|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 25, No 21 (July 20), 2007: pp. 2996-2998 © 2007 American Society of Clinical Oncology. DOI: 10.1200/JCO.2007.11.3241
Laparoscopic and Open Surgery for Colorectal Cancer: Reaching Equipoise?Department of General Oncologic Surgery, City of Hope National Medical Center, Duarte, CA This editorial will serve as both a technical review for Jayne et al's article published in this issue on the long-term results of the United Kingdom Medical Research Council (UK MRC) Conventional Versus Laparoscopic-Assisted Surgery in Colorectal Cancer (CLASICC) trial1 and a discussion of evaluations of new technologies. Studies of laparoscopic compared with open surgical techniques are paradigms for the numerous situations we as oncologists (medical, surgical, and radiation) must analyze. The subtleties are often lost in the panorama and panache of study results. Furthermore, the biases of the particular interest groups are expressed with the support of the data, with the same data used to spin results in opposite directions. In preparing this editorial, I challenged myself, on behalf of the readership, to analyze the results of this and other similar trials on open compared with laparoscopic-assisted large bowel cancer surgery as the devil's advocate and his/her adversary. The 3-year results of a randomized trial designed to compare laparoscopic-assisted with open resections for large bowel cancer are reported in this issue. This multi-institution, UK MRC–supported trial report adds to the oncologic, technical, and quality of life (QoL) outcome data on this subject. Two recent meta-analyses collect information from the Barcelona,2 Clinical Outcomes of Surgical Therapy (COST),3 and Colon Cancer Laparoscopic or Open Resection (COLOR) trials,4 and the first reporting of the CLASICC trial,5 and provide the reader not familiar with the topic an excellent synopsis of the structure and comparability and an overall perspective on the questions at hand.6,7 The initial short-term reporting of the CLASICC trial was an intent-to-treat report that documented an equal number of patients with Duke's C2 (stages IIIB and IIIC) and final pathologic diagnoses. The hospital mortality was 5%. Of note, patients requiring conversion from laparoscopic assisted to open had more complications, and those with laparoscopic low anterior resections for rectal cancer had an increased number of involved resection margins. The current report expands on the early results of the 794 patients from the CLASICC trial, with an extended follow-up on survival, disease-free survival, surgical morbidity, and QoL end points. This was a study performed with care and clarity, structured with precision, and reported within the confines of the data. Despite equivalent survival and local control, two findings are worrisome: an increase in port site/incision local recurrence that was nine times greater in the laparoscopic group (nine in the laparoscopic and one in the open), and a positive margin rate for laparoscopic rectal resections that was twice that for open. These findings may be a result of technical issues that are the burden of the laparoscopic approach. Whether they can be overcome with greater care in specimen delivery and rectal resection remains an important element in the consideration of the nonequivalence of the procedures. It is of considerable interest that the patient-reported QoL scores were nearly equal in both groups. The long-term variation that was seen in the laparoscopic group was a prolonged or persistent negative impact on financial status, physical functioning, and social functioning. This may focus us on improving the immediate variables for the open resection patient and recognizing that leaving the hospital sooner and requiring fewer pain medications while in the hospital may have their own opportunities for improvement. Many readers may skip the details and quickly read the conclusion that laparoscopic-assisted bowel surgery is equivalent to open surgery for the cancer patient. However, a more detailed reading will expose some of the confounding factors that will be important in deciding whether the results are unique to this study, or can be generalized. Let us examine the following on the reported outcomes: patient selection, site selection, surgeon selection, market forces, and investigator bias and the effect of a nonblinded trial. PATIENT SELECTION An inherent challenge in any randomized, prospective clinical trial is patient accrual. The patient is screened with criteria that had been established during trial development. In addition, there is an undocumented patient selection bias introduced by patient, physician, and health care organization. Clearly, between July 1996 and July 2002 the number of eligible patients far exceeded the 794 randomly assigned. This is not unique to this trial, but we must accept that this is not a randomly selected subset of the general population of patients newly diagnosed with colon or rectal cancer. Patients involved in this trial must be interested in the laparoscopic approach, be willing not to have it (ie, willing to be randomly assigned), and have manageable access to a surgeon and research site. Their decision to participate may be influenced by their economic status (inclusion in a trial may provide cost-free care and the possibility of the desired laparoscopic procedure), access to public or private transportation, or personal impressions of the clinical trial process. There is excellent documentation of the issues of human experimentation as it influences decision making in clinical trials. SITE SELECTION Consider this with regard to access: this situation is exemplified in the Barcelona trial,2 in which access is maximally limited to a single center (geographically situated in the northeast corner of Spain). This may be contrasted to a more liberal access to 66 credentialed surgeons at 48 institutions, albeit spread across the United States, in the COST trial. The sites qualified to put patients on trial must have the technology, instrumentation, visual tools, and operative nursing and anesthesia teams to carry out the procedures. This is coupled with adequate operating room time to accommodate the longer surgical procedures. Thus, to be a laparoscopic site also requires a multidimensional commitment at the level of the health care delivery resource. SURGEON SELECTION There are two processes that actively impact the selection of participating surgeons. The first is the inherent surgical skill of the participating surgeon in both the laparoscopic and open techniques. It is interesting that none of the prospective, randomized trials comparing laparoscopic with open resections required a systematic evaluation of the surgeon's open skills. With regard to laparoscopic skills, the intraoperative conversion rate may be used as a surrogate for surgeon skill with laparoscopy. In the CLASICC trial, there was clearly an acquisition of laparoscopic skills reflected in conversion rate reduction from 38% (year 1 in 1996) to 16% (year 6 in 2002). The high number of converted operations and increase in complications in the converted patients suggest that the open method would have been preferred as an initial approach. This opens the discussion for patient entry criteria (weight, preoperative imaging, and location of primary) as a key element for deciding between the open or laparoscopic approach. The lack of difference in local recurrence and survival suggests the correctness of conversion in achieving the primary oncologic end points. This global measurement of surgeon skill with the procedure is reflected in a similar manner in the 21% conversion rate in the COST trial,3 17% in the multicenter (29 hospitals) COLOR trial,4 and 11% in the Barcelona trial. Also, it might be reasoned that the surgeons who participated were, in general, either skilled proponents of laparoscopic surgery or willing and able to include laparoscopy in their practices if the study outcome was interpreted favorably for laparoscopy. Pure open surgeons or those without motivation to incorporate this new procedure would never be interested or qualify to participate in the trials. Thus, there may be a bias against skilled open surgeons enrolled onto the trials. MARKET FORCES Although not addressed in the scientific comparison of the laparoscopic and open procedure, the influence of the health care marketplace is an integral component of the adoption of new technology. In locations where the health services are centrally controlled, the trial results may have significant impact, but their incorporation into the therapeutic armamentarium would be presumably independent of third parties, such as manufacturers and consumer groups. However, in a competitive marketplace, the results of the trial and the reported data will be interpreted for the consumer. Direct-to-consumer marketing to achieve a competitive advantage for the centers offering laparoscopic colectomy is, therefore, a legitimate concern. Understanding how the data will be used highlights the importance of stating, rather than overstating, the conclusions. The enthusiasm to generalize should be controlled, as care is taken to elucidate the absolute data-based findings and the methods of reaching the studies’ conclusions. INVESTIGATOR BIAS Finally, we must consider the impact of investigator bias on the outcome. This would be most evident in the surgeon-controlled variables (eg, initiation of oral intake, need for pain medications, or day of discharge). Consider the situation where the surgeon must decide when to feed. The decision to initiate oral feeding after gastrointestinal surgery is steeped in the customs of surgery. These patterns of waiting for bowel sounds or patient hunger were initially challenged when reduction in length of stay in acute care was an important end point. Many studies have documented that the time of feeding is more based on the training of the surgeons than the ability of a patient's GI tract to tolerate food. In the initial report of the CLASICC trial, patients had equivalent time to first bowel movement and resuming of normal diet, yet hospital length of stay was 2 days longer for open surgeries. This may not be an issue of medical necessity, but rather a surrogate marker toward a more aggressive discharge posture for patients who have undergone laparoscopy. This is potentially a point where surgeon or health care system bias may play a role in the outcomes. More subtle measurements would be patient selection for trial entry and conversion rates. Other trials have attempted to blind the investigators or at least remove them from the evaluation phase of the trials (such as the recording of complications and completing QoL survey form). This would clearly aid in reducing the risk of investigator and system bias. So what do CLASICC and other prospectively, randomized clinical trials comparing laparoscopic to open colectomy tell us? (1) The surgeon must be experienced in laparoscopic surgery to offer the option. (2) Even with a reasonable laparoscopic experience (more than 30 to 40 cases), the chance, in general, of not having a laparoscopic resection is approximately 20%. (3) The skills necessary to complete a laparoscopic-assisted rectal cancer resection may be different than a colon cancer resection. (4) From the patient's perspective, on a global basis, there is little difference in the short term (2-week) or long-term (3-year) QoL outcomes. (5) As a patient, it is not necessary to find a surgeon who performs laparoscopic resections as either open or laparoscopic assisted are nearly equivalent. (6) The current standard of care has not changed from open to laparoscopic assisted/laparoscopic treatment of large bowel cancers (ie, open is not now below the standard). Perhaps this trial should be evaluated in a different manner. Given the oncologic equivalence (disease-free survival and overall survival) and near equivalence of QoL and morbidity, is there a need to incorporate laparoscopic resection into oncologic practice? Surgeons who prefer the open approach can take comfort in knowing that the open technique has been validated. Those surgeons with the skill and necessary experience in the laparoscopic approach can perform the cancer surgery with similar outcomes. This sounds like a classic win-win-win (open surgeon-laparoscopic surgeon-patient) environment. AUTHOR'S DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST The author(s) indicated no potential conflicts of interest. REFERENCES
1. Jayne DG, Guillou PJ, Thorpe H, et al: Randomized trial of laparoscopic-assisted resection of colorectal carcinoma: 3-year results of the UK MRC CLASICC trial Group. J Clin Oncol 25:3061-3068, 2007 2. Lacy AM, Garcia-Valdecasas JC, Delgado S, et al: Laparoscopy-assisted colectomy versus open colectomy for treatment of non-metastatic colon cancer: A randomised trial. Lancet 359:2224-2229, 2002[CrossRef][Medline] 3. The Clinical Outcomes of Surgical Therapy Study Group: A comparison of laparoscopically assisted and open colectomy for colon cancer. N Engl J Med 350:2050-2059, 2004 4. Veldkamp R, Kuhry E, Hop WC, et al: Laparoscopic surgery versus open surgery for colon cancer: Short-term outcomes of a randomised trial. Lancet Oncol 6:477-484, 2005[CrossRef][Medline] 5. Guillou PJ, Quirke P, Thorpe H, et al: Short-term endpoints of conventional versus laparoscopic-assisted surgery in patients with colorectal cancer (MRC CLASICC trial): Multicentre, randomised controlled trial. Lancet 365:1718-1726, 2005[CrossRef][Medline] 6. Bonjer HJ, Hop WC, Nelson H, et al: Laparoscopically assisted vs open colectomy for colon cancer: A meta-analysis. Arch Surg 142:298-303, 2007 7. Jackson TD, Kaplan GG, Arena G, et al: Laparoscopic versus open resection for colorectal cancer: A metaanalysis of oncologic outcomes. J Am Coll Surg 204:439-446, 2007[CrossRef][Medline]
Related Article
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2007 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
