Biceps Tenodesis: Because Apparently Every Painful Tendon Needs a New Home
We took a painful shoulder, removed a structure that contributes to shoulder stability, drilled into the humerus, relocated the tendon, told the patient not to load it for a while—and then acted surprised when some of them still hurt.
I first learned about the biceps tenodesis procedure in graduate school. As it was explained to us, since the biceps tendon is often the cause of shoulder pain you should cut it and move it. Like most overwhelmed graduate students we accepted it as fact and tried to remember that for a test.
Fast forward a few years, when I was practicing in a clinic that saw a lot of shoulders. We often saw patients both before and after surgery.
I started noticing something interesting.
Before surgery, many of these patients certainly had shoulder pain and limitations—but the pain often wasn't particularly anterior. Direct pressure over the long head of the biceps didn't reliably reproduce their complaint, and neither did many of the so-called special tests for biceps tendinopathy.
Then they came back after surgery, which frequently included a biceps tenodesis.
Now they had anterior shoulder pain. Usually a lot of it.
It commonly persisted for at least six weeks, and I also began seeing patients develop the classic postoperative "Popeye" deformity.
And then I noticed something even stranger.
Some of the patients who appeared to lose their tenodesis fixation and develop a Popeye deformity actually reported less anterior shoulder pain afterward.
That's an anecdotal observation, not evidence that a tenodesis needs to fail to feel better.
But it sure made me start asking questions.
That eventually sent me looking into the origins, biomechanics, research, and outcomes of biceps tenodesis.
The biceps tenodesis procedure has become an increasingly common answer to anterior shoulder pain, long-head biceps pathology, and SLAP lesions.
The concept sounds simple enough.
The long head of the biceps tendon hurts. Detach it from its origin, move it out of the shoulder, attach it somewhere on the humerus, let it heal, and the painful structure is no longer in the joint.
Problem solved.
Except biology rarely works that cleanly.
Because biceps tenodesis doesn't simply "remove a pain generator."
It takes a functioning tendon, removes its proximal attachment, changes its mechanical role, creates an entirely new tendon-to-bone interface, changes where and how force enters the humerus, and requires a period of postoperative protection while this new construct heals.
Maybe that trade is worthwhile in the right patient.
But we should probably stop pretending it's mechanically insignificant.
First: The Long Head of the Biceps Isn't Just Taking Up Space
The long head of the biceps tendon (LHBT) originates from the superior glenoid/superior labral region, crosses the glenohumeral joint, passes through the bicipital groove and transitions into the biceps muscle.
We tend to think about the biceps primarily as an elbow flexor and forearm supinator.
But the long head crosses the shoulder.
And it appears to do something while it's there.
Biomechanical studies have demonstrated that loading the LHBT can reduce humeral-head translation under certain conditions. One classic cadaveric experiment found that applying tension to the LHBT significantly reduced anterior, inferior and superior translation of the humeral head. Other experimental work has similarly demonstrated a stabilizing contribution (Pagnani et al., 1996; Alexander et al., 2013).
More recent work makes the story more nuanced. With an intact rotator cuff, the stabilizing contribution of the LHBT may be relatively small. But when the subscapularis or infraspinatus is deficient, its contribution appears to become more important. In a 2023 biomechanical study, tenotomy increased humeral-head translation under several simulated cuff-deficient conditions (“The Long Head of the Biceps Has a Stabilizing Effect,” 2023).
That's an important distinction.
I'm not arguing that every biceps tenodesis creates a clinically unstable shoulder.
The evidence doesn't support that.
I'm arguing that calling the LHBT an expendable structure with no meaningful shoulder function is equally difficult to defend.
"Not essential" and "does nothing" are not the same thing.
Wait—Didn't That Tendon Help Stabilize the Painful Shoulder?
This is where the logic starts getting a little strange.
We have a painful shoulder.
The cuff may not be functioning perfectly. The patient may already have degenerative changes, altered loading strategies, weakness or impaired force production.
And our solution may be to remove the proximal attachment of a structure capable of contributing to humeral-head control (Pagnani et al., 1996; Alexander et al., 2013).
Again, that doesn't mean the shoulder suddenly becomes grossly unstable after tenodesis.
But it should at least make us hesitate before describing the LHBT as nothing more than an annoying pain generator.
Particularly in a cuff-deficient shoulder, there is a reasonable biomechanical question:
Are we removing a structure that was helping compensate for the very shoulder dysfunction we're trying to treat?
Recent biomechanical evidence suggests that question may be particularly relevant when the subscapularis or infraspinatus is deficient (“The Long Head of the Biceps Has a Stabilizing Effect,” 2023).
That question deserves more consideration than it usually gets.
A Tendon Is Not Just a Rope
Tendon isn't an interchangeable piece of cable.
Its extracellular matrix is highly organized, with collagen bundles predominantly oriented longitudinally to transmit tensile forces.
That architecture develops in response to the mechanical environment.
Change the mechanical environment and the tendon must adapt.
And biceps tenodesis changes that environment substantially.
The tendon is detached from its native proximal attachment and secured to a new location on the humerus using an interference screw, cortical button, suture anchor, soft-tissue technique, or another fixation construct.
We have now created a tendon-to-bone interface that didn't previously exist.
What About Collagen Fiber Orientation?
This is one aspect of the procedure that has always bothered me biomechanically.
The native LHBT is predominantly organized to transmit longitudinal tensile force along its collagen fibers.
After tenodesis, force has to transition through a surgically created fixation site into the humerus. Depending on fixation position and technique, the local tendon can experience combinations of tension, bending, compression and shear around that interface.
It is tempting to take that observation one step further and say:
"The collagen fibers are now pointing in the wrong direction."
I don't think the human evidence allows us to say that yet.
Tendons remodel. We don't have good longitudinal histologic evidence demonstrating that abnormal collagen orientation after biceps tenodesis causes long-term clinical failure.
But that doesn't make the mechanical question disappear.
The tendon is being asked to function in an environment that is not identical to its native one.
Healing is not necessarily restoration.
A tendon attaching successfully to bone does not mean we recreated normal tendon architecture, normal excursion or normal loading.
And Then There Is the Pain
This is where my skepticism increases.
Some patients hurt after biceps tenodesis.
And not merely because they are three weeks postoperative.
Patients can present with persistent anterior shoulder or upper-arm pain, focal tenderness, cramping, discomfort with resisted elbow flexion or supination, weakness, stiffness or difficulty returning to higher-level loading (McCrum et al., 2019; Ergün et al., 2022).
A large retrospective study examining 1,526 shoulders found persistent anterior shoulder pain in 10.8% of patients whose tenodesis was performed below/outside the groove and 12.9% when tendon remained in the groove. That difference wasn't statistically significant, but the bigger clinical point is difficult to ignore (McCrum et al., 2019):
Roughly one in nine patients continued to have anterior shoulder pain.
For a procedure designed in large part to address anterior shoulder pain, that number deserves a little more attention.
A systematic review comparing suprapectoral and subpectoral tenodesis likewise documented persistent bicipital pain following both techniques. In the pooled complication data presented by the review, persistent bicipital pain occurred in 88 of 1,404 reported subpectoral cases and 46 of 587 suprapectoral cases (Ergün et al., 2022).
So yes, you can successfully move the biceps.
That doesn't guarantee you successfully move the pain.
Sometimes We Don't Eliminate the Pain. We Relocate It.
Before surgery, the patient has anterior shoulder pain attributed to the biceps.
After surgery, the fixation can be completely intact and the patient can still have:
anterior shoulder or arm pain,
focal tenderness,
cramping,
weakness,
stiffness,
discomfort with pulling or carrying,
or reduced tolerance to loading (McCrum et al., 2019; Ergün et al., 2022).
The surgeon can therefore have a technically successful operation while the patient has a clinically disappointing result.
Which brings us to one of my biggest problems with how these procedures are sometimes discussed:
"The fixation held" is a remarkably low bar for success.
How Often Does the Tenodesis Actually Fail?
This is where my own clinical observation needs to be separated from the published evidence.
I have seen what appears to be a disproportionate number of problems with tenodeses fixated higher in or around the bicipital groove. Clinically, distal fixation has appeared to behave better to me.
But that is an observation.
The published evidence does not demonstrate that a large percentage of proximal tenodeses simply pull out.
True mechanical failure appears relatively uncommon.
In one systematic review, reported tendon-to-bone fixation failure/re-rupture occurred in 8 of 996 subpectoral procedures (0.8%) and 15 of 1,073 suprapectoral procedures (1.4%) in the studies contributing those data. The difference was not statistically significant (Ergün et al., 2022).
So if we're going to criticize tenodesis, mechanical pullout probably isn't our strongest argument.
Persistent symptoms are much more interesting.
But Where You Put the Biceps May Still Matter
The study of 1,526 shoulders provides a fascinating finding.
When the tendon remained in the bicipital groove, the revision rate was:
1.51%.
When fixation was performed below or outside the groove, the revision rate was:
0.60%.
That's roughly a 2.5-fold difference in revision frequency (McCrum et al., 2019).
But here's where we need to be careful.
Popeye deformity was almost identical: 4.72% versus 4.62% (McCrum et al., 2019).
Persistent anterior shoulder pain also wasn't statistically different between locations (McCrum et al., 2019).
So we cannot simply conclude:
"Proximal tenodeses pull out more."
They apparently required revision more often in that cohort, but the reason is more complicated than fixation failure alone (McCrum et al., 2019).
And that may actually be more interesting.
The Bicipital Groove May Still Be Part of the Problem
A proximal tenodesis may leave more tendon and associated tissue within the bicipital groove or tunnel.
That creates the possibility that residual tendon, synovium or extra-articular pathology remains symptomatic despite technically intact fixation.
This is one proposed reason that some failed proximal tenodeses are revised to a more distal/subpectoral location.
Again:
The anchor doesn't have to fail for the operation to fail clinically.
The tendon can remain exactly where the surgeon placed it.
The fixation can look beautiful.
And the patient can still hurt.
Systematic reviews generally find similar functional outcomes between suprapectoral and subpectoral approaches, although complication profiles differ. Some reviews have reported more persistent bicipital pain or Popeye deformity with suprapectoral fixation, whereas subpectoral approaches introduce their own risks, including nerve injury and wound-related complications (Ergün et al., 2022).
So I'm not arguing:
Low fixation good. High fixation bad.
I'm arguing something more basic:
Where you surgically relocate a tendon probably matters, and "the anchor held" doesn't tell us everything we need to know.
The Rehabilitation Paradox
Then comes rehab.
We take a load-bearing tendon.
Cut it from its native attachment.
Drill into the humerus.
Fix the tendon to its new home.
And then immediately tell the patient:
Don't load it. We're worried it might not hold.
To be clear, those postoperative precautions are biologically reasonable. We just created a healing tendon-to-bone interface.
But that's precisely the point.
We took a functioning tendon and intentionally converted it into a healing structure with temporarily reduced loading capacity.
Resisted elbow flexion and supination are commonly restricted early.
Strength declines.
Loading is delayed.
Then eventually rehabilitation has to rebuild the very capacity that surgery temporarily removed.
Again, that doesn't make surgery wrong.
But the reason for doing it should probably be pretty compelling.
Are We Sure the Biceps Was the Problem?
This may be the most important part of the entire discussion.
Anterior shoulder pain is not particularly specific.
Neither are many of the traditional orthopedic tests used to diagnose long-head biceps pathology.
Imaging abnormalities don't automatically establish symptom causation either.
And even the surgical literature has historically struggled to consistently define who actually needs a biceps tenodesis.
A systematic review of surgical indications found considerable variability in why surgeons performed the procedure, including partial tearing, instability, tenosynovitis, SLAP pathology, clinical examination findings and subjective shoulder pain. The authors highlighted the need for clearer indications (Slenker et al., 2015).
That should bother us.
Because once the tendon is cut, there is no "undo" button.
Maybe the Biceps Is the Victim, Not the Criminal
This is where rehabilitation needs to do a better job.
If a patient's biceps is irritated, why?
Has loading changed?
Is the rotator cuff functioning well?
What is the subscapularis doing?
What happens under meaningful pulling and carrying loads?
Is the neck contributing?
Is the patient's "biceps pain" actually reproducible with biceps loading?
Does changing shoulder position change it?
Can we progressively increase tendon capacity?
Has the patient actually undergone a serious loading program—or did they receive three weeks of band exercises before somebody ordered an MRI?
Finding an abnormal tendon near the location of pain doesn't establish causation.
The biceps may be the pain generator.
But it may also be the structure doing its best to manage a shoulder that isn't functioning particularly well.
Cutting it doesn't answer that question.
Before We Cut the Biceps, Make the Diagnosis Earn the Surgery
Before performing an irreversible procedure for pain, I would want a convincing case that the LHBT is actually responsible for the patient's symptoms.
"There's some fraying on the MRI."
I want to know:
Can we consistently reproduce the patient's complaint?
Does biceps loading reproduce it?
Have competing sources been meaningfully evaluated?
Has the patient completed an appropriately dosed progressive rehabilitation program?
Can the patient tolerate progressively heavier elbow flexion, supination, pulling, carrying and shoulder loading?
Did rehabilitation actually fail?
Or did we simply run out of patience?
Because orthopedic rehabilitation should have taught us this lesson by now:
Do not assume the structure closest to the pain is the structure generating the pain.
So Is Biceps Tenodesis a Terrible Idea?
Sometimes?
Maybe.
Always?
No.
The majority of patients reported in surgical series improve after appropriately indicated biceps tenodesis. There are legitimate reasons to perform the operation, including substantial tendon tearing, instability or subluxation, symptomatic SLAP/biceps-labral pathology in selected patients, and persistent symptoms despite appropriate conservative management (Slenker et al., 2015).
That's not really the target of this argument.
The target is the casual assumption that the long head of the biceps is a disposable pain generator and relocating it is a mechanically trivial solution.
It isn't.
The LHBT can contribute to glenohumeral stability (Pagnani et al., 1996; Alexander et al., 2013).
Its native mechanical environment is altered by tenodesis.
The procedure creates a new tendon-to-bone interface.
The patient must temporarily protect and subsequently reload that interface.
Mechanical fixation failure appears relatively uncommon, but persistent anterior shoulder/bicipital pain is very real (McCrum et al., 2019; Ergün et al., 2022).
And fixation location may influence revision and complication patterns even if we don't yet completely understand why (McCrum et al., 2019; Ergün et al., 2022).
That's a lot of biological disruption for a structure we had better be pretty damn sure was causing the problem.
The Bottom Line
Maybe biceps tenodesis is exactly what a particular shoulder needs.
But it should have to earn its way into that shoulder.
Before giving every painful biceps tendon a new home, make the diagnosis earn the surgery.
Because surgery doesn't simply remove a painful structure.
It trades one mechanical situation for another.
We remove the tendon from its native attachment.
We eliminate at least some of its potential contribution to glenohumeral stability (Pagnani et al., 1996; Alexander et al., 2013).
We create a new fixation site.
We create a new healing interface.
We temporarily decrease loading capacity.
We introduce new potential sources of postoperative pain (McCrum et al., 2019; Ergün et al., 2022).
And then we judge success partly by whether the tendon stayed where we screwed it.
I'd argue the standard should be higher.
The anchor can hold.
The tendon can heal.
The MRI can look great.
And the patient can still hurt.
That's not necessarily a successful shoulder.
Sometimes it's just a successful fixation.
-the pissed-off PT- share if you care, or don’t-
References
Alexander S, Southgate DFL, Bull AMJ, Wallace AL. The role of negative intraarticular pressure and the long head of biceps tendon on passive stability of the glenohumeral joint. J Shoulder Elbow Surg. 2013;22(1):94-101.
Pagnani MJ, Deng XH, Warren RF, Torzilli PA, O'Brien SJ. Role of the long head of the biceps brachii in glenohumeral stability: a biomechanical study in cadavera. J Shoulder Elbow Surg. 1996.
The long head of the biceps has a stabilizing effect on the glenohumeral joint in simulated infraspinatus or subscapularis but not supraspinatus rotator cuff deficiency: a biomechanical study. Arthroscopy. 2023.
McCrum CL, et al. Complications of biceps tenodesis based on location, fixation, and indication: a review of 1,526 shoulders. J Shoulder Elbow Surg. 2019.
Ergün S, et al. Clinical outcome comparison of suprapectoral and subpectoral tenodesis of the long head of the biceps with concomitant rotator cuff repair: a systematic review. Shoulder & Elbow. 2022.
Slenker NR, Lawson K, Ciccotti MG, Dodson CC, Cohen SB. Surgical indications for long head biceps tenodesis: a systematic review. Clin Orthop Relat Res. 2015.