Lymph Node Removal Surgery

Free consultation

Ask how many nodes are actually planned, and why

Send your biopsy report with the exact cancer type and grade, every scan report together with the images themselves, any report describing the lymph nodes on ultrasound or on a PET scan, the results of any needle sample taken from a node, and a note of any treatment you have already had including chemotherapy and radiotherapy. A surgeon reviews the file and tells you which nodes would be sampled, which would be left alone, and what the current trial evidence says about that choice in your particular cancer. No fee, no obligation, and a coordinator replies in your own language, usually within the same working day.

The direction of travel

Why nodes were taken out in the first place

Lymph nodes sit along the drainage channels that carry fluid out of every tissue in the body, and cancer cells that leave a tumor tend to arrive in them before they arrive anywhere else. For most of the twentieth century that observation supported a simple plan, which was to remove the tumor and then remove the whole group of nodes it drained into, on the reasoning that any cancer already sitting in them would otherwise be left behind to spread. Whole operations were built on that logic. Clearing the armpit in breast cancer, clearing the neck in mouth cancer, clearing the pelvis in cancer of the womb lining and clearing the groin in cancer of the vulva were all standard, and were all done to almost everyone with the relevant diagnosis. Selection barely came into it. The plan was coherent and it was never really tested, because for decades nobody thought it needed testing.

What the testing found

Then it was tested, cancer by cancer, in randomized trials that assigned people either to the full clearance or to something smaller. The results have been remarkably consistent. Finding and removing more nodes changes the paperwork, because it moves people into higher stages and produces information nobody had before, and in trial after trial it has not changed how long they live. What it reliably does change is the drainage of the limb below the operation, which is why the argument has shifted away from cancer control and toward a straightforward trade of information against lasting swelling.

This does not mean node surgery is finished. It means the question is now specific instead of automatic.

One trial in this collection found the opposite, and it is described further down instead of buried, because a page that only reports the results pointing one way is not much use to someone deciding what to have done.

Melanoma went first

Melanoma is where the modern approach was invented and where it was then dismantled, and the first large trial, which randomized 2,001 patients, compared taking a sample of the first draining node against watching the area and operating only if something came up. It did not show an overall survival benefit. Within the group whose melanoma was of intermediate thickness, ten year disease free survival was 71.3 percent with the biopsy against 64.7 percent with observation, and the trial established that knowing the node status early was worth something. The second trial asked the harder question, which was what to do once that first node came back positive.

Among 1,934 patients with a positive sentinel node, three year melanoma specific survival was 86 percent whether the rest of the nodes were removed or not. Removing them gave better control of disease in the nodes themselves, at 92 percent against 77 percent. Lymphedema affected 24.1 percent of those who had the clearance and 6.3 percent of those who did not.

A European trial had reached the same conclusion by a different route, randomizing 483 patients with a positive sentinel node and reporting three year survival free of distant spread of 77.0 percent with observation against 74.9 percent with clearance. Its final analysis at a median of six years put five year survival free of distant spread at 67.6 percent against 64.9 percent, with a hazard ratio of 1.08, and severe complications in 13 percent of the surgical arm, twenty of whom had lymphedema. The authors were candid that their trial did not accumulate enough events to be conclusive on its own, yet read together with the larger American trial the message is not ambiguous, and completion clearance after a positive sentinel node in melanoma has largely stopped. What replaced it is close surveillance of the node area with ultrasound, which finds the small number of people who do go on to need surgery there without committing everyone else to an operation they were never going to benefit from.

The four breast trials

Breast cancer has the densest evidence, and four trials between them have moved the standard three separate steps down. The first randomized 891 women with a small tumor and one or two positive sentinel nodes, all of whom were having a lumpectomy followed by radiotherapy to the whole breast, either to clearance of the armpit or to nothing further. At ten years, overall survival was 86.3 percent in the group who had no clearance and 83.6 percent in the group who did, and between years five and ten there was a single recurrence in the nodes across both arms combined.

The second trial took a different approach to the same problem by treating the armpit with radiotherapy instead of surgery, and across 4,806 eligible women, 1,425 had a positive sentinel node and were assigned to surgical clearance or to radiation. Recurrence in the armpit was rare either way, at four events among 744 after surgery and seven among 681 after radiation, and at ten years the cumulative figures were 0.93 percent against 1.82 percent. Overall survival did not differ. The reason this trial matters is what it reported about the arm rather than the cancer, because lymphedema at the updated five year analysis affected 24.5 percent of the surgical group and 11.9 percent of the irradiated group. Trading one treatment for another is not the same as having no treatment at all, and the honest way to read that trial is that radiation to the armpit does the same job to the cancer while asking a different price of the arm.

A third trial then asked whether clearance could be dropped in women having a mastectomy and in those with larger tumors, populations the earlier trial had excluded. It enrolled 2,766 people across five countries, almost ninety percent of whom received radiotherapy aimed at the nodes, and at a median follow up of just under four years five year survival free of recurrence was 89.7 percent without clearance and 88.7 percent with it. Its authors are careful to note that overall survival is the trial's real endpoint and has not yet been reported, and that the result applies to a group who mostly had their nodes irradiated. The fourth trial went further still and asked whether the sentinel node needs sampling at all when the tumor is small and the ultrasound of the armpit is clean. Among 1,405 women with tumors up to two centimeters, five year survival free of distant disease was 97.7 percent with the sampling and 98.0 percent without it, and the conclusion was hedged in a way worth repeating, which is that skipping it is reasonable only where the missing pathology would not have changed the treatment given afterward.

Four trials, three separate reductions, and the same finding each time. It is the clearest run of evidence in the whole field.

Where more surgery helped

One trial in this collection points firmly the other way, and it involves mouth cancer, so ignoring it would give a false impression of a field that is moving in one direction everywhere, which it is not.

What a sentinel node is

The idea behind it

Fluid leaving a tumor does not spread out evenly in all directions. It follows particular channels to particular nodes, and the first node on that path is the one that would receive any escaping cancer cells first. If a tracer is injected near the tumor it travels the same route, which makes that first node findable during surgery, so a surgeon can take it out, examine it thoroughly, and let its condition stand in for the condition of the whole group. That single idea is what made it possible to stop removing dozens of nodes from people who turn out not to need it, and it is the technical foundation under every trial described on this page. None of that reasoning holds if the tracer never reaches a node, which is why detection rates matter as much as accuracy, and why the section that follows gives them tracer by tracer instead of as a single number.

How well the tracers work

Two tracers are used, sometimes together. One is a weakly radioactive liquid detected with a handheld probe, the other is a dye that stains the node visibly, either blue or fluorescent green under a special camera. In the largest breast trial, sentinel nodes were successfully removed in 97.2 percent of 5,536 patients, rising to 98.9 percent when a hot spot could be found before the incision. A Cochrane review of 33 studies in cancer of the womb lining put the average detection rate at 86.9 percent, with detection on both sides of the pelvis in 65.4 percent, and found no difference in accuracy between the tracer combinations once the studies were compared directly. Blue dye alone performed worst on detection, at 77.8 percent, and a randomized trial in breast cancer after chemotherapy compared adding fluorescent dye to the radioactive tracer against the tracer on its own and found identification rates of 98.3 percent and 93.8 percent, a difference that did not reach statistical significance in 122 patients.

Cancer by cancer

The table gathers the randomized comparisons of more node surgery against less, one line per cancer, so you can find yours and see what was actually measured.

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

How the mapping is done

When wider went worse

Bladder cancer produced the least comfortable result in the whole set. A trial across 27 centers randomized 592 people having their bladder removed either to the standard node dissection or to an extended one reaching higher into the abdomen, and the extended operation delivered no benefit on any cancer measure, with five year survival free of disease of 56 percent against 60 percent and overall survival of 59 percent against 63 percent. Severe adverse events occurred in 54 percent of the extended group and 44 percent of the standard group, and death within ninety days of surgery occurred in nineteen patients after the extended operation against seven after the standard one.

European investigators had already found the same absence of benefit in a trial of 401 patients, with a median of nineteen nodes removed in the limited arm against thirty one in the extended arm and no significant difference in recurrence or survival. Fluid collections severe enough to need intervention were more common with the extended operation, though the trial did not publish a rate for them. A trial that finds no benefit is easy to set aside as inconclusive, and a trial that finds no benefit alongside a higher rate of early death is a different kind of finding, because it shifts the burden of proof onto the wider operation.

Prostate surgery tells a quieter version of the same story. A trial of 1,440 men randomized by surgeon rather than by patient compared extended and limited templates and found a difference in positive node rate of under two percentage points, no difference in biochemical recurrence, and grade two or three complications in 7.3 percent against 6.4 percent. Its authors made an unusual admission, which is that the two operations turned out to be more alike than intended, and that what the field really needs is a trial of node dissection against none at all. A second trial of 300 men found no overall benefit either, although within the subgroup with the most aggressive biopsy grade the extended operation looked better, a finding its authors explicitly labeled as a hypothesis instead of a result. Taken together the prostate evidence says less about how wide the template should be than about how little either template appears to change what happens to the men inside it.

What to ask your surgeon

Arm swelling by the numbers

Lymphedema is the harm that drives the whole de-escalation argument, so it deserves precise figures rather than a warning. It is swelling caused by fluid that can no longer drain past the point where the nodes were removed, it usually appears in the first two years, and once established it is managed rather than cured.

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

The other lasting problems

Swelling gets the attention, and it is not the only thing that persists, because fluid collections, numbness, wound breakdown and shoulder weakness all follow node surgery at rates worth knowing in advance, and several of them are avoidable by technique instead of luck.

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

Trying to prevent swelling

If swelling has started

The standard treatment, and what part of it works

Established lymphedema is usually treated with a package of compression bandaging, a fitted garment, skin care, exercise and a specialized massage technique. A randomized trial of 103 women tested whether the massage and bandaging component adds anything on top of a compression garment alone, and at six weeks the reduction in excess arm volume was 29.0 percent with the full package against 22.6 percent with the garment, a difference of 6.4 percentage points whose confidence interval ran from minus 6.8 to 20.5. Absolute volume loss did favor the full package, at 250 milliliters against 143. Quality of life and arm function did not differ. The trial's authors attribute the null primary result partly to its size, and the fair summary is that compression is the part carrying the load while the manual technique is unproven rather than disproven. Anyone being asked to commit several hours a week to a treatment program deserves to know which part of it carries the evidence, and in this case the answer is the garment.

Surgery for the swelling itself

Two operations exist for swelling that has not responded to compression. One connects blocked lymphatic channels to tiny veins, the other transplants a small package of healthy nodes with its own blood supply into the affected limb. Pooling 52 studies, clinical improvement was reported in 36.46 percent of upper limbs and 34.16 percent of lower limbs, with the transplant slightly ahead of the channel operation in the arm at 41.66 percent against 29.44 percent. Almost none of that evidence comes from randomized trials, and the reviewers state that how best to combine the two procedures is still undecided, so anyone offered either should hear those numbers first, because a third of patients improving is a real result and it is not a cure.

The womb and vulva trials

Gynecological cancer produced two of the sharpest results. In cancer of the womb lining, a British trial randomized 1,408 women to standard surgery or to standard surgery plus removal of the pelvic nodes, and the extra surgery did not help. There were 88 deaths in the standard arm and 103 in the lymphadenectomy arm, a hazard ratio of 1.16 favoring the smaller operation, and the difference in five year survival was one percentage point. An Italian trial of 514 women found the same absence of benefit while quantifying the cost, since a median of thirty nodes were removed and complications occurred in 81 patients against 34.

In vulvar cancer, 623 groins were studied in 403 women. Comparing sentinel node surgery alone against sentinel node surgery followed by full groin clearance, leg swelling occurred in 1.9 percent against 25.2 percent, wound breakdown in 11.7 percent against 34.0 percent, and skin infection in 4.5 percent against 21.3 percent. Groin recurrence among those with a negative sentinel node was 2.3 percent.

That vulvar study was a controlled multicenter study rather than a randomized trial, which is worth stating plainly, and its authors restricted their recommendation to selected patients treated by a quality controlled team. A follow up study of 1,535 women then tested radiotherapy instead of surgery when the sentinel node contained only a tiny deposit, and among the 160 women in that category two year groin recurrence was 1.6 percent. Where the deposit was larger than two millimeters the same approach performed worse than surgery, at 22 percent against 6.9 percent, and that comparison was not randomized. Sentinel node mapping in cancer of the womb lining was validated separately in a study of 385 women, where it identified metastases in 35 of the 36 patients who had both a mapped node and node positive disease. Reading those two studies in order shows how a field moves, since the first established that a smaller operation was safe when the node came back clear, and the second tested how far that permission could be extended once the node did not.

What the sentinel node misses

No test is perfect, and a page arguing for less surgery has an obligation to state what the smaller operation gets wrong. The honest figures are these.

• About one in ten in the breast, in the trial that measured it. Among 766 patients in the large breast trial who had both the sampling and a clearance, the sampling reported a clear result while the clearance found disease in 9.8 percent, with a confidence interval of 7.8 to 12.2 percent. Overall accuracy in that group was 97.1 percent. The trial also identified what makes the miss rate worse, namely where the tumor sits, whether it was removed by excision beforehand, and how many sentinel nodes were taken.
• About one in thirty in the womb lining, on the better study. The 385 patient study reported sensitivity of 97.2 percent and a negative predictive value of 99.6 percent, and its authors put the practical meaning bluntly, which is that the technique will not identify disease in 3 percent of patients who have it.
• Worse in a second study of the same cancer. A French study of 125 women found three false negatives, giving per patient sensitivity of 84 percent with a confidence interval of 62 to 95, and a negative predictive value of 97 percent. All three misses were in the more aggressive type of the disease. Quoting only the flattering study would be misleading, so both are here.
• Pooled across 33 studies, sensitivity of 91.8 percent. That Cochrane figure covers 2,237 women and 409 with node involvement, and the reviewers rated the methodological reporting of most included studies as poor, producing a large number of unclear risk of bias judgments, while their meta-regression found no difference in sensitivity between the tracer combinations.
• Sometimes the hottest node is the wrong node. In the melanoma series of 1,184 patients, 13.1 percent of positive drainage areas would have been called clear if only the single most radioactive node had been removed. That is an argument for taking a small number instead of exactly one, and it is the reason a surgeon may remove three or four nodes and still describe it as a sampling.

What to send us

Any useful opinion about node surgery depends almost entirely on documents rather than on a conversation, and the list below is what a surgeon needs before saying anything specific about your case.

• The full pathology report, including everything past the summary line. The cancer type, its grade, its size in millimeters, and for melanoma the thickness. These are what determine which trial applies to you, and a one line diagnosis is not enough to place you in one.
• Every report that describes your nodes, and the images. Ultrasound of the armpit or groin, computed tomography, magnetic resonance imaging, and any positron emission scan. Send the image files alongside the written reports, because a radiologist here will want to look rather than read.
• Any needle sample already taken from a node. A positive needle sample changes the plan more than almost anything else in the file, because it moves you out of the group most of the de-escalation trials studied.
• Details of any treatment already given. Chemotherapy before surgery changes what the nodes look like and how reliably they can be mapped. Previous radiotherapy or previous surgery in the same area changes the drainage itself.
• Measurements of both limbs if anyone has taken them. Rare, and valuable. If nobody has, say so, and it will be done here at your first appointment.

The week around surgery

Reading your surgical plan

Written plans use a small vocabulary that hides large differences. The table translates the terms you are most likely to see, so you can tell from the paperwork which operation is actually being proposed.

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

Questions to take with you

Take this list to any consultation about node surgery, wherever it happens. The answers are more informative than any brochure.

1. Ask how many nodes are expected to come out, and whether that counts as a sampling or a clearance.
2. Find out which randomized trial covers a patient like you, and what it found about survival.
3. If the sentinel node is positive, what is your plan, and was that plan tested in a trial?
4. Radiotherapy to the nodes may be an alternative in your case, so find out what it would change about your risk of swelling.
5. Request a measurement of both limbs before surgery, and establish who reviews those measurements afterward.
6. Check whether the nerve supplying sensation to the inner arm is preserved, and whether the arm's own drainage channels are mapped and spared.
7. Immediate lymphatic reconstruction is not offered everywhere, so raise it by name and see whether it is recommended for you.
8. Pin down the drain. Ask how long it stays in, and what the plan is if fluid keeps collecting once it comes out.
9. Name a contact for home before you leave, and confirm how quickly a limb that starts to feel tight can be assessed.
10. If I decline the wider operation, what specifically changes about my treatment and my follow up?

Coming to Istanbul

How the stay is arranged

Most people having a sentinel node sampling as part of their cancer operation stay five to seven days in total, which allows two days before surgery for assessment and mapping, and enough time afterward for the pathology result to come back and be discussed in person instead of emailed later. A full clearance usually means seven to ten days, because a drain needs to be managed and removed before a long flight is sensible. Where the operation involves the groin the stay is longer again, since wound problems there are common and are much easier to treat while you are still here, and nothing about that timing is unusual, since it is set by the wound and the drain instead of by the cancer. Booking a return flight before the pathology report is back tends to create pressure at exactly the wrong moment, so it is worth leaving the date open or choosing a changeable ticket even where everything is expected to be straightforward.

Follow up after you return home

You go home with your pathology report, your limb measurements, a written description of exactly which nodes were removed, and a named contact. Swelling that develops later needs assessment where you live, and a specialist there will want the operative note and the node count, which is why those documents are given to you instead of kept. If the pathology changes the plan, that conversation happens before you leave. Ask for everything in writing and in a language your own doctors read.

Lymph node removal surgery FAQ