Lymph Node Dissection

Free consultation

Send your pathology report and we will read the node count with you

Send the full pathology report including the synoptic summary, the operative note if you have it, every scan report with the images themselves, and any earlier biopsy result. A surgeon and a pathologist here read the report together and tell you how many nodes were examined, whether that number meets the standard for your cancer, whether the stage on the page is trustworthy, and what a second opinion would change. No fee, no obligation, and a coordinator replies in your own language, usually within the same working day.

Not one operation

Lymph node dissection covers a family of operations that share a name and very little else. Three of them get grouped together on consent forms, and the differences between them determine everything that follows.

The number to look for

Different cancers have different published thresholds, and they were arrived at in different ways, some by looking for the point where survival curves separate in large registries and some by expert agreement. None of them is a law. All of them are worth knowing before you open your own report.

Wide table. Swipe or drag it sideways on a narrow screen, because it scrolls instead of shrinking.

Why the number moves

Four things push the count up or down, and only one of them is about you.

What the count changes

Your count is not a scoreboard. It feeds directly into three decisions that shape the rest of your treatment.

After the specimen leaves

What the laboratory can do

Several methods exist to make nodes easier to find, and they have been studied properly enough to say what they do and what they do not do.

• Blue dye pushed into the specimen's artery. In a randomized study of 60 colon specimens, the stained group yielded an average of 35 nodes against 17 in the unstained group, which the authors describe as a highly significant difference, and insufficient harvests occurred in eight unstained cases against one stained case. When the leftover fat was cleared and searched a second time, upstaging happened only in the unstained group, meaning the first pass had missed positive nodes that the dye would have shown.
• A written protocol, even without dye. One department compared its historical practice against a defined handling protocol and then against the protocol plus dye. Average yields were 20.6, then 37.1, then 47.6. Under the old practice the twelve node minimum was missed in 15.9 percent of patients and an eighteen node target in 44.6 percent, while every single patient in both newer groups exceeded eighteen. That comparison was against a historical series rather than a randomized control, which the authors state plainly, and the size of the jump is still hard to explain away. A doubling of the average count without a single change to the operation being performed is the most striking illustration in this literature of how much of the number is decided after the specimen leaves the surgeon's hands, and it is worth holding in mind whenever a count is presented as a measure of surgical quality. Read it as one department's experience.
• Fat clearing solutions. These dissolve the fat chemically so nodes stand out, and they raise the count reliably, though a rectal cancer study comparing cleared and conventionally fixed specimens concluded that the extra nodes retrieved were not associated with any survival benefit and that the method is too slow to be worth doing routinely. The comparison was not randomized and several figures in its published summary do not reconcile, so the conclusion is worth more here than any individual number in it.
• Dye again, in ordinary hospitals. A multicenter series from 2024 found the dye significantly improved node identification and significantly reduced the number of cases falling under twelve nodes, while the number of cancer containing nodes found was unchanged, and that last part is the honest half of the finding, since it recurs across this literature.
• Where all of them stop. Every one of these methods finds more nodes. None has been shown to find more cancer in a way that changed what happened to patients afterward. Hold on to that distinction, because the whole argument of this page sits inside it.

The uncomfortable finding

The usual explanation for why higher node counts go with better survival is that finding more nodes finds more cancer, which moves people into a higher stage where they get treated more aggressively, and two large American studies tested that explanation directly and it did not hold up.

Across 86,394 colon cancer patients between 1988 and 2008, the proportion with at least twelve nodes examined rose from 34.6 percent to 73.6 percent. The proportion found to have positive nodes went from 40 percent to 42 percent, a difference that was not statistically significant. Twenty years of counting harder did not move the stage distribution.

Researchers went at it from the hospital end in a second study, ranking hospitals by how often they examined twelve nodes or more and following 30,625 patients. Hospitals in the top quarter and the bottom quarter were equally likely to find a node positive tumor, gave chemotherapy at almost identical rates of 26 percent and 25 percent, and the adjusted hazard ratio for death comparing the highest to the lowest quarter was 0.95 with a confidence interval of 0.88 to 1.03. Hospitals with the best node counts tended to treat lower risk patients and to do far more operations, which is a more parsimonious explanation for their better raw results than the counting itself. Its authors concluded that efforts by payers and professional bodies to push node examination rates up may have limited value as a public health intervention, and named the twelve node standard specifically, which is a strong claim, and it comes from the people who tested it.

Both studies are observational, and neither says the count is meaningless to you personally, and what they say is that the count works better as a marker of who was operated on than as a lever anyone can pull.

D1 and D2 in the stomach

Stomach cancer has the longest running argument about how wide the field should be, and it is worth following because it shows how a question gets settled over thirty years rather than in one trial.

Another wide one. Swipe or drag it sideways on a narrow screen, because it scrolls instead of shrinking.

Taken together these trials did not conclude that D2 is wrong. They concluded that D2 done without removing the spleen and pancreas, in hands that do it often, gives fewer cancer deaths and fewer local recurrences at acceptable risk, and that everything beyond D2 has been tested and has failed. The British trial's own authors wrote that their negative result could not rule out D2 without pancreas and spleen removal being better than D1, and that sentence deserves to travel with every citation of that trial.

What a wider field costs

Every extra node has a price measured in operating time, blood, and specific complications that a narrower operation does not produce, and the figures set out below come from randomized trials and from pooled analyses of those trials.

Wide again. Drag the table sideways on a narrow screen, because it scrolls instead of shrinking.

Reading your own report

Pathology reports are written for other doctors and they use a compact vocabulary. Here is what the node section is telling you.

This table is wide too. Swipe it sideways on a narrow screen, because it scrolls instead of shrinking.

The colon question

Where the number twelve came from

Seventeen studies covering 61,371 patients were pulled together in a systematic review published in 2007, and sixteen of the seventeen found that examining more nodes went with better survival in stage two disease. That is a strong and consistent signal. It is also a signal built entirely from observational data, and the thresholds proposed across those studies ranged from six nodes to forty, so twelve was chosen as a workable point inside that range rather than discovered as a biological boundary, and the review's authors said so.

Whether a wider colon operation helps

Chinese investigators randomized 1,072 patients with right sided colon cancer between a wider mesenteric dissection and standard surgery. At three years the hazard ratio for disease free survival was 0.74 with a confidence interval of 0.54 to 1.02, and at five years overall survival gave a hazard ratio of 0.74 with a confidence interval of 0.51 to 1.07, neither of which reached statistical significance. The trial's conclusion was that standard dissection should remain routine and the wider operation should be reserved for patients with obvious node involvement. Its five year report did find larger effects inside subgroups, with node positive patients at the higher category showing a hazard ratio of 0.25, and subgroup findings of that kind generate hypotheses rather than settle them. The trial is also a good illustration of how a result gets used, since a hazard ratio of 0.74 with an interval that just crosses one will be read by some surgeons as a signal worth acting on and by others as a null finding, and both readings are defensible from the same numbers.

What the extra plane actually contains

Russian investigators reported on the first 100 patients randomized between the standard and the wider field, and found mean node yields of 26.6 and 27.8, which is almost no difference at all. Three patients in the wider group had cancer in the extra tier of nodes, and in none of them was that tier the only place cancer was found. Complications at thirty days were 47 and 48 percent with no deaths in either arm. That report is an interim safety analysis of a trial that was still recruiting and it contains no survival data, so it settles nothing about whether the wider operation helps. What it does show is how little the extra territory contained. Three patients out of roughly fifty in the wider arm had disease in that extra tier, and every one of them had disease elsewhere as well, which means that on this evidence the wider dissection changed the pathology report for nobody and changed the treatment plan for nobody either.

The ratio beats the count

Why a proportion behaves better than a total

If two people each have three positive nodes, but one of them had eight nodes examined and the other had forty, those are not the same situation, and staging systems built on the absolute number of positive nodes treat them identically. The proportion of examined nodes that were positive corrects for that, and it does so automatically, without anyone needing to know how hard the laboratory looked, which is why the ratio keeps appearing as an independent predictor in studies that also measure the raw count.

What it looks like in numbers

In 134 patients who had a Whipple operation for pancreatic cancer, five year survival was 21.7 percent where fewer than fifteen percent of examined nodes were positive and 5.2 percent where more than fifteen percent were, and the ratio remained an independent predictor after adjustment while several other variables did not. In a registry analysis of 36,712 node positive colon cancer patients the ratio was independently associated with survival, with hazard ratios running from 1.24 to 5.12. A melanoma series of 189 patients having groin dissections found the ratio was the strongest predictor of outcome on multivariate analysis, ahead of whether the nodal disease had been visible or microscopic. None of those three studies was randomized and none of them set out to test the ratio against the count in a head to head way, so the pattern is best described as a consistent observation across different cancers and different research groups rather than as a settled measurement.

Why it has not replaced the count

A ratio still needs a denominator you can trust. If four nodes were examined and one was positive, the ratio of twenty five percent is arithmetically identical to ten positive out of forty, and it is nowhere near as informative, and the authors of the colon registry analysis made exactly this point, arguing that their own findings still imply a minimum node count is necessary for accurate staging. So the ratio is a better summary of a good report, and it does not rescue a thin one.

What the letter D promises

How often the assigned operation was performed

Dutch investigators ran a formal quality control program alongside their gastric trial, comparing what came out of each patient against the anatomical stations the protocol required, and then published the results of that audit. Major non-compliance, meaning stations that should have been cleared and were not, occurred in 15.3 percent of the D1 group and 25.9 percent of the D2 group, while contamination, meaning nodes turning up from outside the assigned territory, occurred in 22.9 and 23.5 percent. Those numbers come from a nationwide trial with Japanese supervision, dedicated coordinators and an explicit protocol, which is about as controlled as surgery ever gets. If a quarter of the patients assigned to the wider operation did not fully receive it under those conditions, the reasonable inference for an ordinary hospital operating outside a trial is that the gap between the name of an operation and its execution is at least as large there, and probably larger. Nobody audits it, so nobody knows.

Whether tightening the supervision fixed it

Tightening the supervision did not fix it. The trial's own report states that intensifying quality control produced only a marginal improvement in protocol adherence and in the number of nodes detected, and it found no association between how closely a surgeon followed the protocol and complications, hospital deaths, or long term survival. Read that carefully, because it cuts both ways. Deviating from the template did not visibly harm anyone in this trial, which undermines the idea that precise field boundaries are what matter, and it equally undermines any confidence that a label on a consent form describes what actually happened in the operating room. Both of those conclusions are uncomfortable for the way surgical quality is usually discussed, and the trial published them anyway, which is the main reason this particular audit is worth more than the many quality programs that never reported whether they worked. That is unusual, and it is why the finding survives.

A footnote about how these trials were designed

British investigators sized their gastric trial on a survey of twenty six surgeons who expected the wider operation to improve five year survival by around nine or ten percentage points. A later analysis by the trial's own statisticians recalculated what would have been needed and put it at 800 to 1,200 patients rather than the 400 recruited. That paper is a methodological survey rather than a trial, though it is indexed confusingly. Its point stands regardless, which is that a famous negative result rested on an expert consensus about effect size that turned out to be wrong. A trial that recruits half the patients it needed will usually report a difference that fails to reach significance whether or not a real difference exists, and once that report is published it tends to be cited for decades as though the question had been answered. The distinction between underpowered and negative gets lost quickly.

Where the count misleads

When the tumor decides the number

Node yield rises with younger age, with certain tumor locations, with higher grade and with deeper invasion, all independently of anything the surgical team does. That is why nearly three quarters of the variation in meeting the twelve node standard traced back to tumor factors in the largest analysis of the question. A low count in an older patient with a small left sided tumor is a different thing from a low count in a young patient with a bulky right sided one, and treating both as evidence of poor surgery would be wrong in one of the two cases. This is the strongest single reason to resist reading your own report as a report card on your surgeon, because the factors that put your count where it is were mostly fixed before anyone picked up an instrument, and no amount of care in the operating room would have changed them. The count describes the specimen more than it describes the surgery.

When the operation makes the count meaningless

In melanoma, one groin template yields a median of eleven nodes and a wider one yields a median of 21.5, on the same anatomy in the same patients at the same institution, where ninety percent of the narrower operations produced at least eight nodes and none of the wider operations produced ten or fewer. Comparing a count across two people who had different templates tells you which operation each of them had. It tells you almost nothing about how thoroughly either was done.

When the count has been tested and did not predict anything

One single center series of 54 patients having their voice box removed found no association between node yield and either survival or disease free survival, and found that a higher yield did not raise the probability of the stage changing in either direction. Seven patients were moved up a stage by the dissection and eight were moved down, and that series is small enough that a real effect could easily have been missed, so it does not overturn the larger head and neck analysis showing worse survival below eighteen nodes. It is a reminder that the relationship is a tendency across populations rather than a rule that holds in every clinic. Both findings can be true at once, since a threshold derived from 572 patients pooled across two randomized trials describes what happens on average across populations, while a series of 54 patients at one hospital describes what one team saw and is far too small to detect a moderate effect.

Before you agree

Most of what determines your node count is decided before the operation starts, in the choice of template and in the laboratory protocol the hospital happens to use. Both are legitimate subjects for a conversation and neither is usually raised.

Ask which named field is planned. Not whether the glands will be checked, which every surgeon says yes to, but which anatomical territory is coming out and what the expected node yield is for that territory in that surgeon's hands. A team that audits its own numbers will answer immediately. A team that has never looked will say it depends, and that answer is informative too.

Then ask what the pathology department does with the specimen, because that is where a large part of your number is decided and it is a question almost nobody asks. Whether the laboratory uses a written protocol for node retrieval, whether it uses a dye or a clearing method, and how long it takes to report are all things the surgical team knows and can tell you. Nothing about this conversation is confrontational. It is closer to asking a builder what materials are going into a wall.

One more question is worth putting plainly. If the count comes back low, ask what happens next, because in some teams the answer is a second look at the specimen by the laboratory, in others it is a discussion about chemotherapy that would not otherwise have happened, and in some it is nothing at all. Knowing which of those three you are dealing with beforehand is far better than discovering it afterward.

Where wider was worse

Pancreatic surgery is the clearest case of a wider dissection producing a worse result rather than a neutral one, and it is worth stating because the pattern elsewhere is mostly neutrality.

Pooling five randomized trials and 724 patients, extending the dissection during a Whipple operation added about an hour of operating time and raised postoperative complications, with no difference in median survival overall or in either the node positive or node negative group. Where the extended operation stripped the nerve plexus around the artery, the odds of persistent diarrhea rose more than twelvefold.

Cochrane reviewers covering seven trials and 843 patients reached a compatible position from a different angle. The wider operation retrieved about eleven more nodes, added roughly fifty minutes and 137 milliliters of blood, and produced no detectable survival difference, with the evidence for survival rated as very low quality, and its authors chose their words carefully and wrote that current evidence neither supports nor refutes the effect of the extended operation. That is a more honest summary than either side of the argument usually offers, and it is the position a patient should be given. Very low quality evidence is a formal grading rather than a criticism of the researchers, and it means the true effect could plausibly sit some distance from the pooled estimate in either direction, which is exactly the situation where a patient's own preferences about time in the operating room and long term bowel function should carry real weight.

Eleven extra nodes for an hour of surgery, a unit of blood in some patients, and a twelvefold rise in one specific lasting side effect. Written out that way, the trade is easy to evaluate.

Questions to take with you

Take these to any consultation, before or after your operation. They are ordered so the first four work in a pre-operative conversation and the rest work once a report exists.

1. Which named field or template are you planning, and what does that territory include?
2. Roughly how many nodes does that operation usually yield in your hands?
3. Does the pathology department here use a written protocol for finding nodes, and does it use a dye or a clearing method?
4. Tell me what happens next should the count come back below the standard for my cancer.
5. How many nodes were examined in my report, and does that meet the published figure for this cancer?
6. Were the nodes submitted grouped by anatomical station, or all together?
7. Work out the proportion of my examined nodes that were positive, and tell me what that ratio means here.
8. Does my report mention growth through the capsule of a node, and does that change the plan?
9. Would a second pathologist reading the same slides be likely to reach a different node count?
10. Would you accept the staging on this report from another hospital, or repeat any of the work?

What to send us

Second opinions about a node count are a document exercise. Send these and it can be done properly.

• Your complete pathology report. Send the report itself and everything the summary letter your doctor wrote leaves out, including the synoptic table at the end if there is one, since that is where the node fraction, the stations and the extranodal findings are recorded.
• Your operative note. This says which territory the surgeon intended to clear, which is the only way to tell whether a low count reflects a narrow operation or a thorough one that found few nodes. Reports and operative notes disagree more often than you would expect.
• Every scan report with its images. Send the image files themselves. A radiologist here will want to look at the nodal areas rather than read somebody else's description of them.
• Any earlier biopsy, including needle samples from a node. A node sampled before surgery changes how the final report should be read, and it is frequently left out of the file people send.
• Slides or blocks, if you can get them released. This takes longer to arrange and it is the only route to an independent recount. Ask your hospital what its release procedure is at the same time as you send everything else.

Coming to Istanbul

If you are traveling for the operation itself, the length of stay is set by the organ involved and by whether a drain is needed, while the node dissection itself adds little. An abdominal cancer operation with a formal nodal clearance usually means seven to ten days in Istanbul, an operation on the neck or the breast rather less, and anything involving the groin longer, because wound problems there are common and are far easier to manage before a flight than after one.

Timing of the pathology report is the part specific to this page. A full node count takes days rather than hours, and where a dye or a clearing protocol is used it takes longer still. Booking a return flight that lands before the report arrives creates pressure at the worst possible moment, so leave the date flexible or buy a changeable ticket. The conversation about what the report means is the reason you came, and it is worth being in the room for.

Follow up after you return home is arranged before you leave, with a named contact here, copies of every document your own oncologist will need, and a review of the pathology once the final report is issued. If you are traveling only for a second opinion on an operation already done elsewhere, nothing needs booking at all until the documents have been read. Send them first.

After the report arrives

Lymph node dissection FAQ