June 2007, VOLUME 188

Recommend & Share

June 2007, Volume 188, Number 6

Head and Neck Imaging


Parathyroid Imaging: Technique and Role in the Preoperative Evaluation of Primary Hyperparathyroidism

+ Affiliations:
1Department of Radiology, University of Pittsburgh Medical Center and School of Medicine, 200 Lothrop St., 3950 CHP/MT, Pittsburgh, PA 15213.

2Department of Surgery, University of Pittsburgh Medical Center and School of Medicine, Pittsburgh, PA 15213.

Citation: American Journal of Roentgenology. 2007;188: 1706-1715. 10.2214/AJR.06.0938

Next section

OBJECTIVE. This article discusses the commonly used techniques for imaging the parathyroid glands and their role in the preoperative evaluation of patients with primary hyperparathyroidism.

CONCLUSION. The importance of sonography and sestamibi scintigraphy in the preoperative evaluation of patients with primary hyperthyroidism has increased with the adoption of minimally invasive parathyroidectomy techniques at most medical centers. When the results of these studies are concordant, the cure rates of minimally invasive surgery equal those of traditional bilateral neck exploration.

Keywords: head and neck imaging, hyperparathyroidism, nuclear medicine, sonography, thyroid gland

Previous sectionNext section

Primary hyperparathyroidism, whether caused by an adenoma or hyperplasia, can be cured surgically with a high rate of success. When performed by experienced surgeons, traditional surgical therapy—bilateral four-gland exploration—is successful in more than 95% of cases [1]. The development of unilateral and focused surgical approaches over the past decade, however, has made it even more imperative for imaging to accurately locate abnormal parathyroid glands before surgery. With optimized preoperative mapping, the success rate of these less invasive techniques equals that of the traditional bilateral approach [27].

The purpose of this article is to review the imaging techniques and rationale for the preoperative localization studies that are frequently used before parathyroidectomy. After a brief review of relevant anatomy and the physiology of primary hyperparathyroidism, we present current surgical approaches. The commonly used noninvasive imaging techniques—sonography, scintigraphy, CT, and MRI—will be discussed. Sonography and 99mTc-sestamibi scintigraphy will be emphasized because these have clearly emerged as dominant, and potentially complementary, techniques in the preoperative evaluation of primary hyperparathyroidism.

Embryology and Anatomy
Previous sectionNext section

Autopsy series show two superior and two inferior parathyroid glands in most individuals (Fig. 1). Supernum erary glands are seen approximately 3–5% of the time, and fewer than four glands are found in up to 3% of patients [8, 9]. The superior glands are derived from the fourth branchial pouch along with the lateral lobes of the thyroid; the inferior glands arise from the third branchial pouch along with the thymus gland. These embryologic relationships help to explain the normal—and variable—anatomic locations of the superior and inferior parathyroid glands.

View larger version (98K)
Fig. 1 Diagram shows posterior view of typical locations of paired superior (white arrows) and inferior (arrowheads) parathyroid glands and their relationship to thyroid gland and surrounding structures. Note close relationship parathyroid glands have with recurrent laryngeal nerves (black arrows), illustrating why nerve injury is a significant concern of endocrine surgeons, particularly with four-gland explorations.

View larger version (107K)
Fig. 2 Sonogram of 25-year-old woman with possible thyroid enlargement (thyroid was normal). Note subtle isoechoic parathyroid gland inferior to lower pole of thyroid (arrows). Normal parathyroid glands are uncommonly seen on sonography because of their small size.

The superior glands tend to be more consistent in location with most (> 90%) glands located deep in relation to the mid portion of the superior pole of the thyroid near the cricothyroid junction. Infrequently, the superior glands can be seen more inferiorly, deep in relation to the mid pole of the thyroid lobes (4%), or they may be located at or above the most superior aspect of the thyroid (3%). Rarely, the superior glands can be found in the retropharyngeal (1%) or retroesophageal (1%) spaces or in the thyroid gland itself (0.2%) [8, 9].

The inferior parathyroid glands are more variable in location but most commonly are located inferior, posterior, or lateral to the lower thyroid pole (69%). Because of their common origin with the thymus gland, they can also commonly be found more inferiorly in the neck, in the thymic tongue, or in the cervical portion of the thymus (26%). Very rarely, the inferior parathyroid glands can fail to descend with the thymus and may remain cephalad to the superior glands. Inferior glands can also be found in the anterior mediastinum with the thymus (2%) or even inferior to the thymus gland in the mediastinum (0.2%) [8, 9].

The average size of a normal parathyroid is 5 × 3 × 1 mm; normal glands weigh between 40 and 50 mg. They are thus infrequently identified at imaging [10] (Fig. 2). Adenomas, on the other hand, are considerably larger; they have a mean mass of greater than 10 times the normal parathyroid gland and are thus often identified at cross-sectional imaging [11]. Hyperplastic glands can be quite variable in size, both within the same patient and among populations, but they tend to have a total gland volume comparable to that of adenomas [12].

Pathophysiology of Primary Hyperparathyroidism
Previous sectionNext section

The parathyroid glands are responsible for calcium homeostasis via the production of parathyroid hormone (PTH). PTH raises the serum calcium level by promoting the renal tubular absorption of calcium, decreasing tubular reabsorption of phosphate, and stimulating osteoclasts. In addition, PTH stimulates vitamin D production, which, in turn, raises serum calcium by promoting its absorption by the gastrointestinal tract. Primary hyperparathyroidism is considered to be present when serum calcium is elevated and PTH is increased or inappropriately normal. The condition is most commonly diagnosed in the fifth through seventh decades of life and is a common endocrine disorder affecting approximately one in 500 women and one in 2,000 men. There are numerous, often nonspecific, clinical manifestations of hypercalcemia. The most common presenting symptoms include fatigue, hypertension, bone pain, muscle weakness, and psychiatric illness [13].

Most cases of primary hyperparathyroidism are caused by a single parathyroid adenoma (89%). Other causes include hyperplasia of all four glands (6%), double adenomas (4%), and, rarely, parathyroid carcinoma [2]. In most instances, parathyroid adenomas are sporadic. There is an increased incidence of parathyroid hyperplasia in multiple endocrine neoplasia, type I and multiple endocrine neoplasia, type IIA, although the incidence of these disorders is not sufficiently high to justify screening in all instances of primary hyperparathyroidism [14]. Another rare cause of primary hyperparathyroidism is familial hypocalciuric hypercalcemia, an autosomal dominant condition that produces PTH-dependent hypercalcemia. It is associated with mild parathyroid hyperplasia, but subtotal parathyroidectomy is not an effective treatment and is contraindicated in these cases.

Current Surgical Approach
Previous sectionNext section

Preoperative imaging is less critical for the endocrine surgeon when a traditional bilateral cervical dissection is used for parathyroid exploration. A transverse cervical incision made approximately 2 cm above the sternal notch, just below the cricoid cartilage, permits identification of all four glands and allows access to nearly all ectopic parathyroid sites. In solitary adenoma cases, all four glands are sampled and the adenomatous gland is resected. Treatment success using this approach exceeds 95%. Imaging is used only after failed initial surgery. For multigland disease, the approach involves either a subtotal 3.5-gland resection with cryopreservation or a total parathyroidectomy with autotransplantation into the forearm or neck muscle.

Over the past decade, the development of minimally invasive approaches to parathyroid surgery has made accurate preoperative localization of parathyroid disease absolutely critical for effective surgical treatment. Currently practiced minimally invasive techniques include unilateral open parathyroidectomy, video-assisted parathyroidectomy, and videoscopic parathyroidectomy. All of these techniques involve a more focused approach, limiting the dissection to the abnormal parathyroid gland and thus substantially decreasing operative time. The unilateral open technique involves a small unilateral transverse or lateral incision positioned for access to the parathyroid adenoma. The video-assisted technique is performed without insufflation, using a 5-mm endoscope and small conventional instruments through a 1.5-cm midline incision to identify and resect the abnormal gland or glands. Videoscopic techniques use carbon dioxide insufflation to create a working space for dissection and use three small incisions and endoscopic instruments to remove the abnormal gland, usually from a lateral cervical approach.

View larger version (64K)
Fig. 3A 44-year-old woman with hyperparathyroidism due to right inferior parathyroid adenoma. Resected gland weighed 629 mg, nearly 15 times weight of a normal gland (40–50 mg). Sonogram shows typical hypoechoic adenoma (arrows) deep in relation to lower pole of thyroid.

View larger version (72K)
Fig. 3B 44-year-old woman with hyperparathyroidism due to right inferior parathyroid adenoma. Resected gland weighed 629 mg, nearly 15 times weight of a normal gland (40–50 mg). Color Doppler sonogram shows peripheral feeding vessel (arrow) characteristic of parathyroid adenomas. Also note typical arc or rim vascularity.

The extent of minimally invasive parathyroidectomy is often guided by rapid intraoperative PTH testing. A greater than 50% drop in the serum PTH to normal or near-normal levels 10 minutes after parathyroid resection accurately predicts removal of all offending parathyroid tissue. Radioguided parathyroidectomy, which involves preoperative injection of 99mTc sestamibi, may also aid the endocrine surgeon in gland localization and may improve success rates, although this approach is controversial and is used in very few centers.

Accurate preoperative localization of parathyroid adenomas, particularly when combined with intraoperative PTH assays, enables use of these less invasive techniques; numerous studies comparing unilateral with bilateral approaches have shown similar success rates when preoperative imaging is highly suggestive of single-gland disease [7, 1518]. Cases of suspected multiglandular disease and those with equivocal preoperative localization still require the traditional bilateral cervical approach because imaging studies have been shown to have low sensitivity for the detection of hyperplasia and double adenomas. In addition, the surgical treatment for hyperplasia requires a bilateral approach; moreover, double adenomas are bilateral in most cases. Thus, the primary benefit of preoperative imaging studies is the accurate determination of uniglandular disease to help select patients most appropriate for unilateral and minimally invasive approaches.

Imaging is also important in the small percentage of cases of failed initial parathyroidectomy. Failures are most commonly due to undetected multiglandular disease but may also be due to ectopic glands or incomplete resection of a parathyroid tumor. Imaging studies can help to locate the offending tissue in these cases, facilitating a more focused approach and fewer surgical complications. In addition, these patients may benefit from MRI or CT of the mediastinum to search for ectopic parathyroid glands.

Parathyroid Imaging
Previous sectionNext section

Sonography and 99mTc-sestamibi scintigraphy are the dominant imaging techniques for preoperative location of parathyroid adenomas. Numerous studies comparing these techniques suggest similar sensitivities and specificities for solitary adenoma detection. Localization accuracy is also improved when both studies are obtained preoperatively [2, 1922]. Contrast-enhanced CT and MRI can also effectively locate parathyroid adenomas but are less commonly used for preoperative location and are more commonly used in the setting of failed parathyroidectomy for the detection of suspected ectopic—often mediastinal—glands. Rarely, cross-sectional imaging will be used if the findings at sonography and 99mTc-sestamibi scintigraphy are discordant.

Previous sectionNext section

The patient should be scanned supine with a pillow beneath the shoulders to slightly hyperextend the neck. Gray-scale imaging should be performed with a high-frequency linear transducer; the study should include longitudinal images extending from the carotid artery to midline and transverse images extending from the hyoid bone superiorly to the thoracic inlet inferiorly. Scanning at 12–15 MHz is possible with currently available higher level platforms. Having the patient swallow under real-time observation may help show inferior glands located deep in relation to the clavicles [23]. The thyroid is also imaged and nodules, particularly those with worrisome sonographic features (e.g., microcalcifications, lobulation), are documented. The size and volume of both lobes are also recorded.

Gray-scale imaging is supplemented by color and power Doppler imaging to look for feeding vessels and vascularity of suspected adenomas shown at initial gray-scale imaging. Graded compression of subcutaneous tissues and strap muscles may also be helpful in difficult-to-scan patients.

Imaging Findings

Parathyroid adenomas are nearly always homogeneously hypoechoic to the overlying thyroid gland on gray-scale imaging and are commonly detected using gray-scale imaging alone when they are larger than 1 cm in diameter (Fig. 3A). Hypoechogenicity may be a result of the marked, compact cellularity that is characteristic of adenomas at sectioning. They are usually oval or bean-shaped, but larger adenomas can be multilobulated. Color and power Doppler imaging commonly show a characteristic extrathyroidal feeding vessel (typically a branch off the inferior thyroidal artery), which enters the parathyroid gland at one of the poles (Fig. 3B). Internal vascularity is also commonly seen in a peripheral distribution. The artery feeding the adenoma tends to branch around the periphery of the gland before penetrating deeper, resulting in a characteristic arc or rim of vascularity. In addition, color Doppler sonography of the overlying thyroid gland may show an area of asymmetric hypervascularity that may help to locate an underlying adenoma [2426].

View larger version (96K)
Fig. 4A 55-year-old woman with primary hyperparathyroidism due to large left superior adenoma. Sonogram shows hypoechoic nodule suspected of being parathyroid medial to common carotid artery (arrow).

View larger version (96K)
Fig. 4B 55-year-old woman with primary hyperparathyroidism due to large left superior adenoma. Graded compression sonogram increases conspicuity of adenoma (arrows).

When gray-scale and Doppler imaging fail to show an abnormal parathyroid gland, graded compression may show a relatively incompressible gland to help differentiate an adenoma from surrounding soft tissue [26] (Fig. 4A, 4B). This technique is often helpful in detection of ectopic glands in the lower neck.

Normal parathyroid glands average 5 × 3 × 1 mm and are uncommonly seen with sonography; the histology of the normal parathyroid gland—chief cells, fibrovascular stroma, and adipocytes—may account for the isoechogenicity of the gland relative to adjacent thyroid. Hyperplasia is also a more difficult diagnosis to make sonographically because the single gland size is generally much smaller than an adenoma [12] (Fig. 5A, 5B, 5C, 5D). However, as with adenomas, compact cellularity may render hyperplastic glands hypoechoic relative to the overlying thyroid.

Cervical lymph nodes can commonly be mistaken for parathyroid glands. Central compartment lymph nodes may be particularly prominent when the patient has coexistent lymphocytic thyroiditis (Figs. 6A and 6B). Several features may help in distinguishing lymph nodes from adenomas, however. An echogenic fatty hilum usually indicates a benign lymph node. At color Doppler examination, lymph nodes are supplied by small hilar vessels, whereas a polar, peripheral distribution of color flow is commonly seen with parathyroid adenomas [25, 27] (Fig. 6C).

Concomitant thyroid disease also contributes to imaging pitfalls. Enlarged multinodular thyroid glands can limit the sonographic evaluation of parathyroid adenomas: Anatomy is distorted and posterior nodules may mimic parathyroid disease. Posterior thyroid nodules can have a similar sonographic appearance to intracapsular parathyroid glands, although the typical vascular pattern of a parathyroid adenoma is uncommonly seen in thyroid nodules (Fig. 7). The rare intrathyroid parathyroid gland is difficult, if not impossible, to distinguish from a thyroid nodule.

Some ectopic glands can be difficult to detect sonographically, particularly those in the retrotracheal region, because of the poor acoustic window caused by the tracheal air column. Sonography has poor sensitivity for detecting ectopic glands in the mediastinum as well.

Sonographically Guided Percutaneous Biopsy and Ethanol Ablation
Previous sectionNext section

In cases in which a more definitive preoperative localization is necessary and sonographic findings are equivocal, sonographically guided fine-needle aspiration (FNA) of the suspected gland with subsequent PTH assay has been shown to be a safe and highly specific technique to confirm whether the suspected lesion is parathyroid tissue [2830]. A study by Stephen et al. [28] of 57 preoperative sonographically guided FNAs in 54 patients found a specificity of 100% and no false-positive findings. The three false-negative FNAs occurred when small glands were sampled. Although PTH aspiration is rarely necessary for routine cases, this procedure can be particularly helpful in cases in which prior thyroid or parathyroid surgery has distorted neck anatomy or when preoperative confirmation of unusual-appearing or ectopic glands is needed. A theoretic complication of the procedure is seeding of abnormal parathyroid tissue along the biopsy track, resulting in an entity known as parathyromatosis. This has been described in cases of recurrent hyperparathyroidism after parathyroidectomy in which the capsule of the parathyroid gland was disrupted during the procedure. However, a study by Kendrick et al. [31] followed 81 patients after percutaneous parathyroid FNA for a mean of 5.8 years and found no cases of parathyromatosis.

Although it is primarily used as a therapy to treat tertiary hyperparathyroidism (most often in the setting of renal failure), sonographically guided parathyroid ablation with ethanol is a potential treatment option for patients with primary hyperparathyroidism. The procedure involves placement of the needle in the parathyroid gland and slowly injecting a volume of ethanol equal to approximately 50% of the volume of the gland until the absence of blood flow to the gland is confirmed on color Doppler sonography [32]. Because surgery is more effective at achieving a long-term cure for primary hyperparathyroidism, this procedure is mostly used in the setting of prohibitive risk due to medical comorbidities or a high risk of surgical morbidity in reoperative cases [32, 33].

Effectiveness of Preoperative Localization and Role in Surgical Planning
Previous sectionNext section

Reported sensitivities for the detection of solitary parathyroid adenomas with preoperative sonography range from 72% to 89% in recent large series [3437]. A meta-analysis performed by Ruda et al. [2] encompassing 54 studies performed between 1995 and 2003 using sonography for preoperative localization in primary hyperparathyroidism calculated sonographic sensitivities for the detection of solitary adenoma, hyperplasia, and double adenoma to be 79% (95% confidence interval, 77–80%), 35% (95% confidence interval, 30–40%), and 16% (95% confidence interval, 4–28%), respectively.

View larger version (84K)
Fig. 5A 15-year-old girl with hyperparathyroidism due to parathyroid hyperplasia. Sonograms show four slightly enlarged parathyroid glands (arrows): right superior (A), right inferior (B), left superior (C), and left inferior (D). Patient subsequently underwent four-gland exploration and subtotal parathyroidectomy, leaving portion of right superior gland. Largest of resected hyperplastic glands weighed only 322 mg. Relatively small size of typical hyperplastic glands decreases sensitivity of sonography.

View larger version (88K)
Fig. 5B 15-year-old girl with hyperparathyroidism due to parathyroid hyperplasia. Sonograms show four slightly enlarged parathyroid glands (arrows): right superior (A), right inferior (B), left superior (C), and left inferior (D). Patient subsequently underwent four-gland exploration and subtotal parathyroidectomy, leaving portion of right superior gland. Largest of resected hyperplastic glands weighed only 322 mg. Relatively small size of typical hyperplastic glands decreases sensitivity of sonography.

View larger version (84K)
Fig. 5C 15-year-old girl with hyperparathyroidism due to parathyroid hyperplasia. Sonograms show four slightly enlarged parathyroid glands (arrows): right superior (A), right inferior (B), left superior (C), and left inferior (D). Patient subsequently underwent four-gland exploration and subtotal parathyroidectomy, leaving portion of right superior gland. Largest of resected hyperplastic glands weighed only 322 mg. Relatively small size of typical hyperplastic glands decreases sensitivity of sonography.

View larger version (91K)
Fig. 5D 15-year-old girl with hyperparathyroidism due to parathyroid hyperplasia. Sonograms show four slightly enlarged parathyroid glands (arrows): right superior (A), right inferior (B), left superior (C), and left inferior (D). Patient subsequently underwent four-gland exploration and subtotal parathyroidectomy, leaving portion of right superior gland. Largest of resected hyperplastic glands weighed only 322 mg. Relatively small size of typical hyperplastic glands decreases sensitivity of sonography.

Although much of the literature on this topic is retrospective, Siperstein et al. [37] recently published one of the largest prospective studies, involving 350 patients with primary hyperparathyroidism who underwent preoperative sonography. This study was unique in that initially a focused unilateral parathyroidectomy was attempted based on preoperative localization studies, and intraoperative PTH assays were performed irrespective of the success of the initial focused parathyroidectomy. Thereafter, all patients underwent a four-gland exploration to determine where there was any additional parathyroid abnormality that was not suspected on preoperative evaluation. In this manner, the success of a focused parathyroidectomy based on preoperative imaging studies could be estimated, but there was pathologic analysis of all glands in each patient. Given preoperative sonographic findings suggestive of a single adenoma, a unilateral approach would yield a single abnormal gland at the correct site in 74% of patients. If it was assumed that the operation would be converted to a bilateral approach if no adenoma was found on the abnormal side, the surgical success rate was improved to 90%. This is somewhat lower than the cure rate of 95–97% reported in retrospective series of focused techniques. A selection bias may explain the discrepancy: Prior retrospective series of focused techniques have included stratified patient cohorts with diagnostic preoperative mapping, whereas in the study by Siperstein et al., all subjects were ultimately subjected to traditional surgical exploration.

Another prospective study by Rickes et al. [34] involving 98 patients with primary hyperparathyroidism suggests that particular sonographic findings—namely, a polar feeding vessel—significantly increase specificity for detecting parathyroid adenomas. Of those suspected adenomas that showed a feeding vessel using color Doppler sonography, the abnormal gland was correctly identified 93% of the time. For the 40% of suspected adenomas without a visualized feeding vessel, localization was correct only 39% of the time. These findings are similar to those in the retrospective study of Lane et al. [25], which found that a prominent feeding artery increased accuracy of localization from 73% to 88%.

Series in which endocrine surgeons performed parathyroid sonography before surgery have reported similar results. Solorzano et al. [36] retrospectively studied 226 patients with primary hyperparathyroidism and found preoperative sonography correctly identified all abnormal glands 77% of the time. The study of Milas et al. [38] involved 350 patients, and sonography correctly identified the site in 72% of patients. These studies suggest the importance of a detailed knowledge of cervical anatomy and of operator experience in the successful use of sonography for preoperative localization.

View larger version (127K)
Fig. 6A 25-year-old woman with Hashimoto's thyroiditis. Sonograms show how prominent central compartment lymph nodes (arrows) may mimic adenomatous parathyroid glands.

View larger version (119K)
Fig. 6B 25-year-old woman with Hashimoto's thyroiditis. Sonograms show how prominent central compartment lymph nodes (arrows) may mimic adenomatous parathyroid glands.

View larger version (100K)
Fig. 6C 25-year-old woman with Hashimoto's thyroiditis. Color Doppler sonogram may aid in differentiating between lymph nodes and adenomas: Lymph nodes are supplied by a central hilar vessel (arrow), whereas vessels that supply adenomas typically enter either pole.

View larger version (82K)
Fig. 7 67-year-old woman with hyperparathyroidism and left tracheoesophageal groove adenoma that could easily be mistaken for posterior thyroid nodule. Peripheral, polar vascularity seen on color Doppler sonogram helps to identify this as adenoma. Subsequent parathyroidectomy preformed at time of total thyroidectomy revealed this to be a supernumerary hyperplastic parathyroid gland.

Multiple studies confirm the poor sensitivity of sonography for the detection of double adenomas. In the study of Haciyanli et al. [39], 21 of 287 consecutive patients were found to have double adenomas. A retrospective analysis of preoperative sonography showed a 40% sensitivity for detection of both adenomas [39]. Similarly, Sugg et al. [40] found 23 cases of double adenoma in 233 consecutive patients, and sonography had a 23% sensitivity. The similarly low sensitivities seen with 99mTc sestamibi, a relatively operator-independent technique, might suggest double adenomas are an inherently different disease process than the typically easily shown solitary adenoma.

Given the significant operator dependence of parathyroid sonography, comparisons between studies is particularly difficult. In addition, substantial advancements in sonography technology over the past 10 years likely will improve the ability of sonography to detect very small structures with specific vascular patterns, such as adenomas. In our recent experience, a high-quality sonographic examination is often sufficient for preoperative mapping of adenomas; nuclear medicine studies, although still routinely performed before attempted focused resection, often add little, although they may be helpful when sonography is nondiagnostic (e.g., with ectopic parathyroid adenomas). A well-performed sonographic examination also aids the surgeon because the adenoma location is precisely described. Moreover, preoperative evaluation of thyroid nodules (with sonographically guided FNA) can often be performed in the same setting. Clearly, an up-to-date assessment of the ability of state-of-the-art sonography to identify and locate parathyroid adenomas as a stand-alone technique is needed.

Parathyroid Scintigraphy
Previous sectionNext section

Sestamibi with 99mTc is the most commonly used radiotracer for imaging the parathyroid glands and has been extensively studied in the setting of primary hyperparathyroidism. Sestamibi is taken up by both the thyroid and parathyroid glands, but adenomatous and hyperplastic parathyroid tissue shows more avid uptake of the radiotracer and often retains the radiotracer longer than adjacent thyroid tissue. Thus, initial planar images obtained shortly after the administration of radiotracer will show both thyroid and parathyroid tissue. Asymmetric foci of increased radiotracer uptake on early images can be seen, representing abnormal parathyroid tissue superimposed on the normal thyroid. Delayed images, obtained approximately 2 hours after radiotracer administration, are acquired to look for foci of retained radiotracer characteristic of hyperfunctioning parathyroid tissue (Fig. 8A, 8B).

View larger version (84K)
Fig. 8A 52-year-old woman with hyperparathyroidism and right superior parathyroid adenoma. Early-phase 99mTc-sestamibi SPECT image shows physiologic uptake in salivary glands and thyroid gland, with focus of more intense uptake overlying superior pole of right thyroid lobe (arrow).

View larger version (106K)
Fig. 8B 52-year-old woman with hyperparathyroidism and right superior parathyroid adenoma. Two-hour delayed SPECT image shows radiotracer retention in adenoma (arrow) but clearing of tracer from overlying thyroid.

SPECT of the neck through a pinhole collimator can help to differentiate parathyroid activity from the overlying thyroid and has been shown to increase the sensitivity of scintigraphic parathyroid imaging [4143]. Subtraction techniques have also been used to image the parathyroid glands by administering a second radiotracer taken up only by the thyroid gland. Either 123I or 99mTc pertechnetate can be used in conjunction with 99mTc sestamibi to produce subtraction images of the parathyroid glands, although these techniques have not been clearly shown to be diagnostically superior to nonsubtracted techniques.

The reported sensitivity of 99mTc-sestamibi scintigraphy is similar to that of sonography. Recent large series using SPECT show sensitivities for detection of solitary adenomas in the range of 68–95% [37, 4446]. The meta-analysis of Ruda et al. [2] included 96 studies using 99mTc-sestamibi scintigraphy between 1995 and 2003 in the setting of primary hyperparathyroidism and calculated the sensitivity for detection of solitary adenomas at 88% (95% confidence interval, 87–89%). Similar to sonography, sensitivities for detection of hyperplasia and double adenomas were low, calculated at 44% (95% confidence interval, 41–48%) and 30% (95% confidence interval, 2–62%), respectively [2].

Civelek et al. [44] reported their experience with 338 consecutive patients undergoing preoperative 99mTc-sestamibi SPECT studies in which the scans were interpreted with the reviewers blinded to the clinical status of the patient or results of other imaging tests such as sonography. In this setting, 99mTc-sestamibi SPECT precisely located 90% of solitary adenomas, 73% of double adenomas, and 45% of hyperplastic glands.

In the large prospective study of Siperstein et al. [37] discussed previously, a similar analysis using preoperative 99mTc sestamibi alone to direct a focal approach resulted in proper identification of the site 68% of the time. As with the results regarding preoperative sonography, this is somewhat lower than similar retrospective studies.

Jones et al. [47] also prospectively interpreted preoperative sestamibi scans and found a high sensitivity for adenomas > 500 mg, with 93% (91/98) accurately located versus only 51% (18/35) for those < 500 mg. This shows the disadvantage of the limited resolution of scintigraphic imaging in the setting of smaller adenomas.

Combined Sonography and Scintigraphy for Preoperative Evaluation of Hyperparathyroidism
Previous sectionNext section

A preoperative approach that combines both the anatomic information of sonography and the physiologic information of scintigraphy has been shown to predict the presence and location of solitary adenomas more accurately than either technique alone. Lumachi et al. [48] retrospectively reviewed preoperative sonography and 99mTc-sestamibi findings in patients with proven solitary adenomas and found a combined sensitivity of 95% versus 80% for sonography and 87% for scintigraphy alone. Similarly, Siperstein et al. [37] predicted 79% surgical success in their prospective study combining both techniques versus 74% for sonography and 68% for scintigraphy alone. Solorzano et al. [36], who advocate preoperative sonography as the only preoperative location test, found that, used separately, sonography and scintigraphy each correctly predicted uniglandular disease in 77% of patients, but this increased to 90% when the techniques were combined. Using concordant results on sonography and scintigraphy to plan a focal surgical approach has resulted in high cure rates similar to the traditional bilateral approach. Sonography has the advantage of being more specific regarding the site of an adenoma in relation to the thyroid gland. Scintigraphy clearly has an advantage in the detection of ectopic glands, particularly in the mediastinum.

Both techniques remain similarly insensitive for the detection of multiglandular disease and double adenomas. Sugg et al. [40] retrospectively studied preoperative imaging findings in 23 of 233 patients found to have multiglandular disease and found that even when combining both techniques, multiglandular disease was correctly predicted in 30% of patients, single-gland disease was incorrectly predicted in 30%, abnormal parathyroid glands were not located in 30%, and discordant results were provided in 10%. Haciyanli et al. [39] specifically studied double adenomas and found combined techniques were only 60% sensitive. Even with the addition of intraoperative PTH assays of the abnormal glands, double adenoma was correctly predicted in 80% of patients, emphasizing the need for a traditional bilateral approach in patients with suspected multiglandular disease [39]. Given that the operation of choice for both multiglandular disease and double adenomas is a traditional bilateral approach, some endocrine surgeons have advocated that equivocal, negative, or discordant results on both preoperative studies warrant a nonselective approach because a high proportion of these patients will have multifocal disease. Thus, the role of the radiologist is to provide highly accurate location of singlegland disease to facilitate focal surgical approaches. When findings are equivocal, suggestive of multiglandular disease, or discordant with other imaging tests, this should be clearly stated as well because these patients will more likely be treated with a traditional bilateral approach.

View larger version (85K)
Fig. 9 40-year-old woman who presented with recurrent hypercalcemia and hyperparathyroidism after resection of both left-sided glands. Contrast-enhanced CT scan shows brisk enhancement of 8-mm soft-tissue nodule (arrow) in mediastinum that correlated anatomically with focus of radiotracer retention in mediastinum on prior sestamibi SPECT. This was found to be a hyperplastic right inferior parathyroid gland.

View larger version (172K)
Fig. 10A 39-year-old woman with left superior adenoma showing typical MRI signal characteristics. T2-weighted MR image shows increased T2 signal in adenoma (arrow) relative to thyroid gland and surrounding soft tissues.

View larger version (149K)
Fig. 10B 39-year-old woman with left superior adenoma showing typical MRI signal characteristics. Axial T1-weighted MR image shows typical intermediate T1 signal (arrow) seen in adenomas.

View larger version (173K)
Fig. 10C 39-year-old woman with left superior adenoma showing typical MRI signal characteristics. Gadolinium-enhanced T1-weighted image with fat suppression shows intense enhancement typical of adenomas (arrow). These imaging characteristics can be indistinguishable from those of lymph nodes and thus must be interpreted in clinical context and in concert with other imaging techniques.

Previous sectionNext section

Axial contrast-enhanced thin-collimation CT images through the neck show intense enhancement in the typical locations for parathyroid tissue in the setting of adenoma. Scanning from the skull base through the mediastinum has the additional advantage of detecting most ectopic glands (Fig. 9). In cases of failed initial parathyroidectomy, artifacts from surgical clips placed in the neck often limit the diagnostic quality of CT. Reported sensitivities of CT range from 46% to 87% [49]. Combined studies of sonography and CT suggest that supplemental CT will detect few additional adenomas over sonography alone. Thus, CT is usually reserved for cases of failed parathyroidectomy or in cases of altered anatomy, in which CT may aid in operative planning [5052].

Recent studies have combined 99mTc-sestamibi SPECT with coregistered CT in an attempt to improve sensitivity by combining anatomic and functional information, but results from these initial studies are conflicting in their conclusions about the added usefulness of CT. More study is required before the appropriate usefulness of this combined technique is established [5355].

Previous sectionNext section

Although less commonly used for preoperative localization than sonography and scintigraphy, MRI provides similar sensitivity to other techniques in the detection of abnormal parathyroid tissue [56, 57]. More commonly, MRI is used in patients with persistent or recurrent hyperparathyroidism, in whom it has been shown to be effective in locating remaining abnormal parathyroid tissue [58].

Images of the neck are generally obtained with an anterior neck surface coil from the hyoid bone to the sternal notch. Axial T1- and T2-weighted fast spin-echo sequences are commonly acquired. In instances in which no neck lesion is identified or an ectopic parathyroid gland is suspected, ECG-gated axial images of the mediastinum can be effective.

The T1 and T2 characteristics of abnormal parathyroid tissue are variable. The most common tissue characteristics are intermediate- to low-intensity T1 signal and high-intensity T2 signal (Fig. 10A, 10B, 10C). Less commonly, fibrosis or old hemorrhage can cause low signal intensity on T1- and T2-weighted images. Subacute hemorrhage into adenomas can cause high signal intensity on both T1- and T2-weighted images [59].

The acquisition of gadolinium-enhanced T1-weighted images with fat suppression has not been shown to significantly increase detection of adenomas when they exhibit T2 hyperintensity. However, false-negative studies are most commonly associated with adenomas that are isointense on T1 and T2 sequences; the addition of contrast-enhanced images can increase sensitivity for these cases [57].

Abnormal parathyroid tissue cannot be diagnosed on MRI by signal characteristics alone because cervical lymph nodes have similar signal characteristics. Therefore, accurate MRI diagnosis depends on knowledge of the typical morphology and location of the parathyroid glands and common sites of ectopic glands.

Previous sectionNext section

Over the past decade, the surgical treatment of primary hyperparathyroidism has changed from predominantly a bilateral approach with four-gland exploration in all cases to unilateral and focused approaches guided by preoperative imaging showing single adenomas. Sonography and 99mTc-sestamibi scintigraphy have assumed dominant roles in preoperative location of solitary adenomas, and focused approaches based on concordant findings from both techniques have cure rates equal to that of the traditional approach. Sonography is particularly diagnostic when a typical polar feeding vessel and peripheral vascularity are shown. Preoperative identification of multiglandular disease and double adenomas remains poor; thus, patients with negative, discordant, or equivocal results will commonly undergo bilateral procedures because of the high incidence of multifocal disease in patients without definitive preoperative imaging findings. MRI and CT are generally reserved for cases of failed initial surgery or recurrent hyperparathyroidism because the prevalence of supernumerary and ectopic glands is much higher in this select population.

Address correspondence to M. E. Tublin.

CME This article is available for CME credit. See www.arrs.org for more information.


This article is available for CME credit. See www.arrs.org for more information.

We thank Eric Jablonowski for his illustration of normal parathyroid anatomy.

Previous sectionNext section
1. Sosa JA, Powe NR, Levine MA, Udelsman R, Zeiger MA. Profile of a clinical practice: thresholds for surgery and surgical outcomes for patients with primary hyperparathyroidism—a national survey of endocrine surgeons. J Clin Endocrinol Metab 1998; 83:2658 –2665 [Google Scholar]
2. Ruda JM, Hollenbeak CS, Stack BC Jr. A systematic review of the diagnosis and treatment of primary hyperparathyroidism from 1995 to 2003. Otolaryngol Head Neck Surg 2005; 132:359–372 [Google Scholar]
3. Baliski CR, Stewart JK, Anderson DW, Wiseman SM, Bugis SP. Selective unilateral parathyroid exploration: an effective treatment for primary hyperparathyroidism. Am J Surg 2005; 189:596 –600 [Google Scholar]
4. Burkey SH, Snyder WH 3rd, Nwariaku F, Watumull L, Mathews D. Directed parathyroidectomy: feasibility and performance in 100 consecutive patients with primary hyperparathyroidism. Arch Surg 2003; 138:604 –608 [Google Scholar]
5. Udelsman R, Donovan PI, Sokoll LJ. One hundred consecutive minimally invasive parathyroid explorations. Ann Surg 2000; 232:331 –339 [Google Scholar]
6. Grant CS, Thompson G, Farley D, van Heerden J. Primary hyperparathyroidism surgical management since the introduction of minimally invasive parathyroidectomy: Mayo Clinic experience. Arch Surg 2005; 140:472 –478 [Google Scholar]
7. Udelsman R. Six hundred fifty-six consecutive explorations for primary hyperparathyroidism. Ann Surg 2002; 235:665 –670 [Google Scholar]
8. Akerstrom G, Malmaeus J, Bergstrom R. Surgical anatomy of human parathyroid glands. Surgery 1984; 95:14–21 [Google Scholar]
9. Wang CA. The anatomic basis of parathyroid surgery. Ann Surg 1976; 183:271 –275 [Google Scholar]
10. Grimelius L, Bondeson L. Histopathological diagnosis of parathyroid diseases. Pathol Res Pract 1995; 191:353–365 [Google Scholar]
11. Yao K, Singer FR, Roth SI, Sassoon A, Ye C, Giuliano AE. Weight of normal parathyroid glands in patients with parathyroid adenomas. J Clin Endocrinol Metab 2004; 89:3208 –3213 [Google Scholar]
12. Mollerup CL, Bollerslev J, Blichert-Toft M. Primary hyperparathyroidism: incidence and clinical and biochemical characteristics—a demographic study. Eur J Surg 1994; 160:485 –489 [Google Scholar]
13. Uden P, Chan A, Duy QY, Siperstein A, Clark OH. Primary hyperparathyroidism in younger and older patients: symptoms and outcome of surgery. World J Surg 1992; 16:791–797 [Google Scholar]
14. Muhr C, Ljunghall S, Akerstrom G, et al. Screening for multiple endocrine neoplasia syndrome (type 1) in patients with primary hyperparathyroidism. Clin Endocrinol 1984; 20:153 –162 [Google Scholar]
15. Ryan JA Jr, Eisenberg B, Pado KM, Lee F. Efficacy of selective unilateral exploration in hyperparathyroidism based on localization tests. Arch Surg 1997; 132:886–890 [Google Scholar]
16. Bergenfelz A, Lindblom P, Tibblin S, Westerdahl J. Unilateral versus bilateral neck exploration for primary hyperparathyroidism: a prospective randomized controlled trial. Ann Surg 2002; 236:543 –551 [Google Scholar]
17. Kacker A, Komisar A. Unilateral versus bilateral neck exploration in parathyroid surgery: an assessment of 55 cases. Ear Nose Throat J 2001; 80:530 –532, 534 [Google Scholar]
18. Russell CF, Dolan SJ, Laird JD. Randomized clinical trial comparing scan-directed unilateral versus bilateral cervical exploration for primary hyperparathyroidism due to solitary adenoma. Br J Surg 2006; 93:418 –421 [Google Scholar]
19. Scheiner JD, Dupuy DE, Monochik JM, Noto RB, Cronan JJ. Pre-operative localization of parathyroid adenomas: a comparison of power and colour Doppler ultrasonography with nuclear medicine scintigraphy. Clin Radiol 2001; 56:984–988 [Google Scholar]
20. Purcell GP, Dirbas FM, Jeffrey RB, et al. Parathyroid localization with high-resolution ultrasound and technetium Tc 99m sestamibi. Arch Surg 1999; 134:824–830 [Google Scholar]
21. De Feo ML, Colagrande S, Biagini C, et al. Parathyroid glands: combination of (99m) Tc MIBI scintigraphy and US for demonstration of parathyroid glands and nodules. Radiology 2000; 214:393 –402 [Google Scholar]
22. Kebapci M, Entok E, Kebapci N, Adapinar B. Preoperative evaluation of parathyroid lesions in patients with concomitant thyroid disease: role of high resolution ultrasonography and dual phase technetium 99m sestamibi scintigraphy. J Endocrinol Invest 2004; 27:24–30 [Google Scholar]
23. American Institute of Ultrasound in Medicine. AIUM practice guideline for the performance of a thyroid and parathyroid ultrasound examination. J Ultrasound Med 2003; 22:1126 –1130 [Google Scholar]
24. Wolf RJ, Cronan JJ, Monchik JM. Color Doppler sonography: an adjunctive technique in assessment of parathyroid adenomas. J Ultrasound Med 1994; 13:303 –308 [Google Scholar]
25. Lane MJ, Desser TS, Weigel RJ, Jeffrey RB Jr. Use of color and power Doppler sonography to identify feeding arteries associated with parathyroid adenomas. AJR 1998; 171:819–823 [Abstract] [Google Scholar]
26. Reeder SB, Desser TS, Weigel RJ, Jeffrey RB. Sonography in primary hyperparathyroidism: review with emphasis on scanning technique. J Ultrasound Med 2002; 21:539 –552 [Google Scholar]
27. Ahuja A, Ying M, King A, Yuen HY. Lymph node hilus: gray scale and power Doppler sonography of cervical nodes. J Ultrasound Med 2001; 20:987 –992 [Google Scholar]
28. Stephen AE, Milas M, Garner CN, Wagner KE, Siperstein AE. Use of surgeon-performed office ultrasound and parathyroid fine-needle aspiration for complex parathyroid localization. Surgery 2005; 138:1143 –1150 [Google Scholar]
29. Sacks BA, Pallotta JA, Cole A, Hurwitz J. Diagnosis of parathyroid adenomas: efficacy of measuring parathormone levels in needle aspirates of cervical masses. AJR 1994; 163:1223 –1226 [Abstract] [Google Scholar]
30. Gooding GA, Clark OH, Stark DD, Moss AA, Montgomery CK. Parathyroid aspiration biopsy under ultrasound guidance in the postoperative hyperparathyroid patient. Radiology 1985; 155:193–196 [Google Scholar]
31. Kendrick ML, Charboneau JW, Curlee KJ, van Heerden JA, Farley DR. Risk of parathyromatosis after fine-needle aspiration. Ann Surg 2001; 67:290 –293 [Google Scholar]
32. Harman CR, Grant CS, Hay ID, et al. Indications, technique, and efficacy of alcohol injection of enlarged parathyroid glands in patients with primary hyperparathyroidism. Surgery 1998; 124:1011 –1019 [Google Scholar]
33. Cercueil JP, Jacob D, Verges B, Holtzmann P, Lerais JM, Krause D. Percutaneous ethanol injection into parathyroid adenomas: mid- and long-term results. Eur Radiol 1998; 8:1565–1569 [Google Scholar]
34. Rickes S, Sitzy J, Neye H, Ocran KW, Wermke W. High-resolution ultrasound in combination with colour-Doppler sonography for preoperative localization of parathyroid adenomas in patients with primary hyperparathyroidism. Ultraschall Med 2003; 24:85 –89 [Google Scholar]
35. Haber RS, Kim CK, Inabnet WB. Ultrasonography for preoperative localization of enlarged parathyroid glands in primary hyperparathyroidism: comparison with (99m) technetium sestamibi scintigraphy. Clin Endocrinol 2002; 57:241 –249 [Google Scholar]
36. Solorzano CC, Carneiro-Pla DM, Irvin GL 3rd. Surgeon-performed ultrasonography as the initial and only localizing study in sporadic and primary hyperparathyroidism. J Am Coll Surg 2006; 202:18 –24 [Google Scholar]
37. Siperstein A, Berber E, Mackey R, Alghoul M, Wagner K, Milas M. Prospective evaluation of sestamibi scan, ultrasonography, and rapid PTH to predict the success of limited exploration for sporadic primary hyperparathyroidism. Surgery 2004; 136:872–880 [Google Scholar]
38. Milas M, Stephen A, Berber E, Wagner K, Miskulin J, Siperstein A. Ultrasonography for the endocrine surgeon: a valuable clinical tool that enhances diagnostic and therapeutic outcomes. Surgery 2005; 138:1193 –1200 [Google Scholar]
39. Haciyanli M, Lal G, Morita E, Duh QY, Kebebew E, Clark OH. Accuracy of preoperative localization studies and intraoperative parathyroid hormone assay in patients with primary hyperparathyroidism and double adenoma. J Am Coll Surg 2003; 197:739–746 [Google Scholar]
40. Sugg SL, Krzywda EA, Demeure MJ, Wilson SD. Detection of multiple gland primary hyperparathyroidism in the era of minimally invasive parathyroidectomy. Surgery 2004; 136:1303 –1309 [Google Scholar]
41. Slater A, Gleeson FV. Increased sensitivity and confidence of SPECT over planar imaging in dualphase sestamibi for parathyroid adenoma detection. Clin Nucl Med 2005; 30:1–3 [Google Scholar]
42. Spanu A, Falchi A, Manca A, et al. The usefulness of neck pinhole SPECT as a complementary tool to planar scintigraphy in primary and secondary hyperparathyroidism. J Nucl Med 2004; 45:40–48 [Google Scholar]
43. Lorberboym M, Minski I, Macadziob S, Nikolov G, Schachter P. Incremental diagnostic value of preoperative 99mTc-MIBI SPECT in patients with parathyroid adenoma. J Nucl Med 2003; 44:904–908 [Google Scholar]
44. Civelek AC, Ozalp E, Donovan P, Udelsman R. Prospective evaluation of delayed technetium-99m sestamibi SPECT scintigraphy for preoperative localization of primary hyperparathyroidism. Surgery 2002; 131:149 –157 [Google Scholar]
45. Moka D, Voth E, Dietlein M, Larena-Avellaneda A, Schicha H. Technetium 99m-MIBI-SPECT: a highly sensitive diagnostic tool for localization of parathyroid adenomas. Surgery 2000; 128:29–35 [Google Scholar]
46. Melton GB, Somervell H, Friedman KP, Zeiger MA, Cahid Civelek A. Interpretation of 99mTc sestamibi parathyroid SPECT scan is improved when read by the surgeon and nuclear medicine physician together. Nucl Med Commun 2005; 26:633 –638 [Google Scholar]
47. Jones JM, Russell CF, Ferguson WR, Laird JD. Preoperative sestamibi-technetium subtraction scintigraphy in primary hyperparathyroidism: experience with 156 consecutive patients. Clin Radiol 2001; 56:556 –559 [Google Scholar]
48. Lumachi F, Zucchetta P, Marzola MC, et al. Advantages of combined technetium-99m-sestamibi scintigraphy and high-resolution ultrasonography in parathyroid localization: comparative study in 91 patients with primary hyperparathyroidism. Eur J Endocrinol 2000; 143:755 –760 [Google Scholar]
49. Gotway MB, Higgins CB. MR imaging of the thyroid and parathyroid glands. Magn Reson Imaging Clin N Am 2000; 8:163 –182 [Google Scholar]
50. van Dalen A, Smit CP, van Vroonhoven TJ, Burger H, de Lange EE. Minimally invasive surgery for solitary parathyroid adenomas in patients with primary hyperparathyroidism: role of US with supplemental CT. Radiology 2001; 220:631–639 [Google Scholar]
51. Spieth ME, Gough J, Kasner DL. Role of US with supplemental CT for localization of parathyroid adenomas. Radiology 2002; 223:878 –879 [Google Scholar]
52. Gross ND, Weissman JL, Veenker E, Cohen JI. The diagnostic utility of computed tomography for preoperative localization in surgery for hyperparathyroidism. Laryngoscope 2004; 114:227–231 [Google Scholar]
53. Profanter C, Wetscher GJ, Gabriel M, et al. CT–MIBI image fusion: a new preoperative localization technique for primary, recurrent, and persistent hyperparathyroidism. Surgery 2004; 135:157 –162 [Google Scholar]
54. Gayed IW, Kim EE, Broussard WF, et al. The value of 99mTc-sestamibi SPECT/CT over conventional SPECT in the evaluation of parathyroid adenomas and hyperplasia. J Nucl Med 2005; 46:248–252 [Google Scholar]
55. Krausz Y, Bettman L, Guralnik L, et al. Technetium-99m-MIBI SPECT/CT in primary hyperparathyroidism. World J Surg 2006; 30:76 –83 [Google Scholar]
56. McDermott VG, Fernandez RJ, Meakem TJ 3rd, Stolpen AH, Spritzer CE, Gefter WB. Preoperative MR imaging in hyperparathyroidism: results and factors affecting parathyroid detection. AJR 1996; 166:705 –710 [Abstract] [Google Scholar]
57. Lopez Hanninen E, Vogl TJ, Steinmuller T, Ricke J, Neuhaus P, Felix R. Preoperative contrast-enhanced MRI of the parathyroid glands in hyperparathyroidism. Invest Radiol 2000; 35:426–430 [Google Scholar]
58. Gotway MB, Reddy GP, Webb WR, Morita ET, Clark OH, Higgins CB. Comparison between MR imaging and 99mTc MIBI scintigraphy in the evaluation of recurrent or persistent hyperparathyroidism. Radiology 2001; 218:783 –790 [Google Scholar]
59. Auffermann W, Guis M, Tavares NJ, Clark OH, Higgins CB. MR signal intensity of parathyroid adenomas: correlation with histopathology. AJR 1989; 153:873 –876 [Abstract] [Google Scholar]

Recommended Articles

Parathyroid Imaging: Technique and Role in the Preoperative Evaluation of Primary Hyperparathyroidism

Full Access, , , ,
American Journal of Roentgenology. 2005;185:406-417. 10.2214/ajr.185.2.01850406
Abstract | Full Text | PDF (3458 KB) | PDF Plus (3627 KB) 
Full Access, , ,
American Journal of Roentgenology. 2009;193:687-695. 10.2214/AJR.08.2286
Abstract | Full Text | PDF (1080 KB) | PDF Plus (1145 KB) 
Full Access, , ,
American Journal of Roentgenology. 2010;194:S26-S37. 10.2214/AJR.07.7141
Citation | Full Text | PDF (1605 KB) | PDF Plus (1441 KB) 
Full Access
American Journal of Roentgenology. 2017;209:826-835. 10.2214/AJR.17.17973
Abstract | Full Text | PDF (978 KB) | PDF Plus (1060 KB) 
Full Access, , ,
American Journal of Roentgenology. 2003;181:1695-1704. 10.2214/ajr.181.6.1811695
Citation | Full Text | PDF (28306 KB) | PDF Plus (1683 KB) 
Full Access, , , , ,
American Journal of Roentgenology. 2013;200:W581-W592. 10.2214/AJR.12.9531
Abstract | Full Text | PDF (1148 KB) | PDF Plus (973 KB)