In colorectal cancer, a lymph node debate

May 2007
Feature Story

Anne Paxton

The classic conundrums of quality assurance never really change: If you can’t measure performance, how do you know you have met high standards? But are precise benchmarks always warranted when assessing “quality”? And should fingers be pointed when the benchmarks aren’t met?

In this perennial debate, the question of how many lymph nodes pathologists should retrieve in evaluating colorectal cancer patients has become a flashpoint. There is consensus on the main issue, but not on the details. Says Lawrence Burgart, MD, staff pathologist at Abbott Northwestern Hospital and professor of pathology at the University of Minnesota College of Medicine, Minneapolis: “It’s simple and complex at the same time.”

“There are a lot of papers in the literature comparing patient outcomes with the number of lymph nodes retrieved, and most report that whether the nodes are positive or negative, the more nodes you find, the better overall the patient outcome. Even for stages II or III it holds true,” says Kay Washington, MD, PhD, professor of pathology and director of surgical pathology at Vanderbilt University Medical Center, Nashville.

Part of this effect is due to stage migration, she notes. “The more lymph nodes you find, the better the chances of identifying a positive node, and that patient will be correctly staged as stage III. But even if you weren’t dealing with stage III patients, the more lymph nodes you find, the better those patients do.”

Variations among patients do have an impact on the ease of finding lymph nodes and patient outcomes, Dr. Washington says. “But the number found is undoubtedly due, in part, to the diligence of the pathologist looking for them.”

Lymph nodes range in size from one centimeter down to two or three millimeters, and the task of retrieving them can be difficult because they are hidden in the fat around the colon and rectum, and the specimen can range from 10 cm up to a whole colon.

“They’re little white or tan nodules embedded in fat. You look for them visually or else palpate the fat. So if there’s a lot of fat, as in obese patients, they’re just harder to find, because you’re looking for something in a larger haystack,” Dr. Washington says.

“So it’s a daunting task when the nodes are small. But it’s been shown over and over again that even nodes less than 5 millimeters in colon cancer can have metastases, so it’s very important to find the small ones.”

An experienced pathology assistant can perform a lymph node retrieval in 15 minutes or so, depending on the amount of fat in the specimen, but a neophyte may easily take more than an hour, Dr. Washington says.

The College does not have official guidelines on lymph node retrieval, but it does have a consensus statement (Compton CC, et al. Arch Pathol Lab Med. 2000;124:979–994), and the CAP cancer checklist for colon cancer includes the following comment on the number of lymph nodes: “It has been shown that 12 to 15 negative lymph nodes predict for regional node negativity; therefore, if fewer than 12 are found then additional enhancement techniques should be considered.”

“There is, I would say, a wide variation in practices, and it’s a challenge to educate colleagues on these recommendations,” Dr. Washington says.

But among surgeons and pathologists there is no unanimity about the wisdom of the number 12. Nancy N. Baxter, MD, PhD, assistant professor of colorectal and general surgery at St. Michael’s Hospital at the University of Toronto, is one of those who has raised questions.

“It’s absolutely demonstrated that we’re not achieving what people are saying is the optimal number of lymph nodes,” she says. “People assume that means if we find more nodes, patients will do better, but this is unknown.” She believes the CAP standards are well meaning and probably good, but not based on hard evidence.

In one of her own studies, reported in 2005, Dr. Baxter found that only 37 percent of colorectal cancer patients received adequate lymph node evaluation (Baxter NN, et al. J Natl Cancer Inst. 2005;97:219–235). That number seems to be improving, but, she says, “I’m not aware of any country in the world on a population basis that’s meeting the CAP standard.”

A very recent review of 17 studies on the issue reported that all but one of them demonstrated improved survival as the number of lymph nodes evaluated increased in patients with stage II colon cancer (Chang GJ, et al. J Natl Cancer Inst. 2007;99:433–441). In addition, four of six studies reported a positive association between lymph node number and survival among patients with stage III colon cancer. “The authors conclude that given the evidence, lymph node evaluation deserves consideration as a quality measure for colon cancer,” says Dr. Baxter, who co-wrote an editorial that accompanied the review (J Natl Cancer Inst. 2007;99:414–415). “But the whole problem is we have no idea if taking or finding an increased number of lymph nodes will improve patient outcomes.

“It’s very likely,” she contends, “that the number of lymph nodes reflects an interaction between the tumor and the host—i.e. how much the host person is reacting to their tumor. You can increase the number you find all you want, but you’re not going to change that relationship. You may be spending a lot of resources in money and time to improve something that isn’t going to affect patient outcomes.”

Raouf E. Nakhleh, MD, professor of pathology at the Mayo Clinic, Jacksonville, Fla., agrees there is no formula to define how many lymph nodes should be in one location in the body. “We know anatomically everyone has a heart and two lungs, but there is no research saying you must have X number of lymph nodes. With disease conditions lymph nodes come and go, and there’s no defined number.” He reports going back as many as three times to look at a colon specimen and, in some situations, still can’t get to 12 nodes.

Dr. Nakhleh contrasts the colon cancer lymph node retrieval guidelines with practices in breast cancer. “In the past, surgeons would do a standard operation with axillary lymph node dissection for breast cancer and we examined a much higher number of lymph nodes, but there was higher morbidity associated with that procedure.

“Today, in the majority of situations, a sentinel node is examined at frozen section. If the sentinel node is positive, then more lymph nodes are taken. In the breast we’re looking at fewer lymph nodes. With breast cancer, the absolute number of lymph nodes examined does not seem to matter. That type of examination is not done with colon cancer,” and the number is important. But “the number 12 is not magical; it’s arbitrary,” he notes. “There’s no anatomic basis to say there should be 12.”

Some pathology practices that use certified pathology assistants to perform grossing report having few difficulties in finding 12 nodes, Dr. Washington says. “In general, certified PAs are very experienced at gross examination of specimens, and it’s easy to give them feedback about the adequacy of their dissection.” But the same is not true of academic centers.

“When you’re teaching residents, you have a moving target. Every month there’s someone new in there, and it takes diligence on the part of the faculty. If a new resident comes in and finds three nodes and the faculty doesn’t tell them to do a better one, the case may get signed out and you don’t learn about it until the specimen is destroyed and your chance to retrieve more is gone. So you almost have to audit their reports, and sit down with the residents to make sure they are finding a sufficient number of lymph nodes.”

Medical schools do not really address the nitty-gritty of pathology examination of surgical specimens like this, Dr. Washington says. “In our lectures to medical students we talk about the staging system and indicate positive nodes would be stage III. We really don’t talk about stage migration.”

There are some cases, pathologists report, where lymph node retrieval is difficult because the mesocolon is too short. “That is related to surgical technique if they are not taking much of the mesentery,” Dr. Washington says, stressing that this is another example of how lymph node retrieval is not only pathologist-dependent but also patient- and surgeon-dependent.

“Our surgeons get very worked up if we don’t find enough nodes, and they’ll call us,” she says. “We have a multidisciplinary tumor board and we don’t like to go before that board if we found fewer than 12 nodes because it’s embarrassing, obviously. So I think our interactions with the surgeons drive us a little harder.”

Often, the second look may include the use of clearing solutions. These are solvents that basically dissolve the fat out of the specimen and make the nodes easier to find. “That’s the upside, but the downside is the added expense and disposal of the solvents,” Dr. Washington notes. “Some private practices use them routinely and love them. We try to teach the residents to find lymph nodes without them, but I certainly would not fault someone for using them. If they help you, that’s great.”

Simply keeping the pathology staff informed about their retrieval counts has proved effective, Dr. Washington has found. “We’ve done a batch of sequential reports on colon cancer resection and given the faculty and residents basically ‘report cards’ on how they’re doing. That’s improved our counts, and 85 percent or more of our cases are now reporting 12 or more lymph nodes. A feedback mechanism like this is just part of an effective quality assurance program.”

The guidelines on lymph node retrieval were novel and controversial when CAP introduced them in 1999, Dr. Burgart says. “But in terms of the patients’ prognosis correlating with the number of lymph nodes recovered from colorectal specimen collection, the data has become incontrovertible.”

However, he emphasizes, lymph node retrieval is multifactorial. “It’s more complex than just the thoroughness of pathological dissection of specimens.” Probably a majority of dissections are performed by residents or pathology assistants, so there will be variability and you cannot generalize, he notes. “But it is absolutely true—it is the pathologist who is responsible for the dissection, regardless of who makes it.”

At Abbott Northwestern Hospital, the guideline is to retrieve all the lymph nodes. “If the yield is low, if it’s less than 12 to 15, we then use a clearing agent and re-dissect it. To tell the truth, it doesn’t increase our yield, because we’re very thorough at dissecting the first time; we just do it to say we’ve done everything we possibly can,” Dr. Burgart says.

The resections themselves are relatively standard and do not show a lot of variability. However, all colon resections are not the same, he notes. “Resection of the right side is much different from the sigmoid and certainly from rectal, so it varies depending on where the cancer is located in the colon. As you get more distal, sigmoid, and rectal, resections vary more. There are also biologic factors in terms of tumor features and the immune system of the patient which are poorly understood.”

At least one study, conducted in Sweden, goes so far as to suggest that the quality of a pathology department can be measured by the number of lymph nodes retrieved (Jestin P, et al. Eur J Cancer. 2005;41:2071–2078). But Dr. Burgart, who is immediate past chair of the CAP’s Surgical Pathology Committee, disagrees with that premise.

“It’s one factor. There is U.S. data showing the high-volume centers have a higher node yield than low-volume centers, indicating experience matters, and the data indicates the volume of the surgeon and hospital correlates with patient outcome. But the quality of the dissection is only one of several factors.”

Connecting pathology department quality to staging accuracy is a real stretch, Dr. Nakhleh agrees. “To say pathology quality is related to or defined by one measure such as the median number of lymph nodes retrieved—that’s an overstatement. There are so many factors that determine quality including accuracy, timeliness, report completion, and your understanding of what your customer needs.”

Dr. Burgart emphasizes that the 12-to-15 guideline represents only breakpoints. “The recommendation is not to find this many nodes, but absolutely to find all the nodes possible. The more nodes, the better the prognosis. And it’s very important not to stop at 12 to 15.”

“But I think there is a general misunderstanding that there are always going to be 15 or more and it’s just a matter of finding them. That is clearly not the case,” Dr. Burgart says. “Some resection specimens, again because of patient immune system factors or tumor factors or maybe limited resection due to patient anatomy, just will not have that many.” Almost all rectal cancers now get neoadjuvant therapy before surgery, he adds, which also affects the number of lymph nodes.

In the 20 months that Luca Stocchi, MD, has been a staff surgeon at the Cleveland Clinic in Ohio, only one case has not met the clinic’s standards of 12 lymph nodes—and in fact he has sent congratulatory e-mails to the pathology staff praising their thoroughness.

“In my opinion,” Dr. Stocchi says, “this is an issue where it’s necessary to have collaboration between the surgeon and the pathologist. The surgeon needs to be very precise and take enough specimen, but the pathologist also needs to be very precise. It’s a shared responsibility if there is not a sufficient harvest.”

An insufficient number of nodes makes it more difficult to establish whether someone needs chemotherapy, Dr. Stocchi notes. “If you have six that are negative, you don’t know if the patient really has node-negative tumor because there is data suggesting this might lead to understaging.” However, just how many nodes would be sufficient is a matter of controversy.

“In the literature, some papers suggest 12 would not be enough and propose other numbers like 14 or 15, and at least one paper suggests the more the better. So it is an open-ended question,” Dr. Stocchi says. At the Cleveland Clinic, “there has not been a substantial difference in survival for patients with negative lymph nodes if we harvested more than 12, but there was certainly a difference if we harvested less.” But that is the experience of just one center, he adds, “and it’s still controversial.”

Dr. Stocchi emphasizes the importance of increased awareness of lymph node retrieval recommendations. “The number of nodes retrieved can drastically change the prognosis of the patient, and it can eliminate the patient from the possibility of receiving beneficial chemotherapy.”

There is an added wrinkle to take into account when considering lymph node retrieval, says Perry Shen, MD, associate professor of surgery at Wake Forest University Baptist Medical Center, in the surgical oncology service. “Many health plans are mandating sampling numbers, or they are aware of the number of lymph nodes being obtained. So this is something that will affect reimbursement as well as patients’ health. If you are documenting in pathology reports that you are getting so many lymph nodes, you may get special preference as a referral source; there are some health plans that are looking for that.”

It’s important to keep in mind, he agrees, that both the surgeon and the pathologist are responsible. “Whatever institution is doing this kind of surgery—whether it’s smaller or larger, private or academic—there needs to be a dialogue between surgeons and pathologists to work together toward the same goal.”

In the U.S., the pay-for-performance paradigm is the way things are heading, Dr. Baxter agrees—but the trend is not necessarily a good one in her view. “For example, if you achieve a certain proportion of patients who have their antibiotics on time or their DVT prophylaxis on time, you’ll get paid more. Now they’re talking about setting lymph nodes as a quality benchmark.”

“In a hospital where the surgical pathology team routinely finds 18 nodes but finds only five nodes in one patient, it probably has something to do with the patient, not the team. Are you going to pay these people less for only finding five nodes in a patient? That doesn’t make sense; there are probably only five nodes there.”

“We as a community can make substantial improvements in lymph node evaluation and staging, but I think there’s a limit beyond which we’re not going to improve care, and I don’t think pushing everyone to achieve some kind of benchmark will lead to the kind of improvement people are projecting,” Dr. Baxter says.

The CAP and other professional organizations have a tremendously important role in sifting through scientific and political factors to make practice recommendations to their members, Dr. Burgart says. “I think the cancer protocols from the College’s Cancer Committee are an excellent example of that. The commentary in that cancer protocol about this issue, and the references with that commentary as well, give pathology practice guidelines and credibility with surgeons and the specialties to work together for quality improvement.”

“This is a great example of professional organizations working together to cull through scientific and political data, make recommendations, and put together educational pieces to help medical centers follow those recommendations over time.”

“The studies have shown there’s a correlation between nodes retrieved and patient outcomes,” Dr. Burgart adds. “But I think some people have taken the misconception from that that if they don’t meet a certain number, the pathologist is automatically doing a bad job. That’s not always the case. It’s a good time to talk about it, but there are multiple explanations for a low lymph node count in a given specimen.”