Clinical Faceoff: Hip Osteoarthritis in the Setting of Adult Spinal Deformity

Degenerative hip pathology has been noted in more than one-third of patients who undergo spine surgery [17]. The management of concurrent hip and spine pathology has been the subject of considerable debate because of what is commonly called overlapping symptomatology [8]. As both spine and hip pathology affect spinopelvic kinematics [3], the treatment of one influences the other. Hip and spine surgeons must collaborate to avoid potentially catastrophic complications like THA implant malposition, instability, and dislocation [14, 20]. Hip pathology in the setting of adult spinal deformity is particularly challenging to treat because altered hip motion changes the compensatory mechanisms for maintaining global sagittal alignment from pelvic tilt to thoracic hypokyphosis [4, 7, 11]. Restoration of normal lumbar lordosis and sagittal alignment has been associated with reduced compensation through the hip and pelvis [5]. However, prior lumbar fusion, particularly a longer fusion such as that performed for adult spinal deformity, has been associated with a higher risk of prosthetic dislocation [6, 29].

Chavarria et al. [6] recommended that surgeons manage adult spinal deformity before THA to permit appropriate acetabular implant placement in the setting of a realigned, less-compensated spine. Vigdorchik et al. [26] validated a spine-driven classification of concurrent hip-spine pathology to identify those at risk of dislocation. Such patients may benefit from dual-mobility components to decrease the risk of THA instability if the THA is performed before spinal surgery. However, such a purely mechanical solution does not completely address the relative benefits of the “order of operations” that should be considered to appropriately treat hip and spine pathology, which involves patient-specific decision-making. This is reflected by Devin et al. [8], whose proposed treatment algorithm for concurrent hip and spine pathology firmly recommends that surgeons perform spine surgery first only in the setting of progressive neurologic deficits.

In this “Clinical Faceoff,” I will discuss the management of a patient with concurrent adult spinal deformity and hip osteoarthritis with two clinical experts on the topic: Han Jo Kim MD, of the Hospital for Special Surgery (HSS) in New York, NY, USA, and Adolph J. Yates MD, of the University of Pittsburgh Medical Center (UPMC) in Pittsburgh, PA, USA. Dr. Kim completed a fellowship in spine and scoliosis surgery at Washington University in St. Louis, and he is the David B. Levine Chair in Scoliosis at HSS, specializing in adult and pediatric deformity surgery. Dr. Yates completed a fellowship in adult reconstruction at West Virginia University, and he has extensive experience with complex hip and knee arthroplasty as the Chief of Orthopaedic Surgery at UPMC Shadyside and Vice Chair for Quality of UPMC Presbyterian-Shadyside.

Clinical Scenario

The patient is a 60-year-old female with well-controlled Type 2 diabetes without neuropathy. She has a history of bilateral slipped capital femoral epiphyses (SCFE) that was treated when she was a child, which resulted in multiple subsequent surgeries on each hip. The patient presents with 1 year of claudication, which causes leg pain that arises after 3 minutes of standing. When symptomatic, the patient complains of 50% low back pain and 50% leg pain, with the right leg being more painful. The patient has to lean forward on a cane/walker to ambulate. Her gait is affected by a substantial (10 cm) limb length discrepancy; her right side is shorter than her left. Her right hip has a 30° flexion contracture with a maximum flexion of 60°, and it is irritable on log roll. Her imaging demonstrates an increased pelvic incidence-lumbar lordosis (PI-LL) mismatch (Fig. 1), central stenosis at L3-5 and foraminal stenosis at L3 and L5 bilaterally (Fig. 2). Dedicated hip imaging shows endstage osteoarthritis with a dysplastic right hip secondary to failed SCFE treatment (Fig. 3). Symptoms at this point have severely limited the patient’s quality of life. Her symptoms have persisted despite supervised exercises with a physical therapist. She has had multiple transforaminal epidural steroid injections at L4-5 with no improvement at all, even within a few minutes of injection.

Fig. 1.

These f-length standing AP and lateral radiographs as well as a supine lumbar radiograph are of a 60-year-old female patient with diabetes who had claudicatory leg pain.

Fig. 2.

This figure shows magnetic resonance cuts of the midpoint and left/right foramina of the patient’s lumbar spine.

Fig. 3.

These figures show (A) AP and (B) frog leg lateral radiographs of the right hip. (C) An axial CT cut through the femoral heads demonstrates a prior slipped capital femoral epiphysis on the right.

Mitchell S. Fourman MD, MPhil: The patient in this case presents with clinical and radiographic evidence of what appears to be symptomatic hip pathology and adult spine deformity. What are your immediate thoughts when working up this patient? Does your examination and workup change when there is evidence of concurrent hip/spine pathology?

Han Jo Kim MD: This is a complex story, with clear spine, hip, and knee pathology. I would need to know more details about the clinical presentation of this patient with respect to not only her history, but also her physical exam. Any neurologic deficits localized to the spine will take priority in her treatment. If there are no deficits and she has primarily back pain with or without neurogenic claudication, that can also change the treatment plan. The presence or absence of hip flexion contractures is a big consideration [4], as is the case with this patient’s right hip flexion contracture. A thorough radiographic workup is paramount to appropriate operative decision-making. This includes dynamic radiographs of the lumbar spine in addition to full-length skull to foot (EOS) images and supine scoliosis radiographs (34” to 36” cassettes). CT scanograms would be valuable as well [27], although their value depends on whether the spine or the lower extremity pathology is to be addressed first. If the spine is addressed second, a scanogram will invaluable when a fusion of S1-ilium is performed.

Adolph J. Yates MD: The patient presents with a complex history and resultant deformities. The offered short history is consistent with spinal stenosis causing neurogenic claudication, especially given that it is bilateral despite the more preserved joint space on the left. It is disconcerting that the epidural injections did not improve the patient’s symptoms. In looking at the patient’s standing films as well as her supine AP pelvis radiograph, her pelvic coronal obliquity stands out and at first appears paradoxical given that the patient’s more arthritic hip is on the high side. This obliquity is an important part of the patient’s limb length deformity; ideally how that was measured would be offered because the hips are actually about the same length if the ischial tuberosity lines are compared with the trochanters. That same film suggests a hip or knee flexion contracture on the left, which gives the false impression of the right tibia being longer. It would be helpful to know if there is an adduction contracture on the right or an abduction contracture on the left. The obliquity of the pelvis is probably fixed; scoliosis films and EOS, if available, would be helpful. One cut of the right hip on CT does not rule out a functional fusion; careful review of all of the images and an exam that demonstrates irritability would help make that distinction. Further evaluation might demonstrate otherwise, but at this point it would be important to set expectations that total hip replacement might not correct a limb length discrepancy secondary to fixed pelvic obliquity.

As to the second question, the history and physical exam does not change; the possibility of the spine contributing to the patient’s hip pain should always be considered. Further workup should be based on the patient’s history/exam and initial films.

Dr. Fourman: In your responses, both of you mentioned the importance of assessing spinopelvic mobility. The mobility of the lumbar spine and pelvis mean different things for the spine and reconstructive surgeon. Spinopelvic parameters such as pelvic tilt, sacral slope, and changes in positional lumbar lordosis can yield important information about the degree to which the patient is compensating for their pathology. In contrast, spinopelvic hypermobility, particularly after spine fusion, may impact the dislocation risk after THA [11]. To what end do spinopelvic measurements influence your preoperative planning and intraoperative management of this patient?

Dr. Yates: Spinopelvic parameters have been clearly demonstrated to contribute to prosthetic hip instability that results in revision surgery [16]. Vigdorchik et al. [26] published a Hip-Spine Classification that helps to predict how to revise instability in relation to anterior or posterior dislocations as well as to plan primary THA in the face of spinopelvic deformity [9, 24]. This classification is based on spine deformity and spinopelvic mobility and presents four scenarios based on AP and sitting/standing lateral radiographs. The first three—1-A, 1-B, and 2-A—represent normal anatomy/normal mobility, normal alignment and poor mobility, as well as flatback deformity and normal mobility, respectively; these can be addressed with attention to cup placement, in particular anteversion [9, 24, 25]. The fourth scenario, 2-B, is a flat-back spine coupled with stiffness that leads to posterior pelvic tilt and an outlet view appearance on the AP radiograph. Although a dual-mobility implant is tempting in the other three, it is this fourth scenario in which I agree with the authors that a dual-mobility implant is advised [24]. The coronal tilt that is not going to be corrected through the THA calls for adjustment of the lateral inclination of the cup to avoid too open or too closed of an angle.

Dr. Kim: Lumbar spine flexibility is key to evaluating adult spinal deformity [10, 18, 19] and predicting the usage for osteotomies to achieve the correct sagittal alignment during corrective surgery [12]. Flexibility of the lumbar spine can be identified using supine radiographs, in which the PI-LL mismatch will normalize compared with standing radiographs. If supine radiographs are not available (such as in this patient’s story), I would evaluate the CT scout or MRI images, which have been shown to be good surrogates for supine radiographs [1]. If there is no improvement in PI-LL mismatch on supine images, I would consider a fusion over isolated laminectomies in this patient. Lastly, I look to the patient’s response to epidural steroid injections when assessing fixed versus stenosis-related flattening of the lumbar spine. In patients with stenosis-related flattening, an epidural steroid injection would result in transient improvement in standing radiograph sagittal alignment [13]. This would consequently favor isolated laminectomies as the preferred treatment approach. Although this patient did not claim a subjective pain response to her epidural steroid injections, we did not assess her imaging changes after the injection.

Dr. Fourman: How have recent advances in surgical planning and intraoperative adjuncts, in particular surgical navigation, changed your management of patients with concurrent hip/spine pathology? Does the order of surgical management influence your use of these technologies?

Dr. Kim: I personally do not utilize surgical navigation in these types of reconstructions. However, surgical planning is critical to ensuring that alignment goals are achieved [15]. It is also important to consider both overall and regional alignment. Where historically we thought of lumbar lordosis as extending from L1 to S1, we now know that the majority of our lumbar lordosis (approximately 60%) is derived from the L4 to S1 region [2, 23]. As our understanding of alignment improves, so has our understanding that regional differences are important in preventing adjacent segment degeneration and minimizing junctional kyphosis [22, 28], which is a big problem for us as adult spinal deformity surgeons. All of these considerations could result in clinically important sagittal plane realignment, which can in turn affect cup position, ultimately increasing the risk of postoperative instability or dislocation [5]. For this reason, particular focus and careful surgical planning are needed to ensure that alignment goals are met in the most precise way possible, thereby decreasing the risk of dislocation.

Dr. Yates: I’m in agreement with Dr. Kim: The biggest advance in THA planning, in my opinion, is the recognition that some form of radiologic evaluation of the spine and spinopelvic motion is warranted, be it EOS or lateral sitting and standing views with plain films. Knowing in advance what adjustments should be performed to achieve the correct cup position is aided by clear visualization of the native acetabulum in surgery to guide adjustments. It is not clear that computer navigation/robotics adjusts for spine deformity/stiffness, as these technologies largely rely on the anatomic location of the native cup. The algorithms target capturing that location, not the dynamic position that pelvic tilt might impose [21]. The next step for robotics should be the utilization of positioning/dynamic studies and artificial intelligence/machine learning data derived from large databases. At this time, the order of surgical management does not influence my utilization of intraoperative navigation/robotics.

Dr. Fourman: So, verdict time. Should this patient present to your practice, what surgery would you recommend that she undergo first?

Dr. Yates: The complexity of the case leads to a recommendation that will seem confounding to some: I think we should address the spine first. In patients with concurrent symptomatic spinal stenosis and hip osteoarthritis, it would be reasonable to restore hip motion and correct any flexion contracture. However, in this patient, the hip deformity has been present for decades with corresponding adaptive chronic deformity of the spine at several levels. It is not clear that restoring hip motion will lead to posture changes that are adequate to alleviate the patient’s symptoms. Her symptoms are more consistent with neurogenic claudication, supported by her multilevel stenosis. Total hip replacement here would be the equivalent of reversing a previous hip arthrodesis, which would help with secondary back pain if it were caused by early changes at a single level. But that isn’t what we are seeing here.

Dr. Kim: I agree with Dr. Yates. I think that this patient’s claudication is her predominant symptom. There’s always a question in spine deformity surgery about how much should be performed. This patient’s bending over appears to be more related to her spinal stenosis than structural sagittal malalignment. It is therefore possible that a more limited surgical procedure would yield a completely satisfactory outcome for a patient who certainly is at risk for complications, a possibility that would be supported by improved symptoms and sagittal alignment after an epidural steroid injection. A laminectomy at the compressed levels would be a reasonable intervention with minimal morbidity that would solve most of this patient’s complaints and disability. A decision could then be made about whether to perform more extensive spine surgery (probably a T10-pelvis fusion here) if her alignment parameters were clinically bothersome enough to intervene. It would be easier to plan the patient’s THA with her long-term spinal alignment clear.

This recommendation for isolated intervention for lumbar stenosis alone in the setting of adult spinal deformity stands in contradistinction to patients whose adult spinal deformity is more structurally prominent and fixed. I present as an example of this a prior patient with severe spinal stenosis at L2-3, severe adult spinal deformity (PI-LL mismatch > 50°) caused by iatrogenic flatback and proximal junctional kyphosis after a prior L3-S1 fusion (Fig. 4), right L5 radiculopathy, and endstage hip osteoarthritis. Although this patient’s hip irritability was the most prominent symptom, the hip surgeon was concerned that the patient’s lumbar spine would become markedly more symptomatic postoperatively, limiting the patient’s rehabilitative capacity and jeopardizing his implant after correction. The patient’s substantial sagittal malalignment because of his prior surgery resulted in a more extensive operation: a T11-pelvis revision posterior spinal fusion with L2-L3 decompression, L4 pedicle subtraction osteotomy, and L5-S1 R-sided foraminotomy (Fig. 5). Correction of the patient’s sagittal alignment and stenosis allowed him to maintain a more upright posture and did in fact improve his ambulatory capacity before a navigated THA without dual-mobility implants 3 months later.

Fig. 4.

AP and lateral radiographs from the cervical spine to hips of a second patient with adjacent segment disease and proximal junctional kyphosis above a prior lumbar fusion.

Fig. 5.

AP and lateral radiographs from the cervical spine to hips of the second patient after a T11-pelvis revision posterior spinal fusion with L2-L3 decompression, L4 pedicle subtraction osteotomy, and L5-S1 right-sided foraminotomy.