. It is too early to examine disease-free survival or overall survival. The pathology has been reviewed for 59 patients. Sinusoidal changes occurred in 48% of the treated group and 11% of the control group, while severe sinusoidal changes were seen in 41% and 0%, respectively (p = .0032). There has been no difference between groups in surgical complications thus far [78].

Management of Hepatic Metastases After a Complete Response to Chemotherapy
When patients undergo preoperative chemotherapy, the aim is to decrease the size of the tumor, but what happens when there is a complete response and the tumor is no longer visible? If the tumor is not visible by scan or intraoperatively, but is present microscopically, there can later be a recurrence at that site. Benoist et al. [79] prospectively followed 586 patients treated for liver metastases from colorectal cancer, and 38 patients (6%) obtained a complete response of at least one lesion; 66 sites disappeared on imaging. Surgery was done 4 weeks after the imaging. Of the 66 disappearing sites, 20 had macroscopic disease present at surgery, and in 46 sites, no lesion was found at surgery. Fifteen of the 46 disappearing lesions were resected, and 12 (80%) still had viable tumor. Thirty-one sites where previous tumor had been seen were left in place and followed, and in situ recurrence was seen in 23 (74%). Therefore, 55 of 66 sites (83%) were not cured (Fig. 7). The authors concluded that medical oncologists should refer patients prior to getting a complete response in order not to miss "dormant" metastases. In the Adam et al. [23] trial of 1,104 patients, of whom 138 (12.6%) became resectable, 80% had tumor recurrence, and 71% of these recurrences were in the liver.

View larger version (17K): [in this window] [in a new window]   Figure 7. Complete response after chemotherapy: Does it mean cure? From Benoist S, Brouquet A, Penna C et al. Complete response of colorectal liver metastases after chemotherapy: Does it mean cure? J Clin Oncol 2006;24:3939–3945.

Elias et al. [81] also demonstrated less recurrence after preoperative HAI therapy. They examined 104 patients who underwent liver resection for colorectal cancer over a 4-year period and found that 15 patients had a dramatic response to chemotherapy and a complete disappearance of at least one lesion by imaging. In four patients (27%), the lesions were found at laparotomy and removed, while in 11 patients the tumors could not be found and were not removed ("missing lesions"). Eight of these 11 patients did not have a recurrence, with a median follow-up of 31 months. There is no description of the type of chemotherapy given preoperatively to the whole group, but the authors do describe the therapy for those with missing lesions. Six of the 11 patients with missing lesions had received hepatic arterial therapy, either before (four patients) or afterward (two patients). Four of the eight patients who had received HAI never had a recurrence of the missing lesions [82].

	CONCLUSION

Top Learning Objectives Abstract Introduction How to Define Resectable... Initial Chemotherapy in... Synchronous Liver Metastases Neoadjuvant Therapy for Patients... Conclusion Disclosure of Potential... Acknowledgments References

 
For many years, the diagnosis of liver metastases from colon cancer portended a dismal outcome. The development of therapeutic agents that produce response rates in the 50% range has led to longer survival times, even in patients with unresectable disease. One of the advantages of these new agents is that they allow initially unresectable patients to become resectable, thus providing a chance of cure or longer survival. A challenge over the next few years will be to decide when these different modalities are best used. If chemotherapy can decrease the tumor size to a point where lesions cannot be seen but are not completely eradicated, then preoperative chemotherapy may be a disservice to initially resectable patients. Additionally, if preoperative chemotherapy produces more liver toxicity and thereby increases the risk for surgery, then preoperative chemotherapy may not prove to be advantageous for resectable patients.

In patients who are clearly unresectable, there is no question that chemotherapy is the appropriate treatment. However, for those who are clearly resectable, there may be harm in giving preoperative chemotherapy. Until studies demonstrate that preoperative chemotherapy is useful in the resectable patient, those patients should be offered resection first and then be treated with postoperative chemotherapy (systemic therapy or HAI plus systemic therapy) [82–86]. A multidisciplinary approach is essential to develop an effective treatment plan for patients with liver metastases from colorectal carcinoma.

	DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST

Top Learning Objectives Abstract Introduction How to Define Resectable... Initial Chemotherapy in... Synchronous Liver Metastases Neoadjuvant Therapy for Patients... Conclusion Disclosure of Potential... Acknowledgments References

 
The author indicates no potential conflicts of interest.

	ACKNOWLEDGMENTS

Top Learning Objectives Abstract Introduction How to Define Resectable... Initial Chemotherapy in... Synchronous Liver Metastases Neoadjuvant Therapy for Patients... Conclusion Disclosure of Potential... Acknowledgments References

 
Thank you to Deirdre B. Casey for manuscript preparation.

	REFERENCES

Top Learning Objectives Abstract Introduction How to Define Resectable... Initial Chemotherapy in... Synchronous Liver Metastases Neoadjuvant Therapy for Patients... Conclusion Disclosure of Potential... Acknowledgments References

 

1. Kemeny N, Kemeny MM, Lawrence TS. In: Abeloff MD, Armitage JO, Niederhuber JE et al, eds. Clinical Oncology. Liver metastases. Third Edition. Philadelphia: Elsevier, 2004:1141-1178.
2. Andre T, Boni C, Mounedji-Boudiaf L et al. Oxaliplatin, fluorouracil, and leucovorin as adjuvant treatment for colon cancer. N Engl J Med 2004;350:2343–2351.[Abstract/Free Full Text]
3. Douillard JY, Cunningham D, Roth AD et al. Irinotecan combined with fluorouracil compared with fluorouracil alone as first-line treatment for metastatic colorectal cancer: A multicentre randomised trial. Lancet 2000;355:1041–1047.[CrossRef][Medline]
4. de Gramont A, Figer A, Seymour M et al. Leucovorin and fluorouracil with or without oxaliplatin as first-line treatment in advanced colorectal cancer. J Clin Oncol 2000;18:2938–2947.[Abstract/Free Full Text]
5. Tournigand C, Andre T, Achille E et al. FOLFIRI followed by FOLFOX6 or the reverse sequence in advanced colorectal cancer: A randomized GERCOR study. J Clin Oncol 2004;22:229–237.[Abstract/Free Full Text]
6. Cunningham D, Humblet Y, Siena S et al. Cetuximab monotherapy and cetuximab plus irinotecan in irinotecan-refractory metastatic colorectal cancer. N Engl J Med 2004;351:337–345.[Abstract/Free Full Text]
7. Hurwitz H, Fehrenbacher L, Novotny W et al. Bevacizumab plus irinotecan, fluorouracil, and leucovorin for metastatic colorectal cancer. N Engl J Med 2004;350:2335–2342.[Abstract/Free Full Text]
8. Fong Y, Cohen AM, Fortner JG et al. Liver resection for colorectal metastases. J Clin Oncol 1997;15:938–946.[Abstract/Free Full Text]
9. Scheele J, Stangl R, Altendorf-Hofmann A. Hepatic metastases from colorectal carcinoma: Impact of surgical resection on the natural history. Br J Surg 1990;77:1241–1246.[Medline]
10. Bismuth H, Adam R, Lévi F et al. Resection of nonresectable liver metastases from colorectal cancer after neoadjuvant chemotherapy. Ann Surg 1996;224:509–522.[CrossRef][Medline]
11. Fong Y, Fortner J, Sun RL et al. Clinical score for predicting recurrence after hepatic resection for metastatic colorectal cancer: Analysis of 1001 consecutive cases. Ann Surg 1999;230:309–318; discussion 318–321.[CrossRef][Medline]
12. Nordlinger B, Guiguet M, Vaillant JC et al. Surgical resection of colorectal carcinoma metastases to the liver. A prognostic scoring system to improve case selection, based on 1568 patients. Association Francaise de Chirurgie. Cancer 1996;77:1254–1262.[CrossRef][Medline]
13. Adam R, Aloia T, Figueras J et al. LiverMetSurvey: Analysis of clinicopathologic factors associated with the efficacy of preoperative chemotherapy in 2,122 patients with colorectal liver metastases. the 2006 American Society of Clinical Oncology Annual Meeting; June 4, 2006; Atlanta, GA, Presented at.
14. Poston GJ, Adam R, Alberts S et al. OncoSurge: A strategy for improving resectability with curative intent in metastatic colorectal cancer. J Clin Oncol 2005;23:7125–7134.[Abstract/Free Full Text]
15. Elias D, Liberale G, Vernerey D et al. Hepatic and extrahepatic colorectal metastases: When resectable, their localization does not matter, but their total number has a prognostic effect. Ann Surg Oncol 2005;12:900–909.[Abstract/Free Full Text]
16. Nordlinger B, Vaillant JC, Guiguet M et al. Survival benefit of repeat liver resections for recurrent colorectal metastases: 143 cases. Association Francaise de Chirurgie. J Clin Oncol 1994;12:1491–1496.[Abstract]
17. Khatri VP, Petrelli NJ, Belghiti J. Extending the frontiers of surgical therapy for hepatic colorectal metastases: Is there a limit? J Clin Oncol 2005;23:8490–8499.[Abstract/Free Full Text]
18. Makuuchi M, Thai BL, Takayasu K et al. Preoperative portal embolization to increase safety of major hepatectomy for hilar bile duct carcinoma: A preliminary report. Surgery 1990;107:521–527.[Medline]
19. Makuuchi M, Mori T, Gunven P et al. Safety of hemihepatic vascular occlusion during resection of the liver. Surg Gynecol Obstet 1987;164:155–158.[Medline]
20. Couinaud C. Le Foie, Études Anatomiques et Chirurgicales. Paris: Masson & Cie, 1957.
21. Curley SA, Izzo F, Delrio P et al. Radiofrequency ablation of unresectable primary and metastatic hepatic malignancies: Results in 123 patients. Ann Surg 1999;230:1–8.[CrossRef][Medline]
22. Giacchetti S, Itzhaki M, Gruia G et al. Long-term survival of patients with unresectable colorectal cancer liver metastases following infusional chemotherapy with 5-fluorouracil, leucovorin, oxaliplatin and surgery. Ann Oncol 1999;10:663–669.[Abstract/Free Full Text]
23. Adam R, Delvart V, Pascal G et al. Rescue surgery for unresectable colorectal liver metastases downstaged by chemotherapy: A model to predict long-term survival. Ann Surg 2004;240:644–657; discussion 657–658.[Medline]
24. Pozzo C, Basso M, Quirino M et al. Long-term follow-up of colorectal cancer (CRC) patients treated with neoadjuvant chemotherapy with irinotecan and fluorouracil plus folinic acid (5-FU/FA) for unresectable liver metastases. the 2006 American Society of Clinical Oncology Annual Meeting; June 3, 2006; Atlanta, GA, Presented at.
25. Alberts SR, Horvath WL, Sternfeld WC et al. Oxaliplatin, fluorouracil, and leucovorin for patients with unresectable liver-only metastases from colorectal cancer: A North Central Cancer Treatment Group phase II study. J Clin Oncol 2005;23:9243–9249.[Abstract/Free Full Text]
26. Ho WM, Ma B, Mok T et al. Liver resection after irinotecan, 5-fluorouracil, and folinic acid for patients with unresectable colorectal liver metastases: A multicenter phase II study by the Cancer Therapeutic Research Group. Med Oncol 2005;22:303–312.[CrossRef][Medline]
27. Falcone A, Masi G, Murr R et al. Biweekly irinotecan, oxaliplatin, and infusional 5FU/LV (FOLFOXIRI) versus FOLFIRI as first-line treatment of metastatic colorectal cancer (MCRC): Results of a randomized, phase III trial by the Gruppo Oncologico Nord Ovest (GONO). the 2006 Gastrointestinal Cancers Symposium; January 28, 2006; San Francisco, CA, Presented at.
28. Giacchetti S, Bjarnason G, Garufi C et al. First line infusion of 5-fluorouracil, leucovorin and oxaliplatin for metastatic colorectal cancer: 4-day chronomodulated (FFL4–10) versus 2-day FOLFOX2. A multicenter randomized phase III trial of the Chronotherapy Group of the European Organization for Research and Treatment of Cancer (EORTC 05963). the 2004 American Society of Clinical Oncology Annual Meeting; June 5, 2004; New Orleans, LA, Presented at.
29. Rougier P, Raoul JL, Van Laethem JL et al. Cetuximab+FOLFIRI as first-line treatment for metastatic colorectal CA. J Clin Oncol 2004;22(14 suppl):3513.
30. Breedis C, Young G. The blood supply of neoplasms in the liver. Am J Pathol 1954;30:969–977.[Medline]
31. Ensminger WD, Rosowsky A, Raso V et al. A clinical-pharmacological evaluation of hepatic arterial infusions of 5-fluoro-2'-deoxyuridine and 5-fluorouracil. Cancer Res 1978;38:3784–3792.[Abstract/Free Full Text]
32. Kemeny N, Jarnagin W, Gonen M et al. Phase I/II study of hepatic arterial therapy with floxuridine and dexamethasone in combination with intravenous irinotecan as adjuvant treatment after resection of hepatic metastases from colorectal cancer. J Clin Oncol 2003;21:3303–3309.[Abstract/Free Full Text]
33. Kemeny N, Jarnagin W, Paty P et al. Phase I trial of systemic oxaliplatin combination chemotherapy with hepatic arterial infusion in patients with unresectable liver metastases from colorectal cancer. J Clin Oncol 2005;23:4888–4896.[Abstract/Free Full Text]
34. Clavien PA, Selzner N, Morse M et al. Downstaging of hepatocellular carcinoma and liver metastases from colorectal cancer by selective intra-arterial chemotherapy. Surgery 2002;131:433–442.[CrossRef][Medline]
35. Kemeny N. Management of liver metastases from colorectal cancer. Oncology (Williston Park) 2006;20:1161–1176; 1179; discussion 1179–1180, 1185–1186.[Medline]
36. White RR, Munoz JA, Schwartz LH et al. Time to maximal tumor reduction with hepatic arterial infusion plus systemic therapy versus systemic therapy alone for colorectal liver metastases: A retrospective analysis. the 2007 Gastrointestinal Cancers Symposium; January 21, 2007; Orlando, FL, Presented at.
37. Bolton JS, Fuhrman GM. Survival after resection of multiple bilobar hepatic metastases from colorectal carcinoma. Ann Surg 2000;231:743–751.[CrossRef][Medline]
38. Tanaka K, Shimada H, Matsuo K et al. Outcome after simultaneous colorectal and hepatic resection for colorectal cancer with synchronous metastases. Surgery 2004;136:650–659.[CrossRef][Medline]
39. Martin R, Paty P, Fong Y et al. Simultaneous liver and colorectal resections are safe for synchronous colorectal liver metastasis. J Am Coll Surg 2003;197:233–241; discussion 241–242.[CrossRef][Medline]
40. Weber JC, Bachellier P, Oussoultzoglou E et al. Simultaneous resection of colorectal primary tumour and synchronous liver metastases. Br J Surg 2003;90:956–962.[CrossRef][Medline]
41. Hao CY, Ji JF. Surgical treatment of liver metastases of colorectal cancer: Strategies and controversies in 2006. Eur J Surg Oncol 2006;32:473–483.[CrossRef][Medline]
42. Huitzil-Melendez F, Capanu M, Haviland D et al. Evaluation of the impact of systemic (SYS) neoadjuvant chemotherapy (neoadj) in patients (pts) with resectable liver metastasis (mets) from colorectal carcinoma (CRC) treated with adjuvant hepatic arterial infusion (HAI) and SYS chemotherapy. the 2007 Gastrointestinal Cancers Symposium; January 21, 2007; Orlando, FL, Presented at.
43. Kemeny N, Eid A, Stockman J et al. Hepatic arterial infusion of floxuridine and dexamethasone plus high-dose mitomycin C for patients with unresectable hepatic metastases from colorectal carcinoma. J Surg Oncol 2005;91:97–101.[CrossRef][Medline]
44. Sauer R, Becker H, Hohenberger W et al. Preoperative versus postoperative chemoradiotherapy for rectal cancer. N Engl J Med 2004;351:1731–1740.[Abstract/Free Full Text]
45. Adam R, Pascal G, Castaing D et al. Tumor progression while on chemotherapy: A contraindication to liver resection for multiple colorectal metastases? Ann Surg 2004;240:1052–1061; discussion 1061–1064.[CrossRef][Medline]
46. Allen PJ, Kemeny N, Jarnagin W et al. Importance of response to neoadjuvant chemotherapy in patients undergoing r