P
Patient population
Unresectable, metastatic colon cancer with an asymptomatic primary tumor
I
Intervention
Primary tumor resection (colectomy) followed by 1st line chemotherapy,
C
Comparator
1st line chemotherapy with primary tumor resection only if/when patient becomes symptomatic
O
Outcomes
Overall survival, Hazard Ratio
Search Strategy
A detailed search of the Embase-Medline databases was conducted for current medical literature published from 2010 to 2015. The following search terms were employed to identify relevant articles: (“colon” OR “colorectal”) AND (“cancer” OR “carcinoma” OR “adenocarcinoma”) AND (“metastatic” OR “Stage IV” OR “Stage 4”) AND “asymptomatic” AND (“surgery” OR “colectomy” OR “resection”). Duplicate articles were excluded. We included 14 articles, published from 2010 to 2015, that were identified in a Cochrane review and meta-analysis on this specific topic, and were not identified in our initial literature search [21, 22]. The title and abstracts of English-language articles were assessed for relevance. We excluded articles for the following reasons: not relevant, no comparator group (trend analysis), review/opinion articles without primary data, and systematic literature reviews/meta-analyses. A total of 30 articles met the inclusion criteria for full review. Full-text articles were excluded if they were limited to an abstract/poster, contained data duplicated in a different journal, or reported on ongoing trials without reporting any preliminary data. Fifteen manuscripts remained for analysis, five of which were identified from the Cochrane Review and meta-analysis. The literature review process, following PRISMA guidelines, is detailed in Fig. 14.1. Selected articles were abstracted for several variables including study design, time interval, patient population, chemotherapy, survival, and quality (Table 14.2).
Fig. 14.1
PRISMA diagram, systematic literature search results
Table 14.2
Literature search results
First author | Year | Study design | Interval | Patient population | Chemotherapy regimen(s) | Patients (n) | Median overall survival (mo) | Survival | Acute surgery (NR) | Quality of evidence | ||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
R | NR | R | NR | HR | % | |||||||
Seo | 2010 | Retrospective cohort | 2001–2008 | umCRC + APT | FL ± Ox/Iri ± Bev/Cetux | 196 | 83 | 22 | 14 | – | 8.4 | Low |
Chan | 2010 | Retrospective cohort | 2000–2002 | mCRC | Not reported | 286 | 125 | 14 | 6 | – | – | Very low |
Venderbosch | 2011 | Retrospective review of RCT (CAIRO I) | 2003–2004 | mCRC | CAPOXIRI | 258 | 141 | 16.7 | 11.4 | 0.63 | – | Very low |
2011 | Retrospective review of RCT (CAIRO II) | 2005–2006 | mCRC | CAPOX/Bev ± Cetux | 159 | 289 | 20.7 | 13.4 | 0.65 | – | Very low | |
Karoui | 2011 | Retrospective cohort, propensity scored | 1998–2007 | umCC ± APT | FL ± Ox/Iri ± Bev/Cetux | 85 | 123 | 30.7 | 21.9 | – | 19 | Low-Mod |
Verberne | 2011 | Retrospective cohort | 2002–2006 | mRC ± APT | Not reported | 26 | 21 | 26 | 17 | 0.5 | – | Very low |
Cetin | 2013 | Retrospective cohort | 2006–2010 | umCRC + APT | CAPOX/IFL/FOLFIRI ± Bev | 53 | 46 | 23 | 17 | – | 4.4 | Low |
Boselli | 2013 | Retrospecitve cohort | 2010–2011 | umCRC + APT | FOLOX ± Bev | 17 | 31 | 4 | 5 | – | – | Very low |
Ferrand | 2013 | Retrospective review of RCT (FFCD 9601) | 1997–2001 | mCRC (mCC) | FL | 156 | 56 | 16.3 (15.2) | 9.5 (11.1) | – | 7 | Low |
Yun | 2014 | Retrospective cohort, propensity matched | 2000–2008 | umCRC + APT umCRC + APT | ± FL ± Ox/Iri ± Bev/Cetux | 113 (286) | 113 (198) | 17.2 | 14.4 | – | 4.5 | Moderate |
Watanabe | 2014 | Retrospective cohort | 2002–2009 | umCRC + APT | FL ± Ox/Iri ± Bev/Cetux | 46 | 112 | 19.9 | 19.0 | – | 21 | Low |
Yoon | 2014 | Retrospective cohort, propensity matched | 2000–2007 | umCRC ± APT | FL/Cape ± Ox/Iri ± Bev/Cetux | 51 (195) | 51 (66) | 16.5 | 12 | 0.68 | – | Low-Mod |
Matsumoto | 2014 | Retrospective cohort | 2005–2011 | umCRC + APT | Fl ± Ox/Iri ± Bev/Cetux | 41 | 47 | 23.9 | 23.6 | 0.72 | 25.5 | Low |
Tsang | 2014 | Retrospective cohort (SEER) | 1996–2007 | mCRC | Not reported | 8599 | 3117 | 21 | 10 | – | – | Very low |
Tarantino | 2015 | Retrospective cohort, propensity scored (SEER) | 1998–2008 | mCRC | Not reported | 22858 | 17575 | – | – | 0.40 | – | Very low |
Ahmed | 2015 | Retrospective cohort | 1992–2005 | mCRC + APT | Not reported | 521 | 313 | 18.0 | 8.1 | 0.52 | – | Very low |
Results
Our literature search identified 15 recently published retrospective studies, in which patients received primarily multi-drug chemotherapy. No prospective observational or randomized controlled trials (RCTs) have been published to date, and secondary analyses of these trials are limited.
Four RCTs were initiated to address this question, although two have already closed due to poor accrual [23, 24]. The Dutch Colorectal Cancer Group (CAIRO4) and the German SYNCHRONOUS trial group have opened multicenter, randomized, superiority trials comparing primary tumor resection + fluoropyrimidine-based regimens with targeted therapy, vs. fluoropyrimidine-based regimens with targeted therapy alone [25, 26]. The results of these trials are not anticipated for several years, but will obviously have a significant impact on surgical decision-making. Until that time, the data from our literature search represents the body of knowledge available on which surgeons may base decisions. Many studies report upfront resection vs. no resection, but do not address the more important question of upfront surgery and chemotherapy vs. chemotherapy alone. Among both groups, there were limited or no data regarding chemotherapy received, and need for acute surgery while on chemotherapy. Recognizing these limitations, a review of the current literature does provide some guidance to the practicing surgeon who is attempting to decide whether or not to resect the primary colon tumor before initiating chemotherapy in patients with unresectable metastatic disease.
Overall Survival
Twelve of the 15 studies identified in our search demonstrated better OS with primary tumor resection vs. no resection in patients with metastatic colon cancer, with a median survival benefit of 7 months. At face value, these results suggest superior OS with primary tumor resection prior to the development of symptoms in patients with unresectable metastatic CRC. Yet on closer analysis, there are serious limitations to these findings, and they should therefore be interpreted with ample skepticism.
To reduce the impact of selection bias and potential confounders, four studies used propensity score modeling, with variable results. Two groups from Korea used propensity scores to match patients who underwent initial primary tumor resection + chemotherapy vs. chemotherapy alone. The smaller of these studies found a statistically significant OS benefit with primary tumor resection (16.5 vs. 12 months, p = 0.048) [21], whereas the other, much larger study found no statistical difference in OS between these groups (17.2 vs. 14.4 months, p = 0.27) [27, 28]. In the remaining two studies, the data were not sufficiently defined or granular, and the results are less compelling [29, 30]. A French publication reported that OS was superior with primary tumor resection + chemotherapy vs. chemotherapy (30.7 vs. 21.9 months, p = 0.031) even after propensity analysis; however, a significant proportion of patients in both groups had obstructive primary tumors at initial presentation (38.8 % vs 26.5 %), and it remains unclear if the survival advantage was from primary resection vs. stent placement or chemotherapy [29]. Similarly, a large U.S. SEER Database analysis demonstrated a dramatic survival advantage with primary tumor resection vs. no resection (HR = 0.40, p < 0.001) even after propensity matching (for such factors as age, grade, baseline carcinoembryonic antigen level), but potential confounders such as chemotherapy, performance status, comorbidity, and metastatic disease extent/resectability were not reported [30]. In the end, after controlling for confounding and sufficiently defining the target population, the data did not suggest a significant survival advantage with upfront surgery in asymptomatic patients.
Chemotherapy and Survival
A central criticism of earlier retrospective studies comparing primary tumor resection vs. initial chemotherapy has been the reliance on 5-fluorouracil (5-FU)/leucovorin monotherapy, which was the only chemotherapy agent available prior to the early 2000s. To restrict our literature search to modern chemotherapeutic regimens, we limited our investigation to studies published from 2010 to the present. Despite these efforts, seven of the selected articles included patients receiving 5-FU/leucovorin or the oral 5-FU pro-drug capecitabine alone [27–29, 31–34], and five studies failed to report whether chemotherapy was even administered [30, 35–38]. Furthermore, considerable heterogeneity in chemotherapy regimens existed in all but three of the studies. In these three studies, patients received only irinotecan- or oxaliplatin-based chemotherapy, with or without the monoclonal antibodies bevacizumab (anti-vascular endothelial growth factor) and cetuximab (anti-epidermal growth factor receptor) [39–41]. One of these studies retrospectively analyzed two RCTs and demonstrated a 5-month survival benefit with primary tumor resection and subsequent palliative chemotherapy; however, this study was limited by the fact that patients in the chemotherapy-only group had a statistically greater metastatic disease burden [39]. In the two remaining studies, primary resection followed by chemotherapy vs. chemotherapy alone did not significantly improve OS [40, 41].
Related posts:
IBD: Management of a Painful Anal Fissure and Skin Tags in Patients with Crohn’s Disease
Deciding on an IRA vs. IPAA for FAP
Benign Anal Disease: When to Operate on the Patient with an Anal Fissure
Anal Fissure: Recurrence After Lateral Internal Sphincterotomy
Which Patients with Fecal Incontinence Require Physiologic Workup?
Conventional vs Single Port Approaches to Laparoscopic Colectomy
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
