Adenoid cystic carcinoma (ACC) is a rare type of epithelial tumor mostly originated from salivary glands, accounting for 1% of total head and neck cancers [1, 2]. Histologically, ACC comprises tubular, cribriform, and solid patterns, and it is generally recognized that a solid growth pattern indicates an advanced tumor grade and a worse prognosis [3, 4]. ACC exhibits the characteristics of slow growth, extensive invasion, frequently local relapse, and a relatively high probability of distant metastases [5]. At present, the main treatment for ACC is surgical resection, yet ACC tends to spread along nerve tracts, involving vital structures and organs in the surgical field, which presents challenges to achieve complete radical resection. Recurrent tumor requires re-surgery or local radiation therapy, which has become a clinical routine treatment. Even so, the rate of 5-year distant metastasis is as high as 52% [6,7,8]. Chemotherapy and targeted therapy are not effective against ACC so far. Therefore, once a patient is diagnosed with metastatic ACC, the prognosis is poor, with a median survival of 20–32 months [2, 9, 10]. Hence, early accurate diagnosis, staging, and effective adjuvant treatment are crucial to the management of ACC patients and improve the prognosis.

In the past few decades, remarkable advances have been made in precision medicine based on positron emission tomography (PET) imaging, and the significance of 18F-fluorodeoxyglucose (18F-FDG) PET/computed tomography (CT) in the diagnosis and staging of various tumors is well recognized. However, not all ACC lesions exhibit identifiable FDG uptake [11, 12]. Prostate-specific membrane antigen (PSMA), also known as folate hydrolase I or glutamate carboxypeptidase II, is overexpressed by tumor cells or neovascular endothelial cells, such as prostate cancer (PCa), ACC, renal cell carcinoma, and hepatocellular carcinoma [13,14,15,16,17,18,19]. In most ACC lesions, PSMA expression is observed on cytomembrane of tumor cells rather than the vasculature [13, 20, 21]. Some previous studies of immunohistochemistry of primary, local recurrent, and distant metastatic ACC confirmed PSMA expression in these tumors [22, 23]. Van Boxtel et al. reported the percentage of PSMA-positive tumor cells for primary ACC and metastatic lesions was 7.5% (range 0–90%) and 5% (range 0–80%). Besides, tumor-associated neovasculature exhibited no PSMA expression [20]. Another research enrolled 9 patients revealed that PSMA expression was seen in all patients, mainly in cytoplasmic or concentrated at the luminal side of the cell membrane, varied widely between 5 and 90%, and a median of 30% of the primary tumor cells (IQR 15–70%) demonstrated PSMA expression [13]. Some studies have demonstrated that PSMA PET/CT is a valuable modality to detect and visualize ACC lesions and proposed the possibility of radioligand therapy (RLT) in ACC patients [13, 20]. Up to now, PSMA-targeted RLT against PCa has achieved encouraging beneficial effects [24,25,26]. One of the most widely studied PSMA radiopharmaceuticals is 177Lu-PSMA-617. As a diagnostic tracer, PSMA-617 is cleared quickly from the blood. Therefore, PSMA RLT based on 177Lu-PSMA-617 requires higher doses, which may cause obvious systemic toxicity. We modified PSMA-617 by conjugating a truncated Evans blue (EB) molecule and a DOTA chelator and then labeled it with 177Lu to synthesize 177Lu-EB-PSMA-617, the molecular structure of which is shown in Fig. 1. EB can bind to albumin to slow down its plasma clearance rate. Hence, EB-PSMA-617 could increase the tumor accumulation and reduce the total dosage of 177Lu, thereby precisely focusing as much radiation as possible on the tumor and improving the utilization rate of 177Lu. A previous study confirmed that the accumulated radioactivity of 177Lu-EB-PSMA-617 in tumor was about threefold higher than that of 177Lu-PSMA-617. However, the absorbed doses of 177Lu-EB-PSMA-617 in the red bone marrow and kidneys were also significantly higher than those of 177Lu-PSMA-617 [27]. Clinical studies have demonstrated the remarkable efficacy of 177Lu-EB-PSMA-617 in the treatment of PSMA-positive PCa [27, 28], which has led to the question whether 177Lu-EB-PSMA-617 could also achieve satisfactory therapeutic efficacy in ACC. It is essential to carry out a prospective trial of PSMA RLT in ACC patients.

Fig. 1
figure 1

Structural formula of PSMA-617 (A) and EB-PSMA-617 (B)

This pilot study was designed to further evaluate the diagnostic performance of 68 Ga-PSMA-617 PET/CT in ACC in a head-to-head comparison with18F-FDG PET/CT and to preliminarily assess the safety of and therapeutic response to PSMA RLT in patients with ACC.

Materials and methods


This study was approved by the institutional review board of Peking Union Medical College Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College (no. ZS-2532), and registered at (NCT04801264).

Patients with pathologically diagnosed ACC were prospectively recruited to undergo 68 Ga-PSMA-617 and 18F-FDG PET/CT. Written informed consent was obtained from each subject.

Regarding the inclusion criteria for PSMA RLT, ACC lesions with high PSMA uptake confirmed by 68 Ga-PSMA-617 PET/CT, which was defined as a baseline uptake value at most of tumor involvement of at least 1.5 times the average standardized uptake value (SUV) of the liver, were eligible [28]. The following exclusion criteria were used: white blood cell count < 2.5 × 109/L, hemoglobin count < 9.0 g/dL, platelet count < 75 × 109/L, serum creatinine > 150 μmol/L, serum albumin > 3.0 g/dL, total bilirubin > 60 μmol/L, cardiac insufficiency, and claustrophobia [28].

Synthesis of 68 Ga-PSMA-617, 18F-FDG, and 177Lu-EB-PSMA-617

The radiolabeling of 68 Ga-PSMA-617 and 177Lu-EB-PSMA-617 was conducted as previously described [27]. 18F-FDG was synthesized in-house with an 11-MeV cyclotron (CTI RDS 111; Siemens).

PET/CT acquisition and interpretation

Within 1 week, both 68 Ga-PSMA-617 and 18F-FDG PET scans were conducted using a dedicated PET/CT scanner (PoleStar m660; SinoUnion Healthcare Inc., Beijing, China). For 68 Ga-PSMA-617 PET/CT, the images were acquired at 50–60 min after the administration of 68 Ga-PSMA-617 (1.8–2.2 MBq [0.05–0.06 mCi]/kg) [29]. For 18F-FDG PET/CT, the patients were instructed to fast for at least 6 h. PET/CT images were obtained at 60–80 min after the intravenous injection of 18F-FDG (5.55 MBq [0.15 mCi]/kg). All patients started with a low-dose CT scan (120 keV; 50 mAs) from head to proximal thigh for attenuation correction and anatomical localization, followed by a PET scan at 2 min/bed position. The acquired data were reconstructed using ordered subset expectation maximization (SinoUnion PoleStar: 2 iterations; 10 subsets; Gaussian filter of 4 mm in full width at half maximum; 192 × 192 image size).

The images were transferred to MIM software (Version 7.1.4, MIM Software Inc., Cleveland, USA) and were interpreted independently by two experienced nuclear medicine physicians blinded to the result of another tracer and relevant clinical information. The volume of interest of tumor was segmented using PET Edge, a gradient-based segmentation algorithm [30]. Any focal accumulations of 68 Ga-PSMA-617 and 18F-FDG that were higher than the surrounding background activity and could not be explained by physiological or benign tracer uptake were interpreted as tumors. The number and SUVmax of tumors were recorded.

Treatment regimen and follow-up

The 177Lu-EB-PSMA-617 radiopharmaceutical was diluted into 100 mL of normal saline and slowly administered intravenously to the patients for 25–30 min. Before that, the patients were infused with normal saline for 30 min for intravenous hydration, and salivary glands were cooled with an ice pack for 30 min to minimize dry mouth syndrome. The patients received up to 3 cycles of PSMA RLT, at 8–10-week intervals.

The clinical data and laboratory profiles, including patients’ subjective health complaints, routine blood examination results, hepatic and renal function indicators, were recorded every 2 weeks. Adverse events were categorized according to the Common Toxicity Criteria for Adverse Events 5.0. The therapeutic effect was evaluated by 68 Ga-PSMA-617 and 18F-FDG PET/CT at 8 weeks after RLT based on the modified PERCIST 1.0 criteria [31].

Statistical analysis

All statistical analyses were conducted using SPSS 26.0 software (IBM Corp., Armonk, NY, USA). The quantitative data were presented as the mean ± standard deviation. For data analysis, two-sided Student’s t test was applied to compare the SUVmax of 68 Ga-PSMA-617 and 18F-FDG PET/CT. Statistical comparison of the tumor numbers was made using Wilcoxon signed-rank test and Friedman’s rank test. The correlation analysis was performed using Spearman correlation coefficient. A P value < 0.05 was considered statistically significant.


Characteristics of the enrolled patients

We enrolled 30 patients with ACC (15 males and 15 females; average age, 43.0 ± 12.2 years; range, 23–66 years; median, 43 years), including a primary ACC patient, 9 patients with local recurrence, 2 patients with intracranial metastasis, 8 patients with bone metastasis, 5 patients with liver metastasis, 23 patients with lung metastasis, and a patient with axillary lymph node metastasis. The characteristics of the patients are summarized in Table 1. Finally, a total of 4 patients (no. 4, 9, 10, and 11) received 177Lu-EB-PSMA-617 treatment with approximately 1.85 GBq (50 mCi). No adverse events were reported or observed in any patient during the radiopharmaceuticals administration.

Table 1 Clinical features of 30 ACC patients

Diagnostic performance of 68 Ga-PSMA-617 and 18F-FDG PET/CT

Comparison of tumor detectability

68 Ga-PSMA-617 exhibited PET-positive lesions as follows: 1 primary maxillary sinus neoplasm, 9 recurrent tumors, 8 intracranial lesions, 91 bone metastases, 47 liver metastases, 1 lymph node metastasis and 202 lung metastases, for a total of 359 lesions. As a contrast, 18F-FDG identified 1 primary tumor, 7 recurrent tumors, 4 intracranial metastases, 86 bone metastases, 42 liver metastases, 1 lymph node metastasis and 301 lung metastases, for a total of 442 lesions. Regarding bone metastases, there were 11 PSMA + /FDG- lesions and 6 PSMA-/FDG + lesions; the combination of two scans can detect 97 bone lesions. For lung metastases, there were 5 foci of PSMA + /FDG- and 104 PSMA-/FDG + , respectively. It is worth noting that CT can exhibit 358 pulmonary nodules, which were interpreted as tumors. The details are shown in Tables 2 and 3.

Table 2 Number of PET-positive lesions detected on 68 Ga-PSMA-617 and 18F-FDG PET/CT
Table 3 Number of PET-positive lesions detected by 68 Ga-PSMA-617 PET/CT and 18F-FDG PET/CT

In short, 68 Ga-PSMA-617 exhibited more PET-positive extrapulmonary tumors (157 vs. 141, P = 0.016) than 18F-FDG. The number of PET-positive pulmonary lesions detected by 68 Ga-PSMA-617 was less than 18F-FDG (202 vs. 301, P = 0.001). The combination of 68 Ga-PSMA-617 and 18F-FDG can detect 469 PET-positive lesions, which was superior to each alone (469 vs. 359 vs. 442, P < 0.001).

Comparison of tumor uptake

68 Ga-PSMA-617 PET/CT exhibited higher tumor uptake than 18F-FDG PET/CT in a primary ACC tumor (SUVmax: 9.8 vs. 6.3) and 9 recurrent lesions (SUVmax: 10.4 ± 3.8 vs. 6.3 ± 5.9, P = 0.135), as shown in Figs. 2 and 3. For patients with distant metastases, 68 Ga-PSMA-617 PET/CT demonstrated lower tumor SUVmax than 18F-FDG PET/CT (4.1 ± 3.6 vs. 5.0 ± 3.9, P = 0.016), as shown in Fig. 4. Recurrent tumors revealed higher 68 Ga-PSMA uptake than metastatic lesions (10.4 ± 3.8 vs. 4.1 ± 3.6, P < 0.001), whereas the difference of 18F-FDG uptake in recurrent tumors and metastases was not statistically significant (6.3 ± 5.9 vs. 5.0 ± 3.9, P = 0.445).

Fig. 2
figure 2

A 53-year-old female patient with primary ACC. Anterior maximum intensity projection (MIP) and axial 68 Ga-PSMA PET/CT (AC) showed increased tracer uptake of the tumor in the left maxillary sinus (green arrow, SUVmax 9.8). 18F-FDG PET/CT (DF) showed a lower uptake (blue arrow, SUVmax 6.3)

Fig. 3
figure 3

A 62-year-old man was diagnosed with local recurrence and distant metastases 21 months after surgical removal of a right maxillary sinus ACC. 68 Ga-PSMA PET/CT (AE) revealed a PSMA-avid tumor in the right maxillary sinus (green arrow, SUVmax 11.2), multiple bone metastases (red arrow, SUVmax 16.2), and liver metastases (blue arrow, SUVmax 8.8). 18F-FDG PET/CT (FJ) showed negative recurrent and metastatic foci

Fig. 4
figure 4

A 49-year-old man was diagnosed with local recurrence and distant metastases 74 months after surgical removal of a left maxillary ACC. 68 Ga-PSMA PET/CT (A–C) showed multiple lung metastases with negligible uptake (green arrow). 18F-FDG PET/CT (DF) revealed FDG-avid pulmonary lesions (blue arrow, SUVmax 7.1)

On lesion-based analysis, for extrapulmonary tumors, 68 Ga-PSMA-617 PET/CT depicted higher tumor uptakes (8.8 ± 3.6 vs. 6.4 ± 4.2, P = 0.027) than 18F-FDG PET/CT. Regarding pulmonary lesions, 68 Ga-PSMA-617 PET/CT illustrated significantly lower SUVmax than 18F-FDG PET/CT (3.1 ± 3.0 vs. 4.2 ± 3.9, P < 0.001).

The SUVmax of tumors, both on 68 Ga-PSMA-617 and on 18F-FDG PET/CT, was not correlated with patients age, sex, pathological type, history of treatment, or the time interval from diagnosis to PET/CT scan.

Safety of and therapeutic response to 177Lu-EB-PSMA-617 in a patient with ACC

Patient no. 11 accepted three cycles of PSMA RLT, and Patients no. 4, 9, and 10 only accepted one cycle of therapy due to the impact of COVID-19 pandemic.

Clinical Symptoms and safety evaluation

The subjective symptoms of pain reported by all 4 patients were improved, with the reduced visual analogue scale (5.0 ± 1.4 for pre-therapy vs. 2.8 ± 1.3 for post-therapy, P = 0.125).

Patient no. 11 suffered from grade 2 anemia. Patient 10 had been experiencing mild hepatic insufficiency (ALT 75 U/L; AST 68 U/L) and was treated using hepatinica before PSMA RLT. Hence, this patient had no significant liver dysfunction. Routine blood examination, liver and renal function examinations of other 2 patients demonstrated no noticeable fluctuations within therapy. Besides, patients 9, 10, and 11 experienced Grade 1 nausea and fatigue during the observation period.

Molecular imaging response

For PSMA PET response, patient 4 showed encouraging therapeutic effect and the SUVmax of meningeal metastasis decreased from 7.0 to 1.1 (equivalent to the background activity), which achieved CR, as shown in Fig. 5. Patient 11 also demonstrated positive therapeutic response, with reduced tumor uptakes (12.0 ± 3.2 for pre-therapy vs. 7.9 ± 3.5 for post-therapy, P = 0.031), which reached PR. The therapeutic responses of patients 9 and 10, however, were heterogeneous. Of them, recurrent tumors, lung metastases, and liver metastases showed reduced tumor uptakes (recurrent tumors: 10.9 vs. 9.5; lung metastases: 3.4 ± 2.3 vs. 1.8 ± 1.5, P = 0.036; liver metastases: 8.9 ± 1.3 vs. 8.0 ± 1.4, P = 0.012). Bone metastases demonstrated increased SUVmax of tumors (9.2 ± 3.3 vs. 10.6 ± 2.3, P = 0.001).

Fig. 5
figure 5

A 41-year-old female patient was diagnosed with a left frontal meningeal metastasis 20 months after surgical removal of the left lacrimal gland ACC. Pre-therapy 68 Ga-PSMA-617 PET/CT (AC) demonstrated intense PSMA uptake of tumor (green arrow, SUVmax 7.0). 68 Ga-PSMA-617 PET/CT reexamination at 9 weeks after the 1st cycle of PSMA RLT (DF) and 8 weeks after the 3rd cycle of PSMA RLT (GI) revealed significantly decreased tracer uptake of tumor (red arrow, SUVmax 3.5; blue arrow, SUVmax 1.1), which reached the level of CR according to modified PERSIST criteria. However, 18F-FDG PET/CT (JL) always exhibited no positive lesions (orange arrow)

For FDG PET response, patient 11 had no FDG-positive lesions. The results of FDG PET response for others were similar to PSMA. Patient 4 depicted reduced uptake of 18F-FDG in tumors (2.5 ± 0.6 vs. 1.5 ± 0.3, P = 0.250). Patients 9 and 10 also exhibited lower SUVmax of tumors after therapy (recurrent tumors: 4.1 vs. 3.4; lung metastases: 2.2 ± 0.8 vs. 2.0 ± 0.5, P = 0.036; liver metastases: 4.7 ± 0.5 vs. 1.9 ± 0.2, P = 0.002), except for bone metastases (4.0 ± 2.2 vs. 5.6 ± 1.9, P = 0.006), as shown in Fig. 6.

Fig. 6
figure 6

A 56-year-old male patient was diagnosed with local recurrence and multiple metastases 12 months after surgical removal of nasal ACC. Pre-therapy 68 Ga-PSMA-617 (AD) and 18F-FDG PET/CT (IL) depicted neoplasm recurrence, multiple liver, bone, and lung metastases. 68 Ga-PSMA-617 PET/CT (EH) and 18F-FDG PET/CT (MP) reexamination at 8 weeks after the 1st cycle of PSMA RLT revealed the decreased SUVmax of lung metastases (red arrow) and liver metastases (blue arrow); and CT showed liquefied necrosis occurred inside the lung nodule (red arrow). The uptakes of bone metastases (yellow arrow), however, were significantly increased


This is a prospective head-to-head comparison of detection capability between 68 Ga-PSMA-617 and 18F-FDG PET/CT in the same group of ACC patients and the first clinical study of 177Lu-EB-PSMA-617 therapy in ACC.

We found that 68 Ga-PSMA-617 PET/CT is superior to 18F-FDG PET/CT in detecting extrapulmonary lesions. As previously mentioned, a negative surgical margin plays a decisive role in the prognosis of primary ACC patients, which requires the preoperative diagnosis of the location and extent of tumor to be as accurate as possible. In our study, 68 Ga-PSMA-617 PET/CT revealed higher tumor uptake and a larger tumor boundary than 18F-FDG PET/CT, which may be a potential advantage over 18F-FDG and need to further confirm in a larger sample of patients. For intracranial metastases, it was reasonable that 68 Ga-PSMA-617 PET/CT showed better diagnostic performance than 18F-FDG PET/CT due to the high physiological accumulation of FDG in the brain. In patients with recurrent tumor, bone metastases, and liver metastases, the diagnostic value of 68 Ga-PSMA-617 was also potentially superior to that of 18F-FDG. For lung metastases, there was a relatively poor diagnostic effect of 68 Ga-PSMA-617, which may be partly attributed to insufficient PSMA uptake in small lung tumor volumes [13, 20]. Besides, we suspect that adenoid cystic carcinoma of the lung contains numerous mucinous secretions within their lumens that may cause relatively low PSMA expression and chronic inflammation of the lungs may also be important reasons. Subsequent studies are needed to confirm the above conjecture [32]. It could be a less significant factor because CT can detect extra pulmonary diseases, which can compensate for the low PSMA PET detection efficiency. We found that the tumor uptake was not correlated with the time interval from diagnosis to PET scan and pathological subtypes, possibly due to the small sample size and heterogeneity of cohort, which will need further confirmation in future studies.

All the above findings are of significance. In fact, for ACC patients after therapy, contrast-enhanced MRI cannot always distinguish between mucosal swelling, inflammatory response, and tumor infiltration [33]. Ruhlmann et al. reported that whole-body FDG PET/CT illustrated high sensitivity in detecting residual/recurrent and regional metastatic spread ACC tumors, which was also superior to that of MRI for local staging and restaging [33]. Furthermore, some studies have revealed that PSMA PET/CT might be useful for detecting lymph node or distant metastases but of limited value for identifying primary tumor or local recurrence [13, 20, 34]. However, in our study, the diagnostic performance of 68 Ga-PSMA-617 PET/CT was not inferior to that of 18F-FDG PET/CT in patients with primary ACC and local recurrence. The reason for these divergences may be that there have been relatively few head-to-head comparative studies on ACC, and it is impossible to draw generalized, clear conclusions. In our study, 68 Ga-PSMA-617 PET/CT combined with 18F-FDG PET/CT can achieve better detection efficiency for ACC than each alone and provide more valuable information for the accurate staging, restaging, and treatment of patients.

Another highlight of this study is the exploration of ACC treatment. Because of the lack of effective treatment against ACC, once the patients develop distant metastases, the prognosis is poor. With the successful application of PSMA RLT in PCa [35], this therapy has attracted some attentions in ACC, which also expresses PSMA. Duygu Has Simsek et al. reported that a case with metastatic ACC received PSMA RLT, which achieved a significant pain relief after the administration of 7.5 GBq of 177Lu-PSMA [36]. Unfortunately, that patient died in malignancy-induced hypercalcemia without 2nd cycle of 177Lu-PSMA therapy. As a new radiopharmaceutical, 177Lu-EB-PSMA-617 ensured an excellent therapeutic effect in metastatic castration-resistant prostate cancer [27, 28]. In this study, 2 of 4 patients received satisfactory therapeutic effects, in terms of improvement in both clinical symptoms and imaging response. The other two patients achieved noticeable beneficial results in recurrent foci, liver and lung metastases. But the uptakes of bone lesions were significantly increased, which is unclear whether it is true tumor progression or nonspecific bone uptake (flare phenomenon) [37]. Regrettably, the two patients were not able to continue RLT due to the COVID-19 pandemic. Anyway, all these cases demonstrated that PSMA RLT is a potentially promising treatment for ACC, which would probably benefit more ACC patients.

There are some limitations to our study. The most remarkable issue is the limited number of studied cohorts, especially patients with primary ACC and local recurrence. In addition, only 1 patient underwent 3 cycles of RLT and the others underwent single RLT cycle. Nevertheless, we found obvious clinical significance in the diagnosis and treatment of these patients targeting PSMA. Another limitation is the lack of immunohistochemical PSMA confirmation as a reference standard. Since recurrent and distant metastases are rarely biopsied, it is difficult to obtain tissue samples for immunohistochemistry. However, as previously mentioned, quite a few studies have confirmed the expression of PSMA in ACC. Therefore, our results of PET/CT and PSMA RLT are reliable.


68 Ga-PSMA-617 PET/CT is a valuable imaging modality for the diagnosis and staging of ACC. When combined with 18F-FDG PET/CT, they can achieve better diagnostic value for identifying ACC lesions than each alone. PSMA RLT based on 177Lu-EB-PSMA-617 may be a promising treatment for ACC. These findings need to be confirmed in further studies with larger cohorts of ACC patients.