Background

Osteochondritis dissecans is an injury of subchondral bone involving fracture and ischemia [1]. This disorder has biological risk factors such as genetics, ossification center defects, and endocrine abnormalities (e.g., vitamin D deficiency) [1]. Mechanical risk factors include repetitive microtrauma, tibial spine impingement, and altered biomechanics [1]. This lesion is classified as the juvenile form of osteochondritis dissecans when it presents in a femur with open distal physes [2]. Knee juvenile osteochondritis dissecans (JOCD) is most common between the ages of 11 and 15 [3] with an incidence during this age of 39 per 100,000 for males and 16 per 100,000 for females [3].

The management of knee JOCD depends, in part, on it being stable or unstable [2, 4]. Patients with unstable lesions have a worse prognosis with non-operative treatment and are more likely to undergo surgery [5, 6]. Signs of instability include a line of high signal intensity between the lesion and adjacent bone on T2-weighted MRI, a focal osteochondral defect, an articular fracture in the subchondral bone plate [6], and the presence of multiple CLL or a large CLL > 5 mm in diameter, which have a high specificity (~ 100%) for instability [7].

The healing probability of knee JOCD is individualized as it depends on clinical and radiological variables. Estimates of what percentage of cases heal vary from 30 to 95% over 6–48 months [5, 6, 8,9,10,11,12,13,14]. Variables associated with a greater probability of healing include a smaller lesion size [6, 11, 12, 14], age < 16 [6, 10, 11], absence of mechanical symptoms (e.g., giving-way, swelling, locking, or clicking) [12], median CLL size < 1.3 mm on MRI [11], and time to consultation < 6 months [13].

Two published nomograms can be used to predict healing of knee JOCD. Both use clinical and radiological variables to estimate the probability of imaging signs of healing at 6 months. Krause et al.’s nomogram uses age, normalized lesion width, and median CLL size, which has the greatest validity of any marker predicting healing of knee JOCD [11]. Wall’s nomogram uses normalized lesion width (Fig. 1), length, and symptom type [12].

Fig. 1
figure 1

Normalized lesion width calculation. a The maximal medial to lateral dimension in millimeters of the OCD lesion on a coronal MRI was divided by b the maximal medial to the lateral dimension of the femoral condyles on a coronal MRI according to the method of Krause et al. [11], which gave a value of 18%

Full size image

This case illustrates that a non-operative approach for knee JOCD in an athlete can be supported by serial improvements on MRI. Currently, there is no standard rehabilitation protocol for JOCD [2], and activity restriction prior to follow-up imaging is common [6, 11,12,13]. In this case, innovative rehabilitation and treatment strategies enabled a return to play despite a low initial healing probability.

Case Presentation

A 15-year-old soccer player presented with a 1-month history of gradually worsening left knee pain, aggravated by a recent knee-to-knee collision. He had prior episodes of mild bilateral anterior knee pain during soccer, diagnosed as patellar tendinitis. There was no family history of JOCD. Examination of the left knee revealed infrapatellar swelling and tenderness, pain with inferior patellar glide, and passive knee range of motion, but retained ability to achieve full range of motion. Knee radiography showed a subchondral lucency in the left lateral femoral condyle, and an MRI confirmed the presence of a JOCD lesion in that region. The lesion had five CLL, the largest of which was 5.9 mm in diameter. The patient’s orthopedist referred him for conservative therapy with the intention of having a follow-up MRI in 3 months.

Two weeks after the left knee JOCD diagnosis, blood flow restriction training (BFRT) was initiated twice per week and was continued until 31 weeks. An 11.5-cm-wide pneumatic cuff device (Delfi Medical Innovations) was applied to the patient’s proximal left thigh, which automatically calculated the limb occlusion pressure (LOP). Pressures from 128 to 179 mmHg, which equaled 80% of the LOP, were used while the patient exercised. The patient rode an exercise bike for 10 min then performed five exercises that varied depending on the day for another 10 min. Initially, these consisted of non-weight-bearing, low-intensity exercises such as clam shells and active straight leg raises. Between appointments, the patient performed BFRT at home for two 10-min sessions on a stationary bicycle with a 7.5-cm-wide elastic wrap around the proximal thigh (BFR Bands®) each day.

Three weeks after the left knee JOCD diagnosis, near-infrared light and nutritional therapies were initiated. A Class IV laser (Diowave™) was applied to the left knee at 45 W for 100 s, twice per week. Between appointments, the patient used a near-infrared bulb (RubyLux NIR-A Therabulb) for 30 min, 6″ from the knee at home daily. The patient complied with advice to take orally 2000 IU vitamin D3 and 100 micrograms vitamin K2 daily.

Seven weeks after the left knee JOCD diagnosis, the patient’s reduction in pain to a mild level and absence of swelling supported a progression to low-impact weight-bearing exercises including step-downs, step-ups, monster walks, and body weight split squats. At 12 weeks, the normalization of outcome assessments supported a progression to sled pushes and pulls.

During the 14th week after left knee JOCD diagnosis, the patient noted a gradual re-aggravation of right anterior knee pain during exercises, of which he had similar episodes 1 year prior. Examination revealed tenderness of the tibial tuberosity. Radiography demonstrated an ossicle at the distal patellar tendon consistent with Osgood-Schlatter disease and a subchondral plate irregularity at the lateral femoral trochlea. Magnetic resonance imaging showed a small stable osteochondral lesion 9 mm in diameter without CLL. Because the patient was restricted from sports due to his left knee JOCD, he elected to have the ossicle related to his Osgood-Schlatter removed from the right knee. Surgery was completed 16 weeks after his left knee JOCD diagnosis. Right knee symptoms related to post-surgical healing subsided over another 16 weeks.

From 16 to 20 weeks after the left knee JOCD diagnosis, exercises regressed to non-weight-bearing due to demands of post-surgical recovery of the right knee. At week 20, step-ups, step-downs, and monster walks were re-introduced. At week 22, pulsed electromagnetic field (PEMF) therapy (Pulse XL Pro™) was applied to both knees for 30 min at 7.8 Hz, once per week for 6 weeks. Following a reduction in post-surgical pain and swelling, the patient was progressed at week 23 to walking lunges and running on an anti-gravity treadmill (AlterG®) at 60% bodyweight.

Discussion

The patient showed clinical and symptomatic improvement within the first 6 weeks of treatment. His Activity Measure for Post-Acute Care improved from an initial 36 to a maximum of 39 at 11 weeks after left knee JOCD diagnosis and his left Knee Outcome Survey (KOS) increased from an initial 46 to 100% at 12 weeks (Fig. 2). He progressed through rehabilitation without re-aggravating the left knee.

Fig. 2
figure 2

Knee Outcome Survey (KOS) and Activity Measure for Post-Acute Care (AM-PAC) scores during treatment. At 22 weeks, the patient’s AM-PAC score dropped due to having surgery for the R knee Osgood-Schlatter lesion (OSD)

Full size image

Clinical, symptomatic, and radiological signs of improvement supported a gradual return to playing soccer. At 28 weeks after the left knee JOCD diagnosis, the patient began light passing and kicking of the soccer ball and running 2 miles twice per week at 70% body weight on the AlterG®. He was cleared for soccer drills with a trainer at 30 weeks. The patient was fully cleared for soccer matches at 36 weeks. At 4-month follow-up the patient did not have a recurrence of knee symptoms. The patient’s outcome and short healing time frame were considered optimal given the low initial probability of healing (20%).

Serial MRI showed progressive healing throughout treatment (Fig. 3). The number and size of CLL reduced from 5, with a median diameter of 3.7 mm at the initial MRI, to 2 with a median diameter of 3.1 mm at the 3-month MRI. At the 6-month MRI, the lesion had one 1.8-mm CLL. The final MRI at 8 months revealed no CLL.

Fig. 3
figure 3

Signs of healing of the left knee JOCD lesion on serial MRI. a Initial MRI, b 3-month, c 6-month, and d 8-month follow-up MRIs. Cyst-like lesion (*) number and size reduced the most from a to b, filling in of the bony defect (#) was most apparent between b and c, and obscuring of the lesion-parent bone interface (arrowheads) was most apparent from c to d

Full size image

In our case, a short follow-up MRI at 3 months helped determine if the knee JOCD lesion was worsening or healing. The presence of multiple CLL with a maximum diameter > 5 mm indicated instability was likely. No change or worsening of imaging findings such as increased lesion size or CLL size or number would have been indications for surgery while improvement supported ongoing conservative treatment. There is no nomogram score that absolutely indicates surgery however a score of < 48% has a positive predictive value of 65% for predicting the need for surgery [11].

We used Krause et al.’s nomogram to predict healing probability because it accounted for signs of instability such as multiple and large CLL (Fig. 4). Cyst-like lesion size has the best validity of any healing marker for knee JOCD [11] (Fig. 5). According to Krause et al.’s nomogram, the probability of our patient’s JOCD healing increased from a baseline of 20 to 26%, 40%, and 70% at the 3, 6, and 8-month follow-up MRIs, respectively. We did not track the coalescence of bone, which is not easily measured, or changes on radiography, which lack standardization [16].

Fig. 4
figure 4

Improvements in healing probability from the a initial MRI to follow-up MRI at b 3 months, c 6 months, and d 8 months as predicted by the Krause et al.’s nomogram [11]. The normalized lesion width remains a constant 18% in nomogram a, b, and c, and decreases to 15% in nomogram d. The median CLL size reduces from a 3.7 to b 3.1, to c 1.8, to d 0. Reprinted with permission from Krause et al. viaRightsLink—the order license number is 4746220915007

Full size image
Fig. 5
figure 5

Reduction in cyst-like lesion size (the most valid measure of healing) and amount from the initial MRI (a, b, c) to the 3-month follow-up MRI (d, e, f). All images are T2-weighted MRIs. Coronal slices b and c correspond with the left and right dashed lines in image a. Coronal slices e and f correspond with the left and right dashed lines in image d. Measurements are taken at the maximum CLL diameter in any imaging plane [7, 15]

Full size image

Previous authors have defined healing of JOCD using clinical and/or radiological features. Clinical signs include an absence of pain or functional limitations [6] and the ability to return to sport without recurrence of symptoms [14]. Radiographic signs include a resolution of the boundary between the JOCD lesion and parent bone and increasing radiodensity of the lesion [17]. Magnetic resonance imaging signs include coalescence [8], also called consolidation [10] or fusion [9] of viable bone, return of articular cartilage to normal width and intensity [8], increased T2 signal of the JOCD lesion from low to intermediate [9], obscuring of the lesion-bone interface [9], and reduction of lesion size by 15% [11].

Aside from restriction of strenuous physical activity [2], there is no standardized protocol for conservative management of JOCD of the knee in the published literature [2]. Evidence suggests repetitive stress with activities like running and jumping should be avoided but quadriceps strengthening is beneficial [2]. Patients may continue to perform their daily activities however total rest and immobilization are unnecessary [2]. In our case, we followed these general guidelines with emphasis placed on quadriceps strengthening.

Blood flow restriction training is a method of exercising with a cuff or wrap around a proximal extremity which restricts venous return. This technique allows for rehabilitation using low loads by making exercise more metabolically demanding. It may even increase bone turnover and benefit bone health [18]. Evidence is limited to one case that showed improvements in an adult OCD lesion on radiography after 6 weeks of BFRT [19].

We recommended vitamins D3 and K2 to aid in bone ossification. Although no studies have examined the treatment of JOCD using D3 and K2, this combination has been found to safely increase bone density in postmenopausal women [20]. One study that identified vitamin D deficiency in a majority of JOCD patients suggested early D3 supplementation could prevent the worsening of an OCD lesion [21]. Because we were unable to obtain a baseline serum level of 25-hydroxy vitamin D, we used a dose considered to be safe.

Near-infrared light therapy was used to accelerate the healing of the left knee JOCD using a laser in the clinic and bulb for home use. Near-infrared light may accelerate the healing of bone injuries by stimulating osteoblasts [22]. Evidence is limited to animal studies, such as one animal study that found laser therapy improved signs of healing of osteochondral defects [23].

Pulsed electromagnetic field therapy was used to accelerate the healing of the left knee JOCD and post-surgical recovery of the right knee. Animal studies have shown that PEMF therapy increases mineralization and bone callus density following injury [24]. Evidence in humans is limited to one trial that found PEMF reduced pain and improved function of patients who had surgery for ankle osteochondral lesions [25].

Another treatment option for knee OCD is surgery. The most common indication for surgery is a failure of 6 months of non-operative treatment [4]. Other indications include a lack of radiographic improvement and lesion instability [4]. The surgical techniques most frequently used for knee JOCD are transarticular drilling for stable JOCD and bioabsorbable pin fixation for unstable JOCD [26].

One review found that surgery had a success rate of over 90% after 38 to 95 months [4]. However, there is a lack of data regarding post-surgical outcomes of pediatric patients once they reach skeletal maturity years later [4, 26]. Complications of surgical intervention include post-operative pain during sport [26] and damage to articular cartilage [4].

Other therapies are emerging for knee JOCD but are limited to case reports. These include stem cell injections [27], platelet-rich plasma injections [28], and extracorporeal shock wave therapy [2]. Future research should continue to evaluate if changes in short-term follow-up MRI and the corresponding nomogram score help decision-making for knee JOCD. Cohort studies or clinical trials could also examine if BFRT, near-infrared light, vitamins D3 and K2, and PEMF accelerate healing of JOCD in comparison with activity restriction alone.

One limitation in this case is any confounding effect of a 4-week period of activity restriction in the middle of the course of care, due to an unrelated contralateral knee surgery. This rest period could have incidentally aided in the healing of the left knee JOCD. Another limitation is the patient did not undergo arthroscopy, which could have confirmed or refuted the imaging signs of JOCD instability seen on MRI. Being a case report, these results may not be generalizable to a larger population.

Conclusions

An athlete with radiographically unstable knee JOCD was monitored carefully with follow-up MRIs during a course of conservative care until a successful return to play. The high resolution of MRI displays the pathology and interval changes which enables measurement of the progression of JOCD. Objective measurements with high validity such as CLL size may be tracked during conservative care to recalculate healing probability. Improvements in healing probability support continued non-operative management. It is unclear if conservative therapies such as BFRT, near-infrared light, vitamins D3 and K2, and PEMF accelerate healing compared with activity restriction alone, and these therapies may deserve further study in JOCD populations.