Pain, whether it is low back pain, joint, pain post-traumatic or post surgical pain, is collectively now the number one clinical problem in the world, and yet, current imaging methods to correctly identify pain generators remain woefully inaccurate. The lack of reliable diagnostic tools leads to misdiagnosis, mismanagement, reliance on the use of opioids, unhelpful and often unnecessary surgeries, and ultimately, therapeutic failures.
Relatively recent developments in clinical molecular imaging (MI) afford the opportunity to pinpoint the exact site(s) of pain generation due to advances in biomarker discovery, imaging technology and radiotracer design. Scientists and clinicians at Stanford University are attempting to develop better approaches to identify pain generators by employing techniques in clinical molecular imaging using PET/MR.
Case 1
Patient history
A 20-year-old male volleyball athlete with left posterior knee pain and unable to play the sport. The patient’s knee pain has been extensively worked up with various studies including MR, electromyography/nerve conduction study, vascular and multiple diagnostic injections. Etiology of the pain remained unclear.
The patient received an ultrasound-guided injection into the popliteus tendon sheath with 4 mL of 0.25% bupivacaine and 4 mg of dexamethasone. He noted that this injection provided him with roughly 80% relief. A repeat injection was performed under ultrasound guidance, where 7 mL of ropivacaine 0.5% and 40 mg of Kenalog were injected in the popliteal tendon sheath; however, the patient reported that this injection gave him only about 10% relief.
Assessment/plan: The patient’s posterior left knee pain continued to have an unclear etiology. He has had quite discordant results from two separate injections into the popliteal tendon sheath.
In order to clarify a diagnosis, it was recommended that he return to the first doctor for a repeat injection of 3 cc lidocaine plus 4 cc ropivacaine (without corticosteroid, given that he had multiple corticosteroid injections) for diagnostic purposes. The hope was that returning the patient to the practitioner who provided the initial injection that gave him substantial relief for a repeat identical injection would hopefully allow us to identify the anatomic source of his pain.
The follow-up injection was not as successful as the first, and the patient presented for FDG PET/MR.
Findings
MR neurography suggested presence of a small neuroma, nerve injury (neuroma in situ), perineurioma or neuropathy in the common peroneal nerve (CPN) at the level of the distal femoral metadiaphysis. Axial DESS (2nd echo)‡ MR showed small focal enlargement and increased signal in the CPN (Figure 1, red arrow) compared to the tibial nerve that appears normal (Figure 1, green arrow).
Figure 1.
MR neurography from PET/MR exam. Axial DESS (2nd echo)‡ MR demonstrates increased signal in the CPN (red arrow). By comparison the tibial nerve appears normal (green arrow).
Axial PET/MR with LAVA Flex (Figure 2A-C) shows abnormal focal FDG uptake in the left peroneal muscle compartment (SUVmax = 1.7, red arrows) and reconstructed coronal LAVA Flex (Figure 2D-E) shows abnormal linear but focal FDG uptake in the left peroneal muscle compartment (SUVmax = 1.7, red arrows). However, it is difficult to discern in either plane if this is in a small nerve or muscle-related neuropathic change. FDG uptake in gastrocnemius muscles is likely normal physiologic uptake. No MR abnormality is seen in the peroneal muscle.
Figure 2.
(A-C) Axial PET/MR with LAVA Flex demonstrates abnormal focal FDG uptake in the left peroneal muscle compartment (SUVmax = 1.7, red arrows). (D, E) Reconstructed coronal PET/MR with LAVA Flex shows abnormal linear but focal FDG uptake in the left peroneal muscle compartment (SUVmax = 1.7, red arrows). (A,D) PET, (B) fused PET/MR and (C, E) MR images.
Normal FDG uptake and grossly normal MR of the left knee joint shows no evidence of inflammation or significant degeneration of the knee joint itself (Figure 3).
Figure 3.
Normal FDG uptake and grossly normal MR of the left knee joint with no evidence of inflammation or significant degeneration of the knee joint itself. (A) PET, (B) PET/MR and (C) MR images.
Normal FDG uptake and grossly normal MR of left popliteus tendon was noted (Figure 4, green arrow). This structure was previously thought to be the culprit, however, there is again no evidence of inflammation or significant degeneration of the knee joint itself.
Figure 4.
Normal FDG uptake and grossly normal MR of left popliteus tendon (red arrow). This structure was previously thought to be the culprit. (A) PET and (B) MR images.
MR neurography suggested focal swelling in the CPN with differential diagnosis including nerve injury, possible small schwannoma, neuroma-in-situ, perineurioma or focal neuropathy. Also, abnormal FDG uptake in the CPN, centered approximately 13 cm distal to the knee joint line, may be due to neuropathic change, muscle strain or other nerverelated changes.
Results
PET/MR data gave the surgeon confidence to explore the CPN, revealing a focal area of injury (Figure 5). The tissue around the CPN was "released" and the patient ultimately returned to normal activities after the surgery.
Figure 5.
Surgical exploration and release of tissue around the CPN resolved the patient’s pain.
Case 2
Patient history
A 37-year-old active and athletic male with remote history of ACL reconstruction and more recent meniscal repair and synovectomy seven months prior to PET/MR. Patient is experiencing severe, persistent left anterior medial knee pain after ACL repair and is unable to exercise; reports pain is 8-10/10. Surgeon refused to re-operate, stating that nothing is abnormal and suggests the pain is "in the patient’s head."
Findings
PET/MR at baseline shows sigma-1 receptor (S1R)-avid fibrosis and synovitis in anterior knee (Figure 6A-B). Another surgeon agrees to perform surgery following initial PET/MR, using the PET/MR images to guide resection of the offending tissue.
Figure 6.
(A, B) PET/MR at baseline shows S1R-avid fibrosis and synovitis in anterior knee. (C, D) Post-surgical PET/MR shows successful removal of offending tissue.
Results
Synovectomy and removal of painful fibrosis results in complete pain relief (0/10 pain) and return to normal athletic activity. Post-surgical PET/MR shows successful removal of offending tissue (Figure 6C-D).
Case 3
Patient history
A 22-year-old male professional soccer player with severe bilateral foreleg-thigh-hip pain. No longer plays soccer due to significant difficulty in ambulating and nearly wheelchair bound.
All workup has been negative including MR. Bilateral fasciotomies of the foreleg have been unsuccessful in alleviating pain. Patient underwent PET/MR and results were compared to an asymptomatic volunteer.
Findings
In the coronal PET/MR acquisition (Figure 7), mild-tomoderate increase in uptake in both quadricep muscles (white arrow). Increased uptake in bilateral anterior foreleg muscle compartments, especially in the extensor digitorum longus and extensor hallucis longus muscles (SUVmax = 5.4, right foreleg, and 3.6, left foreleg, orange arrows). Increased uptake in proximal left tibiofibular joint (SUVmax = 1.5, green arrow). Increased uptake in left popliteus muscle (SUVmax = 2.7, red arrow) and increased uptake in the flexor digitorum tendons (not shown) was noted.
Figure 7.
PET summary view of both legs and knees showing increased uptake in the quadricep muscles (white arrow), bilateral anterior foreleg muscle compartments (orange arrows), tibiofibular joint (green arrow), left popliteus muscle (red arrow) and flexor digitorum tendons (not shown).
In the axial PET/MR acquisition (Figure 8), increased uptake in bilateral anterior compartments of both foreleg muscles (SUVmax = 3.2, right, and SUVmax = 3.5, left, Figure 8A-B, yellow arrows) with the left mildly greater than right and especially pronounced in the extensor digitorum longus muscle (Figure 8C-D, red arrows). Background SUVmax = 0.2-0.5.
Figure 8.
Increased uptake in the bilateral interior compartment of both foreleg muscles (yellow arrows) with the left mildly greater than the right, especially in the extensor digitorum longus (red arrows). (A, B) PET, (C, D) MR and (E, F) PET/MR images.
Results
Patient was treated with Xeomin® (Botox®-like injection) and platelet rich plasma. He reports residual pain that he can work through with pain at rest no longer constant and is hopeful to begin limited training with progression to full training in the near future.
Discussion
Using highly sensitive 18F-labeled PET radiotracers, including tracers for inflammation (FDG), bone remodeling (sodium fluoride (NaF)) and a pain receptor (S1R) with pain generation, Stanford researchers are finding ways to identify the exact source of pain generation and active inflammation. Combined with high-resolution MR imaging, the investigators are able to "see" the source of pain with greater accuracy than previously possible, which has already led to dramatically improved outcomes in some pain sufferers who have failed standard therapy.
Reference
‡ Technology in development that represents ongoing research and development efforts. These technologies are not products and may never become products. Not for sale. Not cleared or approved by the US FDA or any other global regulator for commercial availability.
