Common diagnoses relating to the z-joint include cervical, thoracic, or lumbar facet arthropathy, facet syndrome, and spondylosis. Several other diagnoses often implicate the z-joint as a potential source of pain including whiplash syndrome, cervicogenic headache, and degenerative spondylolisthesis. The z-joints can cause a variety of symptoms including both local axial pain and referred extremity pain. In the cervical spine, the z-joints often cause pain in well-known distributions [12] (Fig. 38.2), with the upper cervical joints capable of causing headaches. Referral patterns from the lumbar z-joints are less specific [11, 13]. No historical features have proven pathognomonic for identification of z-joint pain, but these patients often complain of sharp or aching pain that is worse with loading the joint, such as during spine extension or extension with rotation or with stretching the capsule of the joint, such as extremes of flexion or rotation. Isolated z-joint pain should not cause burning, tingling, numbness, or extremity weakness typical of neuropathic processes. Common diagnoses are described in more detail in the subsections below.

There are currently no historical features, physical exam maneuvers, or radiographic findings that can definitively and specifically identify z-joint pain. History should focus on identification of red flag conditions, such as cancer, infection, and unstable fracture, and can help to distinguish between mechanical pain and inflammatory arthropathies, as well as to differentiate somatic pain from neuropathic pain. At a minimum, the physical exam should attempt to rule out neurological injury, identify biomechanical factors that may predispose to z-joint pain, such as forward-head position or lumbar hyperlordosis, and should include a manual examination of the painful region. Careful segmental deep palpation can identify a joint level of maximal tenderness, especially in the cervical spine. Localizing signs may be absent in the lumbar spine, but the L4-5 and L5-S1 levels are most commonly affected [16].

Diagnostic imaging options include plain X-rays , dynamic bending X-rays, MRI, CT, and single photon emission CT (SPECT) . All of these modalities can demonstrate z-joint osteoarthritis. Dynamic X-rays , including flexion and extension sequences, can show a mobile spondylolisthesis, which may support a specific joint level as being the pain generator. MRI can show increased signal in the peri-facet bone marrow, or in the z-joint itself especially on STIR sequences. SPECT can demonstrate metabolically active z-joint arthropathy that has been linked to injection response in some studies [21, 22], but involves significant radiation exposure and is not recommended for routine use. Studies continue to look at radiographic predictors of response to diagnostic blocks [23, 24]. While patients with z-joint pain are likely to have imaging abnormalities, many asymptomatic individuals also demonstrate these same findings. Furthermore, diagnostic imaging can often overlook microscopic injuries to the joint capsule or cartilage, and cannot be used to definitively rule out a particular joint as a pain generator. This is especially true in the post-whiplash population [10, 15], where imaging may be more helpful in ruling out competing diagnoses.

The current gold standard to diagnose z-joint pain remains controlled diagnostic blocks to the specific joint in question or its nerve supply [13]. A single anesthetic block has a false positive rate as high as 40% [16]. A dual blockade protocol can be used to improve specificity and is recommended in most situations where RFN is considered. For example, on one occasion a short-acting anesthetic such as xylocaine is used, then at a separate time the block is repeated using a longer-acting anesthetic, such as bupivacaine. Patients are blinded as to which anesthetic was used and should complete a pain diary including percentage of relief from their index pain and duration of response. While a well-placed intra-articular injection and medial branch block (MBB) may be equally efficacious for the diagnosis of z-joint pain, MBBs may be a better predictor of response to RFN [25]. Debate continues regarding the most appropriate cutoffs for a “positive response.” Most accept 80% improvement as definitively positive for the lumbar spine, but some argue that 50% is adequate to reduce false-negative responses for selection of patients for RFN [25]. Some studies and insurance companies advocate stricter cutoffs as high as 100% improvement, especially in the cervical spine [15]. A positive concordant response entails a patient demonstrating improvement for a length of time that approximates the duration of action of the anesthetics used. Positive concordant dual blocks reduce false-positives and increase the probability of a successful RFN.

When pursuing diagnoses and interventional treatments for z-joint pain, some common technical standards should be considered. All procedures should be performed only after thorough informed consent, including a discussion of the procedure, risks, benefits, and alternatives. Sedation should be minimized for any diagnostic procedures. For an effective diagnostic block, patients need to have the index pain at time of the procedure or need to be able to consistently trigger the pain with a movement or activity, such that a pre- and post-procedure comparison can be made. Therapeutic procedures can be performed with or without sedation depending on patient and provider preference, as well as availability of cardiopulmonary monitoring and emergency resuscitative supplies. All procedures should be image-guided, most commonly with biplanar fluoroscopy , and should follow strict sterile precautions. Unless contraindicated, radio-opaque contrast dye should be utilized for all diagnostic blocks to ensure accurate placement and to avoid intravascular injection. For diagnostic intra-articular z-joint injections , no more than 1 mL of volume should be used to avoid capsular rupture or extravasation to surrounding structures including the nerve roots and epidural space (Fig. 38.3). For diagnostic MBBs , no more than 0.5 mL should be used to maintain specificity.

After diagnosis, therapeutic interventional options include therapeutic intra-articular injection or RFN . Therapeutic intra-articular injection typically involves instillation of corticosteroid with local anesthetic into the joint itself and can provide primarily short-term pain relief. For chronic z-joint pain, RFN can be considered with goal of more sustained improvement of pain and function. The above procedural setup applies to RFN ; however, rather than relying on local anesthetic to block the joint’s nerve supply, a radiofrequency electrode is used to heat the needle tip immediately adjacent to the nerve target (Fig. 38.4).

Several techniques of RFN are described including pulsed, conventional, or cooled. Pulsed RFN utilizes a lower temperature (~42 °C) for a longer duration (120 s), compared to conventional RFN (~80 °C for 90 s). The rationale for pulsed RFN is to alter neural transmission without fully coagulating the nerve; it may be considered in higher risk areas, such as the upper cervical spine or when there is concern for aberrant anatomy. Conventional RFN aims to coagulate/destroy the target nerve for a longer therapeutic effect. The RF needle tip can sense temperature and shuts off the lesion if temperatures are exceeding 95 °C, a temperature at which surrounding tissues are at risk. Outcomes from conventional RFN are highly dependent on technique, with the best outcomes achieved when the needle tip lies parallel to the target nerve. Cooled RFN is a newer technique that maintains some special advantages over conventional RFN, but carries higher equipment costs that may not be reimbursed by insurance. Cooled RFN utilizes a fluid pump, which circulates sterile water to cool the needle tip and adjacent tissue. With a probe tip temperature of 60 °C, coagulation time is longer (150 s), but the lesion size is larger. Moreover, the lesion extends beyond the needle tip allowing for easier perpendicular needle placement, which is helpful in complex anatomy or SI joint RFN. Added caution is also required with cooled RFN because of the larger lesion size.

Following RFN, the lesioned nerves typically regenerate and can restore nociception to the joint. While duration of effect from RFN varies, positive outcome can be considered at least 50% improvement for at least 6 months.

Serious complications are very rare (<1%) following intra-articular z-joint injections, MBBs, or RFN. All z-joint procedures carry the usual risk of infection, hematoma, allergic or adverse reaction to the medications used, unintended nerve injury, and vasovagal reaction, including syncope. Risk of catastrophic bleeding, which could lead to spinal cord compression, is extremely low as these procedures all remain outside of the spinal canal. RFN carries some additional unique risks including post-RFN neuritis, anesthesia dolorosa, and muscle injury/denervation. Risk to the dorsal and ventral nerve roots is very low when correct technique is utilized. Post-RFN neuritis can occur with perfect technique and typically involves painful dysesthesias in the distribution of the targeted nerve lasting days to several weeks. Anesthesia dolorosa is a rare chronic deafferentation pain, such that an individual experiences neuropathic pain in a region of cutaneous numbness. Both post-RFN neuritis and anesthesia dolorosa are more common in the upper cervical spine (e.g., TON) with neuritis occurring around 20% of the time and lasting 7–10 days on average [26]. Post-RFN neuritis is frequently treated with corticosteroids. An expected side effect of RFN is local, segmental denervation of the multifidi muscles at the levels of the RFN . While multifidi denervation has not been shown to be clinically significant in cases of successful lumbar RFN [27], significant cervical muscle weakness and kyphosis has been reported following multilevel cervical RFN [28]. It is rare to require RFN to more than two joint levels, to achieve clinical benefit and the fewest number of joints possible to achieve good outcome should be targeted.