Shear wave elastography can improve the preoperative diagnosis of parathyroid carcinoma or atypical parathyroid tumor: initial experience

Abstract

Background

Parathyroid carcinoma (PC) and atypical parathyroid tumor (APT) are rare malignant lesions characterized by high recurrence and metastasis rates, and accurate preoperative diagnosis remains a clinical challenge. The purpose of this study was to explore the diagnostic value of preoperative grayscale ultrasound combined with shear wave elastography (SWE) for such lesions.

Methods

In this retrospective case-control study, patients with primary hyperparathyroidism and suspicious parathyroid abnormality on ultrasonography were recruited. All the lesions were assessed by SWE before surgery. Sonographic features and SWE performance along with demographic and related clinical parameters between PC/APT and PA groups were analyzed.

Results

This study included 21 patients with PC or APT matched to 84 patients with PA who underwent parathyroid surgery. The best subset selection method and multiple logistic regression analysis identified the sonographic features of intra-nodal fibrous bands and the ratio of mean shear wave velocity in tumor (SWVmean)/shear wave velocity in thyroid parenchyma (SWVth) as independent factors for PC/APT prediction (odds ratio [OR] = 38.536, 95% confidence interval [CI] = 1.188–1250.066, P = 0.04; OR = 248.783, 95%CI = 1.314–47100.432, P = 0.04, respectively). On receiver operating characteristic curve analysis, the optimal cutoff value of the SWVmean/SWVth ratio for predicting PC/APT was 1.25, with a sensitivity of 83.33% and specificity of 95.56% (AUC = 0.894; 95% CI, 0.783–0.960; P < 0.0001). Combined use of the SWVmean/SWVth ratio and intra-nodal fibrous bands showed significantly improved performance for the identification of PC/APT (C statistic = 0.976, P = 0.02; sensitivity = 1; specificity = 0.889; positive predictive value = 0.706, negative predictive value = 1).

Conclusions

Sonographic features of combined intra-nodal fibrous bands and the SWVmean/SWVth ratio are suitable markers for accurate preoperative prediction of PC or APT.

Clinical trial number

Not applicable.

Background

Parathyroid carcinoma (PC) is a rare endocrine malignancy that accounts for less than 1% of all cases of primary hyperparathyroidism (PHPT) in some western countries, although its incidence is reported to be greater than 5% in China, Japan, and Italy [1–3]. Moreover, recent studies have shown an increase in its occurrence [4, 5]. The recurrence and metastasis rates of PC are very high [6, 7], with more than half of PC cases reported in the literature experiencing recurrence or metastasis, resulting in a 5-year survival rate of less than 50% [7–9]. Uncontrollable hypercalcemia due to recurrence and metastasis of PC is a major cause of poor prognosis and mortality among patients afflicted with this malignancy [10, 11]. Surgical removal remains the most effective and reliable treatment for this malignancy [12, 13]. Unlike local parathyroidectomy for parathyroid adenoma (PA), en bloc resection, which comprises tumor resection, ipsilateral thyroid lobectomy, and adjacent cervical muscle resection, is recommended as the initial surgery for PC to reduce the risk of recurrence and metastasis [6, 14, 15]. However, the rarity of PC has complicated recognition of such lesions. Although some studies have mentioned that the typical manifestations of PC include increased tumor size and significantly elevated serum calcium and PTH levels, these features lack high specificity, and the values reported in various literatures vary greatly without unified criteria [9, 16–19]. Preoperative or intraoperative histological or cytological biopsy of suspected lesions is a possible method for determining whether a parathyroid lesion is benign or malignant. However, puncture can induce the appearance of atypical features, such as envelope infiltration, cellular pleomorphism, fibrous bands, and an increased proliferation index similar to those observed for atypical parathyroid tumor (APT) or PC, which may mislead pathologists in the diagnosis after removal of parathyroid tumors [20]. In addition, the susceptibility of PC to spreading along the needle tract or with implantation and metastasis limits the application of such biopsy techniques [21–23]. Intraoperative diagnosis is based on the surgeons’ findings during surgery, such as the invasive features of the tumors [14], and intraoperative histological examination of frozen sections contributes little to the diagnosis of PC [24]. Therefore, most PC is diagnosed postoperatively by histological examination. As a result, PC is often misdiagnosed as PA preoperatively, resulting in a missed opportunity to perform radical surgery initially. Thus, a method for accurate preoperative detection of PC is needed.

Shear wave elastography (SWE), as an emerging ultrasound technique that uses transverse waves to measure the stiffness of tissues, provides a quantitative method for the objective evaluation of lesion stiffness. It can reflect the parenchymal composition of lesions and further reveal their biological tissue characteristics. The value of SWE in the detection of other tumor types has been previously addressed [25, 26]. Recently, several studies have compared SWE parameters between parathyroid and thyroid glands. The findings demonstrated that the maximum elasticity and mean elasticity values of parathyroid glands on SWE were significantly lower than those of the adjacent thyroid parenchyma [27]. Prior investigations have documented that SWE confers high diagnostic accuracy in the differentiation of parathyroid lesions from adjacent cervical soft tissues [28]. Furthermore, several studies have compared SWE parameters between parathyroid adenomas and parathyroid hyperplasia, revealing that the minimum elasticity, mean elasticity, and maximum elasticity values were significantly higher in adenomas than in hyperplastic lesions [29]. Nevertheless, studies investigating the utility of SWE in parathyroid lesions remain comparatively limited. It is widely recognized that SWE yields improved diagnostic value when used in combination with conventional ultrasound.

The present study aimed to explore the role of preoperative ultrasound combined with SWE in predicting parathyroid tumors which may have poor biological behavior and risk of recurrence preoperatively.

Methods

Patients and biochemical data

Between May 2017 and April 2024, a total of 21 patients who underwent surgical resection at our institution and were postoperatively diagnosed with primary PC or APT on histopathological examination were enrolled in this retrospective study. Of these patients, 19 were pathologically confirmed as PC and 2 as APT. The cohort comprised 14 females and 7 males, with a mean age of 55 ± 17 years. The definitive diagnosis of PC was based on the following clinicopathological features outlined in the World Health Organization 2017 and 2022 criteria [30, 31], including: (i) angioinvasion (vascular invasion) characterized by tumor invading through a vessel wall and associated thrombus, or intravascular tumor cells admixed with thrombus, (ii) lymphatic invasion, (iii) perineural (intraneural) invasion, (iv) local malignant invasion into adjacent anatomic structures, or (v) histologically/cytologically documented metastatic disease. In addition to concerning clinical and biochemical features, atypical parathyroid tumors may exhibit histological abnormalities commonly seen in parathyroid carcinoma, including band-like fibrosis, adherence to adjacent structures, and a Ki-67 labeling index > 5%.The 2022 World Health Organization (WHO) classification of parathyroid tumors renamed “atypical parathyroid adenoma” as “atypical parathyroid tumor” to refer to parathyroid neoplasms of uncertain malignant potential [31], cases treated before these guidelines were published were reviewed and reclassified. In this study, PC and APT were classified into one group because these two entities are often difficult to distinguish definitively without long-term follow-up [32]. Moreover, en-bloc resection is usually warranted for tumors with malignant potential [7].

A selected subset of 84 patients (51 females and 33 males, mean age 51 ± 14 years) with pathologically confirmed parathyroid adenoma (PA) in the same period were enrolled as the control group. The exclusion criteria were: secondary or tertiary hyperparathyroidism, familial syndromes (multiple endocrine neoplasia [MEN] type 1, MEN2A, and hyperparathyroidism-jaw tumor syndrome), initial surgery at another institution, patients without preoperative SWE examination or with invalid SWE measurements due to deep lesion location or choosing the conservative management were excluded from this study.

For all patients, demographic and biochemical data were collected. Biochemical data included preoperative intact parathyroid hormone (iPTH) and serum calcium (Ca), phosphorus (P), and alkaline phosphatase (ALP) levels. The relevant biochemical parameters collected at the latest time point before surgery and without prior pharmacological treatment were used for analysis.

Ultrasonography and SWE

For all patients, preoperative ultrasonic examinations were conducted by sonographers with more than 8 years of experience in neck ultrasound. The ultrasound apparatus used was the Aplio 500 system (Canon Medical, Inc., Otawara, Tochigi, Japan) with a PLT-1005BT 15L4 (5-14 MHz) linear probe, equipped with SWE software. Images from conventional ultrasound of the patient’s thyroid and parathyroid region along with SWE images were acquired. All patients were examined in the supine position using conventional ultrasound to locate the target lesions responsible for PHPT. For the target lesions, the location, size, depth/width (D/W) ratio, shape, echogenicity, homogeneous/heterogeneous classification, presence or absence of thick capsule or infiltration, intra-nodal fibrous bands, calcification, or cystic change were recorded. The D/W ratio was calculated as the ratio of the depth to the wide diameter on a cross-sectional view. The shape was categorized as “oval” or “irregular”. Echogenicity was categorized as “hypoechoic” or “isoechoic.” Intra-nodal fibrous bands were defined by the presence of “thick hyperechoic bands (≥ 1 mm) within the nodule.”

A senior sonographer with systemic training in SWE imaging performed the examination. Guideline stated by WFUMB has been complied with during operation. The candidate lesions stiffness were presented as shear wave velocity in meters per second. In the assessment, the elasticity scale range was set in between 0 and 10 m/s. Upon switching to SWE mode, the probe was kept stable and the sampling frame was adjusted so that the region of interest (ROI) contained the parathyroid gland and adjacent thyroid tissue. After the image had stabilized for 3–5 s, the shear wave was triggered for image acquisition. The criteria for a qualifying SWE image were the absence of visible stress artifacts above the sampling frame and color filling at least three-quarters of the sampling frame. For large lesions with a complex composition, the ROI was placed inside the lesion to avoid areas of cystic change or calcification. For measurement of shear wave velocity (SWV) values, the same ROI size was maintained and different representative areas within the lesion were selected as T1, T2, T3, and T4, while the same depth of adjacent thyroid parenchyma was selected for R. Care was taken to ensure the quality control requirement for the standard deviation (SD, m/s) to be < 20% of the measured value (Ave, m/s) was maintained to ensure the data reliability. The maximum SWV in the tumor (SWVmax), mean SWV in the tumor (SWVmean), SWV in the thyroid parenchyma (SWVth) were recorded and analysed.

Sonographic features of the abnormal lesions enrolled in the study from the PACS system were evaluated by two senior radiologists according to a pre-agreed protocol. In case of disagreement, consensus was achieved through discussion with a specialist with 20 years of experience in thyroid and parathyroid imaging.

This study was approved by the Institutional Review Board of Beijing Chaoyang Hospital, Capital Medical University (2025-ke-26), all methods were carried out in accordance with relevant guidelines and regulations. Informed consent was waived owing to the study’s retrospective nature.

Statistical analysis

Data analysis was performed using SPSS 23.0 (IBM, Armonk, NY, USA), Prism 5.00 (San Diego, CA, USA), and MedCalc 15.8 (MedCalc Software, Ostend, Belgium). The normality of variables was verified using the Kolmogorov–Smirnov test. Data for continuous variables are presented as mean ± standard deviation or median and interquartile range. Data for categorical variables are presented as frequency and percentages. Comparisons between the PA and PC/APT groups were conducted using independent-samples t test (for normally distributed variables) and Mann–Whitney U test (for variables with a skewed distribution). Categorical variables were compared using the chi-square test or Fisher’s exact test. Univariate and multivariate logistic regression techniques were applied to identify independent risk factors related to PC. After disregarding the effect of collinearity, variables with a P value < 0.05 on univariate analysis were included in the multivariate model. Receiver operating characteristic (ROC) curves were constructed, and area under the curve (AUC) values were estimated. The DeLong test was used to compare the diagnostic performance of different models. Values of P ≤ 0.05 or lower indicated statistical significance.

Results

Comparison of demographic, clinical, and ultrasound characteristics between PC/APT and PA groups

A total of 105 patients with PHPT were included in this study (Fig. 1). Of these patients, 65 (61.9%) patients were female, and 40 (38.1%) patients were male. After surgery, 21 of these patients were diagnosed with PC/APT, including 2 cases of PC on one side with concomitant PA on the opposite side. Another case had ectopic PC in the pericardial cavity and was excluded from the statistical analysis because the lesion had not been identified by ultrasound. The control group consisted of 84 PA cases, 2 of which were multiglandular diseases, with 1 case of two PA lesions involving two bilateral parathyroid glands, and 1 case with three PA lesions affecting three bilateral parathyroid glands.

Fig. 1.

Flow chart of patient inclusion

No significant differences in gender and age were observed between the PA and PC/APT groups (P = 0.723 and 0.299, respectively). Also, no significant differences were observed in the blood Ca, iPTH, P, and ALP levels between the two groups (P = 0.655, 0.130, 0.054, and 0.393, respectively; Table 1).

Table 1.

Comparison of demographic, clinical, and ultrasound characteristics between the PC/APT and PA groups

	PC/APTgroup(n = 20)	PA group (n = 84)	t / U / Z /χ2	P
Gender (male/female)	7/13	33/51	0.125	0.723
Mean age (years)	55 ± 18	51 ± 14	1.048	0.299
Serum Ca (mmol/L)	2.70 ± 0.35	2.73 ± 0.22	-0.449	0.655
Serum iPTH (pg/mL)	312.2 (120.2, 473.5)	182.1(118.1,501.1)	-1.515	0.130
Serum P (mmol/L)	0.83 ± 0.22	0.92 ± 0.15	-1.960	0.054
Serum ALP (IU/L)	125 (79, 223)	111 (74, 184)	-0.855	0.393
Size (cm)	2.2 ± 0.7	2.3 ± 1.0	-0.694	0.490
D/W ratio	1.08 (0.80, 1.26)	0.78 (0.67, 0.96)	-2.524	0.012
Location (%)			1.730	0.913
Left (superior)	3 (15.0)	16 (19.0)
Left (inferior)	6 (30.0)	20 (23.8)
Right (superior)	4 (20.0)	17 (20.2)
Right (inferior)	6 (30.0)	25 (29.8)
Intrathyroidal (left)	0 (0)	4 (4.8)
Intrathyroidal (right)	1 (5.6)	2 (2.4)
Shape (%)			7.319	0.007
Oval	13 (65.0)	75 (89.3)
Irregular	7 (35.0)	9 (10.7)
Uniformity (%)			30.522	< 0.001
Homogeneous	2 (10)	64 (72.6)
Heterogeneous	18 (90.0)	20 (27.4)
Echogenicity (%)			7.831	0.051
Hypoechoic	12 (60.0)	73 (86.9)
Isoechoic	8 (20.0)	11(13.1)
Thick capsule or infiltration (%)			49.583	< 0.001
No	5 (25.0)	79 (94.0)
Yes	15 (75.0)	5(6.0)
Intra-nodal fibrous bands (%)			42.451	< 0.001
No	3 (15.0)	73 (86.2)
Yes	17 (85.0)	11 (13.8)
Calcification (%)			17.472	< 0.001
No	16 (80.0)	84 (100)
Yes	4(20.0)	0 (0)
Cystic change (%)			0.623	0.430
No	13 (65.0)	62 (72.8)
Yes	7 (35.0)	22 (26.2)
SWVmax (m/s)	3.82 (2.74, 6.15)	2.23 (1.91, 2.82)	-3.416	0.001
SWVmean (m/s)	3.16 (2.36, 4.88)	1.86 (1.69, 2.25)	-4.004	< 0.001
SWVth (m/s)	2.12 (1.94, 2.30)	2.22 (2.02, 2.49)	-1.067	0.286
SWVmean/SWVth ratio	1.64 (1.29, 2.01)	0.89 (0.76, 0.95)	-4.161	< 0.001

*The relevant parameters shown in Table 1 are from the data of 84 patients with PA and 20 patients with PC/APT. One case of ectopic PC in the pericardium was not shown by preoperative ultrasound and, thus, was not included in this statistical analysis. PC=parathyroid carcinoma, APT=atypical parathyroid tumor, PA=parathyroid adenoma, D/W=ratio of depth to wide diameter on cross-section, SWV=shear wave velocity

The sonographic features of the PC/APT and PA lesions are presented in Table 1; Fig. 2. The D/W ratio was significantly higher in the PC/APT group than in the PA group (P = 0.012). Moreover, irregular shape was observed more frequently in the PC/APT group than in the PA group (P = 0.007). Regarding the uniformity of echogenicity, the incidence of heterogeneous nodules was higher in the PC/APT group than in the PA group (P < 0.001). Also, a thick capsule or infiltration and intra-nodal fibrous bands were observed more frequently in the PC/APT group than in the PA group (P < 0.001), and calcification was more common in the PC/APT group than in the PA group (P < 0.001). Intra-nodule cystic change was present in 22 (26.2%) PA lesions versus 7 (35.0%) PC/APT cases, with no significant difference in the observation of this parameter between the two groups (P = 0.430). Also, no differences in tumor size, location, and echogenicity were observed between the two groups (P = 0.490, 0.913, and 0.051, respectively).

Fig. 2.

Gray-scale sonographic patterns of parathyroid carcinoma (PC/APT) lesions. (a) An irregularly shaped nodule with D/W = 1.20. (b) Heterogeneous echogenicity (arrow indicates nest-like hyperechogenicity). (c) An uneven thick and unsmooth capsule. (d) Tumor breakage through the capsule with infiltration into the adjacent thyroid parenchyma. (e) Hyperechoic intra-nodal fibrous bands. (f) Gross calcification in tumor

The SWV was higher in PC/APT than in the adjacent thyroid parenchyma (Fig. 3), whereas the SWV in PA was slightly lower than that in the adjacent thyroid parenchyma (Fig. 4). Significant differences in SWV also were observed between the PA and PC/APT groups in terms of SWVmax (P = 0.001), SWVmean (P < 0.001), and the SWVmean/SWVth ratio (P < 0.001; Table 1; Fig. 5).

Fig. 3.

Gray-scale sonography and shear wave elastography images from a female patient with parathyroid carcinoma. (a) Irregular tumor breakage through the capsule with infiltration into the adjacent thyroid parenchyma. (b) SWV of PC was higher than that of adjacent thyroid parenchyma, and SWVmax, SWVmean, and the SWVmean/SWVth ratio were 8.15 m/s, 4.60 m/s, and 2.04, respectively. (c) Exposure of PC and adjacent thyroid parenchyma and connective tissue during surgery. (d) PC and peripheral intravenous tumor thrombosis (hematoxylin and eosin, 50×)

Fig. 4.

Gray-scale sonography and shear wave elastography images from a male patient with parathyroid adenoma, whose serum calcium was 2.76 mmol/L, PTH 2000 pg/mL, phosphate 0.72 mmol/L, and ALP 2412 U/L. The patient was followed up for more than 40 months with no recurrence or metastasis. Conventional grayscale sonography showed that the lesion measured 4.0 × 1.7 × 2.3 cm, with regular shape, thin capsule, and homogeneous internal texture (a and b). c. SWV of PC was lower than adjacent thyroid parenchyma, and SWVmax, SWVmean, and the SWVmean/SWVth ratio were 1.86 m/s, 1.67 m/s, and 0.63, respectively (c and d)

Fig. 5.

Comparisons of SWVmax, SWVmean, and the SWVmean/SWVth ratio between the PC/APT and PA groups

Risk factors related to PC/APT

On univariate analysis, shape, uniformity of echogenicity, thick capsule or infiltration, intra-nodal fibrous bands, SWVmax, SWVmean, and the SWVmean/ SWVth ratio were associated with PC/APT. Subsequent multivariate analysis identified intra-nodal fibrous bands and the SWVmean/ SWVth ratio as independent risk factors for PC/APT prediction (odds ratio [OR] = 38.536, 95% confidence interval [CI] = 1.188–1250.066, P = 0.04; and OR = 248.783, 95% CI = 1.314–47100.432, P = 0.04, respectively; Table 2).

Table 2.

Univariate and multivariate binary logistic regression analyses to identify independent risk factors related to PC/APT

Variable	OR (95% CI)	P	OR (95% CI)	P
D/W ratio	4.367 (0.970–19.656)	0.055
Shape	5.750 (1.396–23.687)	0.015	28.248 (0.691–1155.064)	0.078
Uniformity	53.833 (6.470–447.884)	< 0.001	0.948 (0.020–45.060)	0.978
Thick capsule or infiltration	54.833 (10.944–274.736)	< 0.001	1.401 (0.078–25.224)	0.819
Intra-nodal fibrous bands	30.714 (7.029–134.211)	< 0.001	38.536 (1.188–1250.066)	0.040
SWVmax	3.182 (1.511–6.704)	0.002
SWVmean	9.413 (2.283–38.810)	0.002
SWVmean/SWVth ratio	628.387 (16.072–24569.030)	0.001	248.783 (1.314–47100.432)	0.039

PC=parathyroid carcinoma, OR=odds ratio, CI=confidence interval, D/W=ratio of depth to wide diameter on cross-section, SWV=shear wave velocity

On ROC analysis, the optimal cutoff value for the SWVmean/SWVth ratio for predicting PC/APT was 1.25, with a sensitivity of 83.33% and a specificity of 95.56% (AUC = 0.894; 95% CI, 0.783–0.960; P < 0.0001; Fig. 6a). ROC curve analysis also confirmed that the combination of the presence of intra-nodal fibrous bands and the SWVmean/SWVth ratio offered improved performance for predicting PC/APT compared with either parameter alone (Fig. 6b; Table 3).

Fig. 6.

ROC analysis of factors associated with PC/APT and PA. (a) The optimal cutoff value for the SWVmean/SWVth ratio for predicting PC was 1.25, with a sensitivity of 83.33% and specificity of 95.56% (AUC = 0.894; 95%CI, 0.783–0.960; P < 0.0001). (b) ROC analysis of the incremental diagnostic value of the combination of the SWVmean/SWVth ratio with intra-nodal fibrous bands over intra-nodal fibrous bands alone. Model I: intra-nodal fibrous bands, AUC = 0.839; Model II: intra-nodal fibrous bands+SWVmean/SWVth ratio, AUC = 0.976. Model I versus Model II comparison P = 0.02

Table 3.

ROC threshold analysis for diagnostic performance of parameters for identifying PC/APT

Variable	AUC	Optimal cutoff	Sensitivity	Specificity	PPV	NPV
Model I	0.839	1	0.833	0.860	0.682	0.935
Model II	0.894	1.25	0.833	0.956	0.833	0.956
Model III	0.976	0.092	1	0.889	0.706	1

Model I, intra-nodal fibrous bands; Model II, SWVmean/SWVth ratio; Model III, intra-nodal fibrous bands + SWVmean/SWVth ratio. ROC=receiver operating characteristic, PC=parathyroid carcinoma, AUC=area under the curve, PPV=positive predictive value, NPV=negative predictive value

Discussion

Parathyroid carcinoma (PC), atypical parathyroid tumor (APT), and parathyroid adenoma (PA) represent the main parathyroid neoplasms, with markedly different surgical strategies but are frequently indistinguishable preoperatively in clinical practice [33]. Previous scholars believed that diagnosis was only possible after years of initial treatment and subsequent development of metastatic disease [34], and that preoperative diagnosis was almost impossible unless clear pathological manifestations of lymph node or distant metastasis were evident [5]. PC progresses rapidly with a recurrence rate of 40%–60% and a median overall survival of 14.3 years [33, 35]. Early preoperative suspicion or diagnosis of PC/APT as well as en bloc resection during the initial operation may be the key factors for improving the survival rates of affected patients [4]. The results of the present study identified the presence of intra-nodal fibrous bands and the SWVmean/SWVth ratio as independent predictors of PC or APT (P = 0.04) and show that the ability to identify PC or APT is significantly improved when these factors are considered in combination.

Gray-scale ultrasonography can be used to distinguish PC from PA to some extent. It was previously held that PC lesions are generally larger than PA lesions. According to a review of the National Cancer Database in 2000, the mean tumor size was 3.3 cm in a group of 286 PC cases reported over an 11-year period [36]. In the present study, the maximum tumor diameter among the 21 PC patients ranged from 1.1 to 3.2 cm (mean, 2.2 cm), which was smaller than in previous studies, and no significant difference in tumor size was observed between the PC/APT and PA groups (P = 0.490). To further reflect the different growth patterns of PC and PA, the D/W ratio was compared in reference to the D/W ratio proposed by Hara et al. [37]. A significant difference in the D/W ratio was observed between the PC and PA groups in relation to whether the nodule grew upright (P = 0.012). Irregularly shaped tumors were mostly present in the PC group compared with the PA group, but this parameter accounted for only 35.0% of the PC cases. Moreover, the results of this study revealed statistically significant differences in the presence of a thick capsule or infiltration, heterogeneous contexture, and hyperechoic intra-nodal fibrous bands between the PA and PC groups (all P < 0.001), and these findings are in agreement with the results of previous research [38]. An uneven, thick capsule or infiltration may be associated with pathological tumor invasion of the tegument or adhesions to the adjacent tissue, whereas heterogeneous echogenicity and intra-nodal fibrous bands within the tumor may be associated with a decrease in adipose tissue and the proliferation of fibrous tissue or reticular fiber structure destruction within the tumor [39–42]. Liu et al. studied the sonographic manifestations of recurrent or persistent PC and also found extensive hyperechoic fibrous separation in the abnormalities [43]. In the present study, calcification also was more frequently observed in the PC group than in the PA group according to univariate analysis. Multiple previous studies have also highlighted this feature of PC [44, 45]. Despite its low sensitivity (20.0% in the PC group), this feature offers a high specificity of 100%. Cystic change represents hemorrhage, necrosis, or cystic degeneration. However, no difference in the presence of cystic changes was observed between the PC and PA groups, which may be related to the small sizes of tumors among the included cases.

In the present study, SWE was innovatively applied to distinguish PC/APT from PA, and the results demonstrated significant differences in SWVmax, SWVmean, and the SWVmean/SWVth ratio between the PC/APT and PA groups. Moreover, the SWE propagation map specific to this ultrasound device visually demonstrated the heterogeneous changes in tumors. These differences may occur because the pathological composition of PC/APT is predominantly chief cellular, with lamellar, trabecular and nest-like cell growth, and is associated with reduced adipose tissue and fibrous tissue proliferation within the tumor. Up to now, the ability to assess PC/APT by elastography has been rarely addressed in the relevant literature. To our knowledge, only Liu et al. have recently explored the use of SWE in parathyroid neoplasms with malignantdiagnosis or uncertain malignant potential [46]. In their study, SWV values were quantified for PC/APT and PA groups, and the mean SWV values for these groups were 2.4 m/s vs. 1.9 m/s with a cut-off value of 2.35 m/s for detecting PC/APT. The sensitivity and specificity of the selecting cut-off values were 0.56 vs. 0.63 and 0.95 vs. 1.0 (area under the curve [AUC]: 0.813 vs. 0.852 [95% confidence interval (CI): 0.669–0.956 vs. 0.720–0.983]; p < 0.001, p < 0.001; respectively).However, this study had a relatively small sample size, and all enrolled participants were required to have at least one of the following suspicious malignant features: (1) serum intact parathyroid hormone (iPTH) > 500 pg/mL or serum calcium > 2.9 mmol/L; (2) ultrasonographic findings of lesions with irregular shape or infiltrative borders; (3) newly developed lesions following previous parathyroid surgery (suspicious for recurrent parathyroid carcinoma). This selection bias may have resulted in the omission of some PC/APT cases that did not meet the above inclusion criteria. For example, in the present study, the median serum iPTH level of the PC/APT group was 312.2 pg/mL (interquartile range: 120.2–473.5 pg/mL), which did not reach the predefined threshold of 500 pg/mL, yet the postoperative pathological results still confirmed the diagnosis of PC/APT. Morever, it is generally believed that quantitative parameters of SWE cannot be directly extrapolated between different devices. Therefore, in this study, the ratio of SWE measurements to thyroid parenchyma was used as a supplementary index, which may have wider clinical applicability and facilitate standardization. However, when patients have comorbidities such as Hashimoto’s thyroiditis or other inflammatory diseases, or when the lesions are deeply located, these factors may affect the accuracy of shear wave velocity measurements and lead to certain bias in the results.

Several limitations of this study should be considered. First, as a retrospective study with a small cohort size, some subjective bias and sampling variation are possible, and the results of the multivariate logistic regression analysis might have some bias. Secondly, the accuracy of SWE examination is operator-dependent and is limited to some extent by lesion depth. The interobserver variability in the evaluation of parathyroid lesions using ultrasound was not assessed, which requires further research. Also, the SWVmean/SWVth ratio may be affected by thyroid pathology, such as Hashimoto or other inflammatory diseases, which might have an impact on the SWV values and could lead to some bias in the results. Thirdly, the interobserver variability in the evaluation of parathyroid lesions using SWE was not assessed in this study, which warrants further investigation.

Conclusions

In summary, sonographic features of intra-nodal fibrous bands and the SWVmean/ SWVth ratio may provide a useful model for preoperative prediction of PC or APT, which could help surgeons to accurately identify parathyroid malignancy preoperatively and then apply a suitable resection approach during the initial surgical treatment of patients.

Abbreviations

PC

Parathyroid carcinoma

PHPT

Primary hyperparathyroidism

PA

Parathyroid adenoma

APT

Atypical parathyroid tumor

SWE

Shear wave elastography

MEN

Multiple endocrine neoplasia

iPTH

Intact parathyroid hormone

Ca

Calcium

P

Phosphorus

ALP

Alkaline phosphatase

D/W

Depth/width

ROI

Region of interest

SWV

Shear wave velocity

SD

Standard deviation

SWVth

SWV in the thyroid parenchyma

ROC

Receiver operating characteristic

AUC

Area under the curve

OR

Odds ratio

CI

Confidence interval

Author contributions

R.L.: Writing - original draft, Investigation, Methodology, Visualization; T.Z.: Writing - original draft, Software Data curation; D.G.: Writing - original draft, Formal analysis; B.W.: Writing - review & editing, Resources, Validation; W.Z.: Writing - review & editing, Conceptualization, Project administration, Supervision. All authors read and approved the final manuscript.

Funding

Not applicable.

Data availability

Some or all datasets generated during and/or analyzed during the current study are not publicly available but are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This study was conducted in accordance with the Declaration of Helsinki, and approved by the Institutional Review Board of Beijing Chaoyang Hospital, Capital Medical University (2025-ke-26), all methods were carried out in accordance with relevant guidelines and regulations. Informed consent was waived owing to the study’s retrospective nature.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.