Postoperative loading-progression principles shared across microfracture, ACI/MACI, and osteochondral grafting — matching mechanical load to the biology of immature repair tissue
Whether the surgical technique is marrow-stimulation, cell-based grafting, or osteochondral transplantation, every cartilage restoration procedure creates the same early biological problem: a repair site filled with fragile, unorganized tissue that cannot yet tolerate the loads the native joint surface once handled. Rehabilitation begins by reconciling two competing needs — protecting that tissue from destructive shear and compressive load, while still delivering the motion and nutrient exchange it needs to organize into a durable repair.
In the first weeks after any cartilage restoration procedure, the defect is occupied by tissue in transition — a fibrin clot seeded with marrow-derived mesenchymal cells (microfracture), a periosteal- or membrane-covered chondrocyte suspension still adhering and proliferating (ACI/MACI), or a graft-host interface undergoing early bony and cartilage integration (osteochondral transplantation). None of these states resemble mature hyaline cartilage mechanically.
Native articular cartilage is a densely cross-linked type II collagen network reinforced by proteoglycan aggregates that resist compressive and shear stress built up over years of matrix turnover. Early repair tissue has almost none of that organization: collagen fibrils are sparse and randomly oriented, proteoglycan content is low, and the tissue-bone or tissue-defect interface has not yet achieved biomechanical integration. Loading this tissue as if it were mature cartilage — through unrestricted weight-bearing, pivoting, or deep flexion under load — generates shear forces that can shear the clot from the subchondral bone, disrupt cell attachment, or crush a graft before osseous incorporation is complete.
The protection phase is therefore not simply "resting" the joint; it is sequencing load exposure to match the mechanical competence of tissue that is changing week to week.
Because articular cartilage has no direct blood supply, its cells depend entirely on diffusion through synovial fluid for oxygen, nutrients, and growth factor exchange. Static, non-moving joints do not effectively circulate that fluid across the cartilage surface — motion is the pump that drives nutrient delivery to avascular tissue.
CPM devices move the joint passively through a prescribed, gradually increasing arc, without requiring the patient's own muscle contraction and without axial loading. This achieves two things simultaneously:
• Mechanical nutrient circulation — cyclic joint motion generates fluid flow across the repair surface, analogous to how a sponge exchanges fluid when gently flexed, supporting the metabolic needs of the proliferating repair cells • Guided matrix organization — in the near-total absence of a functioning blood supply, mechanical signals from motion appear to help orient collagen fibril deposition along physiologically relevant lines, rather than the disorganized weave that forms under immobilization
CPM protocols typically begin within the first 1–2 postoperative days and continue for several hours daily through the initial weeks, with the motion arc expanded incrementally as the tissue and surgical fixation tolerate it. Motion is prescribed; load is not — the joint moves, but the limb does not bear weight through it.
Weight-bearing status during the protection phase is never one-size-fits-all — it is set by where the defect sits on the joint surface and which procedure created the repair.
Defect location: a defect on a primarily weight-bearing surface (e.g., the femoral condyle in the loaded arc of motion, or the tibial plateau) requires substantially more restriction than a defect on a non- or partially weight-bearing surface (e.g., the trochlea or patella, where compressive contact is more flexion-angle-dependent and can sometimes be partially offloaded through range-of-motion selection rather than crutches alone).
Procedure type: marrow-stimulation techniques create a mechanically simpler repair-bone interface and are sometimes allowed slightly earlier partial weight-bearing; cell-based chondral resurfacing techniques depend on a still-adhering, still-proliferating cell layer with no bony fixation and are typically managed most conservatively; osteochondral grafts must additionally protect a bone-to-bone interface undergoing osseous union before compressive load is safe, which follows its own healing timeline distinct from the cartilage cap above it.
Across all three, the shared principle is the same even though the numbers differ: weight-bearing progression is gated by biology, not by a calendar alone, and premature loading anywhere along that spectrum risks displacing or delaminating repair tissue that has not yet achieved mechanical competence.
The protection phase is not simply immobilization — it deliberately combines controlled passive motion (to feed an avascular tissue that depends on diffusion) with controlled load restriction (to prevent shear on a repair with roughly one-tenth the strength of mature cartilage). Getting either half wrong — too little motion, or too much load — degrades the same outcome from two different directions.
Somewhere between week 6 and week 12, every cartilage restoration protocol turns from protecting the repair to progressively challenging it. This is deliberately gradual, because it overlaps with a phase of tissue biology that is still immature and still remodeling — the same window where under-loading and over-loading both carry real risk, in opposite directions.
Progression off crutches and assistive devices is staged, not switched on all at once. A typical ramp increases weight-bearing allowance in graduated increments — partial, then progressive-partial, then full — over roughly four to eight weeks, guided by procedure-specific protocols rather than a universal calendar.
Bone-based repairs, such as osteochondral grafting, are additionally paced by radiographic and clinical signs of osseous incorporation at the graft-host bone interface; once that bony union is progressing, axial load can be advanced with more confidence than for a repair that depends entirely on soft-tissue and cellular integration. Pure chondral-surface techniques, by contrast, have no bony checkpoint to reference — the surgeon and therapist are progressing load based on time-in-tissue and clinical response alone, which is part of why chondrocyte-based protocols tend to run more conservatively through this window.
In all cases, the ramp is monitored for a give-away sign: increased effusion, pain, or a sense of instability after a loading advancement is treated as a signal to slow the ramp, not push through it.
Once partial weight-bearing begins, therapy introduces closed-chain kinetic exercises — movements where the foot (or hand, for upper-extremity joints) stays fixed against a surface while the rest of the limb moves, such as partial-range leg presses, mini-squats, or stationary cycling with limited resistance.
Closed-chain exercise is favored over open-chain loading (e.g., resisted knee extension against a free-moving lever arm) in this window because it better replicates the compressive, joint-congruent loading pattern the repair will eventually need to tolerate functionally, while allowing the range and resistance to be tightly controlled. Exercises are typically restricted to ranges and loads that keep peak contact stress on the repair site low — for example, limiting knee flexion depth in early closed-chain work when the defect sits in a portion of the joint arc that would otherwise be loaded heavily at deeper flexion angles.
The exercise dose is escalated incrementally: range of motion under load, then resistance, then repetition volume — never all three at once.
Weeks 6–12 sit squarely inside a period where repair tissue is transitioning from a soft, cellular, proteoglycan-poor matrix toward a firmer, more organized one — but has not arrived there yet. This creates a genuinely narrow therapeutic window, with meaningful risk on both sides of it:
• Under-loading — keeping the tissue protected for too long, or advancing exercise too conservatively — risks disuse changes: progressive muscle atrophy, joint stiffness, and a repair matrix that lacks the mechanical stimulation needed to organize its collagen architecture along functionally appropriate lines. Cartilage- and cartilage-adjacent tissue responds to physiological loading as a maturation signal; too little of it can leave the repair mechanically inferior even if it is never overtly damaged.
• Over-loading — advancing weight-bearing or exercise faster than the tissue's maturation curve — risks shear or compressive forces exceeding what the immature repair can withstand, causing delamination (separation of the repair tissue from the underlying bone) or outright mechanical failure of the graft or cell layer.
Because the tissue's mechanical competence is changing continuously through this window rather than crossing a single threshold, load progression is deliberately graded and monitored — a slow ramp rather than a step function — precisely to stay inside that narrow band between the two failure modes.
The weeks 6–12 window is where "protocol" and "clinical judgment" intersect most tightly. A generic calendar-based progression ignores that under-loading and over-loading are both real risks at every point along this ramp — which is why advancement is typically gated by clinical response (effusion, pain, functional signs) as well as elapsed time.
By month three, most repairs have progressed to full weight-bearing and the rehabilitation focus shifts from protecting tissue to rebuilding everything that atrophied or deconditioned while it was being protected — quadriceps (or the relevant muscle group) strength, joint position sense, and the movement patterns needed for daily function and eventually sport.
A well-documented consequence of the protection phase — restricted weight-bearing, limited resistance exercise, and often reflexive inhibition from joint effusion and surgical trauma — is measurable weakness in the muscles crossing the treated joint, most classically the quadriceps in knee cartilage procedures. This is not simply disuse atrophy; effusion and joint pathology can produce arthrogenic muscle inhibition, a reflexive neural suppression of muscle activation that persists even after strength training resumes, independent of true fiber loss.
Addressing this requires more than resistance exercise alone. Neuromuscular re-education techniques — including neuromuscular electrical stimulation, biofeedback-guided activation drills, and careful attention to activating the muscle through its full available range rather than compensating with other muscle groups — are used alongside progressive resistance training to restore both muscle bulk and the nervous system's ability to recruit it effectively.
Joint position sense — the ability to sense where a limb is in space without looking at it — is degraded by the surgery itself, by the period of restricted loading, and by any associated joint effusion. Because proprioceptive input is a key contributor to dynamic joint stability during functional movement, this phase incorporates single-limb balance work, progressively destabilized surfaces, and perturbation training that challenges the neuromuscular system to react to unexpected joint loading — training the reflexive stabilization response that protects the repair during real-world movement, not just planned exercise.
Resistance training is advanced along standard progressive-overload principles — increasing load, range, and complexity in graduated steps — but layered onto the joint-specific loading tolerances established during the earlier phases. By the later part of this window, exercise progresses from purely rehabilitative movements toward sport- or activity-specific movement patterns: multi-directional stepping, controlled deceleration drills, and functional strength work that begins to resemble the demands of the patient's eventual activity goals, still performed at controlled intensity well below full sporting effort.
Gait is also formally re-assessed and normalized during this window — many patients develop compensatory movement patterns during the protected phase that, left unaddressed, can persist as habit long after the biological restriction that caused them has resolved.
Running, cutting, and other impact-loading activities are introduced only once objective strength and functional milestones are met — and the point at which that becomes appropriate is where the different cartilage restoration procedures' biology visibly diverges, because they mature at different rates and toward different long-term durability ceilings.
Impact activities multiply joint contact forces well beyond walking or controlled closed-chain exercise, so their introduction is gated by demonstrated readiness rather than elapsed time alone: adequate strength symmetry between limbs, pain-free full range of motion, absence of effusion with activity, and confirmed tolerance of progressively loaded functional exercise. Programs typically begin with a structured run-walk progression on forgiving surfaces before advancing toward unrestricted running, and further still before introducing cutting, pivoting, or jumping loads.
The three major cartilage restoration approaches do not mature at the same rate, and rehabilitation timelines are adjusted accordingly:
• Microfracture and marrow-stimulation techniques recruit marrow-derived mesenchymal cells that differentiate into a fibrocartilage-predominant repair. This tissue tends to fill in and gain mechanical competence relatively quickly in the first several months, which can allow comparatively earlier progression toward light impact activity in some patients — but fibrocartilage is biomechanically inferior to native hyaline cartilage, with different collagen organization and lower resistance to long-term wear.
• ACI/MACI (cell-based chondrocyte implantation) relies on implanted chondrocytes proliferating, producing matrix, and gradually remodeling toward a more hyaline-like tissue over a much longer time course — commonly cited as continuing to mature well beyond the first year. Because that maturation is slower, protocols are typically more conservative about clearing higher-impact loading in the 6–9 month window, even though the tissue's eventual quality ceiling is generally considered higher.
• Osteochondral grafting transplants a block of native, already-mature hyaline cartilage together with its subchondral bone. The cartilage itself does not need to newly differentiate and organize the way a marrow-stimulation or cell-based repair does — but the graft-host bone interface must achieve osseous union, and the cartilage-cartilage interface at the graft margin must integrate, both of which set their own pacing for safe loading.
It is tempting to treat the technique that allows the earliest return to impact activity as the superior option, but rehabilitation timing and long-term repair durability are separate questions. A repair that matures quickly to a fibrocartilage-dominant tissue may tolerate earlier loading progression while carrying a lower durability ceiling over years of use; a repair that matures slowly toward a more hyaline-like tissue may require more patience early on in exchange for a potentially more durable long-term outcome. Rehabilitation timelines are calibrated to the specific biology of the procedure performed — copying a faster protocol onto a slower-maturing repair does not accelerate the underlying tissue biology, it simply removes the safety margin.
"Faster" and "better" are not synonyms in cartilage repair rehabilitation. The procedure that gets a patient back to jogging soonest is not automatically the one that will hold up best a decade later — timelines are a reflection of each technique's underlying tissue biology, not a ranking of overall quality.
Full sport clearance is the most consequential decision point in the entire rehabilitation timeline, and cartilage restoration procedures apply a longer, more conservative timeline than most other orthopedic surgeries — followed by an ongoing relationship with load management that, for many patients, never fully ends.
Rather than clearing return to sport by calendar date alone, cartilage repair protocols increasingly borrow the objective testing philosophy used in ACL reconstruction return-to-sport (RTS) decision-making: quantified strength symmetry between limbs, a battery of hop tests (single-leg hop for distance, triple hop, crossover hop, and timed hop, each compared side-to-side), and functional movement quality assessment under fatigue. A common benchmark is requiring involved-limb performance to reach roughly 90% or more of the uninvolved limb across these measures before clearing higher-demand sport participation, rather than relying on subjective readiness alone.
Patients and even some clinicians familiar with ACL reconstruction sometimes expect a similar recovery arc for cartilage restoration — but the underlying biology is fundamentally different. Ligament graft healing and remodeling, while itself a multi-month process, involves a single tissue type revascularizing and remodeling along a comparatively well-characterized timeline, commonly supporting return to sport discussions around six to nine months.
Cartilage repair tissue, by contrast, is avascular, depends entirely on diffusion for nutrition, and — depending on procedure — may need to differentiate, proliferate, deposit new matrix, and remodel that matrix toward a mechanically competent architecture from a near-baseline starting point. That process simply takes longer at the tissue level, which is why cartilage restoration return-to-sport timelines commonly extend to nine to twelve months at the earliest, and longer for procedures with slower-maturing repair tissue — a direct reflection of slower underlying biology, not overcaution for its own sake.
Clearance for return to sport is not the end of the rehabilitation relationship — it is the transition into long-term load management. Even a well-matured, hyaline-like repair from cell-based or osteochondral techniques may remain biomechanically inferior to untouched native cartilage indefinitely: subtly different collagen fiber architecture, proteoglycan content, or integration at the repair margin can leave the site somewhat more vulnerable to wear under repetitive high loading than the surrounding native joint surface.
For this reason, patients are typically counseled on ongoing activity modification long after formal rehabilitation ends — awareness of cumulative loading, appropriate cross-training to distribute joint stress across activities rather than concentrating it, weight management where relevant, and monitoring for symptoms that might signal repair-site stress. The goal of rehabilitation was never simply to get the patient back to activity as fast as possible; it was to build and then preserve a repair that can continue functioning for years to come.
Return-to-sport clearance answers the question "is this repair strong enough to load now?" — it does not answer "will this repair last forever under that load?" Long-term durability counseling treats cartilage repair as a structure to be maintained over a lifetime, not a problem that is solved on the day of clearance.