Lumbar Microdiscectomy - Minimally Invasive Disc Surgery

Free consultation

Send the images and the report, and have a spine surgeon read them against your symptoms

Send the image files themselves along with the radiology report, a description of exactly where the pain, numbness or weakness travels in the leg, how long it has been there, whether it is getting better or worse, and a note of everything already tried. A spine surgeon reads the images against your symptoms and tells you which level and which side are responsible, whether an operation is the right question at all, and what the realistic alternatives are. No fee, no obligation, and a coordinator replies in your own language, usually within the same working day.

The hole nobody mentions

What the names mean

Six or seven different words get used for operations that all remove a piece of lumbar disc. Knowing which one is being proposed to you matters less than most websites suggest, but you should still be able to say the word back.

• Open microdiscectomy. A small incision in the midline, a retractor, and an operating microscope. This is the reference operation that every newer technique is measured against, and it is the one with the longest record. Incisions are typically two to three centimeters.
• Tubular microdiscectomy. The same operation delivered through a tube that spreads the muscle instead of stripping it. Sometimes called minimally invasive discectomy or, when a camera is used down the tube, microendoscopic discectomy.
• Full-endoscopic discectomy, interlaminar. A single working channel with a camera at its tip, entered from behind through the gap between two vertebrae. Often done under sedation with the patient awake enough to talk.
• Full-endoscopic discectomy, transforaminal. The same instrument entered from the side, through the opening the nerve exits by, passing behind the abdominal cavity. Marketed hardest, and the version with the most radiation attached to it.
• Unilateral biportal endoscopy. Two small punctures instead of one, a camera in one and instruments in the other, giving the surgeon more room to move than a single channel allows.
• Sequestrectomy. Removing only the loose fragment and leaving the rest of the disc entirely alone. A deliberate choice about how much to take, and a separate question from which door was used.

The door is not the operation

Ten years later, the door made no difference

Korean national insurance records followed 1,856 adults who had a first operation for a lumbar herniated disc between 2005 and 2007, for eight to ten years, using a statistical method that treats death as a competing event. Cumulative reoperation reached 4 percent at one year, 6 percent at two, 8 percent at three, 11 percent at five and 16 percent at ten. Broken down by the technique used the first time, the ten-year figures were 16 percent after open discectomy, 14 percent after laminectomy, 16 percent after percutaneous endoscopic discectomy and 10 percent after fusion, with no significant difference among them. What did differ was the kind of operation people came back for, since open discectomy was the revision operation in 80 percent of patients whose first operation was open and in 81 percent of those whose first operation was endoscopic. Reading that honestly means accepting that the choice of approach, which is the thing patients research hardest and the thing brochures are written about, did not show up a decade later in the outcome most people care about most.

Twelve consecutive null results for the tube

Eight randomized trials and two retrospective studies covering 804 patients were pooled to compare tubular microdiscectomy against conventional microdiscectomy. The pooled analysis found no significant difference in operative time, blood loss, hospital stay, complications during surgery, complications after surgery, dural tear, reoperation, back pain in the short or long term, leg pain in the short or long term, or disability score. Twelve comparisons, twelve null results. The authors themselves urged caution because the total sample was limited, and the honest summary is still that nobody has been able to demonstrate the tube making a measurable difference to anything a patient experiences. A tube that spreads muscle instead of stripping it ought to help, the theory behind it is sound, and the measurements keep declining to confirm it, which is a pattern worth recognizing when the next refinement is announced.

The endoscope, in the round

Twenty-two comparative studies covering 4,068 patients were pooled to set the main minimally invasive routes against each other. Transforaminal endoscopic surgery and microendoscopic discectomy took almost the same operating time, at 64.19 minutes against 66.61, while the transforaminal route lost less blood and kept people in hospital for a shorter time. Recurrence went the other way, at 3.69 percent after microendoscopic discectomy against 6.08 percent after the transforaminal route, reported with no interval and no p value attached to it. Then there is the largest randomized comparison anyone has run, 613 patients across four Dutch hospitals, which found a difference of 7.1 points on a 100-point leg pain scale at twelve months in favor of the endoscope, with repeat surgery within a year at 5 percent against 6 percent. Their own conclusion contains the sentence that matters, which is that the differences were small and may not reach clinical relevance, and it deserves quoting far more often than the headline is. Seven points out of a hundred is the size of the entire advantage that a decade of marketing about keyhole spine surgery has been built on, and it comes from a well-run trial whose investigators went out of their way to say so.

What happens in the room

The first six weeks

How much to take out

Once the fragment is out and the tear is visible, the surgeon faces a genuine choice with a genuine cost on both sides. Taking more disc out through the tear removes material that might escape later. It also removes material that is holding the disc space open and carrying load. An evidence-based review graded the whole literature and named the trade explicitly.

Fair quality evidence supports conservative discectomy giving shorter operative times, a quicker return to work, similar hospital stay, similar pain at discharge, similar function at six months and similar persistent or recurrent back and leg pain at two years. Fair quality evidence also supports conservative discectomy giving a higher incidence of recurrent disc herniation. No level one studies exist on the question.

Read that twice, because it is the shape of the entire argument. Removing less gives you a faster recovery and a slightly higher chance of the problem coming back. Removing more gives you the reverse, and it takes disc height with it. Neither answer is wrong, and a surgeon who presents one of them as obviously correct has skipped the part where the two are weighed against each other for your particular hole and your particular life.

What the trials on that found

The randomized comparison, and how small it was

Eighty-four patients at a single center were randomly allocated to microdiscectomy or microscopic sequestrectomy, with 78 available at two years. Reherniation was 10.5 percent after discectomy and 12.5 percent after sequestrectomy, a comparison involving four events in one group and five in the other, which is far too few to establish that the two are equivalent. Self-rated outcome drifted the other way, deteriorating over the first two years after discectomy and improving after sequestrectomy, with analgesic use and overall outcome at two years favoring the smaller operation. Take the direction seriously and the precision not at all. Small randomized trials of this kind are common throughout spine surgery, and they are a large part of why so many basic questions stay open, because a study with fewer than a hundred patients can only detect a difference so large that nobody would have needed a trial to notice it. That is the honest limit here.

Pooling everything that exists

A systematic review of five studies and 746 participants, of which only one was randomized, found no significant differences in leg pain, back pain, function, complications, hospital stay or two-year recurrence, and graded that evidence Low. It found less painkiller use after sequestrectomy and graded that Very low. Every study included was judged to be at high risk of bias. A second review of seven studies with follow-up between eighteen and eighty-six months reported leg pain improving by 5.6 to 6.5 points after microdiscectomy and 5.5 to 6.6 points after sequestrectomy, with reherniation between 2.3 and 11.8 percent after the first and between 2 and 12.5 percent after the second, and concluded that the two are comparable specifically when a small breach in the posterior ring is found during the operation. Two systematic reviews reaching the same non-answer, one grading its own evidence as Low and the other hedging its conclusion with a condition about the size of the tear, is a fair picture of where this question genuinely stands.

Which puts the condition back where it belongs

Notice the qualifier in that last sentence, because it is the whole point. Taking only the loose fragment appears to be a fair choice when the hole is small. The evidence does not say the same thing about a large defect, and the study that separated patients by exactly that variable found reherniation running from 1 percent to 27 percent across the categories. Every reasonable comparison of how much to remove is really a comparison conditional on something the surgeon can only see once they are already inside.

Questions before you consent

Ten questions, and a competent surgeon can answer every one of them in a sentence. The first four are about your anatomy, the middle three about the operation, and the last three about what happens once you leave.

1. Point at the fragment on my images and confirm the side and the level match where my symptoms actually go.
2. Tell me whether the fragment looks contained or free, and what that implies for me.
3. Say what you expect the annular defect to look like, and promise to record its size in the operative note.
4. Explain what would happen if I waited three more months, in my particular case rather than in general.
5. Name the approach you intend to use and give me the anatomical reason for choosing it.
6. Tell me how much disc you plan to remove, and what makes you lean one way or the other.
7. Say roughly how many of these operations you do in a year, and how long you have been doing this particular approach.
8. Give me your own reherniation and reoperation rates, and tell me how you know them.
9. List which activities you will restrict afterward and what evidence those restrictions rest on.
10. Confirm I leave with the operative note, the implant details if anything is implanted, and my images.

What the scan shows afterward

Two thirds of operated discs still look herniated

The randomized sequestrectomy trial described above also scanned everybody at two years, with a neuroradiologist reading the images blind to which operation each patient had. A protrusion or extrusion of at least four millimeters was present in 66 percent of the discectomy group and 68 percent of the sequestrectomy group, which is not a difference. What matters far more is the second half of the finding, because the presence of that residual bulge did not correlate with back pain or with sciatica. Somebody who has an excellent result and somebody who has a poor one have roughly the same chance of looking abnormal on a scan two years later, and this is the single strongest reason not to order an image for reassurance. Anyone ordering a scan after this operation without a new symptom to explain should expect to find something that looks wrong, and should decide beforehand what they would do about it, because the honest answer is usually nothing at all.

What did differ between the two operations

Loss of disc height appeared in 63 percent of the discectomy group against 38 percent of the sequestrectomy group. Endplate degeneration increased in 47 percent against 14 percent. Both differences reached significance, and endplate change did correlate with low back pain in that cohort, which is the mechanism by which taking out more disc might cost you something years later. That is a single center with 78 patients analyzed, so treat it as a signal about direction and not as a rate you can apply to yourself.

How to read your postoperative report

Words like residual protrusion, scar tissue, enhancing granulation and disc space narrowing appear on almost every scan taken after this operation, and they describe a normal postoperative spine. Radiologists write them because they are accurate, and how alarming they read is a side effect of accuracy. The finding that should trigger a conversation is new compression of a nerve root on the side and at the level of new symptoms, and everything else on the page is context.

Recurrence has a clock

Reherniation is not spread evenly across the years after surgery, and knowing when it clusters changes how you should behave. Among 1,228 patients followed for at least five years after percutaneous endoscopic discectomy at one center, 77 had a recurrence, and 49 of those 77 happened within six months of the operation. Almost two thirds of every recurrence that was ever going to happen had already happened by the half-year mark.

The largest study of what follows a single-level discectomy tracked 308,979 adults in a United States claims database for at least five years, and found that 14.4 percent had further lumbar surgery and 6.1 percent had a lumbar fusion within that window. Among the 67,098 who had already had a revision discectomy, the five-year figures rose to 18.2 percent and 12.4 percent. The survival curves in that study fall most steeply during the first year after both the index operation and the revision, which says the same thing as the smaller study in different units. Claims databases have real limits, and one of them was measured directly in another study where the diagnostic codes for a specific surgical complication picked up only 5.4 percent of the cases that had actually occurred. Figures drawn from insurance records describe what somebody billed for, which overlaps with what happened to patients without ever being quite the same thing, and here the gap between the two is wide enough to change how the number should be read.

Definitions vary more than most people realize. A systematic review restricted to what it called real recurrence, meaning new disc material at the same level and on the same side as the original, found published rates of revision surgery for that specific event between 1.4 and 11.4 percent, with successful outcomes after revision reported anywhere between 60 and 100 percent. A range that wide is not an expectation you can plan around, and it exists because different papers are counting different events under one word.

Two practical conclusions follow. New leg pain in the first six months deserves a phone call rather than a wait-and-see attitude, because that is when it means something. And a figure quoted to you as a recurrence rate is close to meaningless unless whoever quoted it can say which of these things they counted, over what period, and in whom.

When to call the same day

Five things warrant contacting somebody on the day you notice them, in roughly this order of urgency. Everything else can wait for a scheduled appointment.

1. Difficulty passing urine, loss of control, or numbness in the saddle area. In a series of 2,421 microendoscopic operations, bladder or bowel dysfunction occurred in 3.0 percent overall and in 17.9 percent of the patients who had sustained a dural tear during surgery.
2. New or worsening weakness in the foot or leg. Weakness that was there before surgery and is slowly improving is expected. Weakness that appears after you were already improving is not.
3. Leg pain that returns and stays. Especially in the first six months, when almost two thirds of recurrences occur. Pain that comes and goes with position is different from pain that arrives and remains.
4. Clear fluid leaking from the wound, or a severe headache when you sit or stand up. Both point to the dura, and both are far easier to deal with early. In one seven-year cohort, none of the patients who needed further treatment for a dural tear presented within the first 48 hours.
5. Fever, spreading redness or increasing wound discharge. Superficial wound problems are the commonest reason a day-case patient comes back, and they are straightforward when caught early.

The restriction question

Somebody finally randomized it

Two hundred patients having a unilateral microdiscectomy were randomly assigned either to a restricted regime, meaning sitting limited to fifteen to thirty minutes in every two hours, no lifting above five kilograms and no strenuous activity for a fortnight, or to no restrictions at all. Everybody wore an activity monitor for a month. At one year the composite outcome was reached by 41.6 percent of the restricted group and 36.4 percent of the unrestricted group, reherniation occurred in 10.1 percent against 14.1 percent, reoperation in 2.9 percent against 5.5 percent, and back pain fell by 23.5 points against 24.5 points. Not one of those comparisons approached significance, and 200 patients cannot rule out a modest difference in reherniation, so the trial is best read as finding no signal rather than proving there is none. Trials of this kind are rare because they are unglamorous and difficult to fund, which is precisely why one that somebody actually carried out deserves more attention than the twentieth comparison of one endoscope against another.

The number that changes how you read it

Adherence to the restrictions was 10 percent. The monitors showed the restricted group sitting for 4,102 minutes a week and the unrestricted group for 4,140, which is not a difference anyone would notice. So the trial did not really compare restricted living against unrestricted living. It compared being told to restrict against not being told, and found that being told changed nothing, partly because almost nobody complied. That is a genuinely useful finding about instructions, and a weaker finding about behavior.

The same answer, twenty-seven years earlier

Back in 1999, 152 consecutive working patients had a limited open discectomy and were given no restrictions whatsoever, being encouraged instead to resume full activity as fast as they could manage. Followed for an average of 4.8 years, 149 of them returned to work, average work loss was 1.2 weeks, and roughly a third were back within a week, many the day after surgery. Very early return did not correlate with recurrent sciatica, with reoperation for reherniation or with the eventual clinical outcome. That study had no control group and everybody in it was working beforehand, which is the strongest predictor of returning to work there is, so it establishes that the approach is feasible instead of proving it is better. Two studies separated by twenty-seven years, using different methods on different populations in different countries, arrived at the same answer about restrictions, and a conclusion that survives a gap like that is worth more than either study standing alone.

Work and driving

Numbers exist for both questions, and they are more encouraging than the advice most people are given. Read the third column before you plan anything around the second.

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

One counter-intuitive detail deserves pulling out of that table. Braking was faster immediately after surgery than before it, in both operations studied, because the pain itself had been the impairment. People are frequently told to avoid driving for weeks on the assumption that surgery makes them slower, and the measurement points the other way, although patients at thirty days were still slower than healthy volunteers and the study measured a reflex on a simulator instead of competence on a road.

The size of the defect

Four categories, one operation

In the study this page opened with, 187 consecutive patients having a single-level primary discectomy were sorted during the operation into four groups by what the surgeon found. Those with a fragment and a small annular fissure, 89 of them, had a 1 percent reherniation rate and a 1 percent reoperation rate. Those with a contained fragment, 42 people, had 10 percent and 5 percent. Those with an extruded fragment and massive loss of the posterior annulus, 33 people, had 27 percent and 21 percent. A fourth group of 16 with no fragment at all fared worst on symptoms, with 38 percent having recurrent or persistent sciatica. The paper concluded that what was found during surgery mattered more than demographic, socioeconomic or clinical variables, which is a striking finding given how much attention those other variables receive. Nothing about which door the surgeon used appears anywhere in that analysis, because there was only one door and one surgeon throughout, and the twenty-seven-fold spread between the best and worst groups appeared anyway.

What can and cannot be claimed from it

No confidence intervals were published for any of those percentages, the two extreme groups contained 89 and 33 people, and it was one surgeon at one center. The ordering is robust and the precision is not, so nobody should be quoted the figure 27 percent as though it were their personal risk. A separate retrospective study of 188 patients found that the median length of the annular tear was significantly greater in the 21 who reherniated, which supports the direction, and it reported no measurement, threshold or effect size at all, which limits it to exactly that.

Why this changes the conversation

Almost every discussion of disc surgery treats the operation as one thing with one risk profile. The evidence says it is at least two different operations sharing a name, distinguished by something invisible beforehand. That has a practical consequence, since a surgeon who records the annular defect in the operative note has given you and every future doctor a piece of information that no scan can recover afterward, and a surgeon who does not has thrown it away.

Plugging the hole

If the defect governs recurrence, closing the defect ought to reduce recurrence, and somebody built a device to do exactly that and tested it properly. A bone-anchored annular closure device consists of a polymer mesh held across the inside of the tear by an anchor driven into the vertebral body. It is placed at the end of an otherwise ordinary microdiscectomy, and it is offered to patients whose annular defect is measured during surgery as six to ten millimeters wide.

Five hundred and fifty-four patients at twenty-one centers across six countries were randomized between 2010 and 2014 to receive the device or not, which makes this a real trial with a real comparison group, and it is unusual enough for a spinal implant to have one that the results deserve to be taken seriously in both directions. The trial was supported by the manufacturer, two authors received study-specific support of more than ten thousand units of currency a year, eight received less than that, and eleven received none, all of which is printed in the paper itself rather than buried. Declaring it openly is what allows a reader to weigh it properly, and the pattern worth watching for is never sponsorship on its own but sponsorship combined with the absence of anybody outside the trial group ever reproducing the central finding independently. Both halves of that pattern are present here, and the second half is the one that should give you pause.

Reading the results honestly requires separating four things that often get merged. There is what shows up on a scan, there is what a patient feels, there is what leads to another operation, and there is what the device itself does to the bone it is anchored in. The trial reported on all four, and they do not all point the same way.

What that trial showed

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

Weighing that table is a real decision and not an obvious one. Halving symptomatic reherniation is worth having if you are the person it would have happened to. Against it sits an implant anchored in bone that produces endplate changes in five patients out of six, a single-center series outside the trial reporting mesh migration into the canal in almost one patient in five, and the plain fact that patients with the device did not feel better than patients without it over two years. Anyone offering it should be able to state both halves of that without prompting, and anyone who offers only the first half is selling.

Learning curve and volume

How many cases it takes

Published estimates cluster in a narrow band. A cumulative sum analysis of one surgeon's first 55 transforaminal endoscopic cases found operative and fluoroscopy times falling sharply after case 31, with operative time dropping from 85.7 to 62.2 minutes. Another such analysis put the threshold for the same operation at 36 cases and for biportal endoscopy at 27. A study of four surgeons learning biportal endoscopy found that those with prior endoscopic experience reached proficiency at 17 and 18 cases while those without needed 25 and 27. For the tubular microendoscopic approach, a review of one surgeon's first 125 operations recommended 25 to 30 cases, having found operative time falling from 87 to 58 minutes and blood loss from 76 to 33 milliliters between the first and second quartiles. Those thresholds are not interchangeable with one another, since each comes from a single center applying its own definition of proficiency to its own patients, and what survives the comparison is the order of magnitude and not any particular figure in it.

The part experience does not fix

Guide-wire migration, neural injury, dural tear and problems docking the tube all fell significantly after the first thirty cases in that tubular series. Recurrence did not. The authors' own phrasing is that recurrence occurred at any phase, and it is the sharpest sentence in this entire literature, because it separates the two halves of the operation cleanly. Experience fixes the technical problems that belong to the approach. It does not touch the biology of a hole in the annulus, which is indifferent to how skilled the person who made it happens to be. The learning-curve arm of the large Dutch trial found something consistent, since patient-reported outcomes were no worse during the first twenty supervised cases, while two of the three surgeons learning the technique had substantially higher one-year reoperation rates than the senior surgeon or the comparison group. Nobody has ever marketed an operation on the grounds that their surgeon has now done thirty of them, and the distance between what gets advertised and what has actually been measured is the whole subject of this section.

A number you can actually ask about

Analysis of 187,185 lumbar procedures performed by 5,514 surgeons across 178 hospitals in one American state found a volume threshold for discectomy at 40 operations a year, or roughly four a month. Surgeons below that threshold carried a 56 percent higher risk of complications within ninety days. Two cautions belong with it, since a 56 percent increase on a small absolute risk is still a small absolute risk, and the dataset covered inpatient procedures only, which will misclassify a busy day-case surgeon as a low-volume one. The question is still worth asking out loud.

Radiation, measured

Endoscopic approaches that go in from the side are guided by live X-ray, and the dose that involves is rarely mentioned in the brochure, although it has been measured repeatedly, by medical physicists as well as by surgeons, and the numbers are worth seeing.

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

None of this makes the transforaminal route a bad operation. It does mean that the version marketed most aggressively as gentle carries a cost that is invisible in the marketing, that the cost falls substantially as the surgeon gets more experienced, and that simple measures reduce it. Reaming bone through the foramen roughly doubled the surgeon's dose in one study, which is worth knowing if that step is planned for you.

Complications, named

Tearing the dura

The membrane holding the spinal fluid sits directly against everything the surgeon is working around, and it tears sometimes. In 346 consecutive single-level microdiscectomies by one surgeon it happened in 4.9 percent, and in 2,421 microendoscopic operations with video documentation it happened in 6.9 percent, neither of which should be confused with the 14.5 percent reported across 3,684 lumbar decompressions of every kind, a figure that includes far bigger operations. When a tear happens it is repaired during the same operation, and two separate cohorts found no benefit to keeping people in overnight afterward, with one noting that none of the patients who eventually needed further treatment for a tear presented within the first 48 hours.

Bladder and bowel trouble after a tear

Among those 2,421 operations, bladder or bowel dysfunction occurred in 3.0 percent overall and in 17.9 percent of patients who had a dural tear, against 1.9 percent of those who did not. The recovery curve is the useful part for anybody living through it, since the problem persisted in 64 percent at one week, 44 percent at one month, 40 percent at three months, 28 percent at six months and 13.6 percent at one year. Most people recover and the recovery is slow, and roughly one in seven has not fully recovered by a year. Exposure of the cauda equina during the tear carried by far the highest risk, from a very small number of events, so treat the size of that association with caution. Anybody living through that recovery deserves to see the whole curve instead of one reassuring sentence, because knowing that improvement carries on for a full year changes how the third month feels.

Bleeding behind the abdomen

Reaching the disc from the side means passing behind the abdominal cavity, and in 412 consecutive transforaminal endoscopic operations, 4 patients had a symptomatic retroperitoneal hematoma afterward, which is close to one in a hundred. All four complained of groin pain. Average volume was 528 milliliters, two needed open evacuation by general surgeons and two were managed without surgery. Every one of them recovered without lasting nerve damage. The route that causes this is the same route that spares the back muscles, which is a fair summary of most surgical trade-offs.

The rare and the serious

Instruments passing all the way through the front wall of the disc can reach the aorta and the iliac vessels, which lie a couple of millimeters beyond it. A literature review found 159 such injuries published since 1945, most commonly at the L4 to L5 level, and the figure has no denominator because it counts published cases and publication favors the dramatic. Modern management is increasingly a stent rather than open surgery. Nerve root injury and lasting altered sensation are the other approach-specific problems named in a systematic review of 117 studies covering 20,020 endoscopic patients, although that review published no rates at all.

Operating on the wrong level

A national survey of neurosurgeons covering 4,695 lumbar discectomies in one year put self-reported wrong-level lumbar surgery at 12.8 per 10,000 operations. A separate survey found that almost half of responding surgeons had done it at least once across a career, which is an entirely different statistic and must never be presented as a per-operation risk. The contributing factors surgeons named themselves were fatigue, time pressure, emergency operations, unusual anatomy and failure to check the level with an image, which is why step one on this page is an image with a marker in it.

Coming to Istanbul

Traveling for this operation is reasonable when the diagnosis is clear and the plan is agreed before anybody books a flight, so send the images first, because a second reading sometimes changes the plan entirely and it costs you nothing but an email.

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

Two things travel badly and both are worth planning around. Language matters most in the first two days after surgery, when you need to describe a new symptom accurately to somebody who can act on it, so establish before you fly who you will message and in which language. Records matter most years later, when a doctor who was never in the room has to work out what was done, and the operative note describing the level, the side and the size of the annular defect is the document that answers that. Hand both to your own doctor at the first appointment instead of filing them.

Lumbar microdiscectomy FAQ