Portal de Periódicos da CAPES

THE AUSTRALASIAN LAPARAOSCOPIC COLON CANCER STUDY

2008; Wiley; Volume: 78; Issue: 10 Linguagem: Inglês

10.1111/j.1445-2197.2008.04673.x

1445-2197

Randall A. Allardyce, Philip Bagshaw, Christopher Frampton, Francis A. Frizelle, Peter J. Hewett, Nicholas Rieger, Shona Smith, Michael J. Solomon, Andrew R. L. Stevenson,

Cardiac, Anesthesia and Surgical Outcomes

The Australasian Laparoscopic Colon Cancer Study (ALCCaS) is a non-inferiority randomized controlled trial (RCT) comparing laparoscopic and open surgical treatments of right-sided and left-sided potentially curable colon cancer. It is not often that large and complex multicentre surgical trials are jointly undertaken between Australia and New Zealand and ALCCaS methodological coherence with other major international trials adds an unprecedented analytical dimension. This paper emphasizes some of the implications and issues that transcend the nuts, bolts and results of ALCCaS. Similar international studies (e.g., Clinical Outcomes of Surgical Therapy (COST), Colon Cancer Laparoscopic or Open Resection, Conventional versus Laparoscopic-Assisted Surgery in Patients with Colorectal Cancer (CLASICC) and ALCCaS) set against different economic and cultural backgrounds achieve clinical consensus by reducing both acknowledged and unrecognized regional variables. These related studies also allow economic, social and clinical variables to be identified and their contributions to outcomes assessed. In addition, close alignment of methods and outcomes between important RCT increases the evidence base and statistical power to assess combined data that can be stratified beyond the primary and secondary clinical outcomes and quality of life assessments of any single trial. The original accrual target of 1260 patients proved unattainable within a practical time frame. An adjustment to 600, to detect a mortality difference of 11% at 5 years, was approved by the Health Research Council of New Zealand-appointed international Data Monitoring Committee (DMC) and was attained in March 2005 in the expectation of combining ALCCaS with other RCT outcome data. Patient recruitment to surgical trials is a problem that is not confined to ALCCaS. Both the COST and CLASICC trials were required to reduce their accrual targets to provide timely operative and teaching guidance.1,2 The chief factors limiting accrual were the number of surgeons enrolled and their patient recruitment to ALCCaS. This was influenced by the stringent credentialing criteria, surgeon and/or patient preference for a procedure,3 increased laparoscopic colonic resection (LCR) time and reluctance to randomize private patients. There were also delays in gaining ethics approval in individual centres that could have been improved by centralization of the process. Surgeon credentialing, as specified in the COST protocol, was based on an audited videotape of an LCR and 20 operative reports describing oncologically appropriate laparoscopic procedures. It must be said that, at the outset of the COST study (from which ALCCaS derived its protocols) the criteria on which qualified participation and learning curve prediction were based may have been arbitrary and unproved. Whatever criteria are appropriate may not be a simple matter to decide, but arguably what was done for ALCCaS seemed to work. Procedure credentialing and the achievement of short-term results similar to other important RCT raise several points. First, the willingness to undergo close peer scrutiny illustrates the integrity of surgeons and their desire to achieve excellence. Some may also think that it emphasizes the potential for the future adoption of video evidence in vocational examinations and staff appointments. Moreover, the amalgamation of coherent international RCT trial data provides not only information about new therapies, but also gives an historical perspective on well-defined, conventional therapies and outcomes for comparison by surgeons in the future. With regard for that perspective the question might be put: if all the compatible RCT show equivalence or non-inferiority for survival and recurrence and the benefits of minimally invasive surgery are modest and largely subjective, what is the ‘gold standard’? Is it LCR or open colon resection (OCR)? Are the level 1 data to be considered together or separately? In practice, future open operations might be judged against OCR results, whereas new minimally invasive techniques would be judged against LCR with two separate, but statistically equivalent, ‘gold standard’ constructs because the results of each new treatment could be expected to influence both short-term and long-term outcomes differently. The overall results attained in this RCT may prove to be better than ‘real life’ because of credentialing, patient selection, treatment standardization and possibly better care of patients included in high-quality clinical research.4 It is not possible to quantify the individual or combined effects of these factors. It is also difficult to assess the impact of changing practice. For example, the length of stay following open operations may be more influenced by experience of laparoscopic patients and fast-track recovery technique changes than by the method of removal of the cancer.5,6 Acknowledgement, however, that a higher standard is attainable may be required for the future advancement of safe laparoscopic surgery as well as the adoption of other new procedures. Although a voluntary moratorium on LCR for malignant disease outside ALCCaS was recommended,7,8 it was not the case in practice. There was a surge in LCR before and less so since RCT short-term data papers were released. Are RCT needed to introduce new procedures? This probably depends on the type of procedure and information to be gained. Learning new techniques, operations and the use of equipment are technical objectives (e.g., harmonic scalpel) and could be taught traditionally or using a variety of virtual, remote or ‘hand-in-hand’ robotic formats. Biological outcome assessments and comparisons are different and require the full suite comprising procedure standardization, operator competence, unbiased data collection and the analysis of actual rather than surrogate end-points that are encompassed in the RCT trial model.8 The present process for clinical introduction of new operations involves an initial audit of the outcomes and issues surrounding a procedure, so that surgeons know what the questions are and whether they can be answered. Once a procedure has been developed to a point where it has reproducible outcomes, an RCT can help define the way ahead and a temporary moratorium constraining surgical practice can provide a means of mitigating surgical enthusiasm for new techniques with unexpected, undesirable outcomes.9 However, the point at which surgeons believe they have enough conclusive information from short-term or long-term outcome publications from international trials that are published asynchronously remains undetermined and unaddressed. Many and varied institutional infrastructural elements are required to support important multicentre surgical trials in the face of regulatory and ethical issues as well as necessary performance standards. The institutions need structures and so do the participating surgeons. The critical elements are listed in Table 1. It is important to protect the integrity of trial data by coding it and blinding the participating clinicians and data collators to the accumulating results. This practice reduces the influence of observer bias on the recorded data and protects the interests of the participating patients by achieving the most accurate conclusions possible. However, once patient accrual is complete, it has been argued that observer bias can have no significant influence on mortality and tumour recurrence. Also, failure to report short-term, clinically relevant data could prejudice the usefulness of the study to inform practical clinical decision-making and teaching. Supporting that view, it is noted that many surgeons consider short-term results to be definitive end-points in their own right. The publication of short-term data is, however, capable of misleading clinicians and the public to reach unwarranted conclusions regarding the principal long-term outcomes.10,11 Recognizing this to be an important issue, ALCCaS participants formed a publications committee and agreed that publications reporting short-term outcomes will formally reflect this caution and that short-term ALCCaS results would be disseminated following an independent audit of the accuracy and completeness of the recorded data. It became apparent during discussions between ALCCaS surgeons and the DMC regarding RCT conduct that some of the guidelines intended to guard the integrity of medical treatment and drug intervention trials might be of limited applicability to surgical practice. Regardless of the specific arguments relating to the publication of short-term and long-term data, a comprehensive reconsideration of surgical trial data management, publication and trial integrity is called for.11