CHAPTER 3 Cross-sectional Imaging of the Hip
With rapid technical advances over the last two decades, cross-sectional imaging, most notably CT and MRI, have become integral tools in diagnosis and treatment of musculoskeletal disease. Although shoulder and knee MRI have been standard of care for more than a decade, more recently, MRI, MR arthrography, and multidetector CT have played increasingly important roles in diagnosing diseases of the hip. The principal benefit of MRI and multidetector CT over radiographs is that they allow for three-dimensional, multiplanar evaluation of the hip joint. Both modalities have strengths and relative weaknesses, and these inherent characteristics typically favor one modality over the other in evaluation of specific pathologic conditions.
A primary advantage of CT is its wide availability and accessibility. It is generally a succinct and accurate examination that is commonly used when time and availability are the prime considerations. The more recent advent of multidetector CT allows for the simultaneous acquisition of 4, 16, or 64 thin or overlapping tomographic slices, greatly reducing imaging time, decreasing motion artifact, and markedly improving image resolution compared with the predecessors of multidetector CT. High-resolution multiplanar reformats can be performed days after the scan has been performed. For fine bony detail, multidetector CT offers unparalleled resolution advantages compared with MRI or conventional CT. It has no compatibility issues with metallic prostheses or devices such as pacemakers and protocols using multidetector CT have been designed to allow for supreme resolution at the prosthesis-bone interface. CT exposes the patient to varying degrees of ionizing radiation, however, and higher resolution multidetector CT studies tend to increase this radiation dose even more. Also, CT is insensitive to soft tissue injuries around the hip, although it can easily detect a hip effusion.
MRI produces excellent tissue contrast compared with the gray-scale images of CT. It allows evaluation of not only the bony integrity of the hip and abnormalities of the surrounding soft tissues, but also the physiologic state of structures, as in bone marrow edema after a traumatic contusion. Furthermore, MRI makes routine contrast discrimination at tissue-tissue interfaces possible, a trait unique to this imaging modality. This means that fibrocartilage and hyaline cartilage structures may be reliably assessed without subjecting the patient to the ionizing radiation required for CT and radiography.
One disadvantage to MRI is that the lengthier MRI examination (typically 30 to 45 minutes) requires the patient to remain motionless for prolonged periods to obtain optimal images. Also, many patients with cardiac pacemakers and shrapnel near the orbits or spinal cord are not candidates for MRI, and true claustrophobia remains an issue with many MRI systems. Nevertheless, mild claustrophobia or generalized anxiety should not preclude a diagnostic MRI examination. Patients with mild claustrophobia or generalized anxiety should be referred for MRI on newer “open” or “short bore” magnet designs that are tolerated more easily by anxious patients.
A limitation of MRI and CT is artifact generated by orthopedic hardware. Although the “susceptibility artifact” of MRI can be minimized by tailoring the technique, the remaining artifact sometimes precludes optimal evaluation of the area of concern. “Beam hardening” artifact of prostheses in CT was a major problem for years, but multidetector CT protocols have virtually eliminated this problem. At this time, multidetector CT with a metal protocol is the imaging study of choice for indications of periprosthetic lesions, such as component loosening and particle disease.
With both imaging modalities, there are additional considerations to keep in mind, such as contrast administration. Contrast-enhanced examinations with intravenously administered contrast agents are typically reserved for evaluation for infection, inflammatory arthropathies, neoplasms, and vascular lesions.1–4 Rarely, a contrast-enhanced multidetector CT scan should be performed over MRI for the aforementioned indications. In addition, direct MR arthrography and CT arthrography (which involve direct administration of contrast material into the joint) can be used to better evaluate small intra-articular bodies and cartilaginous structures such as the labrum or articular cartilage. Indirect MR arthrography (intravenous administration of contrast material, which readily accumulates in the joint after a short delay) can be used in similar situations. This method cannot, however, achieve adequate joint distention with indirect arthrography in the absence of a preexisting joint effusion. For this reason, we reserve indirect arthrography of the hip for suspected labral tears when a direct arthrogram is logistically impractical and for some postoperative indications. The radiologist generally should have a role in deciding which study is most appropriate before imaging, but intra-articular or intravenous contrast administration often requires an order or prescription from the referring clinician.
Although interpretation of cross-sectional imaging studies of the hip might be best left to the radiologist, orthopedists and emergency medicine clinicians frequently find themselves in a setting where they must provide a preliminary interpretation of CT or MRI examinations. With multidetector CT, identifying pathology reliably on a quality study can be easy for someone comfortable with plain x-ray interpretation; getting interpretable images is the most difficult part. All of the information from the multidetector CT is on one series of axial images, although additional reformatting of this information in coronal and sagittal planes and three-dimensional models can be helpful in confirming pathology. Software applications allowing for accurate three-dimensional reformats are useful in the setting of articular fractures to help quantify the percentage of surface area involvement (Fig. 3-1).
FIGURE 3-1 A and B, Coronal (A) and sagittal (B) reformatted images from 16 detector row multidetector CT (Philips Medical Systems) show a comminuted and displaced posterior column acetabulum fracture (arrows). C, Coronal oblique three-dimensional reconstruction displays displaced acetabular fragments with an intact hip joint (arrows). D and E, After digital subtraction of the femur from the three-dimensional reconstructions, fracture extension to the articular surface is shown (curved arrow on D) along with the degree of displacement of acetabular rim components (straight arrows on E).
In most cases, multidetector CT of a bone or joint may be interpreted in a similar fashion to a radiographic series. Displaced fractures are often readily visible and practically unmistakable, although the chronicity of some fractures can be more difficult to establish. Arthritis on CT looks similar to arthritis on radiographs. The same can be said for specific radiologic findings; for example, a periosteal reaction in the setting of osteomyelitis can be clearly diagnosed by CT.
Interpretation of MRI sequences can be more daunting. For even the most basic interpretations, each MRI sequence must be categorized as fluid-sensitive or fat-sensitive. Fluid-sensitive sequences include all T2-weighted sequences and short tau inversion recovery (STIR) sequences. On these images, all fluids (including water, blood, and edema) are bright, or hyperintense. On fat-sensitive T1-weighted sequences, fluid is dark, but normal bone marrow is bright. With these images, loss of the normal hyperintense bone marrow signal often leads to identification of pathology. When the interpreter is confident about this categorization of the MRI sequences available, basic and preliminary interpretation of pathologies such as fracture and joint effusion is possible for clinicians who have an understanding of the pathologies themselves.5
In the setting of a radiographic examination that is equivocal for hip fracture or negative for fracture but accompanied by a persistent high clinical suspicion for occult fracture, MRI and multidetector CT can be used for further assessment. In our opinion, which is supported by radiology literature, MRI is the imaging study of choice to exclude occult hip fracture. Even a limited, 15-minute MRI protocol is nearly 100% sensitive for occult hip fracture if it is a fluid-sensitive (STIR or T2-weighted fat-suppressed) sequence. In cases of fracture, both of these sequences show hyperintense (bright) bone marrow edema surrounding the fracture site, and an accompanying T1-weighted sequence can be used for description and classification of the fracture using the hypointense (dark) fracture line (Fig. 3-2).6–9
FIGURE 3-2 A and B, Coronal STIR (A) and T1-weighted spin echo (B) MR images acquired on a 0.3-T open system (Hitachi Airis II) show extensive bone marrow edema throughout the femoral neck (arrow) diagnostic of a fracture. The hypointense fracture “line” is more subtle, but confirms the diagnosis. These two sequences, and a T2-weighted fast spin echo image not shown, comprise a fast hip fracture protocol that totals 11 minutes of imaging time and is sensitive and specific for fracture, avascular necrosis, effusion, osteoarthritis, and numerous extra-articular pathologies.
In difficult cases of subtle nondisplaced fracture in an osteopenic patient, the edema on MRI that alerts the radiologist to fracture is not visible on CT. Similarly, subtle stress fractures of the femoral neck, acetabulum, pubic symphysis, and sacrum are common and are best evaluated by MRI for the same reason. Subcapital proximal femur fractures are particularly difficult to diagnose on CT and on conventional radiographic series. MRI of the hip or of the entire pelvis is the standard of care in these cases when there is discordance between physical examination findings and radiographs or CT, or when CT and radiographic studies are equivocal for fracture. Even so, a multidetector CT examination identifies most hip fractures and is a reasonable option to try, especially when the patient is already undergoing CT scanning as part of a trauma workup. If the CT scan is negative but the clinical suspicion for proximal femoral fracture persists, MRI is indicated. In contrast, if even a mediocre-quality MRI examination is negative for hip fracture, there is no acute or subacute hip fracture.10
Complex fractures such as acetabular fractures, severely comminuted hip fractures, and hip dislocations (postreduction) are generally best evaluated by multidetector CT due to its superior resolution and multiplanar capabilities. Small bone fragments can easily be missed on MRI, and small degrees of displacement are difficult to quantify. Our policy is to perform coronal, sagittal, and three-dimensional reformatted imaging by multidetector CT in all cases of isolated acetabular fracture. In contrast, when proximal femoral fractures are identified, they might be more consistently characterized by MRI. MRI findings of bone marrow edema lend insight into fracture extension and vector of biomechanical force. A subcapital fracture that was occult on radiographs and CT would be readily identifiable on noncontrast MRI. In subacute fractures, MRI is extremely sensitive for early femoral head avascular necrosis. Likewise, previously occult femoral neck fractures are easy to distinguish from intertrochanteric fractures on MRI by examination of the bone marrow edema pattern. If the size or state of a hematoma is of concern, MRI is the modality of choice, but if the primary objective is to map out the fracture course, CT might be a better tool.
The preferred technique for imaging the acetabular labrum is direct MR arthrography. Labral tears are diagnosed by identifying paramagnetic contrast material (which is white on most MRI sequences) that undermines or outlines the labral defect or extends directly into the labrum substance (which is normally black on MRI sequences) (Fig. 3-3). Smaller, undersurface tears can be differentiated from normal variations such as sublabral recesses (which are currently a subject of controversy in the radiology literature), by their location and by the configuration of the defect. In younger patients with little joint wear and tear, the normal anterior and superior labrum should be sharply defined; it should be triangular and hypointense on all sequences. There is no recess anteriorly, so a defect in the undersurface of the anterior labrum which alters its triangular morphology should be considered a tear. Signal alteration within the labrum (especially fluid bright defects or findings into which contrast material readily flows) should also raise strong suspicion of a tear.
FIGURE 3-3 A and B, Sagittal (A) and axial (B) T1-weighted spin echo fat-suppressed MR images dedicated to the left hip acquired at 1.5 T (Philips Intera) after direct, intra-articular infusion of dilute gadolinium contrast material (Magnevist; Berlex) show a defect in the undersurface of the anterior acetabular labrum with frank imbibition of contrast material into the labral substance (arrows