Int J Colorectal Dis (2015) 30:1091–1102 DOI 10.1007/s00384-015-2246-2
ORIGINAL ARTICLE
A systematic review and meta-analysis to reappraise the role of adjuvant hepatic arterial infusion for colorectal cancer liver metastases Wei Liu 1 & Qing-Kun Song 2 & Bao-Cai Xing 1
Accepted: 8 May 2015 / Published online: 27 May 2015 # Springer-Verlag Berlin Heidelberg 2015
Abstract Objective The potential benefit of adjuvant hepatic arterial infusion remains unknown for patients with colorectal liver metastases after radical hepatic resection. The principle aim of this study was to investigate the long-term outcome of adjuvant hepatic arterial infusion. Methods Eligible trials were identified from Embase, PubMed, the Web of Science, and the Cochrane library since their inception to June 1, 2014. Patients with colorectal liver metastases, who underwent radical hepatic resection and received adjuvant hepatic arterial infusion, were enrolled. The study outcomes included 5-year disease-free and overall survival rate, respectively. Hazard ratio with a 95 % confidence interval was used to measure the pooled effect according to a random effects model or fixed effects model, depending on the heterogeneity between the included studies. The statistical heterogeneity between trials was detected by I2 test. Sensitivity analyses were also carried out. Results A total of nine studies containing 1057 patients were included. The comparison indicated that the
Wei Liu and Qing-Kun Song contributed equally to this work. * Bao-Cai Xing [email protected] 1
Hepatopancreatobiliary Surgery Department I, Key Laboratory of Carcinogenesis and Translational Research, Ministry of Education, Beijing Cancer Hospital and Institute, Peking University School of Oncology, No. 52, Fu-Cheng-Lu Street, Beijing 100142, People’s Republic of China
2
Beijing Key Laboratory of Cancer Therapeutic Vaccine, Beijing Shijitan Hospital, Capital Medical University Cancer Center, Beijing, China
overall pooled hazard ratio for 5-year overall survival was 0.75 (95 % CI: 0.56–0.99, p=0.048). The hazard ratio for 5-year disease-free survival rate was 0.61 (95 % CI: 0.48–0.79, p=0.001). When compared with systemic chemotherapy alone, adjuvant hepatic arterial infusion plus systemic chemotherapy also improved the long-term survival. Conclusions Adjuvant hepatic arterial infusion improved the 5-year disease-free and overall survival rate, respectively. It should be recommended for patients with a high risk of recurrence, but these findings require prospective confirmation. Keywords Hepatic artery infusion . Colorectal cancer . Metastases . Meta-analysis
Introduction Colorectal cancer is the second leading cause of cancerrelated death in the world [1]. About 25 % of patients present with synchronous liver metastases and 50 % of patients overall develop liver metastases [2]. Hepatic resection is considered to be the only curable treatment associated with a long survival time for colorectal liver metastases (CRLM), and the overall survival rate at 5 years ranges from 43 to 58 % in recent studies [3]. However, the 5-year disease-free survival rate remains around 28 % [4, 5]. The introduction of more effective systemic and regional chemotherapy for CRLM over the past decade has likely contributed to the improvements in survival after hepatic resection for CRLM [6]. Hepatic arterial infusion (HAI) chemotherapy, a local treatment for liver metastases, has been an appealing investigational
1092
method for patients with CRLM confined to the liver, in whom it has reproducibly yielded higher response rates. HAI has presented a significantly higher response rate than systemic chemotherapy in unresectable CRLM patients [7]. However, the adjuvant HAI chemotherapy significantly prolonged survival in only few studies [8, 9]; the current evidence supporting the efficacy of this therapy remains controversial. Therefore, this meta-analysis was performed to investigate the long-term outcome of adjuvant HAI in CRLM after radical hepatic resection. The treatment benefits were critically discussed in terms of 5-year disease-free and overall survival rate.
Materials and methods Literature search A comprehensive search was performed to identify all published studies of adjuvant HAI and systemic chemotherapy in CRLM after hepatic resection. Literature search of Embase, PubMed, the Web of Science, and the Cochrane database was conducted to identify eligible studies up to June 1, 2014. The following keywords were used as the search strategy: Bhepatic artery^ and Bcolorectal^ and Bcancer^ and Barterial infusion^ or BHAI.^
Int J Colorectal Dis (2015) 30:1091–1102
time from allocation to death by any cause. Living patients were censored on the date of last follow-up. The secondary endpoint was 5-year disease-free survival (DFS) rate, which was defined from the allocation until locoregional recurrence or distant metastasis. Quality of studies Two authors independently evaluated the quality of each retrieved trial according to the method described in the Cochrane Handbook for Systematic Review of Interventions (version 5.1.0) [10]. All studies were subjectively reviewed and assigned a value of Bhigh,^ Blow,^ or Bunclear^ to the following domains: random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective reporting, and other bias. Trials with high risk of bias for any one or more key domains were considered as at high risk of bias. Trials with low risk of bias for all key domains were considered as at low risk of bias. Otherwise, they were considered as unclear risk of bias [11]. A modified Newcastle-Ottawa scale (NOS) was used to assess the quality of the nonrandomized studies included in this meta-analysis [12]. This scale, which ranged from 0 to 9 points, consisted of three items that described the patient selection method, the comparability of adjuvant HAI and systemic chemotherapy, and outcome assessment. Articles scored as ≥6 were considered as high-quality studies.
Study inclusion and exclusion criteria Assessment for grading the quality of evidence We included studies based on the following inclusion criteria: (1) CRLM patients were performed hepatic resection and then underwent HAI chemotherapy in study group, which was compared with hepatic resection alone or adjuvant systemic chemotherapy in control group; (2) data from efficacy and/or safety analyses were reported; and (3) the study was reported in English as a full paper. Letters to the editors and conference abstracts were excluded due to the limited data. Studies without a control arm, animal studies, and review articles were also excluded. Data extraction and outcome measures Two reviewers (Wei Liu and Qing-kun Song) assessed articles independently for their eligibility and extracted data from the included studies. Discrepancies were resolved by consensus. Data retrieved from studies included the first author’s last name, publication year, patients’ characteristics, sample size, study design, duration of follow-up, endpoint evaluations, and methodological quality. The quality of the studies was assessed independently by all investigators. The primary outcome was 5-year overall survival (OS) rate defined as the
The overall quality of the evidence and strength of recommendations were evaluated using GRADE [13]. GRADE Working Group grades of evidence were as follows: high quality, further research is very unlikely to change our confidence in the estimate of effect; moderate quality, further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate; low quality, further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate; very low quality, we are very uncertain about the estimate. Statistical analysis For the outcome, the association between OS, DFS, and adjuvant HAI was expressed as pooled hazard ratio (HR). For the time-to-event variables, the hazard ratios of DFS and OS with 95 %CIs were directly extracted or calculated using a calculation sheet as previously described [14]. Statistical heterogeneity was explored by inconsistency (I 2) statistics; an I 2 value
Int J Colorectal Dis (2015) 30:1091–1102
1093
of 50 % or more represented substantial heterogeneity [15]. In the absence of heterogeneity, studies were pooled using a fixed effect model. If heterogeneity was observed, a random effects model was used. An estimate of potential publication bias for primary outcome was carried out by the Egger’s regression test and Begg’s adjusted rank correlation test. The metaanalyses for pooled HR and sensitivity were performed with Stata software 12.0 (Stata Corp., College Station, TX).
Results
studies were excluded due to the irrelevance to the current analysis or their disagreement with the inclusion criteria. Finally, nine studies were eligible for this meta-analysis [8, 9, 16–22] (Fig. 1). The characteristics of the nine included studies are outlined in Table 1. The studies differed somewhat in a number of details regarding HAI chemotherapy. Early studies used some form of fluorouracil (5-Fu) and floxuridine (FUDR), with or without dexamethasone (DXM). Recently, oxaliplatin (OXA) or irinotecan (CPT11) was the main regimen. Moreover, three studies underwent surgery alone and another six studies received systemic chemotherapy. The median follow-up period ranged from 29.0 to 62.7 months (Table 2).
Literature research and characteristic of studies
5-year DFS and OS rate
A total of 358 studies containing 1057 patients were identified based on the initial search. After an independent review, 321
Six studies (724 patients) were identified with outcome measurement of 5-year disease-free survival rate. The pooled HR
Fig. 1 Flow chart for studies selection
Publications identified after retrieval and relevant reviews research (N=358)
Studies excluded after screening title and abstract (N=321) Non-relevant (N=288) Not clinical trials (N=12) Reviews (N=6) Non-English (N=15)
Potentially relevant studies in full text for detailed evaluation (N=37)
Studies excluded (N=25) Lack of relevance (N=15) Duplicates studies (N=7) No survival outcome report (N=3)
Potentially appropriate studies included in (N=12)
Studies excluded (N=3) Study with insufficient data (N=1) Studies with low quality (N=1) Duplicates studies (N=1)
Studies included in meta-analysis (N=9)
2011 2013 2011
1999 2002 2000
1998 2003
2000
Bolton et al. [16]. Goere et al. [8]. House et al. [9].
Kemeny N et al. [17]. Kemeny M et al. [18]. Kusunoki et al. [19].
Lorenz et al. [20]. Onaitis et al. [21].
Tono et al. [22].
RCT
RCT Cohort
RCT RCT RCT
RCT Cohort Cohort
Study design
N/A
N/A N/A
N/A N/A N/A
N/A 7 N/A
Preoperative HAI (n)
N/A
N/A N/A
N/A N/A N/A
N/A 37 69
Preoperative sys (n)
HAI(5-Fu)+Sys(5-Fu)
HAI(5-Fu+FA) HAI(FUDR+ DXM)+Sys/HAI
HAI(FUDR+DXM)+Sys(5-Fu±LV) HAI(FUDR)+Sys(5-Fu) HAI(5-Fu)+Sys(UFT)
HAI(FUDR)+Sys(5-Fu+LV) HAI(OXA/CPT-11)+Sys(5-Fu+LV) HAI(FUDR + DXM)+Sys(5-Fu+OXA/CPT-11)
Regimens of study group
Sys(5-Fu)
None Sys(N/A)
Sys(LV+ 5-Fu) None Sys(UFT)
None Sys(5-Fu+OXA/CPT-11) Sys (5-Fu+OXA/CPT-11)
Regimens of control group
24 weeks
6 N/A
6 8 11 weeks
N/A 6 N/A
HAI duration/ course(s)
49 98 250
156 109 58 226 92
19
Bolton et al. [16]. Goere et al. [8]. House et al. [9].
Kemeny N et al. [17]. Kemeny M et al. [18]. Kusunoki et al. [19]. Lorenz et al. [20] Onaitis et al. [21].
Tono et al. [22].
9 vs. 10
74 vs. 82 53 vs. 56 30 vs. 28 113 vs. 113 21 vs. 71
36 vs. 13 44 vs. 54 125 vs. 125
Sample (study vs. control)
N/A
N/A N/A N/A N/A Isolated CRLM
No. of CRLM≥4 No. of CRLM≥3 N/A
Indication for HAI
26.3±14.9
N/A N/A 25 (10–100) N/A 22±41
N/A 29 (17–40) 28 (2–170)
Tumor size (mm)
N/A
47 (64 %) N/A N/A 49 (46.3 %) N/A
6 (17 %) N/A 76 (61 %)
Multiple tumors (n)
N/A
11.5 (0.9–258.9) N/A 11.7 (1–1020) N/A 11.4±25.4
N/A 4.5 (3–10) 11.7 (1.2–1235)
CEA level (ng/ml)
N/A
12 (57.1 %) N/A N/A 53 (49.5 %) N/A
N/A N/A 84 (68 %)
Primary N stage(+) (n)
4 (44.4 %)
23 (31 %) 15 (28.3 %) 12 (40.4 %) 34 (31.5 %) N/A
N/A 38 (86 %) 62 (50 %)
Synchronous CRLM (n)
62.2
62.7 51 N/A N/A 29
N/A 60 43
Median follow-up (months)
OS+DFS
OS+DFS DFS OS+DFS OS+DFS OS+DFS
OS+DFS OS+DFS DSS
Primary end-point
CRLM colorectal liver metastases, OS overall survival, DFS disease-free survival, DSS disease-specific survival, Study group hepatic resection with adjuvant HAI ± systemic chemotherapy, Control group hepatic resection ± systemic chemotherapy, N/A not available
Patients (n)
Characteristics of clinical trials included in the meta-analysis
Author
Table 2
HAI hepatic arterial infusion, Sys systemic chemotherapy, 5-Fu fluorouracil, FUDR floxuridine, LV leucovorin, UFT uracil tegafur, OXA oxaliplatin, CPT-11 irinotecan, DXM dexamethasone, Study group hepatic resection with adjuvant HAI±systemic chemotherapy, Control group hepatic resection±systemic chemotherapy, N/A not available
Year
Interventions of clinical trials included in the meta-analysis
Author
Table 1
1094 Int J Colorectal Dis (2015) 30:1091–1102
Int J Colorectal Dis (2015) 30:1091–1102
1095
Fig. 2 The pooled HR for 5-year DFS (a) and OS (b) in the included six studies (724 patients) and nine studies (1057 patients), respectively, calculated using the random-effects model
1096
Int J Colorectal Dis (2015) 30:1091–1102
Fig. 3 The pooled HR for 5-year DFS (a) and OS (b) in the included studies (three RCTs (284 patients) and (six RCTs (617 patients), respectively) calculated using the random-effects model
for 5-year DFS in the included studies calculated using the random effects model was 0.61 (95 % CI: 0.48–0.79; p=
0.000; I2 =21.0 %, p=0.275) (Fig. 2a). Nine studies (1057 patients) were identified with outcome measurement of 5-
Int J Colorectal Dis (2015) 30:1091–1102
1097
Fig. 4 Four studies (523 patients) were identified with outcome measurement of 5-year DFS rate. The pooled HR for 5-year a DFS (four studies (523 patients)) and b OS (five studies (581 patients)) in the included studies calculated using the random-effects model
year overall survival rates. The pooled HR for 5-year OS in the included studies calculated using the random effects model
was 0.75 (95 % CI: 0.56–0.99; p=0.048; I2 =37.1 %, p= 0.122) (Fig. 2b).
Int J Colorectal Dis (2015) 30:1091–1102 Total quality score
1098
9
9
√
√ √ √
√
9 √
√ √ √
√ √ √
Adequacy of follow-up duration Assessment of outcome with independency Comparability between patients in different treatment armssecondary factor: tumor size
Subgroup analysis for HAI plus systemic chemotherapy compared with systemic chemotherapy Four studies (523 patients) were identified with outcome measurement of 5-year DFS rate. The pooled HR for 5-year DFS in the included studies calculated using the random effects model was 0.55 (95 % CI: 0.42–0.72; p=0.000; I2 =11.2 %, p=0.337) (Fig. 4a). Five studies (581 patients) were identified with outcome measurement of 5-year OS rate. The pooled HR for 5-year OS in the included studies calculated using the random effects model was 0.68 (95 % CI: 0.50–0.94; p=0.020; I2 =0.40 %, p= 0.404) (Fig. 4b).
√ √ √
√ √
√ √
√ House et al. [9].
√
√ Goere et al. [8].
√
Eight studies reported the HAI-related events (Table 3). Only one study reported the catheter-related event that was pumprelated malfunction (19 %). The most usually reported hematological events were neutropenia (range 3.8–18 %) and thrombocytopenia (range 1.9–6.7 %). The most typical nonhematological side effects were nausea (range 5.7– 55.4 %), diarrhea (range 1.9–29 %), vomit (range 3.8– 11.1 %), stomatitis (range 11.0–57.6 %), hepatotoxicity (range 13.9–30.0 %), and biliary sclerosis (range 2.3–28.6 %). There was no HAI-related death.
√
√
√
Quality of the studies
Onaitis et al. [21]. √
Comparability between patients in different treatment arms primary factor: number of tumor Representativeness Selection of the Ascertainment of of the treatment arm comparative treatment treatment arm regimen
Demonstration that the outcome of interest was not present at the beginning of study
Three RCTs (284 patients) were identified with outcome measurement of 5-year DFS rates. The pooled HR for 5-year DFS in the included studies calculated using the random effects model was 0.52 (95 % CI: 0.37–0.83; p = 0.007; I2 = 16.5 %, p = 0.302) (Fig. 3a). Six RCTs (617 patients) were identified with outcome measurement of 5-year OS rates. The pooled HR for 5-year OS in the included studies calculated using the random effects model was 0.69 (95 % CI: 0.53–0.89; p=0.004; I2 =0 %, p=0.635) (Fig. 3b).
HAI-related adverse events
Author
Comparability Non-RCT studies Selection
Table 3
Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomized trials
Assessment of outcome
Loss to follow-up acceptable (less than 10 % and reported)
Subgroup analysis for 5-year DFS and OS rates
Randomized sequence generation was conducted adequately, and allocation concealment was not described in most of studies. Among all the selected studies, participants and personnel were not blinded. All the selected studies were considered as at high risk of bias (Fig. 5). The quality of the nonrandomized studies was assessed using the NOS, and the score was 9 for each study, indicating that these studies were of high quality (Table 4).
Int J Colorectal Dis (2015) 30:1091–1102
1099
Fig. 5 Risk of bias summary (a) and graph (b)
Assessment for grading the quality of evidence GRADE Working Group grades of evidence were moderate quality for 5-year OS and moderate quality for 5-year DFS. When adjuvant HAI plus systemic chemotherapy compared with systemic chemotherapy, the evidence was considered to be high quality.
overall pooled HR did not change virtually. When adjuvant HAI compared with non-HAI, the pooled HR was 0.69 (95 % CI: 0.53–0.89, p =0.004). The result of Egger’s test and Begg’s funnel plot was 0.655 and 0.835, respectively. It demonstrated that no publication bias existed.
Discussion Sensitivity analyses and publication bias Six RCTs were enrolled in sensitivity analysis. After omitting each individual study one at a time mentioned above, the
To the best of our knowledge, this systematic review is the first to assess the long-term outcomes of adjuvant HAI in CRLM. The present study, which included six
1100 Table 4 HAI-related adverse events
Int J Colorectal Dis (2015) 30:1091–1102
Adverse events
Studies (cited in reference list)
Rate (range) (%)
17, 18, 19, 20, 22
5.7–55.4
16, 17, 18, 20 17, 18, 19
1.9–29.0 3.8–18.0
Chemotherapy-related complication Nausea Diarrhea Neutropenia Vomit
17, 18, 22
3.8–11.1
Stomatitis
16, 17, 20
11.0–57.6
Hepatotoxicity Pain
16, 18, 19 19, 20
13.9–30.0 16.7–29.0
Thrombocytopenia Biliary sclerosis
18, 19 9, 21
1.9–6.7 2.3–28.6
Skin reaction
20
26.9
Alopecia Constipation
20 20
26.9 11.8
Anemia Anorexia
19 19
3.3 10.0
Erythema Chemical hepatitis Infection Arterial pseudoaneurysms
19 16 16 9
10.0 27.8 2.8 4.5
Gastrointestinal ulcerations Catheter-related complications
9
4.5
21
19.0
Pump malfunction
RCTs and three cohort studies, provided a relatively high level of evidence showing that CRLM patients treated with adjuvant HAI chemotherapy exhibited a significantly higher 5-year OS (HR 0.75; 95 % CI: 0.56–0.99; p = 0.048; I2 = 37.1 %, p = 0.122) and 5-year DFS (HR 0.61; 95 % CI: 0.48–0.79; p = 0.000; I 2 = 21.0 %, p = 0.275) than those treated without HAI chemotherapy, demonstrating the significant benefit of adjuvant HAI in terms of survival. Complete resection of the CRLM is the only hope for cure, or more humbly, for prolonged survival. However, a multidisciplinary approach is important for coordinating care of CRLM patients, which should be proposed whenever it is possible [23]. Regional chemotherapy has been developed to ensure greater local concentration of cytotoxic agents because liver metastases derive most of their blood supply from the hepatic artery, whereas normal liver tissue is primarily perfused by the portal vein. With HAI, significantly greater concentrations of chemotherapeutic agents can be administered directly to the tumor, achieving significant response rates [24]. Furthermore, drugs used for HAI have a first-pass hepatic clearance effect, such as FUDR and 5-Fu. These drugs have a short half-life and are primarily metabolized in the liver, allowing
extremely low drug concentrations to be maintained in the peripheral blood, thereby providing a higher exposure of chemotherapy to malignant cells and minimizing the risk of systemic adverse events [25]. 5-Fu has been the preferred agent in Europe (usually delivered via surgically placed ports) [26]. FUDR is converted to 5Fu in the liver, which is the preferred agent because of its short half-life, 95 % first-pass extraction rate, up to 400-fold estimated increase in tumor exposure, and low rate of systemic toxicity [27]. FUDR is probably too toxic, due principally to the possible occurrence of severe sclerosing cholangitis even though this adverse event is less frequent with the concomitant use of dexamethasone and a careful follow-up of liver function tests [28]. Recently, oxaliplatin has been used more widely in HAI therapy with favorable results [29–31]. A response rate of 63 % was obtained with HAI oxaliplatin in a multicenter phase 2 study in 28 patients with unresectable CRLM, even if the local concentration of oxaliplatin is less than that of FUDR [29, 32]. The fact that adjuvant HAI chemotherapy prevents the development of liver metastases in patients with stage II or III colorectal cancer suggests that HAI chemotherapy is a very effective local treatment for micrometastases in the liver [33, 34]. Administering chemotherapy after surgery for disappearing liver metastasis that is left in the hepatic remnant
Int J Colorectal Dis (2015) 30:1091–1102
may be important. It has been associated with a sixfold increased probability of pathological complete response in patients with liver metastases that had disappeared on therapy [35]. In the study reported by Elias et al., none of the disappearing liver metastasis recurred among the 6 patients who received postoperative HAI chemotherapy compared with 8 of 11 patients who received either systemic chemotherapy or no chemotherapy [36]. Moreover, it was also considered to perform HAI in patients with multiple liver metastases [8, 16]. Therefore, adjuvant HAI chemotherapy should be reserved for selected patients at high risk of hepatic recurrence. Despite strong evidence suggesting that HAI chemotherapy is effective at controlling CRLM, enthusiasm and widespread adoption have been limited in part because of the significant event rate associated with this therapy. Adverse events can be technical or toxicities related to chemotherapy. The common catheter-related events include catheter displacement, hepatic artery occlusion, and catheter-related infection [37, 38]. The adverse event rates for these issues have been reported to be lower than 7 % in recent studies, compared to 22–35 % in earlier studies. At meanwhile, the common chemotherapy-related event is gastrointestinal symptoms. Nausea and vomiting can occur in 25–35 % patients [39]. Hepatobiliary toxicity, including elevation of serum hepatic transaminase levels, and hyper bilirubinemia, are also important problems [40, 41]. At the present study, only one study reported the catheter-related event was pump-related malfunction. The most usually reported chemotherapy-related events were neutropenia, nausea, diarrhea, vomit, stomatitis, and biliary sclerosis. There was no HAI-related death. The present meta-analysis provided some information on the efficacy of adjuvant HAI. There are, however, several potential limitations that should be taken into account. The first finding, which strongly emerges from this review, is the consistent heterogeneity concerning protocols of HAI and systemic chemotherapy. It is evident that there is still no consensus regarding the most effective combination of drugs in this setting. Second, colorectal liver metastases represent a heterogeneous disease, in that variations are possible in the number of metastases, size, location, and more importantly, in biological characteristics and protein expression, all factors that are linked with the response rate to chemotherapic drugs (monoclonal antibodies). In addition, age, comorbidities, and patient characteristics (compliance) should be taken into account to maximize the benefits of chemotherapy. Finally, although the present meta-analysis included a large sample, the number of studies included for analysis was limited. All included studies are not blinded, which may result in bias, especially in studies with a small sample size. Furthermore, there is still the possibility of undetected bias due to the nonrandomized trials. These factors may have a potential impact on our results. Larger scale studies and more randomized studies are needed to draw definitive conclusions.
1101
Conclusions Adjuvant HAI improved long-term survival of CRLM after hepatic resection, especially for patients with a high risk of recurrence. However, a properly powered clinical trial should be designed to further examine this topic. Conflict of interest The authors declare that they have no competing interests. Funding This study was not supported by any funding.
References 1. 2.
3.
4.
5.
6.
7.
8.
9.
10. 11.
12.
13.
14.
Siegel R, Ma J, Zou Z et al (2014) Cancer statistics, 2014. CA Cancer J Clin 64(1):9–29 Adam R, De Gramont A, Fig.ueras J et al (2012) The oncosurgery approach to managing liver metastases from colorectal cancer: a multidisciplinary international consensus. Oncologist 17(10): 1225–1239 Rees M, Tekkis PP, Welsh FK et al (2008) Evaluation of long-term survival after hepatic resection for metastatic colorectal cancer: a multifactorial model of 929 patients. Ann Surg 247(1):125–135 Choti MA, Sitzmann JV, Tiburi MF et al (2002) Trends in long-term survival following liver resection for hepatic colorectal metastases. Ann Surg 235(6):759–766 Shah SA, Bromberg R, Coates A et al (2007) Survival after liver resection for metastatic colorectal carcinoma in a large population. J Am Coll Surg 205(5):676–683 Ensminger WD, Rosowsky A, Raso V et al (1978) A clinicalpharmacological evaluation of hepatic arterial infusions of 5fluoro-2′-deoxyuridine and 5-fluorouracil. Cancer Res 38(11 Pt 1):3784–3792 Mocellin S, Pilati P, Lise M et al (2007) Meta-analysis of hepatic arterial infusion for unresectable liver metastases from colorectal cancer: the end of an era? J Clin Oncol 25(35):5649–5654 Goere D, Benhaim L, Bonnet S et al (2013) Adjuvant chemotherapy after resection of colorectal liver metastases in patients at high risk of hepatic recurrence: a comparative study between hepatic arterial infusion of oxaliplatin and modern systemic chemotherapy. Ann Surg 257(1):114–120 House MG, Kemeny NE, Gonen M et al (2011) Comparison of adjuvant systemic chemotherapy with or without hepatic arterial infusional chemotherapy after hepatic resection for metastatic colorectal cancer. Ann Surg 254(6):851–856 Higgins JPT, Green S (2011) Cochrane handbook for systematic reviews of interventions version 5.1.0. The Cochrane Collaboration, Oxford Higgins JP, Altman DG, Gotzsche PC et al (2011) The Cochrane collaboration’s tool for assessing risk of bias in randomised trials. BMJ 343:d5928 Wells G, Shea B, O’connell D, Peterson J, Welch V, et al (2000) The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomized studies in metaanalyses.3rd symposium on systematic reviews: beyond the basics 3–5 Guyatt GH, Oxman AD, Vist GE et al (2008) GRADE: an emerging consensus on rating quality of evidence and strength of recommendations. BMJ 336(7650):924–926 Tierney JF, Stewart LA, Ghersi D et al (2007) Practical methods for incorporating summary time-to-event data into meta-analysis. Trials 8:16
1102 15.
Higgins JP, Thompson SG, Deeks JJ et al (2003) Measuring inconsistency in meta-analyses. BMJ 327(7414):557–560 16. Bolton JS, O'Connell MJ, Mahoney MR et al (2012) Hepatic arterial infusion and systemic chemotherapy after multiple metastasectomy in patients with colorectal carcinoma metastatic to the liver: a North Central Cancer Treatment Group (NCCTG) phase II study, 92-46-52. Clin Colorectal Cancer 11(1):31–37 17. Kemeny N, Huang Y, Cohen AM et al (1999) Hepatic arterial infusion of chemotherapy after resection of hepatic metastases from colorectal cancer. N Engl J Med 341(27):2039–2048 18. Kemeny MM, Adak S, Gray B et al (2002) Combined-modality treatment for resectable metastatic colorectal carcinoma to the liver: surgical resection of hepatic metastases in combination with continuous infusion of chemotherapy—an intergroup study. J Clin Oncol 20(6):1499–1505 19. Kusunoki M, Yanagi H, Noda M et al (2000) Results of pharmacokinetic modulating chemotherapy in combination with hepatic arterial 5-fluorouracil infusion and oral UFT after resection of hepatic colorectal metastases. Cancer 89(6):1228–1235 20. Lorenz M, Muller HH, Schramm H et al (1998) Randomized trial of surgery versus surgery followed by adjuvant hepatic arterial infusion with 5-fluorouracil and folinic acid for liver metastases of colorectal cancer. German cooperative on liver metastases (Arbeitsgruppe Lebermetastasen). Ann Surg 228(6):756–762 21. Onaitis M, Morse M, Hurwitz H et al (2003) Adjuvant hepatic arterial chemotherapy following metastasectomy in patients with isolated liver metastases. Ann Surg 237(6):782–8, discussion 788-9 22. Tono T, Hasuike Y, Ohzato H et al (2000) Limited but definite efficacy of prophylactic hepatic arterial infusion chemotherapy after curative resection of colorectal liver metastases: a randomized study. Cancer 88(7):1549–1556 23. Nordlinger B, Vauthey JN, Poston G et al (2010) The timing of chemotherapy and surgery for the treatment of colorectal liver metastases. Clin Colorectal Cancer 9(4):212–218 24. Liang YH, Shao YY, Chen JY et al (2013) Modern prospection for hepatic arterial infusion chemotherapy in malignancies with liver metastases. Int J Hepatol 2013:141590 25. Sadahiro S, Suzuki T, Tanaka A et al (2013) Clinical significance of and future perspectives for hepatic arterial infusion chemotherapy in patients with liver metastases from colorectal cancer. Surg Today 43(10):1088–1094 26. Kerr DJ, Mcardle CS, Ledermann J et al (2003) Intrahepatic arterial versus intravenous fluorouracil and folinic acid for colorectal cancer liver metastases: a multicentre randomised trial. Lancet 361(9355):368–373 27. Kingham TP, D'Angelica M, Kemeny NE (2010) Role of intraarterial hepatic chemotherapy in the treatment of colorectal cancer metastases. J Surg Oncol 102(8):988–995 28. Kemeny N, Seiter K, Niedzwiecki D et al (1992) A randomized trial of intrahepatic infusion of fluorodeoxyuridine with dexamethasone
Int J Colorectal Dis (2015) 30:1091–1102 versus fluorodeoxyuridine alone in the treatment of metastatic colorectal cancer. Cancer 69(2):327–334 29. Ducreux M, Ychou M, Laplanche A et al (2005) Hepatic arterial oxaliplatin infusion plus intravenous chemotherapy in colorectal cancer with inoperable hepatic metastases: a trial of the gastrointestinal group of the Federation Nationale des Centres de Lutte Contre le Cancer. J Clin Oncol 23(22):4881–4887 30. Boige V, Malka D, Elias D et al (2008) Hepatic arterial infusion of oxaliplatin and intravenous LV5FU2 in unresectable liver metastases from colorectal cancer after systemic chemotherapy failure. Ann Surg Oncol 15(1):219–226 31. Goere D, Deshaies I, de Baere T et al (2010) Prolonged survival of initially unresectable hepatic colorectal cancer patients treated with hepatic arterial infusion of oxaliplatin followed by radical surgery of metastases. Ann Surg 251(4):686–691 32. Kern W, Beckert B, Lang N et al (2001) Phase I and pharmacokinetic study of hepatic arterial infusion with oxaliplatin in combination with folinic acid and 5fluorouracil in patients with hepatic metastases from colorectal cancer. Ann Oncol 12(5):599–603 33. Sadahiro S, Suzuki T, Ishikawa K et al (2004) Prophylactic hepatic arterial infusion chemotherapy for the prevention of liver metastasis in patients with colon carcinoma: a randomized control trial. Cancer 100(3):590–597 34. Xu J, Zhong Y, Weixin N et al (2007) Preoperative hepatic and regional arterial chemotherapy in the prevention of liver metastasis after colorectal cancer surgery. Ann Surg 245(4):583–590 35. Auer RC, White RR, Kemeny NE et al (2010) Predictors of a true complete response among disappearing liver metastases from colorectal cancer after chemotherapy. Cancer 116(6):1502–1509 36. Elias D, Goere D, Boige V et al (2007) Outcome of posthepatectomy-missing colorectal liver metastases after complete response to chemotherapy: impact of adjuvant intra-arterial hepatic oxaliplatin. Ann Surg Oncol 14(11):3188–3194 37. Allen PJ, Nissan A, Picon AI et al (2005) Technical complications and durability of hepatic artery infusion pumps for unresectable colorectal liver metastases: an institutional experience of 544 consecutive cases. J Am Coll Surg 201(1):57–65 38. Barnett KT, Malafa MP (2001) Complications of hepatic artery infusion: a review of 4580 reported cases. Int J Gastrointest Cancet 30(3):147–160 39. Kanat O, Gewirtz A, Kemeny N (2012) What is the potential role of hepatic arterial infusion chemo-therapy in the current armamentorium against colorectal cancer. J Gastrointest Oncol 3(2):130–138 40. Shao YY, Huang CC, Liang PC et al (2010) Hepatic arterial infusion of chemotherapy for advanced hepatocellular carcinoma. Asia Pac J Clin Oncol 6(2):80–88 41. Cohen AD, Kemeny NE (2003) An update on hepatic arterial infusion chemotherapy for colorectal cancer. Oncologist 8(6):553–566
ORIGINAL ARTICLE
A systematic review and meta-analysis to reappraise the role of adjuvant hepatic arterial infusion for colorectal cancer liver metastases Wei Liu 1 & Qing-Kun Song 2 & Bao-Cai Xing 1
Accepted: 8 May 2015 / Published online: 27 May 2015 # Springer-Verlag Berlin Heidelberg 2015
Abstract Objective The potential benefit of adjuvant hepatic arterial infusion remains unknown for patients with colorectal liver metastases after radical hepatic resection. The principle aim of this study was to investigate the long-term outcome of adjuvant hepatic arterial infusion. Methods Eligible trials were identified from Embase, PubMed, the Web of Science, and the Cochrane library since their inception to June 1, 2014. Patients with colorectal liver metastases, who underwent radical hepatic resection and received adjuvant hepatic arterial infusion, were enrolled. The study outcomes included 5-year disease-free and overall survival rate, respectively. Hazard ratio with a 95 % confidence interval was used to measure the pooled effect according to a random effects model or fixed effects model, depending on the heterogeneity between the included studies. The statistical heterogeneity between trials was detected by I2 test. Sensitivity analyses were also carried out. Results A total of nine studies containing 1057 patients were included. The comparison indicated that the
Wei Liu and Qing-Kun Song contributed equally to this work. * Bao-Cai Xing [email protected] 1
Hepatopancreatobiliary Surgery Department I, Key Laboratory of Carcinogenesis and Translational Research, Ministry of Education, Beijing Cancer Hospital and Institute, Peking University School of Oncology, No. 52, Fu-Cheng-Lu Street, Beijing 100142, People’s Republic of China
2
Beijing Key Laboratory of Cancer Therapeutic Vaccine, Beijing Shijitan Hospital, Capital Medical University Cancer Center, Beijing, China
overall pooled hazard ratio for 5-year overall survival was 0.75 (95 % CI: 0.56–0.99, p=0.048). The hazard ratio for 5-year disease-free survival rate was 0.61 (95 % CI: 0.48–0.79, p=0.001). When compared with systemic chemotherapy alone, adjuvant hepatic arterial infusion plus systemic chemotherapy also improved the long-term survival. Conclusions Adjuvant hepatic arterial infusion improved the 5-year disease-free and overall survival rate, respectively. It should be recommended for patients with a high risk of recurrence, but these findings require prospective confirmation. Keywords Hepatic artery infusion . Colorectal cancer . Metastases . Meta-analysis
Introduction Colorectal cancer is the second leading cause of cancerrelated death in the world [1]. About 25 % of patients present with synchronous liver metastases and 50 % of patients overall develop liver metastases [2]. Hepatic resection is considered to be the only curable treatment associated with a long survival time for colorectal liver metastases (CRLM), and the overall survival rate at 5 years ranges from 43 to 58 % in recent studies [3]. However, the 5-year disease-free survival rate remains around 28 % [4, 5]. The introduction of more effective systemic and regional chemotherapy for CRLM over the past decade has likely contributed to the improvements in survival after hepatic resection for CRLM [6]. Hepatic arterial infusion (HAI) chemotherapy, a local treatment for liver metastases, has been an appealing investigational
1092
method for patients with CRLM confined to the liver, in whom it has reproducibly yielded higher response rates. HAI has presented a significantly higher response rate than systemic chemotherapy in unresectable CRLM patients [7]. However, the adjuvant HAI chemotherapy significantly prolonged survival in only few studies [8, 9]; the current evidence supporting the efficacy of this therapy remains controversial. Therefore, this meta-analysis was performed to investigate the long-term outcome of adjuvant HAI in CRLM after radical hepatic resection. The treatment benefits were critically discussed in terms of 5-year disease-free and overall survival rate.
Materials and methods Literature search A comprehensive search was performed to identify all published studies of adjuvant HAI and systemic chemotherapy in CRLM after hepatic resection. Literature search of Embase, PubMed, the Web of Science, and the Cochrane database was conducted to identify eligible studies up to June 1, 2014. The following keywords were used as the search strategy: Bhepatic artery^ and Bcolorectal^ and Bcancer^ and Barterial infusion^ or BHAI.^
Int J Colorectal Dis (2015) 30:1091–1102
time from allocation to death by any cause. Living patients were censored on the date of last follow-up. The secondary endpoint was 5-year disease-free survival (DFS) rate, which was defined from the allocation until locoregional recurrence or distant metastasis. Quality of studies Two authors independently evaluated the quality of each retrieved trial according to the method described in the Cochrane Handbook for Systematic Review of Interventions (version 5.1.0) [10]. All studies were subjectively reviewed and assigned a value of Bhigh,^ Blow,^ or Bunclear^ to the following domains: random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective reporting, and other bias. Trials with high risk of bias for any one or more key domains were considered as at high risk of bias. Trials with low risk of bias for all key domains were considered as at low risk of bias. Otherwise, they were considered as unclear risk of bias [11]. A modified Newcastle-Ottawa scale (NOS) was used to assess the quality of the nonrandomized studies included in this meta-analysis [12]. This scale, which ranged from 0 to 9 points, consisted of three items that described the patient selection method, the comparability of adjuvant HAI and systemic chemotherapy, and outcome assessment. Articles scored as ≥6 were considered as high-quality studies.
Study inclusion and exclusion criteria Assessment for grading the quality of evidence We included studies based on the following inclusion criteria: (1) CRLM patients were performed hepatic resection and then underwent HAI chemotherapy in study group, which was compared with hepatic resection alone or adjuvant systemic chemotherapy in control group; (2) data from efficacy and/or safety analyses were reported; and (3) the study was reported in English as a full paper. Letters to the editors and conference abstracts were excluded due to the limited data. Studies without a control arm, animal studies, and review articles were also excluded. Data extraction and outcome measures Two reviewers (Wei Liu and Qing-kun Song) assessed articles independently for their eligibility and extracted data from the included studies. Discrepancies were resolved by consensus. Data retrieved from studies included the first author’s last name, publication year, patients’ characteristics, sample size, study design, duration of follow-up, endpoint evaluations, and methodological quality. The quality of the studies was assessed independently by all investigators. The primary outcome was 5-year overall survival (OS) rate defined as the
The overall quality of the evidence and strength of recommendations were evaluated using GRADE [13]. GRADE Working Group grades of evidence were as follows: high quality, further research is very unlikely to change our confidence in the estimate of effect; moderate quality, further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate; low quality, further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate; very low quality, we are very uncertain about the estimate. Statistical analysis For the outcome, the association between OS, DFS, and adjuvant HAI was expressed as pooled hazard ratio (HR). For the time-to-event variables, the hazard ratios of DFS and OS with 95 %CIs were directly extracted or calculated using a calculation sheet as previously described [14]. Statistical heterogeneity was explored by inconsistency (I 2) statistics; an I 2 value
Int J Colorectal Dis (2015) 30:1091–1102
1093
of 50 % or more represented substantial heterogeneity [15]. In the absence of heterogeneity, studies were pooled using a fixed effect model. If heterogeneity was observed, a random effects model was used. An estimate of potential publication bias for primary outcome was carried out by the Egger’s regression test and Begg’s adjusted rank correlation test. The metaanalyses for pooled HR and sensitivity were performed with Stata software 12.0 (Stata Corp., College Station, TX).
Results
studies were excluded due to the irrelevance to the current analysis or their disagreement with the inclusion criteria. Finally, nine studies were eligible for this meta-analysis [8, 9, 16–22] (Fig. 1). The characteristics of the nine included studies are outlined in Table 1. The studies differed somewhat in a number of details regarding HAI chemotherapy. Early studies used some form of fluorouracil (5-Fu) and floxuridine (FUDR), with or without dexamethasone (DXM). Recently, oxaliplatin (OXA) or irinotecan (CPT11) was the main regimen. Moreover, three studies underwent surgery alone and another six studies received systemic chemotherapy. The median follow-up period ranged from 29.0 to 62.7 months (Table 2).
Literature research and characteristic of studies
5-year DFS and OS rate
A total of 358 studies containing 1057 patients were identified based on the initial search. After an independent review, 321
Six studies (724 patients) were identified with outcome measurement of 5-year disease-free survival rate. The pooled HR
Fig. 1 Flow chart for studies selection
Publications identified after retrieval and relevant reviews research (N=358)
Studies excluded after screening title and abstract (N=321) Non-relevant (N=288) Not clinical trials (N=12) Reviews (N=6) Non-English (N=15)
Potentially relevant studies in full text for detailed evaluation (N=37)
Studies excluded (N=25) Lack of relevance (N=15) Duplicates studies (N=7) No survival outcome report (N=3)
Potentially appropriate studies included in (N=12)
Studies excluded (N=3) Study with insufficient data (N=1) Studies with low quality (N=1) Duplicates studies (N=1)
Studies included in meta-analysis (N=9)
2011 2013 2011
1999 2002 2000
1998 2003
2000
Bolton et al. [16]. Goere et al. [8]. House et al. [9].
Kemeny N et al. [17]. Kemeny M et al. [18]. Kusunoki et al. [19].
Lorenz et al. [20]. Onaitis et al. [21].
Tono et al. [22].
RCT
RCT Cohort
RCT RCT RCT
RCT Cohort Cohort
Study design
N/A
N/A N/A
N/A N/A N/A
N/A 7 N/A
Preoperative HAI (n)
N/A
N/A N/A
N/A N/A N/A
N/A 37 69
Preoperative sys (n)
HAI(5-Fu)+Sys(5-Fu)
HAI(5-Fu+FA) HAI(FUDR+ DXM)+Sys/HAI
HAI(FUDR+DXM)+Sys(5-Fu±LV) HAI(FUDR)+Sys(5-Fu) HAI(5-Fu)+Sys(UFT)
HAI(FUDR)+Sys(5-Fu+LV) HAI(OXA/CPT-11)+Sys(5-Fu+LV) HAI(FUDR + DXM)+Sys(5-Fu+OXA/CPT-11)
Regimens of study group
Sys(5-Fu)
None Sys(N/A)
Sys(LV+ 5-Fu) None Sys(UFT)
None Sys(5-Fu+OXA/CPT-11) Sys (5-Fu+OXA/CPT-11)
Regimens of control group
24 weeks
6 N/A
6 8 11 weeks
N/A 6 N/A
HAI duration/ course(s)
49 98 250
156 109 58 226 92
19
Bolton et al. [16]. Goere et al. [8]. House et al. [9].
Kemeny N et al. [17]. Kemeny M et al. [18]. Kusunoki et al. [19]. Lorenz et al. [20] Onaitis et al. [21].
Tono et al. [22].
9 vs. 10
74 vs. 82 53 vs. 56 30 vs. 28 113 vs. 113 21 vs. 71
36 vs. 13 44 vs. 54 125 vs. 125
Sample (study vs. control)
N/A
N/A N/A N/A N/A Isolated CRLM
No. of CRLM≥4 No. of CRLM≥3 N/A
Indication for HAI
26.3±14.9
N/A N/A 25 (10–100) N/A 22±41
N/A 29 (17–40) 28 (2–170)
Tumor size (mm)
N/A
47 (64 %) N/A N/A 49 (46.3 %) N/A
6 (17 %) N/A 76 (61 %)
Multiple tumors (n)
N/A
11.5 (0.9–258.9) N/A 11.7 (1–1020) N/A 11.4±25.4
N/A 4.5 (3–10) 11.7 (1.2–1235)
CEA level (ng/ml)
N/A
12 (57.1 %) N/A N/A 53 (49.5 %) N/A
N/A N/A 84 (68 %)
Primary N stage(+) (n)
4 (44.4 %)
23 (31 %) 15 (28.3 %) 12 (40.4 %) 34 (31.5 %) N/A
N/A 38 (86 %) 62 (50 %)
Synchronous CRLM (n)
62.2
62.7 51 N/A N/A 29
N/A 60 43
Median follow-up (months)
OS+DFS
OS+DFS DFS OS+DFS OS+DFS OS+DFS
OS+DFS OS+DFS DSS
Primary end-point
CRLM colorectal liver metastases, OS overall survival, DFS disease-free survival, DSS disease-specific survival, Study group hepatic resection with adjuvant HAI ± systemic chemotherapy, Control group hepatic resection ± systemic chemotherapy, N/A not available
Patients (n)
Characteristics of clinical trials included in the meta-analysis
Author
Table 2
HAI hepatic arterial infusion, Sys systemic chemotherapy, 5-Fu fluorouracil, FUDR floxuridine, LV leucovorin, UFT uracil tegafur, OXA oxaliplatin, CPT-11 irinotecan, DXM dexamethasone, Study group hepatic resection with adjuvant HAI±systemic chemotherapy, Control group hepatic resection±systemic chemotherapy, N/A not available
Year
Interventions of clinical trials included in the meta-analysis
Author
Table 1
1094 Int J Colorectal Dis (2015) 30:1091–1102
Int J Colorectal Dis (2015) 30:1091–1102
1095
Fig. 2 The pooled HR for 5-year DFS (a) and OS (b) in the included six studies (724 patients) and nine studies (1057 patients), respectively, calculated using the random-effects model
1096
Int J Colorectal Dis (2015) 30:1091–1102
Fig. 3 The pooled HR for 5-year DFS (a) and OS (b) in the included studies (three RCTs (284 patients) and (six RCTs (617 patients), respectively) calculated using the random-effects model
for 5-year DFS in the included studies calculated using the random effects model was 0.61 (95 % CI: 0.48–0.79; p=
0.000; I2 =21.0 %, p=0.275) (Fig. 2a). Nine studies (1057 patients) were identified with outcome measurement of 5-
Int J Colorectal Dis (2015) 30:1091–1102
1097
Fig. 4 Four studies (523 patients) were identified with outcome measurement of 5-year DFS rate. The pooled HR for 5-year a DFS (four studies (523 patients)) and b OS (five studies (581 patients)) in the included studies calculated using the random-effects model
year overall survival rates. The pooled HR for 5-year OS in the included studies calculated using the random effects model
was 0.75 (95 % CI: 0.56–0.99; p=0.048; I2 =37.1 %, p= 0.122) (Fig. 2b).
Int J Colorectal Dis (2015) 30:1091–1102 Total quality score
1098
9
9
√
√ √ √
√
9 √
√ √ √
√ √ √
Adequacy of follow-up duration Assessment of outcome with independency Comparability between patients in different treatment armssecondary factor: tumor size
Subgroup analysis for HAI plus systemic chemotherapy compared with systemic chemotherapy Four studies (523 patients) were identified with outcome measurement of 5-year DFS rate. The pooled HR for 5-year DFS in the included studies calculated using the random effects model was 0.55 (95 % CI: 0.42–0.72; p=0.000; I2 =11.2 %, p=0.337) (Fig. 4a). Five studies (581 patients) were identified with outcome measurement of 5-year OS rate. The pooled HR for 5-year OS in the included studies calculated using the random effects model was 0.68 (95 % CI: 0.50–0.94; p=0.020; I2 =0.40 %, p= 0.404) (Fig. 4b).
√ √ √
√ √
√ √
√ House et al. [9].
√
√ Goere et al. [8].
√
Eight studies reported the HAI-related events (Table 3). Only one study reported the catheter-related event that was pumprelated malfunction (19 %). The most usually reported hematological events were neutropenia (range 3.8–18 %) and thrombocytopenia (range 1.9–6.7 %). The most typical nonhematological side effects were nausea (range 5.7– 55.4 %), diarrhea (range 1.9–29 %), vomit (range 3.8– 11.1 %), stomatitis (range 11.0–57.6 %), hepatotoxicity (range 13.9–30.0 %), and biliary sclerosis (range 2.3–28.6 %). There was no HAI-related death.
√
√
√
Quality of the studies
Onaitis et al. [21]. √
Comparability between patients in different treatment arms primary factor: number of tumor Representativeness Selection of the Ascertainment of of the treatment arm comparative treatment treatment arm regimen
Demonstration that the outcome of interest was not present at the beginning of study
Three RCTs (284 patients) were identified with outcome measurement of 5-year DFS rates. The pooled HR for 5-year DFS in the included studies calculated using the random effects model was 0.52 (95 % CI: 0.37–0.83; p = 0.007; I2 = 16.5 %, p = 0.302) (Fig. 3a). Six RCTs (617 patients) were identified with outcome measurement of 5-year OS rates. The pooled HR for 5-year OS in the included studies calculated using the random effects model was 0.69 (95 % CI: 0.53–0.89; p=0.004; I2 =0 %, p=0.635) (Fig. 3b).
HAI-related adverse events
Author
Comparability Non-RCT studies Selection
Table 3
Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomized trials
Assessment of outcome
Loss to follow-up acceptable (less than 10 % and reported)
Subgroup analysis for 5-year DFS and OS rates
Randomized sequence generation was conducted adequately, and allocation concealment was not described in most of studies. Among all the selected studies, participants and personnel were not blinded. All the selected studies were considered as at high risk of bias (Fig. 5). The quality of the nonrandomized studies was assessed using the NOS, and the score was 9 for each study, indicating that these studies were of high quality (Table 4).
Int J Colorectal Dis (2015) 30:1091–1102
1099
Fig. 5 Risk of bias summary (a) and graph (b)
Assessment for grading the quality of evidence GRADE Working Group grades of evidence were moderate quality for 5-year OS and moderate quality for 5-year DFS. When adjuvant HAI plus systemic chemotherapy compared with systemic chemotherapy, the evidence was considered to be high quality.
overall pooled HR did not change virtually. When adjuvant HAI compared with non-HAI, the pooled HR was 0.69 (95 % CI: 0.53–0.89, p =0.004). The result of Egger’s test and Begg’s funnel plot was 0.655 and 0.835, respectively. It demonstrated that no publication bias existed.
Discussion Sensitivity analyses and publication bias Six RCTs were enrolled in sensitivity analysis. After omitting each individual study one at a time mentioned above, the
To the best of our knowledge, this systematic review is the first to assess the long-term outcomes of adjuvant HAI in CRLM. The present study, which included six
1100 Table 4 HAI-related adverse events
Int J Colorectal Dis (2015) 30:1091–1102
Adverse events
Studies (cited in reference list)
Rate (range) (%)
17, 18, 19, 20, 22
5.7–55.4
16, 17, 18, 20 17, 18, 19
1.9–29.0 3.8–18.0
Chemotherapy-related complication Nausea Diarrhea Neutropenia Vomit
17, 18, 22
3.8–11.1
Stomatitis
16, 17, 20
11.0–57.6
Hepatotoxicity Pain
16, 18, 19 19, 20
13.9–30.0 16.7–29.0
Thrombocytopenia Biliary sclerosis
18, 19 9, 21
1.9–6.7 2.3–28.6
Skin reaction
20
26.9
Alopecia Constipation
20 20
26.9 11.8
Anemia Anorexia
19 19
3.3 10.0
Erythema Chemical hepatitis Infection Arterial pseudoaneurysms
19 16 16 9
10.0 27.8 2.8 4.5
Gastrointestinal ulcerations Catheter-related complications
9
4.5
21
19.0
Pump malfunction
RCTs and three cohort studies, provided a relatively high level of evidence showing that CRLM patients treated with adjuvant HAI chemotherapy exhibited a significantly higher 5-year OS (HR 0.75; 95 % CI: 0.56–0.99; p = 0.048; I2 = 37.1 %, p = 0.122) and 5-year DFS (HR 0.61; 95 % CI: 0.48–0.79; p = 0.000; I 2 = 21.0 %, p = 0.275) than those treated without HAI chemotherapy, demonstrating the significant benefit of adjuvant HAI in terms of survival. Complete resection of the CRLM is the only hope for cure, or more humbly, for prolonged survival. However, a multidisciplinary approach is important for coordinating care of CRLM patients, which should be proposed whenever it is possible [23]. Regional chemotherapy has been developed to ensure greater local concentration of cytotoxic agents because liver metastases derive most of their blood supply from the hepatic artery, whereas normal liver tissue is primarily perfused by the portal vein. With HAI, significantly greater concentrations of chemotherapeutic agents can be administered directly to the tumor, achieving significant response rates [24]. Furthermore, drugs used for HAI have a first-pass hepatic clearance effect, such as FUDR and 5-Fu. These drugs have a short half-life and are primarily metabolized in the liver, allowing
extremely low drug concentrations to be maintained in the peripheral blood, thereby providing a higher exposure of chemotherapy to malignant cells and minimizing the risk of systemic adverse events [25]. 5-Fu has been the preferred agent in Europe (usually delivered via surgically placed ports) [26]. FUDR is converted to 5Fu in the liver, which is the preferred agent because of its short half-life, 95 % first-pass extraction rate, up to 400-fold estimated increase in tumor exposure, and low rate of systemic toxicity [27]. FUDR is probably too toxic, due principally to the possible occurrence of severe sclerosing cholangitis even though this adverse event is less frequent with the concomitant use of dexamethasone and a careful follow-up of liver function tests [28]. Recently, oxaliplatin has been used more widely in HAI therapy with favorable results [29–31]. A response rate of 63 % was obtained with HAI oxaliplatin in a multicenter phase 2 study in 28 patients with unresectable CRLM, even if the local concentration of oxaliplatin is less than that of FUDR [29, 32]. The fact that adjuvant HAI chemotherapy prevents the development of liver metastases in patients with stage II or III colorectal cancer suggests that HAI chemotherapy is a very effective local treatment for micrometastases in the liver [33, 34]. Administering chemotherapy after surgery for disappearing liver metastasis that is left in the hepatic remnant
Int J Colorectal Dis (2015) 30:1091–1102
may be important. It has been associated with a sixfold increased probability of pathological complete response in patients with liver metastases that had disappeared on therapy [35]. In the study reported by Elias et al., none of the disappearing liver metastasis recurred among the 6 patients who received postoperative HAI chemotherapy compared with 8 of 11 patients who received either systemic chemotherapy or no chemotherapy [36]. Moreover, it was also considered to perform HAI in patients with multiple liver metastases [8, 16]. Therefore, adjuvant HAI chemotherapy should be reserved for selected patients at high risk of hepatic recurrence. Despite strong evidence suggesting that HAI chemotherapy is effective at controlling CRLM, enthusiasm and widespread adoption have been limited in part because of the significant event rate associated with this therapy. Adverse events can be technical or toxicities related to chemotherapy. The common catheter-related events include catheter displacement, hepatic artery occlusion, and catheter-related infection [37, 38]. The adverse event rates for these issues have been reported to be lower than 7 % in recent studies, compared to 22–35 % in earlier studies. At meanwhile, the common chemotherapy-related event is gastrointestinal symptoms. Nausea and vomiting can occur in 25–35 % patients [39]. Hepatobiliary toxicity, including elevation of serum hepatic transaminase levels, and hyper bilirubinemia, are also important problems [40, 41]. At the present study, only one study reported the catheter-related event was pump-related malfunction. The most usually reported chemotherapy-related events were neutropenia, nausea, diarrhea, vomit, stomatitis, and biliary sclerosis. There was no HAI-related death. The present meta-analysis provided some information on the efficacy of adjuvant HAI. There are, however, several potential limitations that should be taken into account. The first finding, which strongly emerges from this review, is the consistent heterogeneity concerning protocols of HAI and systemic chemotherapy. It is evident that there is still no consensus regarding the most effective combination of drugs in this setting. Second, colorectal liver metastases represent a heterogeneous disease, in that variations are possible in the number of metastases, size, location, and more importantly, in biological characteristics and protein expression, all factors that are linked with the response rate to chemotherapic drugs (monoclonal antibodies). In addition, age, comorbidities, and patient characteristics (compliance) should be taken into account to maximize the benefits of chemotherapy. Finally, although the present meta-analysis included a large sample, the number of studies included for analysis was limited. All included studies are not blinded, which may result in bias, especially in studies with a small sample size. Furthermore, there is still the possibility of undetected bias due to the nonrandomized trials. These factors may have a potential impact on our results. Larger scale studies and more randomized studies are needed to draw definitive conclusions.
1101
Conclusions Adjuvant HAI improved long-term survival of CRLM after hepatic resection, especially for patients with a high risk of recurrence. However, a properly powered clinical trial should be designed to further examine this topic. Conflict of interest The authors declare that they have no competing interests. Funding This study was not supported by any funding.
References 1. 2.
3.
4.
5.
6.
7.
8.
9.
10. 11.
12.
13.
14.
Siegel R, Ma J, Zou Z et al (2014) Cancer statistics, 2014. CA Cancer J Clin 64(1):9–29 Adam R, De Gramont A, Fig.ueras J et al (2012) The oncosurgery approach to managing liver metastases from colorectal cancer: a multidisciplinary international consensus. Oncologist 17(10): 1225–1239 Rees M, Tekkis PP, Welsh FK et al (2008) Evaluation of long-term survival after hepatic resection for metastatic colorectal cancer: a multifactorial model of 929 patients. Ann Surg 247(1):125–135 Choti MA, Sitzmann JV, Tiburi MF et al (2002) Trends in long-term survival following liver resection for hepatic colorectal metastases. Ann Surg 235(6):759–766 Shah SA, Bromberg R, Coates A et al (2007) Survival after liver resection for metastatic colorectal carcinoma in a large population. J Am Coll Surg 205(5):676–683 Ensminger WD, Rosowsky A, Raso V et al (1978) A clinicalpharmacological evaluation of hepatic arterial infusions of 5fluoro-2′-deoxyuridine and 5-fluorouracil. Cancer Res 38(11 Pt 1):3784–3792 Mocellin S, Pilati P, Lise M et al (2007) Meta-analysis of hepatic arterial infusion for unresectable liver metastases from colorectal cancer: the end of an era? J Clin Oncol 25(35):5649–5654 Goere D, Benhaim L, Bonnet S et al (2013) Adjuvant chemotherapy after resection of colorectal liver metastases in patients at high risk of hepatic recurrence: a comparative study between hepatic arterial infusion of oxaliplatin and modern systemic chemotherapy. Ann Surg 257(1):114–120 House MG, Kemeny NE, Gonen M et al (2011) Comparison of adjuvant systemic chemotherapy with or without hepatic arterial infusional chemotherapy after hepatic resection for metastatic colorectal cancer. Ann Surg 254(6):851–856 Higgins JPT, Green S (2011) Cochrane handbook for systematic reviews of interventions version 5.1.0. The Cochrane Collaboration, Oxford Higgins JP, Altman DG, Gotzsche PC et al (2011) The Cochrane collaboration’s tool for assessing risk of bias in randomised trials. BMJ 343:d5928 Wells G, Shea B, O’connell D, Peterson J, Welch V, et al (2000) The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomized studies in metaanalyses.3rd symposium on systematic reviews: beyond the basics 3–5 Guyatt GH, Oxman AD, Vist GE et al (2008) GRADE: an emerging consensus on rating quality of evidence and strength of recommendations. BMJ 336(7650):924–926 Tierney JF, Stewart LA, Ghersi D et al (2007) Practical methods for incorporating summary time-to-event data into meta-analysis. Trials 8:16
1102 15.
Higgins JP, Thompson SG, Deeks JJ et al (2003) Measuring inconsistency in meta-analyses. BMJ 327(7414):557–560 16. Bolton JS, O'Connell MJ, Mahoney MR et al (2012) Hepatic arterial infusion and systemic chemotherapy after multiple metastasectomy in patients with colorectal carcinoma metastatic to the liver: a North Central Cancer Treatment Group (NCCTG) phase II study, 92-46-52. Clin Colorectal Cancer 11(1):31–37 17. Kemeny N, Huang Y, Cohen AM et al (1999) Hepatic arterial infusion of chemotherapy after resection of hepatic metastases from colorectal cancer. N Engl J Med 341(27):2039–2048 18. Kemeny MM, Adak S, Gray B et al (2002) Combined-modality treatment for resectable metastatic colorectal carcinoma to the liver: surgical resection of hepatic metastases in combination with continuous infusion of chemotherapy—an intergroup study. J Clin Oncol 20(6):1499–1505 19. Kusunoki M, Yanagi H, Noda M et al (2000) Results of pharmacokinetic modulating chemotherapy in combination with hepatic arterial 5-fluorouracil infusion and oral UFT after resection of hepatic colorectal metastases. Cancer 89(6):1228–1235 20. Lorenz M, Muller HH, Schramm H et al (1998) Randomized trial of surgery versus surgery followed by adjuvant hepatic arterial infusion with 5-fluorouracil and folinic acid for liver metastases of colorectal cancer. German cooperative on liver metastases (Arbeitsgruppe Lebermetastasen). Ann Surg 228(6):756–762 21. Onaitis M, Morse M, Hurwitz H et al (2003) Adjuvant hepatic arterial chemotherapy following metastasectomy in patients with isolated liver metastases. Ann Surg 237(6):782–8, discussion 788-9 22. Tono T, Hasuike Y, Ohzato H et al (2000) Limited but definite efficacy of prophylactic hepatic arterial infusion chemotherapy after curative resection of colorectal liver metastases: a randomized study. Cancer 88(7):1549–1556 23. Nordlinger B, Vauthey JN, Poston G et al (2010) The timing of chemotherapy and surgery for the treatment of colorectal liver metastases. Clin Colorectal Cancer 9(4):212–218 24. Liang YH, Shao YY, Chen JY et al (2013) Modern prospection for hepatic arterial infusion chemotherapy in malignancies with liver metastases. Int J Hepatol 2013:141590 25. Sadahiro S, Suzuki T, Tanaka A et al (2013) Clinical significance of and future perspectives for hepatic arterial infusion chemotherapy in patients with liver metastases from colorectal cancer. Surg Today 43(10):1088–1094 26. Kerr DJ, Mcardle CS, Ledermann J et al (2003) Intrahepatic arterial versus intravenous fluorouracil and folinic acid for colorectal cancer liver metastases: a multicentre randomised trial. Lancet 361(9355):368–373 27. Kingham TP, D'Angelica M, Kemeny NE (2010) Role of intraarterial hepatic chemotherapy in the treatment of colorectal cancer metastases. J Surg Oncol 102(8):988–995 28. Kemeny N, Seiter K, Niedzwiecki D et al (1992) A randomized trial of intrahepatic infusion of fluorodeoxyuridine with dexamethasone
Int J Colorectal Dis (2015) 30:1091–1102 versus fluorodeoxyuridine alone in the treatment of metastatic colorectal cancer. Cancer 69(2):327–334 29. Ducreux M, Ychou M, Laplanche A et al (2005) Hepatic arterial oxaliplatin infusion plus intravenous chemotherapy in colorectal cancer with inoperable hepatic metastases: a trial of the gastrointestinal group of the Federation Nationale des Centres de Lutte Contre le Cancer. J Clin Oncol 23(22):4881–4887 30. Boige V, Malka D, Elias D et al (2008) Hepatic arterial infusion of oxaliplatin and intravenous LV5FU2 in unresectable liver metastases from colorectal cancer after systemic chemotherapy failure. Ann Surg Oncol 15(1):219–226 31. Goere D, Deshaies I, de Baere T et al (2010) Prolonged survival of initially unresectable hepatic colorectal cancer patients treated with hepatic arterial infusion of oxaliplatin followed by radical surgery of metastases. Ann Surg 251(4):686–691 32. Kern W, Beckert B, Lang N et al (2001) Phase I and pharmacokinetic study of hepatic arterial infusion with oxaliplatin in combination with folinic acid and 5fluorouracil in patients with hepatic metastases from colorectal cancer. Ann Oncol 12(5):599–603 33. Sadahiro S, Suzuki T, Ishikawa K et al (2004) Prophylactic hepatic arterial infusion chemotherapy for the prevention of liver metastasis in patients with colon carcinoma: a randomized control trial. Cancer 100(3):590–597 34. Xu J, Zhong Y, Weixin N et al (2007) Preoperative hepatic and regional arterial chemotherapy in the prevention of liver metastasis after colorectal cancer surgery. Ann Surg 245(4):583–590 35. Auer RC, White RR, Kemeny NE et al (2010) Predictors of a true complete response among disappearing liver metastases from colorectal cancer after chemotherapy. Cancer 116(6):1502–1509 36. Elias D, Goere D, Boige V et al (2007) Outcome of posthepatectomy-missing colorectal liver metastases after complete response to chemotherapy: impact of adjuvant intra-arterial hepatic oxaliplatin. Ann Surg Oncol 14(11):3188–3194 37. Allen PJ, Nissan A, Picon AI et al (2005) Technical complications and durability of hepatic artery infusion pumps for unresectable colorectal liver metastases: an institutional experience of 544 consecutive cases. J Am Coll Surg 201(1):57–65 38. Barnett KT, Malafa MP (2001) Complications of hepatic artery infusion: a review of 4580 reported cases. Int J Gastrointest Cancet 30(3):147–160 39. Kanat O, Gewirtz A, Kemeny N (2012) What is the potential role of hepatic arterial infusion chemo-therapy in the current armamentorium against colorectal cancer. J Gastrointest Oncol 3(2):130–138 40. Shao YY, Huang CC, Liang PC et al (2010) Hepatic arterial infusion of chemotherapy for advanced hepatocellular carcinoma. Asia Pac J Clin Oncol 6(2):80–88 41. Cohen AD, Kemeny NE (2003) An update on hepatic arterial infusion chemotherapy for colorectal cancer. Oncologist 8(6):553–566
