Serendipitous Detection of Hepatocellular Carcinoma in a Patient Undergoing Staging for Prostate Cancer

Introduction

In recent years, prostate-specific membrane antigen (PSMA)–positron emission tomography/computed tomography (PET/CT) has become the one-stop shop imaging in prostate cancer replacing conventional imaging with contrast-enhanced CT (CECT) and bone scans. PSMA is a Type II transmembrane glycoprotein which was discovered in 1987 in metastatic prostate cancer cell lines.[1] It consists of mainly three regions: intracellular, transmembrane, and extracellular. The last, i.e., the extracellular region, is the largest and contains specific enzymatic domains, which are the main target for current PSMA-ligand imaging and therapy.[2] The PSMA transmembrane structure enables it to exhibit a unique internalization functionality by means of endosomal complexes, which is a highly desired feature for theranostic approaches with radiopharmaceuticals.[3,4] PSMA expression is generally seen in normal epithelial prostate cells but is highly upregulated in prostate cancer cells. However, in opposition to the idea induced by its name, PSMA expression is not exclusive to prostate cells and can be found in several other tissues and conditions, including inflammation/infection, on nonprostatic epithelial cells, nonprostatic neoplastic cells, and also on tumor-associated neovascularization. Several studies have been performed to study the PSMA ligand uptake patterns in prostate cancer, nonprostatic malignancies, and other benign entities. Most of these study results have shown generally a consistently lower PSMA ligand uptake and predominantly PSMA expression in the cytoplasm, which is normally less intense than that in the apical membrane expression seen in prostate cancer lesions.[5,6] Despite the consistency of these findings across several studies, till date, no validated threshold of PSMA ligand uptake exists to safely differentiate prostate cancer from nonprostatic lesions, thus highlighting the importance of recognizing specific metabolic patterns of different diseases to produce clinically valuable reports. On the other hand, although nonprostatic PSMA uptakes might hamper the diagnostic performance of PSMA PET imaging, it is a double-edged sword as the receptor expression observed in such malignant tumor cells makes them potential candidates for the application of its use in therapeutic approaches using PSMA-targeted radionuclide therapy.

As literature has accumulated over the years and the indications of PSMA PET/CT have expanded, it has become widely evident that PSMA is not specific to only prostate cancer and shows uptake in a wide range of benign and malignant tumors, including hepatocellular carcinomas (HCCs). This has raised questions about the specificity of this imaging modality. Proper interpretation of findings has thus become challenging and to some extent perplexing. It has become important ever so to correlate imaging findings to relevant clinical history, to the expected natural history of disease spread, and mainly to the anatomic findings on CT and/or magnetic resonance imaging (MRI) to help reinforce the distinction of lesions that are more likely related to prostate cancer from those that are not. Recognition of such false positives is paramount to the subsequent clinical decision-making involved especially in cases which are being treated with curative intent. We report one such example of an oligometastatic prostate cancer who presented with an isolated liver lesion.

Case Report

A 59-year-old male presented with lower urinary tract symptoms of 1-year duration. Digital rectal examination findings suggested Grade 1 benign prostatic hyperplasia with a hard nodule. Biochemical workup revealed an elevated serum prostate-specific antigen level of 93 ng/mL. Prostatic biopsy revealed adenocarcinoma Gleason’s Grade 3 + 4 = 7. A whole body Gallium 68 PSMA PET/CT (68Ga-PSMA PET/CT) revealed primary prostatic lesion, bilateral pelvic nodes [Arrows, Figure 1a], solitary skeletal lesion in the right 3^rd rib [curved arrow, Figure 1a], and a liver lesion in segment V [Arrowhead, Figure 1a]. PSMA expression of liver lesions was significantly higher than primary prostatic lesions or nodal metastases [Figure 1b and d]. Morphologically, CT features revealed an enhancing lesion whose enhancement pattern was very different from the metastatic nodal and primary prostatic lesions. Hence, the possibility of this representing a distinct entity was considered, and a biopsy correlation was sought. A dedicated triple-phase CECT was also performed which showed the presence of arterial phase hyperenhancement with washout in the delayed phases [Figure 1c]. A biopsy was performed from the liver lesion which ultimately revealed a HCC. Serum Alpha Fetoprotein (AFP) level was found to be 12.1 ng/mL (0.89–8.78 ng/mL). The patient subsequently underwent a segment V hepatectomy for HCC and bilateral orchiectomy followed by androgen deprivation therapy for prostate cancer and currently is on follow-up. Postoperative follow-up serum AFP levels were <2.00 (0.89–8.78 ng/mL) [Figure 2: Histopathology showing the tissues prostatic and liver lesions].

Close

Figure 1 Gallium 68 prostate specific membrane antigen PET/CT (68Ga-PSMA PET/CT) for prostate cancer. (a) Maximum intensity projection image shows PSMA expressing prostatic lesion, pelvic nodes (arrows), rib lesion (curved arrow) and a liver lesion (arrowhead). Transaxial fused PET/CT images show PSMA expressing bilateral pelvic nodes (b, arrow). CECT image shows enhancing rounded soft tissue lesion in the liver (c, arrow) and fused PET/CT images show PSMA expression in the same (d, arrowhead).

Export to PPT

Close

Figure 2 Microphotographs (a) Prostate core biopsy showing conventional prostatic adenocarcinoma with a predominant Gleason pattern 3 in the lower half and a component of Gleason pattern 4 with cribriform glands in the upper half (b) Liver lesion core biopsy and (c) Hepatic segment V resection showing neoplastic cells arranged in trabecular, acinar, and sheeted patterns. Tumor cells are immunoreactive for (d) Arginase and (e) Glypican-3 (focally); while are negative for (f) Hep-Par 1 (g) NKX3. 1 and (h) Prostate-specific antigen

Export to PPT

Discussion

Nonprostate cancer pathologies exhibiting PSMA uptake on PET/CT are becoming increasingly frequent as the number of performed scans increases by the day. An important step to differentiate them from metastatic prostate cancer is by getting oneself familiarized with the plethora of conditions presenting with PSMA uptake on imaging, from inflammatory processes and benign tumors to nonprostatic malignant neoplasms. PSMA expression in the apical membrane, which is a typical marker of prostate cells, presents greater tracer uptake on PET/CT imaging than the level of expression shown in the cytoplasm, which is more often observed in nonprostatic entities. However, this finding does not enable a clear differentiation between prostatic and nonprostatic malignancies merely on the basis of imaging. The correlation of PSMA PET to the CT and/or MRI morphology, however, becomes a more assertive analysis tool in this particular regard.

Isolated liver metastasis from prostate cancer is exceedingly rare.[7] As a relatively newer imaging modality, PSMA PET has shown emerging pitfalls in the medical literature. Although it is increasingly being used for staging prostate cancer patients owing to its high diagnostic accuracy,[8] it is a relatively nonspecific tracer and its expression can be seen in many solid tumors such as sarcomas, thyroid, and lung cancers.[9,10] In nonprostate cancer malignancies, PSMA principally localizes to tumor-associated endothelial cell membrane proliferation.[11] Endothelial PSMA expression has been extensively studied and occurs in almost all nonprostatic solid tumors associated with neovascularization. Some promising studies have addressed the diagnostic (and therapeutic) potential of PSMA ligand radiopharmaceuticals in nonprostate cancer malignant neoplasms. Renal cell cancer is the most studied neoplasm within this context of the use of PSMA ligands in nonprostatic malignancies since renal cell carcinoma (RCC) was one of the first recognized neoplasms with high PSMA expression in endothelial tumor neovasculature. A pilot study prospectively compared the PSMA PET and CT findings of 10 patients (36 lesions) with metastatic RCC, showing not only higher sensitivity of PSMA PET (92% vs. 69%) but also changes in intended management.[12] HCCs also have been long known to show PSMA expression. In a study by Goenka et al., a high proportion of HCCs (about 90%) expressed PSMA, and about 40% of the HCC lesions had PSMA expression >30%.[11] In another prospective pilot study on 68Ga-PSMA PET/CT for imaging HCC, 36 of 37 lesions showed increased PSMA uptake.[13] Hence, PSMA expressing isolated liver lesions in a patient with localized or oligometastatic prostate cancer burden presents a unique diagnostic challenge. The careful and systematic review of CT morphology of such lesions is an important (but rarely enlightening) step in the analysis, and features such as the number, the size, its borders, the contrast enhancement and/or the washout patterns, and the presence of liver surface nodularity can provide an invaluable insight into the final interpretation. In particular, one must elicit a careful and detailed history and look for signs and symptoms of chronic liver disease. On imaging per se, one must give due attention to the lesions’ arterial, portal venous, and delayed phase enhancement, and the contrast washout characteristics. Additional indirect and supportive signs of HCC may include liver surface nodularity, the presence of caudate lobe hypertrophy, evidence of arterial phase enhancement in portal veins, and nonvisualization of portal vein or its tributaries. A differential of primary HCC should always be borne in mind in such cases. Thus, in a scenario of nonmetastatic or oligometastatic prostate cancer with an isolated liver lesion that may alter patient management, it should be always clarified by biopsy, regardless of the degree of PSMA ligand uptake. This approach can also prove helpful for nonhepatic, isolated, equivocal lesions.