The Usefulness of Neck Pinhole SPECT as a Complementary Tool to Planar Scintigraphy in Primary and Secondary Hyperparathyroidism

MATERIALS AND METHODS

From May 1997 to December 2002, 110 consecutive patients with HPT underwent P-SPECT in addition to conventional planar double-phase ^99mTc-MIBI or ^99mTc-pertechnetate/^99mTc-tetrofosmin parathyroid scintigraphy. For this study, we selected from this cohort the patients who underwent neck surgery and had definitive histologic diagnosis. The group included 67 patients (52 women, 15 men) who were 22–78 y old.

According to clinical and biochemical data—the latter referring to both serum intact parathormone (PTH: normal range, 15–65 pg/mL) and calcium (Ca: normal range, 8.4–10.2 mg/dL)—48 of the 67 patients had pHPT (PTH, 223.14 ± 154.10 pg/mL; Ca, 11.21 ± 0.63 mg/dL), including 1 patient affected by multiple endocrine neoplasia (MEN) 1a syndrome, whereas 19 had sHPT (PTH, 1,103 ± 654.69 pg/mL; Ca, 10.12 ± 1.04 mg/dL). Sixteen of the latter patients showed chronic renal failure (15 on hemodialysis and 1 on peritoneal dialysis), whereas 3 had had renal transplants. All patients with sHPT had been scheduled for surgery because of their unresponsiveness to oral calcitriol or severe osteodystrophy.

Sixty-three of the 67 patients were at their first parathyroid surgical exploration, whereas 4 patients—2 with pHPT, including the patient with MEN 1a syndrome, and 2 with sHPT—had persistent (n =1) or recurrent (n = 3) HPT after parathyroidectomy, which was associated with total thyroidectomy in the persistent case.

Before scintigraphy, all patients submitted to a careful clinical examination to verify the presence of palpable thyroid nodules and to neck ultrasonography for evaluation of thyroid and parathyroid glands; 7 of 67 patients—5 with pHPT and 2 with sHPT—had a nodular goiter.

Bilateral surgical neck exploration was performed on each patient within 1 mo of scintigraphy by an experienced parathyroid surgeon and abnormal parathyroid glands were found in all cases at surgery. In 3 patients, parathyroidectomy was associated with total (n = 1) or partial (n = 2) thyroidectomy for benign nodular goiter and with the surgical resection of the transplanted kidney affected by chronic pyelonephritis in another case. In all 67 patients, abnormal parathyroid glands were found at surgery.

Postoperative serum PTH and Ca levels were determined in each patient, and surgical success in each patient was defined by their significant reduction postoperatively. PTH values fell to within the normal range in all 48 pHPT patients (mean value, 35 ± 9 pg/mL) and in 17 of 19 sHPT patients (mean value, 39 ± 10.2 pg/mL); Ca also decreased in the former (mean value, 8.9 ± 0.5 mg/dL) and in the latter (mean value, 9.1 ± 0.2 mg/dL) patients. In 2 sHPT patients, HPT persisted after surgery with PTH levels remaining high (835 and 342 pg/mL, respectively) and both of these patients were rechecked by scintigraphy after surgery.

All patients gave their written informed consensus before parathyroid scintigraphy.

RESULTS

Scintigraphic Findings

The results of both planar parathyroid scintigraphy and neck P-SPECT in pHPT patients are reported in Table 2.

View this table:

• View inline
• View popup

TABLE 2

Planar Parathyroid Scintigraphy and Neck P-SPECT Results Related to Histopathologic Findings in Patients with pHPT

Planar scintigraphy was positive in 43 of 48 patients, detecting 43 of 49 lesions. P-SPECT was true-positive in all 48 patients, identifying 48 of 49 lesions and missing only 1 hyperplastic gland, which was also negative on planar imaging.

The 2 pHPT patients with recurrent disease were true-positive with both procedures.

Planar scintigraphy was false-negative in 4 patients with solitary parathyroid adenomas that were <10 mm in size (range, 6.1–8 mm) and <500 mg in weight (range, 261–360 mg), whereas all of these were clearly revealed on P-SPECT (one of these cases is illustrated in Fig. 1). None of these 4 patients had concomitant thyroid nodules. A planar scan was also false-negative in the patient with 2 omolateral eutopic hyperplastic parathyroid glands without associated thyroid nodules, missing both lesions that were each 6 mm in size and 300 mg in weight; in this patient, P-SPECT revealed only 1 of the lesions sited behind the inferior pole of the thyroid left lobe, missing the lesion localized at the top of the same lobe. This was the only P-SPECT false-negative result in this group of patients.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

A 55-y-old female patient affected by pHPT with small parathyroid adenoma (size, 6.1 mm; weight, 261 mg) sited behind inferior pole of right thyroid lobe, negative on planar ^99mTc-pertechnetate (A), ^99mTc-tetrofosmin (B), and subtraction (C) scintigraphy and clearly revealed (arrow) on coronal P-SPECT (D).

Thus, planar scintigraphy showed a lower per-lesion sensitivity than P-SPECT in pHPT patients, but the difference was not statistically significant (P > 0.05; McNemar test), as shown in Table 2. The higher sensitivity of P-SPECT with respect to planar procedures was independent of the different type of method, since it was more sensitive with respect to ^99mTc-MIBI double-phase scintigraphy (94.4% vs. 77.8%) and ^99mTc-pertechnetate/^99mTc-tetrofosmin subtraction scintigraphy (100% vs. 93.5%); however, the difference in sensitivity was higher with respect of the former rather than the latter.

The only pHPT patient with a parathyroid carcinoma was rechecked after surgery, and no residual uptake was ascertained in the neck by either scintigraphic procedure, confirming normal clinical and hormonal data.

The results of both planar parathyroid scintigraphy and P-SPECT in the 19 sHPT patients are reported in Table 3.

View this table:

• View inline
• View popup

TABLE 3

Planar Parathyroid Scintigraphy and Neck P-SPECT Results Related to Histopathologic Findings in Patients with sHPT

Planar parathyroid scintigraphy detected 40 of 51 lesions and missed 11 lesions: 1 solitary adenoma, 15 mm in size and 750 mg in weight, and 10 eutopic hyperplastic glands. Four of the latter were >10 mm in size (range, 14–20 mm), weighed >500 mg (range, 650–1,650 mg), and were sited behind the upper two thirds of the thyroid, whereas the remaining 6 were ≤10 mm in size (range, 6–10 mm), weighed ≤500 mg (range, 210–500 mg), and were sited behind the upper two thirds of the thyroid (n = 4) or behind or at the bottom of the thyroid inferior poles (n = 2).

P-SPECT was true-positive in 47 of 51 lesions and missed 4 eutopic hyperplastic glands, 3 sited behind the upper two thirds of the thyroid and 1 behind the thyroid inferior pole, all also negative on planar scintigraphy. Three of 4 glands were 10 mm in size and ≤500 mg, whereas the fourth gland was larger (size, 20 mm; weight, 1,650 mg).

Only 1 of the sHPT patients with false-negative findings on planar imaging or P-SPECT had concomitant thyroid nodules, but these were not superimposed on the missed hyperplastic glands.

The per-lesion sensitivity of P-SPECT was significantly (P < 0.05) higher than that of planar imaging in sHPT patients (McNemar test), as shown in Table 3. When P-SPECT was compared separately with ^99mTc-MIBI double-phase scintigraphy and with ^99mTc-pertechnetate/^99mTc-tetrofosmin subtraction scintigraphy, we found that the difference in sensitivity was higher with respect to ^99mTc-MIBI double-phase cases (93.3% vs. 73.3%) than ^99mTc-pertechnetate/^99mTc-tetrofosmin subtraction scintigraphy cases (91.7% vs. 80.5%).

Planar and P-SPECT were concordantly positive in 10 of 17 sHPT patients with multigland disease, including 1 with persistent HPT, whereas only P-SPECT correctly defined both the exact number of lesions and their localization in 4 of 17 cases; 1 of these latter cases is illustrated in Figure 2. The 2 procedures, both positive, underestimated the number of lesions in a further 2 of 17 patients, one of whom with recurrent HPT. In the remaining patient, both planar imaging and P-SPECT revealed 2 hot areas that did not correspond to the site of the 2 hyperplastic glands excised at surgery: These were 10 mm in size each and weighed 480 and 500 mg, respectively. Given persistently high PTH serum levels after surgery, the patient repeated scintigraphy, which confirmed the areas of increased uptake seen at the first observation.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

A 67-y-old male patient affected by sHPT with 4 hyperplastic glands sited behind upper two thirds of right thyroid lobe (size, 6 mm; weight, 274 mg), below inferior pole of right thyroid lobe (size, 15 mm; weight, 685 mg), behind upper two thirds of left thyroid lobe (size, 9 mm; weight, 429 mg), and below inferior pole of left thyroid lobe (size, 17 mm; weight, 850 mg). All glands were clearly revealed (arrows) on different multiple coronal P-SPECT slices (D), whereas planar ^99mTc-pertechnetate (A), ^99mTc-tetrofosmin (B), and subtraction (C) scintigraphy identified only the 2 sited below inferior pole of 2 thyroid lobes (arrows), missing the other 2, which were smaller in size and weight and were sited behind upper two thirds of 2 thyroid lobes.

Another sHPT patient with 2 hyperplastic glands, both true-positive on planar imaging and P-SPECT before surgery, was also rechecked after surgery because of persistently high PTH serum levels. Only P-SPECT revealed a small area of increased uptake behind the inferior pole of the right thyroid lobe; this area was contiguous with 1 of the excised glands and it had not been visualized on the first observation, probably because it was obscured by the size of the gland (size, 20 mm; weight, 1,000 mg), which had appeared highly active on scintigraphy.

Up to now, neither of the above 2 patients with persistent HPT has undergone further surgery.

We compared planar imaging and P-SPECT performance for a correct preoperative diagnosis in all 67 patients enrolled in this study. Only P-SPECT revealed hyperfunctioning glands in 6 cases (5 with pHPT, 1 with sHPT), whereas in 4 cases (all with sHPT) it revealed more lesions than planar imaging. In no case was planar imaging more sensitive than P-SPECT.

In addition, in the patients for whom both procedures were concordantly positive, P-SPECT gave a more distinct visualization of the lesions in 21 cases, 14 with pHPT adenomas, including 5 ≤10 mm (1 of these cases is illustrated in Fig. 3) and 9 >10 mm, and in 7 patients with sHPT multigland disease. In all of these cases, a sharp separation of the hyperfunctioning glands from the surrounding normal activity was observed on P-SPECT, which also precisely defined lesion depths. Moreover, neither P-SPECT nor planar imaging had false-positive results in either pHPT or sHPT patients.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

A 65-y-old female patient affected by pHPT with small parathyroid adenoma (size, 6.8 mm; weight, 290 mg) sited behind upper two thirds of left thyroid lobe, positive (arrow) on both early (A) and delayed (B) planar ^99mTc-MIBI scintigraphy and on coronal P-SPECT (C), but more clearly visualized by the latter.

Finally, to date, none of the remaining 53 of 110 patients not included in this study who underwent both planar and neck P-SPECT procedures has had neck surgical exploration; 13 of these 53 patients had pHPT and the remaining 40 had sHPT. One of the 13 pHPT patients, in whom both procedures revealed 1 area of increased tracer uptake, refused surgical treatment; 3 of 13 patients, negative for neck hyperfunctioning parathyroid lesions on both planar imaging and P-SPECT, had an adenoma in the upper mediastinum revealed on planar imaging and have undergone surgery for this; the remaining 9 patients, all positive on P-SPECT for 1 area and 8 also positive on planar imaging, have been scheduled for neck surgery.

Four of the 40 patients with sHPT were positive for 13 areas on P-SPECT and for 11 areas on planar imaging and have been scheduled for surgery; 35 of the remaining 36 patients were positive on P-SPECT for 91 areas and 30 of these were also positive on planar imaging for 72 areas; however, none of these 36 patients has had neck surgical exploration either because they were well controlled with medical treatment (n = 32) or because of high surgical risk (n = 4).

DISCUSSION

P-SPECT is recognized as having very high spatial resolution, superior to that achieved with conventional SPECT with a parallel-hole collimator due to the more favorable geometric properties of the cone beam collimator, as demonstrated in phantom studies (22).

The high spatial resolution of P-SPECT has also been demonstrated in small laboratory animal imaging (23–26) and confirmed in clinical studies focused on thyroid (22,27) and bone (28) diseases and, more recently, on breast cancer axillary lymph node metastasis detection (29,30), in which P-SPECT gave useful anatomic details (28) and increased the sensitivity of conventional planar imaging and SPECT imaging (22,29).

Data regarding the use of P-SPECT in the detection of hyperfunctioning parathyroid lesions have also been reported (17–20) and they are encouraging since the method showed preoperatively higher sensitivity values than conventional planar scintigraphic procedures; however, these data only refer to preliminary studies.

In this study, we used high-resolution 180° P-SPECT of the neck combined with conventional planar ^99mTc-MIBI double-phase scintigraphy or ^99mTc-pertechnetate/^99mTc-tetrofosmin subtraction parathyroid scintigraphy, both of the latter acquired with a large-field-of-view parallel-hole collimator, in both pHPT and sHPT; in all cases, the scintigraphic findings were correlated with histology. At the beginning of the study, we used ^99mTc-MIBI double-phase planar scintigraphy. However, we then chose to use the subtraction technique, since we considered it more sensitive and also time saving as it does not require delayed images and is thus better tolerated by patients, particularly sHPT patients with renal failure.

P-SPECT images were acquired with a tilted detector using a specific software originally developed for the study of thyroid diseases (22). For several years, we have also used this procedure with some personal modifications for the detection of axillary lymph node metastases from breast cancer with ^99mTc-tetrofosmin as the tumor-seeking agent (29,30).

We obtained very high-quality and high-resolution P-SPECT images, also maintaining an acceptable acquisition time for the patients, using the approximate Feldkamp algorithm (21) and the backprojection filter method for image reconstruction.

In our series, P-SPECT proved to be a highly sensitive method for the detection of hyperfunctioning parathyroid glands, since it globally identified 97.9% and 92.1% of pHPT and sHPT lesions, respectively, sited in the neck in either eutopic or ectopic sites. Most of the patients were at their first surgical exploration and some were scheduled for further surgery because of persistent or recurrent disease.

P-SPECT sensitivity was higher than that of planar imaging in the identification of lesions in both pHPT and sHPT patients, independent of the different type of procedure used, ^99mTc-MIBI double-phase scintigraphy or ^99mTc-pertechnetate/^99mTc-tetrofosmin subtraction scintigraphy, with a statistical difference for sHPT. This finding always permitted a more distinct visualization and a more correct localization of hyperfunctioning glands than planar imaging when both were positive, with a sharp separation from the surrounding normal activity in the presence of a single adenoma or multiple hyperplastic glands, especially when small and light-weight or deeply located in the neck.

In confirmation, P-SPECT, which revealed all adenomas, increased planar sensitivity, especially in the detection of those ≤10 mm in size or ≤500 mg in weight, whereas planar scanning missed 30.8% of these. Furthermore, P-SPECT also identified more glands with hyperplasia, with respect to planar imaging, globally detecting 90.7%, whereas planar imaging was true-positive in 77.8%.

Thus, P-SPECT false-negative findings essentially referred to glands with hyperplasia that are generally more difficult to identify than adenomas, as demonstrated by several authors using conventional planar imaging or SPECT procedures, irrespective of the radiotracer and the acquisition protocol used (9,31–35), with small size generally considered the most important limiting factor (8).

However, biologic factors (8), such as low proliferative cell activity and a low quantity of mitochondria, might also be associated with a low radiotracer uptake, thus reducing the possibility of visualizing hyperplastic glands, particularly in sHPT patients in whom biochemical factors may also interfere—such as normal serum Ca combined with only slightly increased PTH levels.

Moreover, the overexpression of P-glycoprotein or other multidrug-resistance proteins could also be probable causes of false-negative findings in sHPT hyperplasia (36), as also demonstrated in adenomas (37,38).

In this study, the hyperplastic glands missed on P-SPECT were relatively small in diameter and weight, except in 1 case. However, it is possible that besides size, certain of the above-mentioned limiting factors could cause interference, since some hyperplastic glands smaller than those missed were detected on P-SPECT, and also on planar imaging, although the latter proved to be more affected by size than P-SPECT.

In confirmation, planar imaging was false-negative in 6 small glands, 3 of which also were missed on P-SPECT, but it was also false-negative in 4 large glands, only 1 of which was missed on P-SPECT. Thus, it could also be hypothesized that P-SPECT could partially overcome low gland uptake in large glands that were false-negative on planar imaging, most probably due to the above biologic or biochemical factors.

Our data seem to suggest that P-SPECT could represent a reliable complementary tool to conventional planar scintigraphy, also giving more information than planar imaging in some cases. In our series, only P-SPECT permitted a preoperative localization of hyperfunctioning lesions in 8.9% of all patients and revealed a higher number of hyperfunctioning sHPT lesions in a further 6% of cases, whereas in no case did planar imaging give more information than P-SPECT. Furthermore, neither P-SPECT nor planar imaging had false-positive findings.

Thus, P-SPECT could represent a useful guide to the surgeon in the localization of abnormal parathyroid glands in patients with pHPT, since in our study it correctly identified all lesions found at surgery except 1 small hyperplastic gland. Planar imaging also proved highly sensitive in the detection of large, but not small, adenomas in our patients, although it did not always give a distinct visualization of the lesions or define their site and depth; thus, the addition of P-SPECT to conventional planar scintigraphy seems particularly indicated in selected patients with pHPT, both when planar imaging is negative, although suspect clinical and biochemical signs of disease are present, and when it is positive but inconclusive for the localization of gland site.

Moreover, P-SPECT correctly predicted the status found at surgery in 82.3% of patients with sHPT with multigland disease, whereas planar imaging correctly predicted the status in 58.8% of cases. These results, although limited to a small number of patients, appear extremely encouraging and suggest the addition of this new procedure as a better guide to the surgeon in sHPT patients than planar imaging alone.

At present, given the difficulty of identifying hyperplastic glands by planar scintigraphy, and also by conventional SPECT, and on the assumption that in sHPT patients 4 hyperplastic glands may be involved, all of which need to be removed, the usefulness of preoperative conventional parathyroid scintigraphy in this group of patients is markedly reduced and its use is limited to cases with recurrent or persistent disease.

However, the combination of a highly sensitive, high-resolution imaging procedure such as P-SPECT and an expert surgeon lead to a significant increase in the number of hyperplastic glands found at surgery and, thus, to the reduction of recurrences due to incomplete surgical treatment. Finally, in our experience, parathyroid P-SPECT proved to be simple, time saving, and easy to read.

CONCLUSION

P-SPECT appears a highly sensitive method in the preoperative localization of hyperfunctioning parathyroid glands sited in the neck in both pHPT and sHPT. In our cases, this procedure increased the sensitivity of conventional planar parathyroid scintigraphy, particularly in small adenomas and in hyperplastic glands, improving the quality and the resolution of images and also permitting a better topographic localization.

Thus, we suggest a more extensive clinical application of this new procedure as a complementary tool to conventional planar imaging in selected pHPT patients in whom the latter is negative or inconclusive, whereas the use of P-SPECT could be widely suggested in sHPT patients since it proved to be a more effective aid than planar imaging to surgical planning. However, these suggestions need to be confirmed by further studies in a larger number of patients.