Intra-arterial Peptide Receptor Radionuclide Therapy for the Treatment of Hepatic Neuroendocrine Tumor Metastases: Hope or Hype?

Abstract

Peptide receptor radionuclide therapy (PRRT) confers significant progression-free survival advantage for patients with small bowel grade 1 and 2 well-differentiated neuroendocrine tumors (WD NET). PRRT may also be clinically beneficial for patients with NET of pancreatic, bronchial, and other sites of origin; patients with paragangliomas; as well as for patients with well-differentiated grade 3 NET. Direct intra-arterial (IA) administration of PRRT into the hepatic artery for patients with NET liver metastases may result in higher radiopharmaceutical dose and longer dwell time in the liver tumors while relatively sparing non-tumor liver tissue and other organs such as the kidneys and bone marrow when compared with intravenous (IV) administration. This review summarizes currently available data on IA and IV PRRT dose distribution, reports safety and efficacy of IA PRRT, and proposes future research questions.

Neuroendocrine tumors (NETs) represent a heterogenous group of malignancies that arise from neuroendocrine cells throughout different parts of the body. NETs commonly metastasize to the liver. If tumors are hormonally active, they may cause a variety of clinical presentations such as carcinoid syndrome, hyper- or hypoglycemia from secretion of glucagon or insulin, or Zollinger-Ellison syndrome from gastrin overproduction. Metastatic disease is associated with 5-year survival rates less than 50%. ¹ Studies report an increasing incidence of NETs over the past few decades. ² First-line therapy for metastatic NETs is usually with somatostatin analogs such as octreotide and lanreotide, which have demonstrated both symptomatic relief from carcinoid syndrome as well as inhibition of tumor growth. ³ Several therapies have been found to be effective in second line including locoregional treatments for liver-dominant disease and systemic agents such as everolimus, sunitinib, capecitabine with temozolomide, and peptide receptor radionuclide therapy (PRRT). Currently, there is no universally agreed upon second-line treatment for patients with advanced disease. PRRT is a form of treatment that has been employed since 1992, in which a somatostatin analog such as octreotide is labeled with a radionuclide and delivered systemically to tumors that express high levels of somatostatin receptors on the cell surface, such as NETs. ^{4 5} Intravenous PRRT (IV-PRRT) with lutetium-177 ( ¹⁷⁷ Lu) DOTATATE markedly improved progression-free survival (PFS) for patients with grade 1 and 2 small bowel NET compared with high-dose octreotide ⁵ but did not result in overall survival benefit. ¹ Long-term toxicities of PRRT may include renal dysfunction, cytopenias, and rarely myelodysplastic syndrome or acute leukemia. ⁶ These toxicities result from renal excretion and deposition of the radioligand in the bone marrow. Many centers throughout Europe and the United States are trying to further refine treatment dosimetry and minimize systemic toxicity. An emerging treatment approach for patients with liver-only or liver-dominant NET has been intra-arterial PRRT (IA-PRRT), which involves delivery directly into the hepatic artery with the goal of maximizing radioligand uptake in liver metastases while reducing systemic exposure to radiation. This article will review the current literature on IA-PRRT, as it compares to IV-PRRT in the treatment of advanced NET, with focus on radiopharmaceutical distribution, treatment safety and efficacy, and directions for future research.

Intravenous Peptide Receptor Radionuclide Therapy

A seminal work evaluating the efficacy and safety of IV-PRRT for the treatment of midgut NETs was the NETTER-1 phase 3 randomized clinical trial. ^{2 5} In this study, patients from eight countries and 41 centers aged 18 years and older, with advanced well-differentiated (Ki67 index ≤20%) midgut NETs with positive uptake on ¹¹¹ In-DTPA-octreotide scintigraphy (OctreoScan) were randomized to receive either four cycles of intravenous ¹⁷⁷ Lu-DOTATATE 7.4 GBq (200 mCi) every 8 weeks and long-acting intramuscular octreotide (117 patients) or high-dose long-acting intramuscular octreotide every 4 weeks alone (114 patients). The primary endpoint of the study was PFS with secondary endpoint of objective response rate. Median follow-up time in both groups was ∼76 months following randomization. The PFS at month 20 in the IV-PRRT group was reported to be 65.2 versus 10.8% in the control group with a radiographic response rate of 18% in the treatment group versus 3% in the control group ( p  < 0.001). Although PFS was improved in the IV-PRRT group, median overall survival in the long-term analysis did not demonstrate statistically significant differences, with the IV-PRRT group being 48 versus 36 months in the control group (hazard ratio: 0.84, p  = 0.30).

Safety of IV-PRRT has also been studied. Bodei et al ⁶ retrospectively evaluated 807 patients who received IV-PRRT between 1997 and 2013. Of the radionuclides used for PRRT, 34.4% of patients received ¹⁷⁷ Lu, 44.4% received yttrium-90 ( ⁹⁰ Y), 19.5% received both ¹⁷⁷ Lu and ⁹⁰ Y, and 2% received other agents. Follow-up assessment of renal and bone marrow toxicity occurred at 30 months following administration. Treatments with ⁹⁰ Y and ⁹⁰ Y/ ¹⁷⁷ Lu DOTA compounds resulted in greater nephrotoxicity than ¹⁷⁷ Lu alone (33.6, 25.5, and 13.4% of patients, respectively; p  < 0.0001). Approximately 35% of patients experienced either transient or persistent nephrotoxicity, with only 1.5% experiencing severe nephrotoxicity. Myelodysplastic syndrome occurred in 2.4% of patients and acute leukemia developed in 1.1% of patients.

Intra-arterial PRRT Radiopharmaceutical Distribution

The concept of administering PRRT directly into the hepatic artery for patients with NET liver metastases has been introduced in an attempt to improve radiopharmaceutical uptake by tumors while minimizing radiation to non-tumor liver, kidneys, bone marrow, and other organs. This approach promises a more targeted delivery of radiopharmaceutical to liver metastases while minimizing radiation dose to the non-tumor liver and other organs. Kratochwil et al ⁷ compared hepatic NET uptake of ⁶⁸ Ga-DOTATATE that was delivered both intravenously and intra-arterially. A total of 15 patients with 122 metastatic DOTA-avid NET liver lesions were enrolled in the study. Each patient received both IA and IV of ⁶⁸ Ga-DOTATATE 4 weeks apart, followed by PET-CT ∼40 ± 10 minutes after injection. IA delivery was performed through a catheter placed in hepatic artery supplying the main tumor burden using 64 to 172 MBq of ⁶⁸ Ga-DOTATATE. IV dose range was 84 to 196 MBq of ⁶⁸ Ga-DOTATATE. Standard uptake value (SUV) maximum of liver tumors, non-tumor liver, pituitary, and kidney were measured for each patient from PET scans following IA and IV administration of the radiopharmaceutical. Compared with IV administration, IA administration resulted in an increased SUV in 117 of 122 (96%) of liver metastases with the average increase in SUV being 3.75-fold higher with IA administration (mean SUV max: 17.7 IA vs. 14.1 IV, p  < 0.001). In non-tumor tissues, lower uptake was demonstrated after IA administration compared with IV administration. In the pituitary, for example, mean SUV max was 3.7 IA versus 4.9 IV ( p  < 0.05), and in kidney mean SUV max was 5.6 IA versus 7.4 IV ( p  < 0.08). In another study when IA and IV ¹¹¹ In-DOTATATE were administered to 15 patients over 20 minutes, activity per unit time in liver metastases was 3.5 times higher after IA administration versus IV administration, with an IA/IV uptake ratio of 2 at 4 hours and 1.3 at 7 hours following administration. The study demonstrated saturation of somatostatin receptors with IA but not with IV PRRT administration. ⁸ These data provided compelling evidence that IA-PRRT may provide higher and more durable exposure of radioligand to NET liver metastases while reducing exposure to non-tumor liver and extrahepatic tissues, possibly leading to a more favorable toxicity profile.

Several subsequent studies evaluated the distribution of IA-PRRT. Limouris et al ⁹ studied 30 patients with gastroenteropancreatic somatostatin-positive NET with liver metastases. Seventeen patients were treated with IA ¹¹¹ In-[DTPA ⁰ ]-octreotide, and 13 patients were treated with IA non-carrier added ¹⁷⁷ Lu-[DOTA ⁰ , Tyr3]-octreotate. Blood samples were collected 2, 4, 8, and 24 hours following administration to assess blood and red marrow pool dwell time. Absorbed doses in different organs were also evaluated using geometric means obtained from planar gamma camera images. While this study did not directly compare IA-PRRT with IV-PRRT, its results shed light on the distribution of IA ¹⁷⁷ Lu-DOTA-octreotate. The study reported a dose per unit activity of 48 mGy/MBq in liver tumors compared with 0.14 mGy/MBq in background liver parenchyma (with tumor/liver dose ratio of 343). Doses per unit activity in kidneys, spleen, and bone marrow were 0.6, 1.4, and 0.04 mGy/MBq, respectively.

Thakral et al ¹⁰ studied 29 patients with metastatic gastroenteropancreatic NETs of whom 15 patients were treated with a single dose of IA ¹⁷⁷ Lu-DOTATATE infused over 5 to 7 minutes through a microcatheter positioned in dominant tumor feeding hepatic artery, while 14 patients received a single dose of IV ¹⁷⁷ Lu-DOTATATE (also infused over 5–7 minutes). In both groups, 200 mCi was administered. Dosimetry calculations were based on SPECT/CT images obtained 2, 24, and 96 hours post treatment administration. In terms of dose per unit activity in tumoral versus hepatic parenchymal tissue, the IA group demonstrated mean dose per unit activity of 4.2 mGy/MBq in liver tumors and 0.17 mGy/MBq in non-tumor liver (tumor to liver uptake ratio: 25). Doses per unit activity in the kidneys, spleen, and bone marrow were 0.3, 0.6, and 0.03 mGy/MBq, respectively. Radiopharmaceutical residence time was significantly longer for liver tumors (7.6 hours for IA delivery vs. 4.2 hours for IV delivery, p  = 0.001), and showed a trend toward shorter residence time for normal liver parenchyma (2.6 hours IA vs. 3.8 hours IV, p  = 0.06), kidneys, (2.2 hours IA vs. 3.2 hours IV, p  = 0.07), and spleen (5.8 hours IA vs. 6.7 hours IV, p  = 0.06).

A pilot study from our institution ¹¹ included 10 patients treated with a single dose of 94.7 ± 5.4 mCi of ⁹⁰ Y-DOTATATE-administered IA over 30 minutes through a microcatheter positioned in the proper hepatic artery or in a dominant lobar hepatic artery for patients with variant arterial anatomy. All patients underwent ⁶⁸ Ga-DOTATATE PET within 1 month prior to IA treatment. The first five patients also received 5.6 ± 0.9 mCi of IA ⁶⁸ Ga-DOTA-TOC followed by PET within 1 hour of completing IA administration of both radiopharmaceuticals. For these five patients, radiotracer uptake on post-treatment PET was compared with pre-treatment ⁶⁸ Ga-DOTATATE PET, for which a similar diagnostic radiopharmaceutical dose was administered IV. No significant improvement of radiopharmaceutical uptake by liver metastases was observed for IA versus IV administration. The median IA to IV SUV max ratio was 0.81 (range: 0.36–2.09) on a per-lesion basis and 0.90 (range: 0.54–0.97) on a per-patient basis. However, extrahepatic metastases and organs uninvolved by tumors demonstrated a decreased median uptake between IA and IV administration in all patients (ratio of 0.73 and range of 0.42–0.87 for extrahepatic metastases and ratio of 0.53 and range of 0.41–0.76 for uninvolved organs).

With regard to dose distribution of IV PRRT as a comparison, Cremonesi et al ¹² conducted a meta-analysis in 2018 that included 126 studies that reported dose–effect outcomes in patients treated with IV PRRT. A total of 10 studies were included in the analysis (590 patients). The meta-analysis reported data for both ⁹⁰ Y and ¹⁷⁷ Lu labeled peptides. Dose per unit activity was 3.4 to 10 mGy/MBq within liver tumors, 0.1 to 2.3 mGy/MBq within liver parenchyma, while kidneys, spleen, and bone marrow received 0.3 to 1.0, 0.5 to 0.9, and 0.02 to 0.07 mGy/MBq, respectively.

Compared with the IA PRRT distribution, it appears that dose per unit activity in kidneys, spleen, and bone marrow were similar between IA and IV PRRT administration, while higher uptake was observed in liver tumors for IA route ( Table 1 ).

Table 1. Dose per unit activity (mGy/MBq) for IA and IV PRRT delivery in liver tumors, non-tumor liver, and extrahepatic organs.

	Intra-arterial Dose per unit activity mGy/MBq		Intravenous Dose per unit activity mGy/MBq
Reference	Limouris et al 9	Thakral et al 10	Cremonesi et al 12
No. of patients	13	15	Meta-analysis
No. of doses	36	15	Meta-analysis
Liver tumor	48	4.2	3.4–10
Liver parenchyma	0.14	0.17	0.1–2.3
Tumor/liver dose ratio	343	25	NC
Kidneys	0.6	0.3	0.3–1
Spleen	1.4	0.6	0.5–0.9
Bone marrow	0.04	0.03	0.02–0.07

Abbreviations: IA, intra-arterial; IV, intravenous; PRRT, peptide receptor radionuclide therapy.

Intra-arterial Peptide Receptor Radionuclide Therapy Efficacy

Efficacy of IA-PRRT has been evaluated by several small prospective and retrospective studies. Kratochwil et al ⁸ reported on 15 patients who received 47 doses of IA-PRRT. Between 2 and 5 doses per patient were administered with activity range of 270 to 864 mCi. Each dose was administered over 3 hours. ⁹⁰ Y-DOTATATE was used for patients with lesions larger than 2 cm, while patients with tumors smaller than 2 cm were treated with ¹⁷⁷ Lu-DOTATATE. Overall response rate (ORR) was 60%, while disease control rate (DCR) was 100%.

Another study by Limouris et al ¹³ included 12 patients who received 2 to 6 IA ¹⁷⁷ Lu-DOTATATE doses (36 total) with total administered activity range of 400 to 1,200 mCi. The authors reported ORR of 67% and DCR of 88%. The largest published study to date by Kolasińska-Ćwikła et al ¹⁴ included 39 patients who received two to four doses of IA ⁹⁰ Y-DOTATATE (activity range: 38–111 mCi) over 20 minutes per dose. ORR was 24% (9 patients), while DCR was 95% (37 patients). In our study, 10 patients were treated with a single dose of ⁹⁰ Y-DOTATATE administered IA over 30 minutes ¹¹ observed to objective responses. DCR was 90% (nine patients).

A range of retrospective studies evaluating efficacy of IV PRRT for WD-NET that included 12 to 310 patients treated with two to six cycles of ¹⁷⁷ Lu or ⁹⁰ Y DOTA (total administered activity of 400–1,200 mCi) reported ORR range of 24 to 67% and DCR range of 55 to 93%. NETTER-1 trial, in which 116 patients were treated with one to four cycles of ¹⁷⁷ Lu-DOTATATE (total activity of 200–800 mCi) reported ORR of 18% ( Table 2 ). ⁵

Table 2. Summary data for IA ¹⁷⁷ Lu and ⁹⁰ Y-DOTATOC PRRT efficacy .

Parameter	Kratochwil et al 8	Limouris et al 13	Lawhn-Heath et al 11	Kolasińska-Ćwikła et al 14
No. of patients	15	12	10	39
No. of doses	47	36	10	NS
Doses/patient	2–5	2–6	1	2–4
177 Lu or 90 Y	Both a	177 Lu	90 Y	90 Y
Activity, mCi	270–864	400–1,200	100	38–111
Rx duration	3 h	NS	30 min	20 min
Response rate
ORR	60%	67%	0%	24%
DCR	100%	88%	90%	95%
PD		12%	10%	5%

Abbreviations: DCR, disease control rate; NS, not specified; ORR, objective response rate; PD, progressive disease; PRRT, peptide receptor radionuclide therapy.

^a90 Y for lesions ≥ 2 cm, ¹⁷⁷ Lu for lesions < 2 cm.

While achievement of ORR ≥60% by two of the available IA-PRRT trials that allowed multiple cycles of PRRT is promising, ORR in this range was also reported by retrospective IV-PRRT studies. ⁴ Prospectively acquired efficacy data comparing IA and IV administration routes are needed.

Safety of Intra-arterial Peptide Receptor Radionuclide Therapy

Our group ¹¹ reported that 5 of 10 (50%) patients experienced no adverse effects (AEs) or mild (CTCAE grade 1) AEs, while 4 of 10 patients developed moderate (CTCAE grade 2) AE. There were three severe non–life-threatening (CTCAE grade 3) AEs observed in three patients, which were deemed unrelated to treatment (cholangitis in a patient who previously had bouts of cholangitis, urinary tract infection, hyponatremia, and hyponatremia and pulmonary embolism in a patient with a history of both). Most grade 1 AEs were considered to be due to amino acid administration. There were no renal toxicities. No CTCAE grade 4 AEs were observed. Two of 10 patients with ∼50 to 75% liver replacement by tumor died during the study. These patients developed transient grade 1 hyperbilirubinemia, infectious complications, venous thromboembolism, and lower extremity edema, which eventually led to clinical deterioration. On review, it was unclear whether deaths were due to PRRT, disease progression, or other causes.

In the study of 15 patients treated with IA ¹⁷⁷ Lu or ⁹⁰ Y PRRT, ⁸ one patient developed grade 3 thrombocytopenia, leukopenia, and anemia. Other AEs were not described in detail. Kolasińska-Ćwikła et al ¹⁴ reported grade 1 anemia at 6 weeks following therapy in 39% of patients. In addition, 32% of patients developed grade 1 leukopenia that also peaked 6 weeks following PRRT. There was one patient who developed grade 3 anemia, which was managed conservatively. There were no liver toxicities reported.

Future Directions

The prospective randomized LUTIA trial comparing safety and efficacy of IA and IV PRRT has been recently completed in the Netherlands. ¹⁵ The study aimed to include 26 patients with bulky (≥ 3 cm) bilobar liver metastases. Patients were randomly assigned to receive four cycles of 200 mCi of IA ¹⁷⁷ Lu-DOTATATE delivered either into the right or into the left liver lobe. The other liver lobe was treated with the circulating radiopharmaceutical, providing a comparison between targeted IA and systemic ¹⁷⁷ Lu-DOTATATE administration. The primary objective was to determine the tumor to normal liver uptake ratio in target versus non-target liver lobes on post-treatment SPECT-CT. Secondary endpoints included absorbed dose in both liver lobes, tumor response, dose–response relationship, toxicity, uptake in extrahepatic lesions, and renal uptake. Results are expected in 2024. There are two additional ongoing single-arm studies that are evaluating safety and efficacy of IA PRRT with ¹⁷⁷ Lu-DOTATATE that aim to enroll 10 patients each (NCT04544098, NCT04837885).

Many unanswered questions remain. When studies evaluating IA PRRT are designed, patient selection parameters such as tumor grade, primary site, lesion avidity on DOTA PET, minimum and maximum liver tissue replacement by tumor, minimum (and maximum) tumor size, and the extent of extrahepatic disease should be considered. The selection of radioisotope may also be important. Beta emitters ¹⁷⁷ Lu and ⁹⁰ Y have been used historically. These radioisotopes, at least theoretically, may have relative dosimetry advantages for lesions >2 cm ( ⁹⁰ Y favored due to higher energy of the β particles and longer β energy penetration in tissue) versus < 2 cm ( ¹⁷⁷ Lu favored). Safety profile of ¹⁷⁷ Lu and ⁹⁰ Y IV PRRT suggests that utilization of ¹⁷⁷ Lu may lead to less renal toxicity. ⁵ Alpha particle emitters such as ²²⁵ Ac and ²¹² Pb may improve both safety and efficacy profiles of PRRT and may also be reasonable agents for IA PRRT delivery.

IA PRRT administration parameters warrant additional work. Receptor saturation within 20 minutes of starting IA delivery has been noted. ⁸ Optimal radiopharmaceutical duration of administration (e.g., 5–7 minutes, 20–30 minutes, 3 hours, 6 hours, 24 hours) remains to be determined. Standard activity of 200 mCi per treatment cycle has generally been used. However, with the ongoing interest in personalized PRRT dosimetry, administered activities may need to be modified based on lesion uptake on DOTA PET and tumor burden. Lastly, the number of PRRT cycles may have an effect on treatment efficacy, with studies that allowed two to six higher numbers of treatments having reported higher radiographic objective response rates than studies that allowed a single IA PRRT administration. ^{8 11 13 14}

Conclusion

IA delivery of PRRT may result in higher tumor to normal liver radiopharmaceutical uptake ⁷ and decreased radiation dose to non-target organs such as the kidneys and the bone marrow, ^{7 11} potentially improving IA toxicity profile. However, data reporting on activity distribution to date have been conflicting with some studies reporting similar tumor to normal dose ratios ¹¹ and similar levels of uptake and residence times in normal liver, kidneys, spleen, and bone marrow. ^{9 10 12} Objective response rate for IA-PRRT may be ≥60% when at least two cycles of PRRT are administered, and when longer radiopharmaceutical administration time is allowed (3 hours vs. 20–30 minutes). Some retrospective studies of IV PRRT have reported similar treatment efficacy. Prospectively acquired studies that focused on dosimetry, optimization of IA PRRT delivery parameters, and comparing safety and efficacy of IV and IA PRRT administration routes are needed.