Inconsistent Detection of Sites of Metastatic Non-Clear Cell Renal Cell Carcinoma with PSMA-Targeted [¹⁸F]DCFPyL PET/CT

Abstract

Purpose:

To investigate the utility of prostate-specific membrane antigen (PSMA)-targeted [¹⁸F]DCFPyL positron emission tomography (PET)/X-ray computed tomography (CT) imaging for the detection of sites of disease in patients with metastatic non-clear cell renal cell carcinoma (RCC).

Procedures:

Eight patients with metastatic non-clear cell RCC underwent imaging with PSMA-targeted [¹⁸F]DCFPyL PET/CT. Imaged RCC histologic subtypes included papillary RCC (n = 3), chromophobe RCC (n = 2), unclassified RCC (n = 2), and Xp11 translocation RCC (n = 1). Using comparison to conventional CT and/or magnetic resonance imaging as reference, two radiologists with expertise in nuclear medicine identified putative sites of disease on [¹⁸F]DCFPyL PET/CT and classified each lesion as having no radiotracer uptake, equivocal uptake, or definitive uptake.

Resuts:

In total, 73 metastatic sites and 3 primary tumors compatible with sites of non-clear cell RCC were identified on conventional imaging. Metastatic sites of disease included lymph nodes (n = 40), venous thrombi (n = 3), pulmonary nodules (n = 10), bone lesions (n = 15), brain lesions (n = 3), and retroperitoneal masses (n = 2). Only 10 of the 73 lesions (13.7 %) were classified as having definitive radiotracer uptake (median SUV_max = 3.25, range = 1.2–9.5), 14 lesions (19.2 %) had equivocal uptake (median SUV_max = 2.85, range = 0.5–6.5), and 49 lesions (67.1 %) had no definitive uptake above background (median SUV_max = 1.7, range = 0.2–3.0). The three primary renal tumors demonstrated lower radiotracer avidity relative to surrounding normal renal parenchyma.

Conclusions:

A small proportion of sites of non-clear cell RCC showed uptake of the PSMA-targeted radiotracer [¹⁸F]DCFPyL. Unlike for clear cell RCC, the results of this study indicate that PSMA-based PET is not appropriate for imaging other RCC subtypes.

Introduction

Renal cell carcinoma (RCC) is the most common primary malignancy of the kidney, with an annual global incidence of 425,000 cases per year [1]. At the time of initial diagnosis, 35 % of patients with RCC will present with metastatic disease, a condition that is near universally fatal. RCC is a general term used to describe a histologically and molecularly heterogeneous group of tumors that arise from the renal nephron. The most common subtypes of RCC are the clear cell (75 % of all RCCs), papillary (15 %), and chromophobe (5 %) histologies [2]. Less common histologic subtypes include collecting duct RCC, renal medullary carcinoma, mucinous tubular and spindle cell carcinoma, and MiT family translocation RCC [3].

At the present time, the detection of metastatic RCC is largely carried out with conventional cross-sectional imaging modalities including X-ray computed tomography (CT) and magnetic resonance imaging (MRI) [4, 5]. However, these anatomical imaging techniques are limited in their ability to characterize small sites of metastatic disease [6]. As an alternative, positron emission tomography (PET) utilizing the metabolic radiotracer 2-deoxy-2-[¹⁸F]fluoro-d-glucose (¹⁸F-FDG) has been studied for imaging metastatic RCC [7–9]. The sensitivity of this imaging test, however, is relatively low in comparison to conventional cross-sectional imaging and its role for monitoring response to treatment remains unclear.

To overcome the limitations of currently available imaging tests, a number of molecular imaging agents have been investigated for RCC [10, 11]. This includes small-molecule inhibitors of prostate-specific membrane antigen (PSMA), a type II transmembrane glycoprotein that is ubiquitously expressed on prostate cancer epithelial cells [12, 13] as well as endothelial cells within the neovasculature of a number of solid malignancies including RCC [14–22]. Owing to the abundant neovascularity and high degree of PSMA expression in clear cell RCC [23, 24], PSMA-targeted PET imaging has been of growing interest for the characterization of this malignancy, with some preliminary indications that PSMA-based PET may be more sensitive for sites of metastatic clear cell RCC than conventional imaging as well as [¹⁸F]FDG PET [21, 22, 25–31].

At the present time, the utility of PSMA-targeted PET imaging of non-clear cell RCCs is unknown. Although PSMA expression is commonly present in cases of non-clear cell RCC, it is found at lesser degree than for the clear cell subtype [32, 33]. Recently, three patients with metastatic non-clear cell RCC were reported in the literature to have been imaged with [⁶⁸Ga]PSMA-11 PET/CT with variable success [25, 29]. Herein, we report the results of a small prospective study in which patients with metastatic non-clear cell RCC were imaged with PSMA-targeted [¹⁸F]DCFPyL PET/CT.

Materials and Methods

Patient Population

Eight patients with histologically proven non-clear cell RCC and conventional imaging findings compatible with metastatic disease were enrolled in a prospective study to broadly investigate the utility of [¹⁸F]DCFPyL PET/CT in patients with RCC (ClinicalTrials.gov Identifier NCT02687139). [¹⁸F]DCFPyL was synthesized as previously described [34], and the imaging protocol was performed in a manner consistent with the method described by Rowe and colleagues [21]. In brief, patients fasted from solid food for 4–6 h prior to intravenous injection of approximately 333 MBq (9 mCi) of the [¹⁸F]DCFPyL radiotracer. One hour following injection, a whole-body PET/CT was performed extending from the mid thighs to the vertex of the skull. In addition to imaging with [¹⁸F]DCFPyL PET/CT, all patients had previously undergone standard-of-care conventional cross-sectional imaging with CT or MRI of the chest, abdomen, and pelvis, with additional imaging of the brain as indicated by the presence of focal neurologic symptoms. Patients had typically been treated with multiple lines of systemic therapy at the time of [¹⁸F]DCFPyL imaging.

Image Analysis

All images were collaboratively reviewed by two nuclear medicine specialists with extensive experience in PET/CT imaging (YY and SPR). For each [¹⁸F]DCFPyL PET/CT, the following features were recorded: lesion location, lesion size, presence or absence of focal radiotracer uptake, and the maximum lean body mass-corrected standardized uptake value (SUV_max). The readers qualitatively categorized each lesion as having either no detectable radiotracer uptake, equivocal uptake, or definitive uptake. For primary renal lesions, tumor-to-background ratios (TBR) were calculated by dividing SUV_max of each lesion by SUV_max of the surrounding normal parenchyma.

Results

Between October 2015 and February 2017, eight patients were enrolled in this prospective study (median age 61, range 43–71 years). This included 3 (37.5 %) patients with papillary RCC, 2 (25.0 %) patients with chromophobe RCC, 2 (25.0 %) patients with unclassified RCC, and 1 (12.5 %) patient with Xp11 translocation RCC. [¹⁸F]DCFPyL PT/CT studies were performed a median of 21 days (range 2 to 44) following conventional imaging. Additional patient characteristics are provided in Table 1.

Table 1..

Patient characteristics and results of [¹⁸F]DCFPyL PET/CT imaging

Patient	Age (years)	Sex	Histologic subtype	Location of metastases	#Lesions	Definitive uptake	Equivocal uptake	No uptake	Median SUVmax (Range)
1	65	Male	Papillary RCC	Subdiaphragmatic mesenteric LNs	2		2		3.2(2.9– 3.5)
2	71	Female	Papillary RCC	Retroperitoneal mass	1		1		3.6(3.6)
3	63	Male	Papillary RCC	Retroperitoneal LNs	9		1	8	2.2(1.8– 4.1)
				Right renal vein thrombus	1		1
4	58	Male	Unclassified RCC	Left parietal and temporal lobes	3	2	1		2.3(0.5– 6.2)
				Retroperitoneal Mass	1	1
5	68	Male	Unclassified RCC	Bones	15		3	12	1.2(0.5–2.1)
6	43	Male	Chromophobe	Left supraclavicular LNs	2		2		1.9(0.8– 2.8)
			RCC	Mesenteric LNs	3			3
				Retroperitoneal LNs	12		1	11
7	57	Male	Chromophobe RCC	Bilateral lower pulmonary lobes	10			10	1.4(0.2– 9.5)
				Retroperitoneal LNs	2		1	1
				IVC thrombus	1	1
				Left renal vein thrombus	1	1
8	44	Female	Xp11 translocation RCC	Bilateral Supraclavicular LNs	5	5			2.7(1.7– 5.1)
				Retroperitoneal LNs	4		1	4
				Retroperitoneal mass	1
Total					73	10 (13.7 %)	14 (19.2 %)	49	(67.1 %)

Among the 8 patients, 73 putative sites of metastatic disease and 3 primary renal lesions (in two patients) were identified on the basis of conventional imaging. Metastatic sites of disease included lymph nodes (n = 40), venous thrombi (n = 3), pulmonary nodules (n = 10), bone lesions (n = 15), brain lesions (n = 3), and retroperitoneal masses (n = 2). In total, 10 of the 73 (13.7 %) suspected metastatic lesions were classified as having definitive radiotracer uptake on [¹⁸F]DCFPyL PET/CT imaging. An additional 14 (19.2 %) sites were classified as having equivocal uptake, and the remaining 49 (67.1 %) lesions were found to have no significant radiotracer uptake above background (Table 1). Overall, the median SUV_max of the 73 metastases was 1.9 (range 0.2–9.5), and the median diameter was 1.6 cm (range 0.6–9.0 cm) (Table 2). No lesions were identified on PET/CT without a corresponding finding on conventional imaging.

Table 2..

SUV_max and tumor diameter of definitive lesions, equivocal lesions, and lesions without radiotracer uptake

Radiotracer uptake (n)	Median SUVmax (range)	Median diameter (cm) (range)
Definitive (10)	3.25(1.2–9.5)	2.7 (1.1–6.2)
Equivocal (14)	2.85 (0.5–6.5)	3.2(0.9–9.0)
No uptake (49)	1.7 (0.2–3.0)	1.5 (0.6–3.9)
Total (73)	1.9 (0.2–9.5)	1.6 (0.6–9.0)

The ten lesions with definitive radiotracer uptake were identified in only 3 (37.5 %) of the 8 imaged patients. Radiotracer-avid lesions had a median diameter of 2.7 cm (range 1.1–6.2 cm) with a median SUV_max of 3.3 (range 1.2–9.5) (Table 2). One patient with unclassified RCC had three identifiable lesions on [¹⁸F]DCFPyL PET/CT, including two brain metastases (SUV_max = 3.4) and a retroperitoneal mass (SUV_max = 6.2) (Fig. 1). A third brain lesion in this patient was classified as having equivocal uptake (SUV_max = 0.5). A second patient with chromophobe RCC was found to have radiotracer uptake in thrombi of left renal vein (SUV_max = 9.3) and inferior vena cava (SUV_max = 9.5) (Fig. 2). This patient also had two retroperitoneal lymph nodes that lacked definitive radiotracer uptake and 10 pulmonary lesions with no uptake of radiotracer (Fig. 2). An additional five lesions with clearly definitive [¹⁸F]DCFPyL uptake (all supraclavicular lymph nodes) were identified in a third patient with Xp11 translocation RCC (median SUV_max = 2.7, range 1.7–5.1) (Fig. 3). This patient also had four sites of retroperitoneal lymphadenopathy and one retroperitoneal mass identified on CT. However, none of these lesions demonstrated significant radiotracer uptake (Fig. 3).

Fig. 1.

Images of a patient with metastatic unclassified RCC. a, b Conventional imaging with CT demonstrated metastatic lesions within the brain as well as c a retroperitoneal mass. Imaging with [¹⁸F]DCFPyL PET/CT demonstrated definitive uptake in d one of the brain lesions and f the retroperitoneal mass (red arrow) with SUV_max values of 3.4 and 6.2, respectively. e Equivocal radiotracer uptake was observed in the other intracranial lesion (yellow arrow, SUV_max = 0.5).

Fig. 2.

Images of a patient with metastatic chromophobe RCC. CT imaging showed a an inferior vena cava thrombus, b a primary renal mass, and c multiple bilateral pulmonary nodules. d The inferior vena cava thrombus demonstrated avidity for [¹⁸F]DCFPyL (red arrow, SUV_max = 9.5), whereas e the renal mass had heterogeneous uptake that was lower than the normal renal parenchyma (white arrow, SUV_max = 9.3, TBR = 0.1). f The pulmonary nodules had no detectable radiotracer uptake on [¹⁸F]DCFPyL PET/CT (blue arrow, median SUV_max = 1.1, range 0.2–1.7).

Fig. 3.

Images of a patient with metastatic Xp11 translocation RCC. On CT, the patient was noted to have a metastatic supraclavicular lymph nodes, b enlarged retroperitoneal lymph nodes, and c a retroperitoneal mass. While the patient’s d supraclavicular lymph nodes were identifiable on [¹⁸F]DCFPyL PET/CT (red arrows, median SUV_max = 2.7, range 1.7–5.1), e the retroperitoneal lymph nodes had no discernible radiotracer uptake (blue arrows, median SUV_max = 2.55, range 2.2–2.9). f The retroperitoneal mass had equivocal uptake (yellow arrow, SUV_max = 4.0).

The 14 lesions with equivocal uptake had a median diameter of 3.2 cm (range 0.9–9.0 cm) with a median SUV_max of 2.9 (range 0.5–6.5), and the 49 lesions with no uptake had a median diameter of 1.5 cm (range 0.6–3.9 cm) with a median SUV_max of 1.7 (range 0.2–3.0) (Table 2). Fifteen bone lesions were identified in one patient with unclassified RCC, and 10 pulmonary lesions were identified in one patient with chromophobe RCC; none of these lesions demonstrated definitive radiotracer uptake (Fig. 4).

Fig. 4.

[¹⁸F]DCFPyL PET/CT images of a patient with metastatic unclassified RCC. Multiple bone metastases were observed that had a no (blue arrow, SUV_max = 0.5) or b equivocal (yellow arrow SUV_max = 2.1) uptake of the [¹⁸F]DCFPyL radiotracer.

The three primary renal lesions (one papillary RCC and two chromophobe RCCs) demonstrated lower uptake of radiotracer compared with the normal renal parenchyma (Figs. 2 and 5). SUV_max of the three primary lesions were 5.8 (TBR 0.3), 9.3 (TBR 0.1), and 11.6 (TBR 0.2).

Fig. 5.

Images of primary renal lesion in a patient with papillary RCC. a The large left renal tumor can be readily seen on CT (white arrow). b The PET and c fused PET/CT images demonstrated radiotracer uptake that was markedly lower than that of the normal left renal parenchyma (white arrows, SUV_max = 5.8, TBR = 0.3).

Discussion

In this study, we observed inconsistent uptake of the PSMA-targeted PET radiotracer [¹⁸F]DCFPyL in eight patients with metastatic non-clear cell RCC. These results suggest that PSMA-targeted PET imaging of non-clear cell RCC is unlikely to play a significant role in this patient population and is inferior to conventional imaging with CT and/or MRI. Our findings are consistent with previously published reports in the pathology literature that showed present yet intermediate expression of PSMA in the neovasculature of non-clear cell RCC subtypes [32, 33].

Although our data suggest little utility of PSMA-targeted imaging for non-clear cell RCC, there is strong preliminary evidence to support a role for this class of radiotracers in patients with clear cell RCC [21, 22, 25–31]. In fact, currently available data suggest a higher sensitivity with PSMA-targeted PET for detecting sites of metastatic disease than with conventional imaging. However, these data require confirmation in large prospective trials and the therapeutic implications of this improvement in sensitivity remain unclear. Furthermore, there may be role for endoradiotherapy in patients with clear cell RCC with PSMA-targeted agents conjugated to α- or β-emitting therapeutic radionuclides [35]. It is worth noting that although PSMA-targeted PET appears to have a limited role in imaging all-comers with non-clear cell RCC, it is conceivable that endoradiotherapy directed against PSMA can be used in select patients with metastatic non-clear cell RCC that have exhausted all potential lines of conventional therapy and show uptake with PSMA-targeted imaging. However, based on our data, we acknowledge that this circumstance is likely to be quite uncommon.

The present study is not without limitations. Most notably, this study was comprised of a small number of patients with multiple non-clear cell RCC histologic subtypes, thus limiting the definitiveness of our observations. Additionally, patients in this study had received a number of different therapies prior to being imaged; therefore, it is unknown whether some of the lesions that lacked radiotracer uptake or had equivocal levels of uptake might reflect some degree of treatment effect. Nonetheless, the difference in lesion detection efficiency between conventional imaging and [¹⁸F]DCFPyL was pronounced and unlikely to have been fully explained by prior treatment.

Conclusions

Although PSMA-targeted radiotracers such as [¹⁸F]DCFPyL have been previously shown to have high detection rates for lesions in patients with metastatic clear cell RCC, the low and inconsistent uptake of these agents in cases of non-clear cell RCC suggest that PSMA-targeted PET is not appropriate for imaging this patient population.