Customize your JAMA Network experience by selecting one or more topics from the list below.
Bilimoria KY, Bentrem DJ, Nelson H, et al. Use and Outcomes of Laparoscopic-Assisted Colectomy for Cancer in the United States. Arch Surg. 2008;143(9):832–840. doi:10.1001/archsurg.143.9.832
Michael J. Stamos, MD, Orange, California: This a very important study due to its large size and because it gives us a snapshot of clinical care in the United States during the time period 1998 to 2002. It is quite interesting to note that this was during the era when there was essentially a moratorium on laparoscopic colectomy for curative cancer outside of clinical trials. Obviously, this was not a mandated moratorium, but it certainly may explain the low rates of utilization of this technique and the relative slow advance, at least during this time period. It also may help explain the lower lymph node harvest rate. I notice that you had a higher number of stage I patients in the laparoscopic arm compared with the open arm. Some of these patients may have been going to operation for polyps that turned out to be early cancers. Although many papers have been published showing the importance of treating a polyp like a cancer when you operate on it, in fact that does not always happen, and it may or may not explain some of the lower lymph node harvest rates. So I would be curious if you have any data to suggest that that may or may not be true.
On a similar note, does this database allow any kind of evaluation of gross data of utilization during the periods from 2002 until now? Do you have any preliminary data to suggest that the 5.2% rate has gone up since that time and where it has gone to?
It is also interesting to note that your highest quintile group of hospitals performed only 9 cases per year. That is not per surgeon; that is per hospital! In fact, these 34 hospitals, with this relatively modest annual volume, accounted for 20% of the total volume in the United States despite the fact that they made up only a little under 3% of the overall number of hospitals reporting to the NCDB. I think this is important to keep in mind, as what we define as a high-volume hospital is obviously a very relative term.
You also noted that patients undergoing laparoscopic colectomy at lowest volume hospitals were 2-fold less likely to have more than 12 nodes resected and examined compared with patients at the highest-volume centers. Did you look at these lowest-volume hospitals and highest-volume hospitals to see whether they were also low and high volume for open colectomy; in other words, were they just low-volume hospitals period, or were they just low volume for laparoscopic compared with open? Further, did you look at the lymph node harvest numbers for their open cases? In other words, what I am trying to say is, is this a pathologist issue or is this a surgeon issue?
The reason I ask that question is that your lymph node harvest rate did not seem to correlate with your observed patient outcome or survival. Others have found similar results so this may just be the confounding nature of lymph node evaluation or the small differences absolutely between the 2 groups.
Perhaps the most important and controversial finding in this study is the improved outcome or cancer-related survival in the laparoscopic cases compared with the open cases. Again, as you pointed out, the COST trial did not show this and neither did the COLOR trial nor the CLASICC [Conventional vs Laparoscopic Surgery in Colorectal Cancer] trial. Antonio Lacey's single-institution trial out of Barcelona, published in 2002, did show improved cancer-related survival, although it was largely confined to stage III patients in contrast to yours where it is mostly stage I and stage II patients. There was also a recent meta-analysis combining the COLOR, the COST, and the CLASICC trials that found no difference in cancer-related outcomes. It is certainly possible that the reason your study found a better outcome was because of its large size compared with these randomized trials, and it may be that these other trials were just simply underpowered and it was not their end point. The other possibility of course is patient selection, as you pointed out, or surgeon selection; that is, maybe the most experienced and most qualified surgeons are doing the cases laparoscopically and perhaps they are [[ldquo]]cherry picking[[rdquo]] the most favorable cases. I would appreciate your thoughts and comments. In particular I would like your opinion as to how we could really answer this question definitively, how big of a study that would take.
Dr Russell: I think before I try to answer some of Dr Stamos' questions, I would like to point out the power of these databases. I think it is really important as surgeons that we have some ability to collect and control data. I think that ACS NSQIP [American College of Surgeons National Surgical Quality Improvement Program] provides good data for examining short-term outcomes, and the NCDB is good for assessing long-term oncologic outcomes.
With respect to some of Dr Stamos' questions, obviously the utilization rate of laparoscopic colectomy is still low, probably a reflection of the moratorium that you mentioned and the complexity of the procedure. Recent studies have shown that, even in 2006, utilization of laparoscopic colectomy remains in the 6% to 8% range. Cancer registries in the United States stopped collecting the laparoscopic vs open field in 2003, so we are going to try to change this so detailed data can be collected regarding the surgical approach. The volume thresholds were defined a priori based on quintiles with equal numbers of patients in each category. This allows for good discrimination between the groups while providing enough statistical power to make reasonable inferences.
If the operation is being performed, it is being performed for a reason. It should be a standard cancer operation each time. An appropriate number of lymph nodes should be resected and examined any time a colectomy is done for cancer. The low node counts are somewhat concerning, and they are related to the hospital case volume. This volume[[ndash]]node count association was seen for both open and laparoscopic cases. We think that lymph node examination needs to continue to be improved. The low node count may be related to the pathologists' interest and diligence. But, nevertheless, this will become more often used, and I think payment may be linked to the count.
There are certainly selection factors influencing the results. Our analysis showed that a more favorable population in terms of patient and tumor characteristics was undergoing laparoscopic colectomy. We attempted to adjust for these, but this may be incomplete.
Financial Disclosure: None reported.
Laparoscopic-assisted colectomy (LAC) has gained acceptance for the treatment of colon cancer. However, long-term outcomes of LAC have not been examined at the national level outside of experienced centers.
To compare use and outcomes of LAC and open colectomy (OC).
Retrospective cohort study.
National Cancer Data Base.
Patients who underwent LAC (n = 11 038) and OC (n = 231 381) for nonmetastatic colon cancer (1998-2002).
Main Outcome Measures
Regression methods were used to assess use and outcomes of LAC compared with OC.
Laparoscopic-assisted colectomy use increased from 3.8% in 1998 to 5.2% in 2002 (P < .001). Patients were significantly more likely to undergo LAC if they were younger than 75 years, had private insurance, lived in higher-income areas, had stage I cancer, had descending and/or sigmoid cancers, or were treated at National Cancer Institute–designated hospitals. Compared with those undergoing OC, patents undergoing LAC had 12 or more nodes examined less frequently (P < .001), similar perioperative mortality and recurrence rates, and higher 5-year survival rates (64.1% vs 58.5%, P < .001). After adjusting for patient, tumor, treatment, and hospital factors, 5-year survival was significantly better after LAC compared with OC for stage I and II but not for stage III cancer. Highest-volume centers had comparable short- and long-term LAC outcomes compared with lowest-volume hospitals, except highest-volume centers had significantly higher lymph node counts (median, 12 vs 8 nodes; P < .001).
Laparoscopic-assisted colectomy and OC outcomes are generally comparable in the population. However, survival was better after an LAC than after an OC in select patients.
Create a personal account or sign in to: