Parathyroid Imaging: Comparison of 2~ Subtraction Scintigraphy, Computed Tomography and Ultrasonography

Shinya KOBAYASHI, Makoto MIYAKAWA, Yoshio KASUGA, Tamotsu YOKOZAWA, Osamu SENGA, Akira SUGENOYA and Futoshi IIDA

ABSTRACT: From 1982 to 1985, twenty-nine patients with suspected hyperparathyroidism were examined using e~ subtraction scinti- graphy (T1-Tc), computed tomography (CT) and uhrasonography (US). For diagnosing neoplasm (adenoma or cancer), the sensitivities of the three proce- dures were 80 per cent or more, with no statistically significant differences. For diagnosing hyperplasia of the parathyroid glands, CT scan had the highest sensitivity (47 per cent). The most frequent source of error was mini- really enlarged glands, weighing less than 500 rag. The second highest source of error was thyroid nodules, such as adenomatous goiter or cancer. Serum calcium and c-PTH levels were significantly higher in those with a parathy- roid neoplasm than in those with hyperplasia (p<0.01, p<0.05, respectively). We concluded that hyperplasia is less easy to detect than neoplasm, and CT scan is superior to T1-Tc or US scan for localizing hyperplasia.

KEY WORDS: hyperparathyroidism, preoperative localization, 2~ 99mTc subtraction, CT scan, US scan

INTRODUCTION

Although some surgeons advocate intra- operative detection of enlarged parathyroid glands, without a preoperative localization study, preoperative identification by imaging techniques aids in locating the glands during surgery. Such procedures as scinti-scanning with 75Se-methionine, arteriography and selective venous sampling for parathyroid hormone were developed by radiologists, but have not been generally utilized because of limitations in imaging the parathyroid

The Second Department of Surgery, Shinshu University School of Medicine, Matsumoto, Japan

Reprint requests to: Shinya Kobayashi, MD, The Second Department of Surgery, Shinshu Uni- versity School of Medicine, 3-1-1 Asahi, Matsu- moto 390,Japan

glands, and because of severe complications. Recently, subtraction scintigraphy using ~01Tl-chloride and 99mTc-pertechnetate (T1- Tc) made possible improved imaging of the glands. Ferlin et al.1 reported a 92 per cent success rate with this technique in pa t i en t s with primary hyperparathyroidism. Com- puted tomography (CT) has been also useful for localizing the parathyroid glands, particu- larly in cases of recurrent hyperparathy- roidism3 The usefulness of ultrasonography (US) for detecting enlarged parathyroids has been documented by Scheible et al. s and Simeone et al. 4 Although clinical experiences of one or two of the three procedures have often been reported, comparative studies are few.

In this study, the accuracies of US scan, CT scan and scinti-scanning by T1-Tc for detect- ing the parathyroid glands were compared in

JAPANESE JOURNAL OF SURGERY, VOL. 17, No. 1 pp. 9-13, 1987

10 Ko bayashi et al. Jpn. J. Surg. January 1987

clinical cases of p r ima ry and secondary hyper- pa ra thyro id i sm.

MATERIALS AND METHODS

From 1982 to 1985, twenty-nine pat ients suspected of having hyperpa ra thyro id i sm were e x a m i n e d by the T1-Tc, CT and US scans to demons t r a t e the loca t ion of the dis- eased glands, preoperat ively. Twenty-seven of the 29 were d iagnosed as having hyperpara - thyroidism and elective surgery was p lanned . Of the 27, 23 were p r imary and 4 were secon- dary cases, while 17 (63 per cent) had neo- p lasm and 10 (37 per cent) hyperplas ia . The total n u m b e r of glands examined in the local izat ion s tudy was 53, consisting of 17 of neoplasm and 36 of hyperplas ia .

Tl- Tc scinti-scanning Two mCi of 201Tl-chloride was injected

intravenously, and the image was recorded 15 min la ter with a Tosh iba g a m m a camera GCA-401. After scanning with emTl, 5 mCi of 99mTc-pertechnetate was injected intra- venously, and the image was recorded 20 min later . Af ter comple t ion of the two scannings, sub t rac t ion was done using a min i -compute r , T I A C OP 500.

C T s c a n The pat ients were scanned using a Siemens

Somatom 2. Twenty to 25 continuous sec- tions, 4 m m thick, were taken f rom the hyoid bone to the aort ic arch with intravenous infu- sion of 30 per cent contras t m e d i u m at 5 m l / m i n . A scan t ime of 10 seconds was used with exposure factors of 125 KVp and 230 Dose.

US scan The pat ients were examined with a high-

resolut ion rea l - t ime scanner (Aloka SSD-256) equ ipped with a 7.5 MHz l inear a r ray trans- ducer . The sonographic examina t ion of the neck was carr ied out in two directions, longi- tud ina l and transverse, f rom the thyroid glands to the level of the clavicles and lateral- ly to the in ternal j ugu la r vein.

Laboratory examination Serum c - P T H was measured using a radio-

immunoassay kit f rom Eiken Co., and serum calc ium was measured in the l abora to ry of our hospital , as a rout ine examina t ion .

RESULTS

The diagnost ic accuracies of the three pro- cedures are summar ized in T a b l e 1. In pa- tients with pa ra thy ro id neoplasm, sensitivities were 80 to 87 per cent, with no significant differences among the three procedures . Spe- cificities were 95 to 100 per cent, again with no significant differences. In case of hyper- plasia of the pa ra thyro id , the sensitivities were 25 to 47 per cent and specificities were 66 to 100 per cent. C T showed the highest specificity for hyperplas ia , and T1-Tc the lowest. C o m p a r i n g the detectabi l i t ies of neo- p lasm and hyperplas ia , the former was clearly more de tec tab le than the lat ter .

The sources of error in false-negative diag- nosis are summar ized in T a b l e 2. The most

Tab le 1. Diagnostic Accuracy of Imaging for Parathyroid Disease

Accuracy Sensitivity Specificity

Neoplasm T1-Tc 87 100 CT 82 96 US 80 95

Hyperplasia T1-Tc 25 66 CT 47 100 US 40 88

All values are percentages.

Table 2. False-Negative Results by Parathyroid Imagings

No. of Glands Misdiagnosed Source of Error Procedure

Total TI-Tc CT US

Less than 500mg 20 20 18 19 Thyroid disease 3 0 2 3 Enlarged parathyroid 2 2 0 0 Ectopic position 1 0 1 1 Unknown 1 0 1 1 Total 27 22 22 24

All values are numbers of glands.

I/olume 17 Parathyroid imaging 1 l Number 1

frequent source of error was a minimally en- larged gland weighing less than 500 rag. Of 20 such glands, none was detected by TI-Tc, 18 were undetected by CT, and 19 by US. The second most frequent source of error was a misdiagnosis of thyroid nodules, such as adenomatous goiter or thyroid cancer. This happened with CT and US in 2 and 3 para- thyroid glands, respectively, but not wi th T1- Tc. Two enlarged para thyroid glands were missed by T1-Tc because of lack of accumula- tion of 2~ but they were not missed by CT and US. One ectopic gland was missed by CT and US, but not by T1-Tc.

The sources of error in false-positive diag- nosis are summarized in Table 3. Two thyroid nodules were misdiagnosed as being parathy- roid glands, f rom the US findings. TI-Tc re- vealed 3 false-positives related to technical difficulties in the subtraction.

The frequencies of visualization of the glands by the three imaging techniques are presented in relation to the largest diameter of the glands in Fig. 1. All para thyroid glands over 3.0 cm in diameter could be detected by all three procedures, but the rate of detection

Table 8. False-Positive Results by Parathyroid Imagings

No. of Glands Misdiagnosed Source of Error Procedure

T1-Tc CT US

Thyroid nodule 0 0 2 Technical error 3 0 0 Unknown 0 2 1 Total 3 2 3 All values are numbers Of glands.

dropped as the size of the glands decreased below this figure. TI-Tc and US rarely imaged glands less than 1.0 cm in diameter.

Serum calcium and c -PTH levels in pa- tients with pr imary hyperparathyroidism are shown in Table 4. These levels were signifi- cantly higher in patients with a neoplasm than in those with hyperplasia (p<0,01 and p<0.05, respectively). Weight and volume of glands were approximately 20 times greater in cases of neoplasm than in hyperplasia. There were thus marked differences between neo- plasms and hyperplasia in both serum cal- cium and c -PTH levels and weight and volume of the parathyroid glands.

DISCUSSION

Controversy still exists over the significance of preoperative localization tests of a hyper-

% 1oo

8o

.~ 40

20

T1-Te ~-]CT ~ U S

<I.0 1.0--2.0

:i:i:! iii!i!i

_ - - - ~ i!!ii!!

!!ii!!~ V//A ~iiiiil

vzz~4 iii!i!

:i:i:i: f//dA

v//wl !:i:i:i ~//~ .,... 2. i--3.0 3.0<

Greatest Diameter of Parathyroid Glands (ern)

Fig. 1. Visualization byimagings and size of parathyroid glands.

~~ subtraction scintigraphy; [] computed tomography; [] ultrasono- graphy

Table 4. Comparison of Serum Calcium and c-PTH levels and Weight and Volume of the Glands between Neoplasm and Hyperplasia in patients with Pri- mary Hyperparathyroidism

Neoplasm Hyperplasia p

No. of cases 17 6 Calcium (mg/dl) 12,3__+ 1.5 11.0+0.3 <0.01 c-PTH (ng/ml) 2.59• 0.88• <0.05

No. of glands 17 20 Weight (rag) 3420• 160• <0.01 Volume (cm~) 3.22• 0.139• <0.01

12 Kobayashi et al. Jpn. J. Surg. January 1987

func t ion ing p a r a t h y r o i d g land . In our experi- ence, the pa r a thy ro id glands can be readi ly de tec ted without unnecessary man ipu la t ion , by preopera t ive imaging, and the opera t ing t ime has thus been reduced. Both the sensi- tivity and the specificity of the imagings were h igher in neoplasms than in hyperplas ia . The r a p i d detec t ion of neoplasms is perhaps due to the size and funct ion of the pa ra thy ro id glands. An analysis of false-negative results revealed thyroid nodules to be an impor t an t source of error . P r imary hyperpara thyro id i sm sometimes coexisted with thyroid nodules, such as adenomatous goiter or thyroid cancer . To different iate t h e pa ra thy ro id glands f rom thyroid nodules and to detect the ectopic pa r a thy ro id glands, sc int igraphy by TI-Tc was more useful than the other two proce- dures. The mechanism of Z0tT1 accumula t ion in the glands is obscure, bu t Strauss et al. 5 specula ted that it may result f rom an increase in local b lood flow. Argument s over pa ra thy- roid sc int igraphy have focused on the prob- lems of high cost and exposure to rad ia t ion . T1-Tc may be most efficient for pat ients who will undergo re -explora t ion for an u n c o m m o n locat ion of the pa ra thy ro id glands.

CT scan is super ior to T1-Tc and US for de tec t ing hyperplas t ic pa ra thy ro id glands, which are usual ly smaller than neoplasms. Wi th US, hypoechogenic i ty might sometimes be a cause of misdiagnosis.

As others have emphasized,6, 7 locat ing the pa ra thy ro id glands by imag ing is easier with larger glands, and is vir tual ly l imi ted to glands la rger than 1.0 cm in greatest d iame- ter. Takag i et al. s r epor ted 88 per cent sen- sitivity for de tec t ing large hyperplas t ic g lands by CT and only 33 per cent for small glands less than 500 rag. Read ing et al. 6 using high- resolution rea l - t ime sonography, ob ta ined 95 per cent de tec tabi l i ty in g lands grea ter than 1,000 m g and 35 per cent in those less than 200 mg.

In the s tudy of serum ca lc ium and c -PTH, cases of secondary hype rpa ra thy rod i sm were exc luded because the et iologic factors under- lying the hyperfunct ion of the g land differ

from the p r imary disease. In those with pri- ma ry hyperpara thyro id i sm, serum calcium and c - P T H levels were h igher in case of neo- p lasm than the those with a hyperplas ia , and the sizes of the g lands were larger in the former than in the lat ter . Therefore , patients with high ca lc ium and c - P T H levels before opera t ion should be first examined using US scan while for o ther pat ients , CT is to be pre- fered.

I f none of the three procedures demon- s trate the glands, more invasive techniques such as a r t e r iog raphy and venous samplingg,a0 should be under taken . Selective venous cathe- ter izat ion m a y be useful when other localiza- t ion tests p roduce negat ive or results are equivocal , Non-invasive techniques for the detec t ion of small pa r a thy ro id glands require fur ther development .

(Received for pub l ica t ion on Dec. 17, 1985)

References

1. Ferlin G, Borsato N, Gamerani M, Come N, Zotti D. New perspectives in localizing enlarged parathy- roids by technetium-thallium subtraction scan. J Nucl Med 1983; 24: 438-441.

2. Roza AM, Wexler MJ, Stein L, Goltzman D. Value of high-resolution computerized tomography in localizing diseased parathyroid glands. Can J Surg 1984; 27:334 336.

3. Scheible W, Deutch AL, Leopold GR. Parathyroid adenoma: accuracy of preoperative localization by high resolution real-time sonography. JCU 1981; 9: 325-330.

4. Simeone JF, Mueller PR, Ferrucci JT, van Sonnen- berg E, Wang CA, Hall DA, Wittenberg J. High- resolution real-time souography of the parathyroid. Radiology 1981; 141: 745-751.

5. Strauss HW, Harrison K, Pitt B. Taltium-201 non- invasive determination of the regional distribution of cardiac output. J Nucl Med 1977; 18: 1167-1170.

6. Reading CC, Charboneau JW, James EM, Karsell PR, Purnell EC, Grant CS, van HeerdenJA. High- resolution parathyroid sonography. AJR 1982; 139: 539-546.

7. Stark DD, Gooding GAW, Moss AA, Clark OH, Ovenfors CO. Parathyroid imaging: comparison of high-resolution CT and high-resolution sono- graphy. AJR 1983; 141: 633-638.

Volume 17 Parathyroid imaging 13 Number 1

8. Takagi H, Tominaga Y, Uchida K, Yamada N, Ishii T, Morimoto T, Yasue M. Preoperative diagnosis of secondary hyperparathyroidism using computed tomography. J Comput Assist Tomogr 1982; 6: 527-528.

9. Dunlop DAB, Papapulos SE, Lodge RW, Fulton AJ, Kendall BE, O'Riordan JLH. Parathyroid venous sampling: anatomic considerations and results in

95 patients with primary hyperparathyroidism, gr J Radiol 1980; 53: 183-191.

10. Clark OH, Okerlund MD, Moss AA, Stark D, Norman D, Newton TH, Duh OY, Arnaud CD, Harris S, Gooding GAW. Localization studies in patients with persistent or recurrent hyperparathy- roidism. Surgery 1985; 98: 1083-1094.