Short-Term Outcomes of the Australasian RandomizedClinical Study Comparing Laparoscopic and Conventional
Open Surgical Treatments for Colon CancerThe ALCCaS Trial
Peter J. Hewett, FRACS,*† Randall A. Allardyce, PhD,‡ Philip F. Bagshaw, FRCS,‡Christopher M. Frampton, PhD,§ Francis A. Frizelle, FACS,‡ Nicholas A. Rieger, FRACS,*†
J. Shona Smith, MA(Hons),¶ Michael J. Solomon, FRACS,�** Jacqueline H. Stephens, MPH,* andAndrew R. L. Stevenson, FRACS††
Background: Laparoscopy has revolutionized many abdominal sur-gical procedures. Laparoscopic colectomy has become increasinglypopular. The short- and long-term benefits and satisfactory surgicaloncological treatment of colorectal cancer by laparoscopic-assistedresection remain topical. The long-term outcomes of all internationalrandomized controlled trials are still awaited, and short-term out-comes are important in the interim.Methods: Between January 1998 and April 2005, a multicenter,prospective, randomized clinical trial in patients with colon cancerwas conducted. Six hundred and one eligible patients were recruitedby 33 surgeons from 31 Australian and New Zealand centers.Patients were allocated to colectomy by either laparoscopic-assistedsurgery (n � 294) or open surgery (n � 298). Patient demographicsand secondary end-points, such as operative and postoperativecomplications, length of hospital stay, and histopathological data,will be presented in this article. Analysis was by intention-to-treat.Survival will be reported only as the study matures.
Results: Histopathological parameters were similar between the twogroups, except in regard to distal resection margins. There was nostatistically significant difference found in postoperative complica-tions, reoperation rate, or perioperative mortality. Statistically sig-nificant differences in quicker return of gastrointestinal function andshorter hospital stay were identified in favor of laparoscopic-assistedresection. A statistically significant increased rate of infective com-plications was seen in cases converted from laparoscopic-assisted toopen procedures but with no difference in reoperation or in-hospitalmortality.Conclusions: Laparoscopic-assisted colonic resection gives signif-icant improvements in return of gastrointestinal function and lengthof stay, with an increased operative time and no difference in thepostoperative complication rate.
(Ann Surg 2008;248: 728–738)
Laparoscopic-assisted colorectal surgery is being performedwith increasing frequency worldwide. One of the reasons
for this has been the perception of short-term recovery ad-vantages using this modality of surgery. Recent collectivereviews and appraisals of the literature of short-term benefitsfor laparoscopic colorectal resection have indicated reducedblood loss, less intense postoperative pain, improved postop-erative pulmonary function, shorter postoperative ileus,shorter length of stay, reduced wound infection rates, and areduced postoperative ileus rate.1–3
In addition, the popularization of laparoscopic colorec-tal cancer resection has increased on the presumption ofequivalence in regard to oncological safety. The AustralianLaparoscopic Colon Cancer Surgical trial (ALCCaS) is aprospective randomized controlled study comparing open(resection by laparotomy) and laparoscopic resection of coloncancer. The primary aim of this study is to determine equiv-alence in 3- and 5-year disease-free survival. Survival resultswill be published as the study matures and the sample size
From the *Division of Surgery, The Queen Elizabeth Hospital, WoodvilleSouth, Australia; †The University of Adelaide Discipline of Surgery,Adelaide, Australia; ‡Department Of Surgery, University of Otago,Christchurch, New Zealand; §Department of Medicine, University ofOtago, Christchurch, New Zealand; ¶The Academy of Endosurgery,Christchurch, New Zealand; �Department of Colon and Rectal Surgery,Royal Prince Alfred Hospital, Sydney, Australia; **Surgical OutcomesResearch Centre (SOuRCe), University of Sydney, Australia; ††Depart-ment of Colon and Rectal Surgery, Royal Brisbane & Women’s Hospital,Herston, Australia.
This research was supported by the following grants from the NationalHealth and Medical Research Council: NH&MRC ID 207815, NH andMRC ID 349381; Health Research Council of New Zealand: HRC97/154 and HRC 04/102; and by Johnson & Johnson Medical NZ,Ethicon Endo-Surgery Inc. (Sydney, Australia), and the CanterburyMedical Research Foundation.
Correspondence: Mr. Peter Hewett, FRACS, Division of Surgery, The QueenElizabeth Hospital, 28 Woodville Road, Woodville South SA 5011,Australia. E-mail: [email protected].
Annals of Surgery • Volume 248, Number 5, November 2008728
significantly passes the 3-and 5-year periods. Patient demo-graphics, in-hospital outcomes, and histopathological datawill be presented in this paper. Cost-effectiveness and qualityof life data will be published in the near future.
METHODSBetween January 1998 and April 2005, we undertook a
multicenter, prospective, randomized parallel-group trialcomparing laparoscopic-assisted and conventional open co-lectomy in patients with colon cancer at centers withinAustralia and New Zealand.4,5
PatientsPatients were eligible for inclusion if they were aged 18
years or older and had a clinical diagnosis of a singleadenocarcinoma of the ascending, descending, or sigmoidcolon. Exclusion criteria were advanced local disease (tumorsize greater than 8 cm on radiologic imaging); metastaticdisease; rectal or transverse colon cancer; emergency presen-tation; morbid obesity defined as body mass index greaterthan 35 kg/m2; an American Society of Anesthesiologists’physical status classification IV or V;6 associated gastroin-testinal disease that required extensive operative evaluationor intervention; pregnancy; or malignant disease in the past 5years (except superficial squamous or basal cell carcinoma ofthe skin or in situ cervical cancer).
Written informed consent was obtained from all pa-tients. Patient registration and randomization were adminis-tered at a centralized randomization center, located at theAcademy of Endosurgery, University of Otago, Christchurch,New Zealand. Randomization occurred at a 1-to-1 ratio.Surgery was scheduled to be performed within 21 days ofpatient registration. The trial received approval from theregional and hospital-based research ethics committees of allthe participating centers.
ProceduresLaparoscopic-assisted resection involved the mobiliza-
tion of the colon, visualization of critical structures, andintracorporeal vascular ligation. A small abdominal incisionwas required to remove the specimen. Each surgeon per-formed the surgical resection and anastomosis according totheir routine practice. The majority of cases were performedutilizing a lateral to medial approach. In this study, anincision longer or different to that planned was used todetermine a conversion. Conversion to open colectomy wasat the discretion of the individual surgeon for concerns ofpatient safety, technical difficulties, or associated unexpectedconditions requiring treatment by laparotomy. Conversionswere recorded and analyzed as part of the laparoscopic arm ofthe study. Open procedures were performed according to thestandard technique of the operating surgeon. Postoperatively,patients were allowed to have fluid and diet on the firstpostoperative day as tolerated.
Outcome MeasuresThe primary outcomes of the trial were disease-free
survival and overall survival 3 and 5 years after surgery, andthese will be reported at the conclusion of the appropriate
follow-up period. Secondary outcomes were intraoperativeand postoperative complication rates, conversion rate, reop-eration rate, postoperative pain, paralytic ileus, transfusionrequirements, length of stay, return to normal function, ac-tivity restriction, quality-of-life assessment, and 30-day mor-tality. Definitions of these outcome measures are in Table 1.Hospital staff was not blinded to the operative procedure inthe postoperative period. All serious adverse events anddeaths were reported to the Data Monitoring Committee.Adequate pathologic margins are defined as a distance of 5cm or more between the edge of the tumor and the nearestresection margin.
Statistical AnalysisPower calculations indicated that to detect an 11%
difference in 5-year survival between the 2 treatments wouldrequire 600 patients to obtain 80% power (� � 0.05). In thedesign considerations, we assumed that 20% of the patientsrandomized to receive laparoscopic surgery would be con-verted. This was an intention-to-treat, noninferiority trial. Weassumed that 48% of patients receiving “curative” opencolectomy survive 5 years after surgery.7
Descriptive analyses (mean, standard deviation, me-dian, range) were performed, as appropriate. Normally dis-tributed data were analyzed with analysis of variance orStudent t test. Given the highly skewed nature of some of theoutcomes, such as the “time-to-event” data, such parameterswere tested with the Mann-Whitney U test. Proportions werecompared with either the Pearson �2 test or the Fisher exacttest, as appropriate. All P values are for 2-tailed analyses.
TABLE 1. Definitions of Outcome Measures
Conversion Defined as making a larger skin incisionthan was originally planned at thecommencement of the operation
Pyrexia Two or more documented patienttemperatures greater than 38°C thatrequire any treatment intervention orthat result in a prolonged inpatient stay
Prolonged ileus Bowel dysfunction that occurs for greaterthan 6 d following surgery or thatrequires intervention, includingreplacement or placement of nasogastrictube, surgery, etc
Recurrent ileus Bowel dysfunction that occurs in a patientthat had been taking enteral nutrition,but who subsequently requires nil bymouth status
Pulmonary, urinary tract,wound, and abdominalinfection
Need antibiotic treatment and/or drainage
Urinary retention Failure to pass urine after removal ofcatheter and/or the necessity torecatheterize because of failure to passurine
Perioperative hemorrhage Requiring blood product transfusion(s) orreoperation. Correction of preoperativeanaemia is not included as a surgicalcomplication
Medical or anestheticcomplications
Result in patient disability or requireintervention
Annals of Surgery • Volume 248, Number 5, November 2008 Short-term Outcomes of the ALCCaS Trial
Analysis was performed using SPSS Version 13.0 (2006SPSS Inc., Chicago, IL).
RESULTSThere were 601 patients enrolled into the trial with 592
eligible for inclusion in the analysis: 294 were randomized tolaparoscopic-assisted colectomy and 298 were randomized tocolectomy via laparotomy (Fig. 1). These patients were re-cruited from 33 surgeons at 31 participating study sites(84.12% patients from Australian sites, 15.88% patients fromNew Zealand sites). Seventeen surgeons recruited more than10 patients and were responsible for the recruitment of 535(90.37%) patients. There was no statistically significant dif-ference found in the majority of the baseline characteristics ofthe 2 patient populations; however, the patients in the lapa-roscopic arm were significantly older than those in the openarm (Table 2). A 6-month log of all eligible patients deter-mined that 45% of patients presenting with colon cancer wereentered in this randomized controlled trial.8
Procedural ParametersThere was no statistically significant difference found
between treatment groups in the types of surgery performed,with the majority being either right hemicolectomy or ante-rior resection (Table 2). Furthermore, there was no statisti-cally significant difference found in the amount of blood lostor units of blood transfused during surgery. Laparoscopicresections required significantly longer anesthetic and surgi-cal time but had significantly shorter surgical incisions (Table2). As seen in Table 2, the ranges of the surgical incisionlengths between the 2 groups seem to be similar. However,laparoscopic-assisted cases that were converted had a largermedian incision length (17.5cm (range, 10.0–30.0)) com-
pared with those cases completed laparoscopically (6.0cm(range, 3.5–14.0)). The spread of the surgical incision lengthscan be seen in Figure 2 and the overlap of ranges is becauseof those cases which were converted from laparoscopic-assisted to open.
Intraoperative ComplicationsThere was no statistically significant difference found
in the number of adverse anesthetic events (Table 3). Of theadverse anesthetic events that occurred in the laparoscopicarm, 2 were pulmonary and 3 were cardiovascular events. Inthe open arm there were 4 cardiovascular events and 1allergic reaction to an administered antibiotic.
The number of patients who had at least 1 intraopera-tive complication was statistically significantly greater in thelaparoscopic arm: 31 (10.5%) versus 11 (3.7%) (�2
(df � 1) �10.5, P � 0.001). There were significantly more hemorrhagesand minor colonic serosal tears in the laparoscopic arm;however there was no statistically statistical difference foundin other types of intraoperative complications (Table 3). The2 hemorrhages in the open arm were caused by spleniccapsule tears; blood loss was 700 mL and 300 mL. However,neither case required a splenectomy or blood transfusion. Ofthe 10 hemorrhages in the laparoscopic arm, 3 were from thesmall bowel mesentery, 2 from the mesocolon, 3 from thespleen, 1 from the abdominal wall, and 1 from the omentum.A splenectomy was required to control one of the splenichemorrhages. Although there is a difference in the number ofinstances of hemorrhage, there is no difference in the numberof blood transfusions or the required units of blood transfusedbetween the two groups (Table 2). This reflects the frequencyof minor hemorrhage when using laparoscopic instruments inlaparoscopic-assisted colonic resections.
FIGURE 1. Trial profile.
Hewett et al Annals of Surgery • Volume 248, Number 5, November 2008
There were 3 major colonic serosal tears all in thelaparoscopic arm: 2 were completed laparoscopically withresultant 5-day and 10-day length of stays; the third wasconverted (because a second tumor was identified in thetransverse colon). This patient had a postoperative woundinfection that required a reoperation for drainage of an ab-scess and resulted in a 13-day length of stay. The case thatincurred a major small bowel serosal tear was in a patientwhose tumor was adherent to the small bowel and had densesmall bowel adhesions. This case was converted early, 2further enterotomies occurred, a partial cystectomy, and anoophorectomy was required. The patient had no postoperativecomplications and was discharged on day 10.
One duodenal injury occurred when a small amount ofduodenal muscle wall was caught within the stapler; how-ever, because it appeared to be a superficial injury, nointervention was required. This case was not converted, hadno postoperative complications, no reoperations and wasdischarged 6 days after resection. There were no uretericinjuries, gas emboli, pneumothoracies, or episodes of hyper-carbia recorded.
ConversionsForty-three cases (14.6%) were converted from laparo-
scopic to open surgery. Reasons for conversion were inabilityto visualize critical structures (14 cases, 4.8%), advanced
TABLE 2. Patient Demographics and Operative Data (n � 592)
Laparoscopic Open P
No. patients 294 298
Sex (male) 139 (47.3%) 143 (48.0%) 0.863*
Age (y)
Mean � SD 71.1 � 10.4 69.4 � 11.4 0.044†
Range 35.9–94.2 34.3–100.1
Weight (kg)
Mean � SD 71.6 � 15.1 72.7 � 15.2 0.417‡
Height (m)
Mean � SD 167.2 � 10.2 167.0 � 9.3 0.888‡
Body mass index (kg/m2)
Mean � SD 25.8 � 4.5 26.0 � 4.3 0.509‡
ASA grade
I 47 (16.0%) 54 (18.1%) 0.678*
II 164 (55.8%) 163 (54.7%)
III 83 (28.2%) 80 (26.8%)
IV 0 1 (0.3%)
Previous abdominal surgery 133 (45.2%) 136 (45.6%) 0.922*
Procedure performed
Right hemicolectomy 174 (59.2%) 171 (57.4%) 0.095*
Left hemicolectomy 11 (3.7%) 11 (3.7%)
Anterior resection 102 (34.7%) 116 (38.9%)
Anterior resection with covering stoma 4 (1.4%) 0 (0.0%)
Other§ 3 (1.0%) 0 (0.0%)
Anesthetic time (min)
Median (range) 170 (25–400) 115 (20–257) �0.001¶
Surgical time (min)
Median (range) 158 (49–365) 107 (45–250) �0.001¶
Estimated surgical blood loss (mL)
Median (range) 100 (0–1400) 100 (0–700) 0.17§
Blood transfusion
n (%) 16 (5.4%) 18 (6.0%)
Blood transfusions (units)
Median (range) 2 (1–4) 2 (1–3) 0.75¶
Longest incision (cm)
Mean � SD 8.0 � 4.8 20.6 � 6.1
Median (range) 6.5 (3.5–30.0) 20.0 (7.5–34.0) �0.001¶
*Pearson �2 test.†Analysis of variance.‡Student t test.§One transverse colectomy, 1 subtotal colectomy, 1 laparotomy with no resection.¶Mann-Whitney U test.
Annals of Surgery • Volume 248, Number 5, November 2008 Short-term Outcomes of the ALCCaS Trial
disease (12 cases, 4.1%), adhesions (5 cases, 1.7%), inabilityto mobilize the colon (5 cases, 1.7%), inadequate surgicalmargins (1 case, 0.3%), associated disease (1 case, 0.3%), andintraoperative complications (5 cases, 1.7%). The number ofconversions per year peaked in 2001 to 23.7% (Fig. 3) anddid not seem to relate to study year.
Postoperative ComplicationsThe number of patients who had at least 1 postoperative
complications was not statistically significant between the 2groups: 111 (37.8%) for laparoscopic surgery versus 135(45.3%) for open surgery (�2
(df � 1) � 3.5, P � 0.062). Therewere significantly more cases of recurrent postoperative ileusin the open colectomy group; however, there were no statis-tically significant differences in the rates of wound infection,pneumonia, or other types of complications (Table 4). Lapa-
roscopic cases that were converted had a higher rate ofpostoperative pyrexia, pneumonia, urinary tract and woundinfections, and other general complications (Table 4).
ReoperationThere was no statistically significant difference in the
rate of reoperation between the 2 groups: 16 (5.4%) forlaparoscopic surgery versus 13 (4.4%) for open surgery(�2
(df � 1) � 0.4, P � 0.543). There were 4 (1.4%) anasto-motic leaks in the laparoscopic arm compared with 10 (3.4%)in the open surgery arm (�2
(df � 1) � 2.6, P � 0.110).
Perioperative MortalityThere was no statistically significant difference in the
number of patients that died while in the hospital: 4 (1.4%)for laparoscopic surgery versus 2 (0.7%) for open surgery(Fisher exact test, P � 0.448). One death in the laparoscopicarm was due to complications arising from an anastomoticleak. The remaining deaths were due to medical complica-tions related to undergoing surgery: 1 death from pulmonaryembolus and 2 deaths from respiratory failure. The 2 deathsin the open arm were due to cardiovascular incidents.
Postoperative FunctionThere were statistically significant differences in the
time from surgery and the patient taking oral fluids, passingflatus, and having a bowel motion (Table 5). There was a1-day difference in the median length of stay between the 2groups. Laparoscopic cases that were converted to opensurgery had significantly longer length of stay (Table 5).
HistopathologyThe majority of tumors were located in either the
cecum, ascending colon, or sigmoid colon (Table 6). The
FIGURE 2. Surgical incision lengths(cm) for laparoscopic and opencolonic resection.
TABLE 3. Intraoperative Complication Rates
Laparoscopic(n � 294)
Open(n � 298) P*
Adverse anesthetic event 5 (1.7%) 4 (1.3%) 0.751
Hemorrhage 10 (3.4%) 2 (0.7%) 0.020
Minor colonic serosal tear 8 (2.7%) 1 (0.3%) 0.019
Major colonic serosal tear 3 (1.0%) 0 (0%) 0.122
Minor small bowel serosal tear 3 (1.0%) 2 (0.7%) 0.999
Major small bowel serosal tear 1 (0.3%) 1 (0.3%) 1.000
predominant tumor type was adenocarcinoma with the ma-jority being moderately differentiated. There were no signif-icant differences in the T-stage or overall TNM staging;however, there were significant differences in the N-stagingof the 2 patient populations (�2
(df � 4) � 8.2, P � 0.004) withmore tumors of the N2 stage in the open group. The numberof resection specimens with lymphovascular invasion was notsignificantly different between the 2 arms; however, therewere significantly more tumors with perineural invasion inthe open arm (�2
(df � 1) � 6.1, P � 0.013). Both patient
populations had similar occurrences of metastases observedin the specimen.
Resection margins were significantly different betweenthe 2 groups. Resection margins were measured from freshspecimens at operation and/or after fixation with formalin. Inthe laparoscopic arm, 22.8% specimens had fresh measure-ments, 38.8% specimens had fixed measurements, 35.0% hadboth, and 3.4% had no measurements available. Similarly, inthe open arm these percentages were 25.2%, 31.2%, 41.6%,and 2.0%, respectively (�2
(df � 3) � 2.2, P � 0.139). Cases
FIGURE 3. Intraoperative conver-sions (%) by year of study.
where there was no resection margin data were resection afterpolypectomy, where the original polypectomy site was uni-dentifiable. The distal resection margin was greater than 50mm in 77.6% laparoscopic cases and 86.6% of open cases(�2(df � 1) � 8.2, P � 0.004). There was one resection marginin the laparoscopic arm that required the patient to return tothe operating theater for a further resection to ensure adequatemargins were obtained.
Fresh median resection lengths were 280 mm (range,105–600) in the laparoscopic arm and 285 mm (range,130–727) in the open arm (P � 0.016) with median distalresection margins of 90 mm (range, 15–370) for laparoscopiccases and 115 mm (range, 20–537) in open cases (P �0.001). Proximal margins were not significantly different:120 mm (range, 13– 470) versus 120 mm (range, 25–500)(P � 0.769).
Fixed median resection lengths were 200 mm (range,74–855) in the laparoscopic arm and 220 mm (range, 60–630) in the open arm (P � 0.016) with median distal resec-tion margins of 70 mm (range, 8–455) for laparoscopic casesand 90 mm (range, 6–480) in open cases (P � 0.0004).Proximal margins were not significantly different: 85 mm(range, 0–217) versus 90 mm (range, 12–435) (P � 0.684).
There was no statistical difference in the proportion ofcases with a distal margin less than 50 mm between laparo-scopic-assisted and open left-sided resections: 26.9% versus21.3%, respectively (�2
(df � 1) � 1.1, P � 0.301). However,there were a significantly more right-sided resection caseswith a distal margin less than 50 mm in the laparoscopic-assisted group compared with the open group: 19.0% versus7.6%, respectively (�2
(df � 1) � 9.6, P � 0.002).
Left Colon ResectionsThere was no statistical difference between the median
total length of the fresh specimens for laparoscopic and openleft-sided resections: 280 mm (range, 105–600) versus 290mm (range, 160–680), respectively (P � 0.093); or betweenthe fresh distal resection margins: 80 mm (range15-370)versus 90 mm (range, 20–380), respectively (P � 0.355).
Furthermore, there was no statistical difference between lapa-roscopic and open left-sided resections in the median lengthof the fixed resection specimen: 210 mm (range, 105–855)versus 230 mm (range, 130–540), respectively (P � 0.164);or the fixed distal margins: 60 mm (range, 8–310) versus 70mm (range, 6–235), respectively (P � 0.262).
Right Colon ResectionsThere was no statistical difference between the median
total length of the fresh specimens for laparoscopic and openright-sided resections: 265 mm (range, 110–522) versus 280mm (range, 130–727), respectively (P � 0.071). However,there was a statistically significant difference in the mediandistal margins of fresh specimens: 100 mm (range, 15–300)versus 140 mm (range, 30–537) (P � 0.000425). The resultsare similar for the fixed specimens. Median total length forlaparoscopic and open resections were 187.5 mm (range,74–650) and 215 mm (range, 60–630), respectively (P �0.061), while the median distal resection margins were 70mm (range, 15–455) and 108 mm (range, 25–480), respec-tively (P � 0.0012).
DISCUSSIONThis prospective randomized trial shows a quicker
return of gastrointestinal function and shorter hospital staybut no difference in intraoperative parameters, surgical, andpostoperative morbidity when laparoscopic colectomy wascompared with open colectomy.1,9–12 A meta analysis ofpublished trials showed a significantly lower rate of woundinfection/abscess and respiratory complications after mini-mally invasive surgery compared with open operations3 butdifferences in these parameters did not reach statistical sig-nificance in this study.
Reducing the length of postoperative hospital stay is animportant benefit to be derived from laparoscopic-assistedsurgery. For instance, Braga et al13had a median length ofstay of 8 days for laparoscopic-assisted cases and 10 days for
open cases. The Multicentre Randomised Controlled trial-Conventional versus Laparoscopic-Assisted Surgery In pa-tients with Colorectal Cancer (MRC-CLASICC) Trial was 9days and 11 days, respectively;14 however, the Clinical Out-
comes of Surgical Therapy Study group (COST) Trial was5.1 days and 5.6 days, respectively.15 The median stay in thisstudy of 7 days for laparoscopic-assisted procedures and 8days for open procedures reflects this trend.
Postoperative length of stay is a variable indicator withsignificant cultural and local institutional influences. Noelet al3 showed mean stays for minimally invasive surgery andopen colectomy for cancer varied considerably between geo-graphic locations, shorter in North America and longer inEurope and Asia, reflecting cultural and possibly financialreimbursement differences in these areas. There is strongfinancial motivation to shorten postoperative hospital stay inAustralia reflecting the pressures in North America.
The operative procedures performed in this trial weremostly right hemicolectomy (59.2%). The reason this numberis out of proportion with the usual left to right ratio of coloncancers (60%:35%) is the exclusion of rectal cancers fromthis study which skews the site frequency to the right.16
Laparoscopic-assisted procedures took approximately one-third longer in the operating theater compared with colectomydone by laparotomy. This is in keeping with other random-ized controlled studies.14,15,17–20 The mean operating time wasshorter in this study than in the Braga et al13 and MRC-CLASICC14 trials but similar to the Colon carcinoma Laparo-scopic or Open Resection (COLOR)18 trial.
The conversion rate in this study was 14.6%. Ourconversion rate from laparoscopic to open surgery is similarto that recorded in the COLOR trial (17%)18 but less thannoted in the COST (25.4%)15 and MRC-CLASICC (29%)14
Trials. The MRC-CLASICC Trial included cases of rectalcancer and advanced stage cancer in the trial population.Thirty-four percent of patients with cancer of the rectumcompared with 25% of patients with cancer of the colonunderwent conversion from laparoscopic to open proce-dure.14 The ALCCaS Trial did not include patients with rectalcancer. It is interesting to note that in the MRC-CLASICCtrial the rate of intraoperative conversions fell by year ofstudy from 38% in year 1 to 16% in year 6.14 This reductionin conversion rate was not seen in the ALCCaS Trial as theinitial participating surgeons were already experienced. Fur-thermore, any small improvement over time was negated bythe introduction of less experienced surgeons through theduration of the trial. The difference in the conversion ratesbetween the COST and ALCCaS Trials may be because theCOST Trial commenced recruitment in 1994,15 whereas theALCCaS Trial commenced in 1998, thus giving the ALCCaSTrial surgeons the benefit of a further 4 years clinical expe-rience and improvement in laparoscopic instrumentation.
Only 1 trial had a lower rate conversion rate of 5%.13
This trial was a single institution randomized controlled trialcompared with the other studies which were multicenter orinternational trials. Furthermore, that rate compares favorablyto rates of 12% and 23% in nonrandomized studies ofconversion.21,22 Conversion may be of importance in termsof patient’s overall long-term survival as Fleshmanet al23 showed a significant difference in 5 year overallsurvival between converted and open surgery. However, thegroups were not comparable as 25% of the converted grouphad advanced malignancy.23
Casillas et al21 found no difference when comparingconversions from laparoscopic procedures to those completedin operative time, morbidity, length of stay, costs, and read-
mission. Whereas, Gonzales et al22 found a greater bloodloss, longer time to first bowel movement, longer length ofstay when converted cases were compared with the casescompleted with the laparoscopic-assisted approach but nodifference when compared with open surgery. In this study,cases converted were of a longer duration than laparoscopic-assisted or open colectomy and led to higher rates of infectivecomplications and longer postoperative hospital stay, but nodifference in reoperation rate. Fingerhut et al2 rightly pointout that the conversion rate is dependant on the definition ofconversion used. The clinical impact of conversion is depen-dant on whether conversion occurs within 30 minutes into theoperation or after that time.24 This information was notcollected in this study. The impact of conversion on long-term survival outcomes is yet to be assessed but will clearlybe important information.
In the last 2-years of the study, only 17 surgeons (50%)were involved in recruitment and a similar number of sur-geons recruited 90% of the patients. As opposed to otherRCTs, standardization of technique and internal validity wasstronger in this trial with smaller numbers of surgeons andthis is reflected in the lower conversion rates compared withother trials.
Histopathological data reveals similar lymph nodeyields but an increased median length of bowel removed byopen resection with significantly more specimens with resec-tion margins greater than 50mm. Korolija et al25 showed astatistically significant difference in the average distal marginin a meta analysis of 16 comparative series was 4.6cm by thelaparoscopic approach and 5.3 cm with the open approach.Schwenk et al1 looked at 2 trials of 134 cancer patients andfound no difference in resection margins between the 2 studygroups. There was no statistically significant difference in theresection margins for left-sided resections; however, the dis-tal resection margin for right hemicolectomy was signifi-cantly less in the laparoscopic arm. This may imply thattumors of the distal ascending colon and hepatic flexure werenot being adequately mobilized laparoscopically before re-section. Whether this will be clinically significant will bedetermined by long-term follow-up.
Critical analysis of the trial design indicates that thefollowing issues may be significant. Firstly, the length of timeof recruitment occurred between 1997 and 2005. Over thattime a number of technical innovations, such as better qualityoptics, improved atraumatic bowel graspers, newer dissectingand vessel sealing systems, and late in the trial, better qualitywound protectors, have been introduced. All of these couldsubtly alter outcomes.
In this trial the patients, their families, the physicians,and nurses were not blinded to the procedure performed. Thisraises the real possibility that bias in favor of the laparoscopicprocedure may have been introduced.
Like the COLOR, MRC-CLASICC, and COST trials,fast-track surgical principles were not standard of care in thistrial. Impressive reductions in length of stay for open colec-tomy have been demonstrated by Kehlet et al26 and a furtherstudy has shown no difference between length of stay andother parameters in laparoscopy and open resection in a
Hewett et al Annals of Surgery • Volume 248, Number 5, November 2008
blinded randomized trial at Kehlet’s institution.27 In contrast,King et al28 showed decreased postoperative hospital stay forlaparoscopic-assisted cases with fast track recovery tech-niques. A systematic review of available literature on thistopic also confirms this finding.29 A recent study by Wich-mann et al30 suggests that fast track recovery techniques ledto better preserved cell mediated immune function probablydue to decreased preoperative starvation and avoidance ofbowel cleansing.
It is of interest to speculate on the role of laparoscopicsurgery if eventually randomized controlled trials show long-term survival outcomes are equivalent. Will the extra oper-ating time and possibly cost be tolerated if similar short-termoutcomes can be shown to be duplicated by the implementa-tion of fast-track postoperative techniques? It is possible thatover the next few years the 2 philosophies will integratetogether to deliver improved outcomes for patients.31
Recruitment of patients into a surgical randomizedcontrolled trials can be difficult. A study performed withinthis trial has indicated that 55% of patients approached didnot choose to participate in this trial.8 The principle reasonwas the preference of 1 form of surgery by either the surgeonor the patient. Other reasons cited were lack of time on thesurgeon’s part, lack of hospital accreditation, lack of staffing,and also equipment problems. Only a small percentage wereconcerned about the patient-physician relationship or aboutcausing the patient unnecessary anxiety.32 The percentage ofpatients not recruited into the trial has not been indicated byother large randomized trials but was predicted to be 45% ina patient preference trial of laparoscopic colorectal cancer.33
In conclusion, some short-term benefits for laparo-scopic-assisted colorectal resections of earlier return gastro-intestinal function and shorter length of hospital stay havebeen identified. Whether these improvements will be negatedor improved if fast track recovery techniques are imple-mented remains to be seen.
ACKNOWLEDGMENTSThe authors thank Mrs. Vicki Allen, RN, for consider-
able data collection at the Queensland institutions. Theauthors thank the clinical trial coordinators at all other sites.
Contributing SurgeonsMembers of the ALCCaS Trial group who contributed
patients, in order of accrual rate:Nicholas A. Rieger; Andrew R.L. Stevenson; Russell
Stitz; Peter J. Hewett; Eva Juhasz; Frank A. Frizelle; BruceStewart; Adrian Polglase; Ian Martin; John Rutherford;Andrew J. Luck; Mathew Morgan; Paul McMurrick; David J.Rodda; Dayashan Perera; Michael J. Solomon; JustinEvans; Nicolas O’Rourke; David Clark; John Lumley; IanFaragher; Philip Bagshaw; Greg Makin; Julian Hayes; LesNathanson; Nigel Barwood; Paul Sitzler; Alex Karatassas;John McCall; Peter Carne; George Fielding; NimalanPathma-Nathan; Garth Poole.
Contributing InstitutionsAustraliaNew South Wales: Fairfield District Hospital, Prai-
riewood; Manly Hospital, Manly; Mater Hospital, North
Sydney; North Shore Private Hospital, St Leonards; Prince ofWales Private Hospital, Randwick; Royal North Shore Hos-pital and Community Health Service, St. Leonards; RoyalPrince Alfred Hospital, Camperdown; St George Hospital,Kogarah; St George Private Hospital, Kogarah; SydneySouthwest Private Hospital, Liverpool; Queensland: Green-slopes Private Hospital, Greenslopes; Holy Spirit NorthsidePrivate Hospital, Chermside; North West Brisbane PrivateHospital, Everton Park; Royal Brisbane Hospital, Brisbane;The Wesley Hospital, Auchenflower; South Australia: Cal-vary North Adelaide Hospital, North Adelaide; Lyell McEwinHospital, Elizabeth Vale; Royal Adelaide Hospital, Adelaide;The Queen Elizabeth Hospital, Woodville South; WesternHospital, Henley Beach; Victoria: Ballarat Base Hospital,Ballarat; Cabrini Health, Malvern; Monash Medical Centre,Clayton; St John of God Hospital, Ballarat; Western Health,Footscray; Western Australia: Fremantle Hospital, Fremantle.
New ZealandAuckland City Hospital, Auckland; Christchurch Hos-
2. Fingerhut A, Toufik A, Chouillard E, et al. Laparoscopic approach tocolonic cancer: critical appraisal of the literature. Dig Dis. 2007;25:33–43.
3. Noel JK, Fahrbach K, Estok R, et al. Minimally invasive colorectalresection outcomes short-term comparison with open procedures. J AmColl Surg. 2007;204:291–307.
4. Allardyce RA, Bagshaw PB, Frampton CM, et al. Australasian Laparo-scopic Colon Cancer Study (ALCCaS): Organisation And Conduct. ANZJ Surg. In press.
5. Allardyce RA, Bagshaw PB, Frampton CM, et al. Australasian Laparo-scopic Colon Cancer Study (ALCCaS): Perspective. ANZ J Surg. In press.
6. Keats AS. The ASA classification of physical status: a recapitulation.Anesthesiology. 1978;49:233–236.
7. Stewart RJ, Robson RA, Stewart AW, et al. Cancer of the large bowelin a defined population: Canterbury, New Zealand, 1970–1974.Br J Surg. 1979;66:309–314.
8. Abraham NS, Hewett P, Young JM, et al. Non-entry of eligible patientsinto the Australasian Laparoscopic Colon Cancer Study. ANZ J Surg.2006;76:825–829.
9. Danelli G, Berti M, Perotti V, et al. Temperature control and recovery ofbowel function after laparoscopic or laparotomic colorectal surgery inpatients receiving combined epidural/general anesthesia and postopera-tive epidural analgesia. Anesth Analg. 2002;95:467–471.
10. Hasegawa H, Kabeshima Y, Watanabe M, et al. Randomized controlledtrial of laparoscopic versus open colectomy for advanced colorectalcancer. Surg Endosc. 2003;17:636–640.
11. Milsom JW, Bohm B, Hammerhofer KA, et al. A prospective, random-ized trial comparing laparoscopic versus conventional techniques incolorectal cancer surgery: a preliminary report. J Am Coll Surg. 1998;187:46–55.
12. Milsom JW, Hammerhofer KA, Bohm B, et al. Prospective, randomizedtrial comparing laparoscopic vs. conventional surgery for refractoryileocolic Crohn’s disease. Dis Colon Rectum. 2001;44:1–9.
13. Braga M, Vignali A, Zuliani W, et al. Laparoscopic versus opencolorectal surgery: cost-benefit analysis in a single-center randomizedtrial. Ann Surg. 2005;242:890–896.
14. Guillou PJ, Quirke P, Thorpe H, et al. Short-term endpoints of conven-tional versus laparoscopic-assisted surgery in patients with colorectalcancer (MRC CLASICC trial): multicentre, randomised controlled trial.Lancet. 2005;365:1718–1726.
Annals of Surgery • Volume 248, Number 5, November 2008 Short-term Outcomes of the ALCCaS Trial
15. Weeks JC, Nelson H, Gelber S, et al. Short-term quality-of-life out-comes following laparoscopic-assisted colectomy vs open colectomy forcolon cancer. JAMA. 2002;287:321–328.
16. Australian Institute of Health and Welfare (AIHW) and AustralasianAssociation of Cancer Registries (AACR) 2003. Cancer in Australia2000. Canberra: AIHW AIHW Cat. No. CAN 18 Cancer Series no. 23.
17. COLOR Study Group. COLOR: a randomized clinical trial comparinglaparoscopic and open resection for colon cancer. Dig Surg 2000;17:617–622.
18. Veldkamp R, Kuhry E, Hop WC, et al. COLOR Study Group,. Laparo-scopic surgery versus open surgery for colon cancer: short-term out-comes of a randomised trial. Lancet Oncol. 2005;6:477–484.
19. Lacy AM, Garcia-Valdecasas JC, Delgado S, et al. Laparoscopy-assistedcolectomy versus open colectomy for treatment of non-metastatic coloncancer: a randomised trial. Lancet. 2002;359:2224–2229.
20. Lacy AM, Garcia-Valdecasas JC, Pique JM, et al. Short-term outcomeanalysis of a randomized study comparing laparoscopic vs open colec-tomy for colon cancer. Surg Endosc. 1995;9:1101–1105.
21. Casillas S, Delaney CP, Senagore AJ, et al. Does conversion of alaparoscopic colectomy adversely affect patient outcome? Dis ColonRectum. 2004;47:1680–1685.
22. Gonzalez R, Smith CD, Mason E, et al. Consequences of conversion inlaparoscopic colorectal surgery. Dis Colon Rectum. 2006;49:197–204.
23. Fleshman J, Sargent DJ, Green E, et al. Laparoscopic colectomy forcancer is not inferior to open surgery based on 5-year data from theCOST study group trial. Ann Surg. 2007;246:655–664.
24. Belizon A, Sardinha CT, Sher ME. Converted laparoscopic colectomy:what are the consequences? Surg Endosc. 2006;20:947–951.
25. Korolija D, Tadic S, Simic D. Extent of oncological resection inlaparoscopic vs. open colorectal surgery for colorectal cancer: meta-analysis. Langenbechs Arch Surg. 2003;387:366–371.
26. Kehlet H, Wilmore DW. Fast-track surgery. Br J Surg. 2005;92:3–4.27. Basse L, Jakobsen DH, Bardram L, et al. Functional recovery after open
versus laparoscopic colonic resection: a randomized, blinded study. AnnSurg. 2005;241:416–423.
28. King PM, Blazeby JM, Ewings P, et al. Randomized clinical trialcomparing laparoscopic and open surgery for colorectal cancer within anenhanced recovery programme. Br J Surg. 2006;93:300–308.
29. Wind J, Polle SW, Fung Kon Jin PH, et al. Systematic review ofenhanced recovery programmes in colonic surgery. Br J Surg. 2006;93:800–809.
30. Wichmann MW, Eben R, Angele MK, et al. Fast-track rehabilitation inelective colorectal surgery patients: a prospective clinical and immuno-logical single-centre study. ANZ J Surg. 2007;77:502–507.
31. Kumar A, Hewett PJ. Fast-track or laparoscopic colorectal surgery? ANZJ Surg. 2007;77:517–518.
32. Taylor KM. Physicians’ reasons for not entering eligible patients in arandomized clinical trial of surgery for breast cancer. N Engl J Med.1984;310:1363–1367.
33. Solomon MJ, Pager CK, Young JM, et al. Patient entry into randomizedcontrolled trials of colorectal cancer treatment: factors influencing par-ticipation. Surgery. 2003;133:608–613.
Hewett et al Annals of Surgery • Volume 248, Number 5, November 2008
