PET Imaging in Renal Cancer

Abstract

Purpose of review:

Worldwide, over 400,000 new cases of kidney cancer were diagnosed and over 175,000 deaths anticipated in 2018. It is ranked as the 14th most common cancer in women and 9th most common in men. Imaging is important for initial detection, staging and monitoring to assist treatment planning, but conventional anatomic imaging is limited. While functional PET/CT has proven helpful in the diagnosis and management of many cancers, its value in renal cell carcinoma (RCC) is still in evolution.

Recent findings:

FDG is probably the most useful radiotracer in RCC, although CAIX imaging can be helpful in clear cell RCC. Current research on PET imaging agents in RCC including 89Zr bevacizumab, 89Zr geruntuximab, 18F fluoro-L-thymidine (FLT), prostate specific membrane antigen (PSMA), 11C choline, 18F sodium fluoride (NaF) and18F fluorodeoxyglucose (FDG) have shown some interesting results for detection and prognosis of the disease.

Summary:

Many innovative radiotracers have been tested in RCC, but robust differentiation of primary disease from normal parenchyma remains elusive for almost all of them. The metastatic setting and response to therapy for this cancer are more favorable PET applications. Continued research in promising molecular tracers will hopefully advance both diagnostic and therapeutic strategies to ultimately eradicate RCC.

Introduction

Worldwide, over 400,000 new cases of kidney cancer were diagnosed and over 175,000 deaths anticipated in 2018. It is ranked as the 14^th most common cancer in women and 9^th most common in men(1). Renal cell carcinoma (RCC) is stratified into subtypes of clear cell (cc) (75–85%), papillary (10–15%), chromophobe (5–10%), oncocytoma (3–7%) and the rare collecting duct variant (2). Clear cell RCC demonstrates a range of aggressive behavior depending on grade. No matter the histology, unfortunately, 20– 30% of patients undergoing surgical resection for local/loco-regional disease develop recurrence and 20–30% of newly diagnosed patients present de novo with advanced metastatic disease (3). Noninvasive imaging modalities such as CT, MRI and ultrasound are important for initial detection, staging and monitoring but do have limitations, especially in terms of low sensitivity for small metastases and poor specificity for renal cancer.

PET imaging can potentially help assess biological aggressiveness, particularly when combined with CT for added morphological correlation. While PET/CT has proven helpful in the diagnosis and management of many cancers, its value in RCC is still in evolution. This may be changing with the new development and advances of specific radiotracers that exploit features characteristic of RCC. This article will highlight recent research on molecular PET imaging agents in RCC.

18F Fluorodeoxyglucose (FDG)

A glucose analog, 18F FDG (FDG) is the dominant PET radiotracer used worldwide in cancer. Depending on the cancer’s propensity to use Warburg physiology, the glucose transporter, GLUT1, is upregulated and other enzymes, particularly lactate dehydrogenase (LDH), are overexpressed. RCC is not a typical Warburg tumor and the results with FDG in RCC have predictably been mixed. A confounding issue is that FDG is excreted via the kidney making differentiation of a renal tumor from background normal kidney, difficult. As a result, FDG is not routinely recommended as an imaging tool in professional practice guidelines such as those issued by the AUA, ESMO and EAU (4–6).

The use of FDG in primary RCC is challenging. In a small meta-analysis, the pooled sensitivity and specificity for FDG was 62% and 88% respectively (7). However, other studies have had better results with FDG in RCC. For instance, Nakhoda et al retrospectively studied 25 malignant renal masses in 19 subjects and found FDG correctly detected 22 tumors resulting in 88% sensitivity(8). Seven of the 22 lesions were non-RCC renal masses (lymphoma, lung metastases) which were positive on FDG. Interestingly, the 3 non-detected masses were non-exophytic localized RCCs compared to the 15 exophytic lesions that were readily seen with FDG. This emphasizes that it may be difficult to distinguish RCC from normal parenchyma with FDG. In the Nakhoda study, standardized uptake value (SUV) parameters were statistically different between primary RCC and renal metastases but not with renal lymphoma. However, Nicolau et al noted that SUVmax was statistically higher in lymphomatous renal lesions than RCC(9). Therefore, SUVmax cannot reliably separate RCC and lymphoma.

FDG metabolism has been correlated with tumor grade in RCC. Takahashi et al demonstrated a correlation between FDG uptake and nuclear grade but were unable to distinguish RCC from benign lesions(10). This was confirmed by Nakajima et al who retrospectively examined 154 malignant renal lesions by FDG PET/CT and similarly noted higher SUVmax in high-grade tumors compared to low-grade tumors in both ccRCC and papillary RCC (11). However, there is considerable overlap and SUVmax cannot be used to discriminate tumors from benign lesions or differentiate subtypes or grades of RCC in individual patients(12). Evaluating different PET metrics, the same authors also studied metabolic tumor volume (MTV) and total lesion glycolysis (TLG) in RCC patients prior to surgical resection. MTV and TLG were inversely associated with progression free survival (PFS)(13). Patients with high preoperative MTV and TLG had shorter overall survival. Although these results are interesting, they have not been persuasive in regard to the routine use of FDG in primary RCC and accordingly FDG is not commonly utilized in this setting (14).

FDG PET/CT is more widely used when there is a high likelihood of metastatic disease. A meta-analysis by Ma et al found pooled sensitivity of 86% and pooled specificity of 88% for detecting RCC metastases, similar to conventional imaging with CT. However, FDG PET/CT was superior to conventional bone scans in identifying skeletal metastasis (15). Alongi et al retrospectively analyzed FDG imaging in patients with mainly clear cell RCC after primary surgery and discovered recurrent disease with a sensitivity of 74% and specificity of 80%(16), confirming the meta-analysis. In this study, PET/CT results influenced patient management in 43% and were prognostic in that positive scans were associated with an inferior overall survival (OS) at 5 years and progression free survival (PFS) at 3 years compared to patients with negative scans. Elahmadawy et al evaluated disease recurrence with contrast enhanced CT compared to FDG PET/CT and noted similarly high sensitivity (93–100%) and specificity (94–99%) for detecting localized and distant metastasis(17). Hence, FDG supplies important ancillary information to conventional imaging in many patients and can provide useful prognostic information.

FDG PET/CT has also been applied in studies of molecular therapies of RCC. The predictive value of FDG PET/CT for assessing outcome with tyrosine kinase inhibitors (TKIs) has been evaluated. For example, Horn et al evaluated patients with metastatic RCC undergoing treatment with sunitinib and imaged with FDG at baseline and within a month of treatment(18). Baseline FDG SUV inversely correlated with overall survival; higher SUVs were associated with worse prognosis. Metabolic changes in tumors did not occur until after at least 2 weeks of therapy on sunitinib. The connection between SUVmax and outcome was also documented by Nakaigawa et al who evaluated 101 patients with metastatic RCC with FDG PET/CT prior to initial therapy with tyrosine kinase inhibitors (TKI), mammalian target of rapamycin (mTOR) inhibitors, and interferon-alpha. Patients with a higher baseline max SUVmax had lower overall survival with a median observation period of 18 months(19). The median OS for patients with max SUVmax <8.8 was 57 months and for those with a max SUVmax above this cutoff, median OS was 13 months. Even after prior treatment with molecular targeted therapy relative FDG uptake can predict OS (20) The same authors went on to investigate response to specific molecular targeted treatments, with FDG uptake showing potential for predicting PFS as well as drug resistance(21). Farnebo et al found that changes in SUV normalized to lean body mass (SULpeak) and TLG before, and as early as 14 to 28 days after TKI treatment, correlated with PFS and OS (22). Finally, Ito et al noted that FDG PET/CT imaging before and one month after everolimus was an independent predictor of PFS and OS(23). Thus, although additional studies are required, evolving data support the concept that FDG can potentially be a reliable response biomarker in RCC.

Radiolabeled monoclonal antibodies

Bevacizumab is a recombinant humanized monoclonal antibody (mAb) that binds to vascular endothelial growth factor A (VEGF-A) thereby inhibiting tumor angiogenesis. It has been used in the treatment of metastatic RCC in addition to other cancers. A radiolabeled form of the mAb, 89Zr bevacizumab, was subsequently developed by Oosting et al (24). They imaged patients with advanced RCC at baseline and 2 and 6 weeks after starting treatment with bevacizumab and interferon-a or sunitinib, which is also an angiogenesis inhibitor. In their small trial of 22 patients, they demonstrated heterogeneous PET uptake and high baseline tumor SUVmax which correlated with longer time to progression, although the SUVmax on later scans was not predictive. The same group also studied 89Zr bevacizumab in 13 patients with metastatic RCC treated with everolimus(25). Focal uptake was seen in 71% of tumors and heterogeneous activity was similarly noted, but the trial size was not large enough to support meaningful associations with quantitative imaging values and everolimus efficacy on baseline or follow-up exams. So, while hopeful, 89Zr bevacizumab has not been developed further and likely will only be used in research.

Carbonic anhydrase IX (CAIX) is a glycoprotein enzyme activated by cell hypoxia that helps to buffer an acidic environment and promote cell survival. Many cancers, including that of the lung, breast, and cervix, and over 95% of clear cell RCC overexpress CAIX, but other RCC subtypes (chromophobe and papillary) and oncocytomas do not (26). Interestingly, CAIX is not normally found in renal tissue therefore CAIX is a compelling target for imaging and therapy of ccRCC. Geruntuximab is an anti-CAIX mAb first labeled with 124I as a PET/CT agent. In one study with 195 patients scheduled for resection of a renal mass, 124I geruntuximab demonstrated a sensitivity of 86% and specificity of 86% compared to a sensitivity of 76% and a specificity of 47% for contrast enhanced CT(27). Focal radiotracer uptake was noted mainly in ccRCC, but some benign and indolent tumors also accumulated radioactivity. Hekman et al evaluated 89Zr geruntuximab in a small study in which 6 primary ccRCC were detected. Ten other lesions without PET uptake were considered indolent and did not progress over a year(28). Metastatic and recurrent ccRCC were also detected in a separate subgroup, demonstrating the radiotracer’s potential for distinguishing aggressive kidney malignancy. While initially promising, 89Zr geruntuximab has not been advanced further as a clinical imaging agent.

In addition to radiolabeled mAbs, small molecule CAIX targeting agents have been developed. Turkbey et al reported on one such radiotracer targeting CAIX, 18F VM4–037, however physiologic excretion through the renal parenchyma contributed to low conspicuity in primary ccRCC but did show excellent depiction of metastasis(29). Because of its rapid pharmacokinetics compared to radiolabeled mAbs, this agent had the advantage of imaging on the same day of the tracer injection as opposed to mAbs that require several days for clearance of background. Moreover, labeling with 18F resulted in superior imaging to 124I or 89Zr. Thus, the family of CAIX imaging agents, including both mAb and small molecule-based agents, show good sensitivity and specificity for metastatic clear cell RCC. Unfortunately, this is a narrow market for PET development. However, if CAIX tracers can be combined with a theranostic agent, they potentially represent a rational approach to diagnosis and treatment of metastatic ccRCC.

18F Sodium Fluoride (NaF)

18F NaF (NaF) is an old radiotracer that exchanges fluorine with calcium in hydroxyapatite crystals in the bone matrix during bone turnover, such as occurs with metastasis. Because it does not directly target tumor but rather the effect of tumor on bone, NaF uptake is a nonspecific marker for bone metastasis and has been most commonly used for imaging osteoblastic metastasis from prostate and breast cancer. Its use in RCC has not been well described. Bone metastases in RCC are often purely lytic which is not ideal for NaF because it depends on osteoblastic response. Nonetheless, Gerety et al prospectively studied 10 patients with RCC bone metastasis and found that NaF PET/CT detected more RCC metastases than 99mTc-MDP bone scintigraphy/SPECT or CT alone(30). NaF PET/CT is highly sensitive for identifying RCC bone metastasis and should be considered when there is a strong clinical suspicion of bone metastases but conventional imaging is negative.

Prostate Specific Membrane Antigen (PSMA)

Prostate Specific Membrane Antigen (PSMA) is a transmembrane glycoprotein found in great abundance in most prostate cancers and in normal proximal renal tubules but is not substantially present in RCC. Instead, like many other cancers, PSMA is overexpressed in the neovasculature of RCC with the highest expression seen in clear cell, and lower expression in chromophobe and papillary RCC(31). Benign oncocytomas also overexpress PSMA in the neovasculature(32). Because of the success of PET imaging using small molecule PSMA inhibitors for prostate cancer, there is interest in exploring their application in other malignancies, including RCC.

Sawicki et al studied 68Ga HBED-CC PSMA (68Ga PSMA) in 6 patients with primary or metastatic RCC and observed that 68Ga PSMA uptake in primary tumors was overshadowed by normal parenchymal activity and therefore was of limited value(33). However, PSMA imaging had more benefit in RCC metastasis detection(34). Rowe et al confirmed these findings in 5 patients with metastatic ccRCC using the radiolabeled PSMA inhibitor 18F DCFPyL, and noted higher detection of metastasis than contrast enhanced CT or MRI(35). The same group went on to investigate 18F DCFPyL uptake in non-cc RCC disease (including papillary and chromophobe subtypes) with discouraging results(36). Definitive focal radiotracer activity was seen in only 13.7% of metastatic lesions and in none of the 3 primary tumors. Even though the trial was limited to only 8 patients, the use of PSMA PET imaging in non-cc RCC does not appear promising. Hence, PSMA PET/CT is likely comparable to FDG imaging and is likely more useful in detecting renal metastases.

18F Fluoro-L-thymidine (FLT)

3′-deoxy-3′-[18F] fluoro-L-thymidine (FLT) is an analog of thymidine and is used as a biomarker for cellular tumor proliferation. Uptake of this agent implies active DNA replication, a surrogate for cell growth. Few studies have been conducted with FLT in RCC. Liu et al studied tumor response to sunitinib with FLT at baseline, peak therapy and off drug in a small group of patients(37). FLT uptake decreased during treatment. Interestingly, when the drug was withdrawn there was a flare of FLT activity which later subsided. Horn et al also investigated FLT in conjunction with sunitinib treatment (22). FLT uptake rapidly decreased on therapy but was not as predictive of outcome as was FDG. FLT may be helpful in future trials to further assess possible mechanisms of biological resistance to drugs such as sunitinib.

11C choline

Choline is incorporated in the lipid bilayer of the cell membrane as phosphatidylcholine and is consumed in large amounts by rapidly growing tumor cells. Thus, radiolabeled choline is also incorporated into dividing cancer cells. Its use in RCC, however, has been sparse. Nakanishi et al reported the use of 11C Choline and FDG in 28 RCC patients with primary and metastatic disease and showed that 11C choline performed better than FDG with a sensitivity of 88% compared to 56% for FDG(38). Physiologic choline activity does limit primary tumor evaluation. 11C choline PET/CT has also been used in Acquired Cystic Disease of the Kidney (ACDK). Such patients have renal failure and a high risk for RCC(39). In a comparison of FDG PET/CT and contrast enhanced CT, 11C Choline detected all RCC lesions in ACKD with a sensitivity of 100% compared to 57.1% for FDG PET/CT and 42.9% for contrast enhanced CT. Therefore, this niche population may be ideal for 11C Choline.

Other Tracers

In the past, researchers have explored the use of 11C acetate, 18F fluciclovine and 18F fluoromisonidazole in RCC, but results have not been encouraging. Consequently, they are not likely to have a major role in RCC imaging.

Conclusion

Advanced RCC affects hundreds of thousands of patients around the world. Unfortunately, conventional imaging is limited, especially for early metastatic disease. PET/CT can be a useful, multifaceted imaging tool that provides insight into tumor biology and cellular mechanisms compared to conventional anatomic imaging. Among the existing agents, FDG is probably the most useful, although CAIX imaging can be helpful in ccRCC. Many innovative molecular radiotracers have been tested in RCC, but robust differentiation of primary disease remains elusive for almost all of them. Continued research in promising nuclear tracers will hopefully advance both diagnostic and therapeutic strategies to improve the outcomes of patients with RCC.

Key points:

• 18F FDG is the most studied radiotracer in RCC but not reliable for detecting primary cancer. In metastatic disease, uptake values have shown prognostic implications in targeted therapy with correlations in progression free survival and overall survival.

• Targeting carbonic anhydrase IX with novel radiotracers is effective in detecting recurrent and metastatic RCC lesions.

• Prostate specific membrane antigen (PSMA) is not significantly found in RCC tumors but is seen in the neovasculature. However, radiolabeled PSMA inhibitors have not demonstrated meaningful detection capabilities in RCC over FDG.

• 11C choline appears to perform better than 18F FDG in limited studies, but physiologic renal activity is still a challenge in primary tumor evaluation.