Predictive Factors of Early ¹⁸F-Fluorodeoxyglucose-Positron Emission Tomography Response to [¹³¹I] Metaiodobenzylguanidine Treatment for Unresectable or Metastatic Pheochromocytomas and Paragangliomas

Pheochromocytomas and paragangliomas (PPGLs) are rare catecholamine-producing neuroendocrine tumours. The 2022 World Health Organization classification defines pheochromocytomas as tumours arising from the adrenal medulla and paragangliomas as tumours arising from extra-adrenal chromaffin tissues [1]. The estimated 5-year survival rate of PPGLs is approximately 75.4–91% and 10–30% for PPGLs which are metastatic [2, 3].

Metaiodobenzylguanidine (MIBG), an analogue of noradrenaline (norepinephrine), is a substance that is selectively concentrated in tissues under strong adrenergic innervation, including tumours of neuroectodermal origin. The European Society of Medical Oncology recommends [¹³¹I] MIBG treatment for unresectable or metastatic PPGL as a first-line treatment [4]; however, response rates to [¹³¹I] MIBG therapy have been inconsistent, ranging from 0% to 83% [5]. The pattern of catecholamine secretion is related to the degree of tumour differentiation, suggesting that patients with elevated dopamine and noradrenaline secretion have not only a higher frequency of metastasis [6] but also poorer prognoses [7, 8]. Thus, the therapeutic response to [¹³¹I] MIBG treatment may differ depending on the pattern of catecholamine secretion.

¹⁸F-fluorodeoxyglucose-positron emission tomography (FDG-PET) is generally used to detect metastatic lesions of various tumours. The maximum standardised uptake value (SUVmax) is the most widely used indicator to describe the intensity of FDG uptake in a tumour; however, this parameter represents only a small part of the tumour [9]. Volume-based semi-quantitative parameters such as the metabolic tumour volume (MTV) and total lesion glycolysis (TLG) have been suggested to be useful for determining the prognoses of various tumours such as oesophageal, lung, and ovarian cancer [10‒12]. Our previous study demonstrated that a high MTV and high TLG before [¹³¹I] MIBG treatment were significant predictors of poor prognosis in patients with PPGL [7]. In addition, the FDG response appeared after [¹³¹I] MIBG treatment at an earlier timepoint compared to the size response expressed on computed tomography (CT), which suggests that FDG-PET may be useful in evaluating treatment efficacy [13]. However, it is uncertain whether a treatment response evaluation based on FDG-PET can accurately determine the prognosis in patients with PPGLs.

Thus, we conducted the present study to investigate the relationships between changes in the MTV and TLG measured on FDG-PET before and after the first [¹³¹I] MIBG treatment and the prognosis of patients with unresectable or metastatic PPGL treated with [¹³¹I] MIBG. We also investigated the factors that would predict the tumour response to a single treatment with [¹³¹I] MIBG as evaluated by these semi-quantitative parameters.

This retrospective study was conducted as per the principles of the World Medical Association Declaration of Helsinki, and it was approved by the Human Research Ethics Committee of the Hokkaido University Graduate School of Medicine and the Institutional Review Board of Hokkaido University Hospital (#022-0329). The requirement for written informed consent was waived because of the retrospective study design.

Between 2001 and 2020, we reviewed the cases of 28 patients with unresectable or metastatic PPGL who could be followed up for ≥1 month after the first [¹³¹I] MIBG therapy session at our institute. We excluded 1 patient with unavailable prognosis information, 3 patients with unavailable FDG-PET data before or after treatment, 2 patients with insufficient clinical data, and 2 patients who underwent other therapeutic procedures before a second FDG-PET examination. The final population for the subsequent analysis consisted of 20 patients who underwent FDG-PET before and after [¹³¹I] MIBG therapy (Fig. 1).

[¹³¹I] MIBG was provided by Izotop (Institute of Isotopes Co., Budapest, Hungary) (n = 3, August 2008 to March 2009) and POLATOM (National Centre for Nuclear Research, Radioisotope Centre, Otwock, Poland) (n = 17, April 2009–2020). A single intravenous dose of [¹³¹I] MIBG (3.7–5.5 GBq) was administered over a 1-h period. Patients receiving [¹³¹I] MIBG were isolated in a room dedicated to radionuclide therapy until they met the release criteria established by applicable Japanese regulations.

The standard protocol at our institute was to administer [¹³¹I] MIBG treatment three times if side effects were tolerable, regardless of the apparent therapeutic effect of the first and second treatment sessions. However, if significant disease progression was observed in the patient’s imaging findings, biochemical data, or symptoms; or if the attending physician considered it difficult to continue treatment, the treatment course was discontinued. Moreover, some patients underwent more than three sessions of [¹³¹I] MIBG, with eight being the highest number of sessions (median 3, range 1–8). Attending doctors judged progressive disease (PD) or stable (non-progressive) disease at the end of [¹³¹I] MIBG treatment based on imaging findings, biochemical data, or symptoms comprehensively; these included the presence of novel lesions on FDG-PET, increased accumulation of lesions, elevated levels of blood and urine catecholamines, increased blood pressure and pulse rate, exacerbated constipation, and intensified pain attributed to bone lesions. The second and subsequent administration of [¹³¹I] MIBG radiotherapy was performed only after the toxicity associated with the previous [¹³¹I] MIBG radiotherapy had been improved. If moderate-to-severe bone marrow suppression was observed, the dose was reduced to 3.7 GBq; otherwise, 5.5 GBq was used as the standard dose. The interval between each [¹³¹I] MIBG radiotherapy treatment was approximately 3 months. Potassium iodide was administered orally to prevent radiation exposure to the thyroid. [¹³¹I] MIBG scintigraphy was performed within 2 days after the end of the isolation period.

The 24-h urinary catecholamine (adrenaline, noradrenaline, and dopamine) levels were measured using 24-h urine collection in hydrochloric acid prior to treatment. Urine was collected at a median of 2 days before [¹³¹I] MIBG treatment (range: 0–58 days).

Since the study period was as long as 20 years, four PET or PET/CT scanners were used for the FDG-PET examinations of this patient series. All the PET or PET/CT scans were performed at our institute. From 2001 to 2006, two PET scanners (ECAT EXACT 47 and ECAT EXACT HR+, Siemens, Munich, Germany) were used for FDG-PET imaging before and after [¹³¹I] MIBG therapy. After 2006, two PET/CT scanners (Biograph 64 True Point, Siemens; GEMINI TF64, Philips, Amsterdam, Netherlands) were used. Patients were imaged by PET or PET/CT according to standard practice. All patients were instructed to fast for at least 6 h before the FDG-PET/CT examination, and PET images were acquired 60 min after an intravenous injection of FDG (4 MBq/kg).

On the stand-alone PET scanner, a 2-min-per-bed emission scan and a 2-min-per-bed transmission scan were performed using a Ge-68/Ga-68 source, followed by image reconstruction using the ordered subset expectation maximisation (OSEM) method (one iteration, 30 subsets). On the PET/CT scanners, emission scans were performed on each bed after CT acquisition for attenuation correction. The reconstruction methods were as follows: Biograph 64 with 3D-OSEM – two iterations, 21 subsets; GEMINI (3D-OSEM) – three iterations, 33 subsets. PET images before and after the first MIBG treatment were acquired using the same scanner to minimise the effects of scanner performance differences.

We used the free software, Metavol, to calculate the MTVs and TLG based on liver accumulation [14]. Briefly, a 3-cm-diameter volume of interest was defined semi-automatically in the right lobe of the liver, and the mean and standard deviation were obtained from the volume of interest. We adopted the method of Wahl et al. [15], who used the mean + 3 × standard deviation as a consistent threshold for determining tumour boundaries. All the voxels above the threshold in the patient’s entire body were automatically extracted and highlighted. A nuclear medicine physician (J.T., 7 years of relevant work experience) then classified each uptake mass as either “tumour accumulation” or “non-tumour accumulation” (e.g., brain, myocardium, urinary tract) based on CT and MRI images. His judgement was confirmed by a second nuclear medicine physician (S.W., 12 years’ experience), and disagreements were resolved via a consensus.

The semi-quantitative FDG-PET parameters for determining the treatment response were defined as complete response (CR), partial remission (PR), stable disease (SD), and PD as described [15, 16]. In cases where the SUVmax, the MTV, or TLG decreased to zero on the second PET, the patients were classified as CR_SUV, CR_MTV, or CR_TLG, respectively. Based on the SUVmax, PR_SUV was defined as a ≥25% decrease, PD_SUV was defined as a ≥25% increase, and SD_SUV was defined as a change between −25% and +25%. Based on the MTV, PR_MTV was defined as a ≥30% decrease, PD_MTV was defined as a ≥30% increase, and SD_MTV was defined as a change between −30% and +30%. Finally, based on TLG, PR_TLG was defined as a decrease of ≥75%, PD_TLG was defined as an increase of ≥75%, and SD_TLG was a change between −75% and +75% (Table 1).

Statistical calculations were performed using JMP^® ver. 16.1 (SAS, Cary, NC, USA). Overall survival (OS) was defined as the time from the date of the first [¹³¹I] MIBG treatment to the patient’s death date or the last confirmed date of patient survival at censoring. The patients were classified into PD and non-PD groups (i.e., SD + PR + CR). The OS was analysed using Kaplan-Meier curves, and the OS differences between groups were evaluated using the log-rank test. We also investigated the relationship between the patients’ responses to [¹³¹I] MIBG treatment and other clinical parameters by performing a univariate logistic regression analysis. The clinical information included sex, age, PPGL, metastatic site, history of prior chemotherapy and external radiation before [¹³¹I] MIBG treatment, 24-h urine catecholamine level, and semi-quantitative parameters of FDG-PET before initiating the treatment. A p value of <0.05 was considered statistically significant.

The patients’ characteristics are summarised in Table 2. Eight (40%) of the 20 patients were male, while the other 12 (60%) were female. Their ages ranged from 22 years to 84 years. Fourteen (70%) patients had pheochromocytoma, and the other six (30%) had paraganglioma. The primary tumours, peritoneal dissemination, and lymph node metastases were resected as extensively as possible before the patients’ [¹³¹I] MIBG treatment; 1 patient was treated with [¹³¹I] MIBG as adjuvant therapy, while the other 19 had imaging evidence of the tumour: residual primary tumour (n = 2), lymph nodes or soft tissue (n = 10), bone (n = 9), liver (n = 7), or lung (n = 10). Of the 20 patients, 16 had at least one lesion on [¹²³I] MIBG or [¹³¹I] MIBG scintigraphy, and the other 4 patients had lesions that showed no accumulation on either image. Six (30%) patients had received treatment with cyclophosphamide, vincristine, and dacarbazine prior to [¹³¹I] MIBG treatment; 1 (5.0%) patient had received [¹³¹I] MIBG therapy at another hospital at a dose of 7.4 GBq; 5 patients (25.0%) received external radiation for bone metastases.

Table 3 summarises the [¹³¹I] MIBG treatment details during the study period. The median number of [¹³¹I] MIBG sessions was three (range: 1–8), and the median total dose was 16.65 GBq (range: 5.55–44.4 GBq). The median follow-up duration was 55.5 months (range: 2–136 months); the median OS has not been reached yet. The status at the end of [¹³¹I] MIBG treatment was disease progression in 10 patients and SD in the 10 other patients. The post-treatment observation period ranged from 2 months to 136 months, during which nine deaths were observed. The overall 5-year survival rate was 51.5% (Fig. 2a).

Twenty-four-hour urine collection was performed at the median of 2 days before [¹³¹I] MIBG treatment (range: 0–58 days). As presented in Table 4, the median urinary adrenaline level was 5.15 μg/day (range: 1.8–1,080 μg/day, interquartile range [IQR]: 4.0–13.6, normal range: 3.4–26.9 μg/day). The median urinary noradrenaline level was 478.3 μg/day (range: 59.1–4,282 μg/day, IQR: 175.3–999.0, normal range: 48.6–168.4 μg/day). The median urinary dopamine level was 778.9 μg/day (range: 249.9–3,187.5 μg/day, IQR: 556.9–1,040.0, normal range: 365.0–961.5 μg/day).

FDG-PET was performed on the same scanner model before and after a patient’s first [¹³¹I] MIBG treatment. The median interval between the pre-treatment PET and the first [¹³¹I] MIBG treatment was 30 days (3–46 days), and the median interval between the first [¹³¹I] MIBG treatment and second PET examinations was 94 days (19–137 days). The results of the semi-quantitative analysis showed the following median (range) pre- versus post-treatment values: SUVmax, 6.81 (range: 0–4.79 × 10) versus 9.83 (range: 0–4.18 × 10); MTV, 1.63 × 10 (range, 0–2.02 × 10²) mL versus 2.12 × 10 (range, 0–8.60 × 10²) mL; and TLG, 6.92 × 10 (range, 0–1.47 × 10²) mL versus 9.59 × 10 (range, 0–4.41 × 10²) mL (Table 5).

Based on the change in the FDG-PET parameter SUV, 2 patients were classified as PR_SUV, ten as SD_SUV, and eight as PD_SUV. Based on the change in the FDG-PET parameter MTV, 2 patients were classified as PR_MTV, ten as SD_MTV, and eight as PD_MTV. Based on the change in the FDG-PET parameter TLG, 12 patients were classified as SD_TLG and eight as PD_TLG. Moreover, in terms of PD versus non-PD classification, MTV and TLG response evaluations were unexpectedly consistent.

The OS was significantly longer in the non-PD group compared to the PD group when PD was defined based on the MTV or TLG (both p = 0.014) (Fig. 2b); however, the SUVmax was not a significant predictor (p = 0.49) (Fig. 2c). Therefore, we used the MTV and TLG as the response evaluation parameters for the further analyses. Figure 3 provides representative cases of non-PD and PD, respectively.

Table 6 summarises the relationships between the FDG-PET response evaluation and other clinical information. Regarding the urinary catecholamine levels, higher urinary dopamine levels were significantly associated with poor treatment response (odds ratio 1.002, p = 0.029). The between-group difference in urinary adrenaline levels was not significant; however, there was a trend of higher adrenaline levels towards better treatment response (p = 0.078). Noradrenaline was not a significant factor (p = 0.885).

As per the results of our analyses, the patients’ treatment response was not significantly associated with sex, patient age, tumour type, prior chemotherapy, prior treatment of bone metastases with external radiation, the site of metastasis, or any of the single semi-quantitative parameters by FDG-PET (SUVmax, MTV, TLG) before treatment. The analysis based on other treatments (i.e., [¹³¹I] MIBG treatment at another institution, radiofrequency ablation, or endovascular therapy) was difficult due to the small number of patients previously treated with these respective therapies.

In this retrospective analysis of 20 patients with PPGL, we investigated the response to a single [¹³¹I] MIBG treatment session based on FDG-PET/CT. The analysis results demonstrated that poor treatment response, as evaluated based on the MTV and TLG, was associated with shorter OS. Our findings also indicated that it may be possible to estimate the MTV and TLG response after [¹³¹I] MIBG treatment by using the patient’s pre-treatment urine dopamine level.

This study shows that the response to [¹³¹I] MIBG treatment as shown by FDG-PET was related to the patients’ prognoses. FDG-PET has been used as an imaging biomarker not only for visual evaluations but also for quantitative evaluations in malignancy and dementia [17‒19]. It is also used to search for tumour metastases in various types of cancer, and semi-quantitative parameters such as the MTV and TLG evaluated using FDG accumulation have been suggested to be useful for prognosis prediction in various types of cancer [10]. FDG-PET is also widely used to detect metastases in patients with PPGL [20, 21], and it has been shown to be useful in determining the patient’s prognosis before treatment [7]. Patient response to the first [¹³¹I] MIBG treatment evaluated using the Response Evaluation Criteria in Solid Tumours (RECIST) criteria was described as a prognostic factor [22].

In the present series, most of the patients had undergone only low-dose CT examinations for absorptive correction, and it was thus difficult to evaluate response based on the RECIST criteria. Besides, sclerotic osseous metastases were unevaluable in the RECIST criteria [23]. Even in such cases, FDG-PET enables the evaluation of the treatment response. Moreover, it has been reported that changes in the accumulation of FDG in response to [¹³¹I] MIBG treatment become evident sooner than changes in tumour size on CT [13]. However, our present findings demonstrated that the changes in the SUVmax were not significantly associated with the PPGL patients’ prognosis, partly because of the images obtained from the four unharmonised PET scanners. In this study, the poor tumour response evaluated using the FDG-PET semi-quantitative parameters, MTV and TLG, was associated with a poor prognosis. Such volumetric indicators could be less sensitive to scanner differences, which makes them appropriate for evaluating the response to [¹³¹I] MIBG treatment; however, larger scale studies are needed to test this new hypothesis.

PPGL is a rare neuroendocrine tumour that produces catecholamines. The biosynthesis of catecholamines proceeds by a series of enzymatic reactions that change tyrosine to the dopamine precursor known as DOPA, dopamine, noradrenaline, and adrenaline [24]. PPGL is classified into adrenergic, noradrenergic, and dopaminergic phenotypes. Phenylethanolamine N-methyltransferase is not expressed in the noradrenergic phenotype, and dopamine-β-hydroxylase is not expressed in the dopaminergic phenotype. These phenotypes are thus regarded as immature compared to the adrenergic type [6, 24, 25]. It has been suggested that patients who present with elevated dopamine and noradrenaline secretion have not only a higher frequency of metastasis [6] but also a poorer prognosis compared to patients without elevated dopamine and noradrenaline secretion [7, 8]. The results of our present investigation suggest that (i) high dopamine-producing PPGLs respond poorly to [¹³¹I] MIBG treatment, and (ii) more effective treatments should be explored, including higher dose [¹³¹I] MIBG treatment [26, 27]. The dopaminergic phenotype probably corresponds to a pseudohypoxia cluster that is characterised by the activation of pathways that mimic hypoxia signalling, such as succinate dehydrogenase, fumarate hydratase, and isocitrate dehydrogenase mutations [28]. This type of PPGL frequently shows lower MIBG avidity and strong somatostatin receptor expression [28]. Therefore, several retrospective studies suggested that peptide receptor radionuclide therapy is one of the most effective clinical therapies for PPGL [27, 28].

This was a single-centre retrospective study, which means the number of study participants (n = 20) was limited. Moreover, the patients’ inconsistent backgrounds made accurate evaluations difficult. It was also not possible to perform a multivariate analysis due to the limited number of study participants. In addition, most of the patients had undergone only low-dose CT examinations for absorptive correction for PET, and it was thus difficult to measure tumour diameters based on the RECIST criteria in this analysis. We did not evaluate the patients’ catecholamine metabolites, metanephrines, normetanephrines, or 3-methoxytyramine. These are more reliable markers than catecholamines themselves [29]. 3-methoxytyramine is generally unavailable in Japan, and urinary metanephrines and normetanephrines were not analysed in this study due to missing data. We did not include [¹²³I] or [¹³¹I] MIBG accumulation on scintigraphy in the analyses because there is no standard evaluation method for these accumulations. A quantitative assessment can be carried out using [¹²⁴I] MIBG-PET [30]; however, this is not widely available in Japan. In addition, the images obtained from the four PET machines were processed using different reconstruction methods and were analysed together without harmonisation. Poor prognostic factors such as a high Ki-67 value and genetic mutations were not evaluated.

The evaluation of treatment response to a single [¹³¹I] MIBG treatment session using FDG-PET/CT was associated with the patient’s prognosis. In this series of patients with PPGL, the response to [¹³¹I] MIBG treatment was associated with urinary catecholamines; more specifically, the dopamine-producing PPGLs had a poor response to MIBG treatment. Other treatment strategies for high-risk patients should be considered.

This retrospective study was conducted as per the principles outlined in the World Medical Association Declaration of Helsinki, and it was approved by the Human Research Ethics Committee of the Hokkaido University Graduate School of Medicine and the Institutional Review Board of Hokkaido University Hospital (#022-0329). The requirement for written informed consent was waived due to the retrospective nature of the study.

The authors have no potential conflicts of interest related to the content of this paper.

This work was partly supported by grants from the Japan Society for the Promotion of Science KAKENHI (Nos. JP20K16781 and JP20K08015).

Junki Takenaka and Shiro Watanabe collected and analysed the data and wrote the manuscript. Takashige Abe, Takahiro Tsuchikawa, Satoshi Takeuchi, Kenji Hirata, Rina Kimura, Naoto Wakabayashi, Nobuo Shinohara, and Kohsuke Kudo critically reviewed the manuscript for important intellectual content and approved the final version of the manuscript. Kenji Hirata significantly contributed to imaging analyses.

All data generated or analysed during this study are present in this article. Further enquiries can be directed to the corresponding author.