A dual-tracer approach using [¹¹C]CH and [¹⁸F]FDG in HCC clinical decision making

Hepatocellular carcinoma (HCC) is a major cause of global morbidity and mortality with rising incidence [1]. Therefore, early detection of recurrent or progressive tumour remains the strongest prognostic factor for survival [2, 3]. Currently, diagnostic imaging with dynamic CT and MRI are currently advised in detecting progression of HCC according to ESMO clinical guidelines [4].

The Barcelona clinic liver cancer (BCLC) staging system currently divides therapeutic options for HCC into: curative treatments (surgical resection, ablation, and liver transplantation) in case of very early or early stage HCC (A), transarterial chemoembolization (TACE) or selective internal radiotherapy (SIRT) for intermediate stage HCC (B), sorafenib for advanced stage HCC (C), and best supportive care for terminal stage HCC (D) [5]. Imaging after treatment with dynamic CT or MRI can be complicated by fibrosis and heterogenous residual tumour appearance and extrahepatic metastasis [4, 6]. Histological analysis of tumour differentiation is often not performed, but is a major predictive factor of post-operative recurrence in HCC [7]. This motivated the use PET imaging as non-invasive marker of tumour differentiation with the potential to be a more sensitive method for localizing tumour recurrence [8].

[¹⁸F]-fluoro-2-deoxy-d-glucose ([¹⁸F]FDG) PET/CT evaluates tumour viability based on glycolytic activity. Success has been achieved in HCC with sensitivity ranging from 36 to 70% after treatment [9‐11]. Less success has been realized for multifocal HCC and non-bone extrahepatic metastases [10, 12]. The sensitivity of [¹⁸F]FDG PET/CT for well-differentiated intrahepatic HCC is similar to conventional imaging [13‐18]. To overcome this lack of sensitivity for certain extrahepatic and intrahepatic metastases, the [¹¹C]Choline ([¹¹C]CH) tracer of cell membrane lipid metabolism is used. HCC may show a high proliferation and increased metabolism of cell membrane components, which will lead to an increased uptake of choline [15]. Clinical studies of [¹¹C]CH show better detection rates than [¹⁸F]FDG PET/CT for well to moderately differentiated HCC lesions (84%) [11].

As both tracers relate to different metabolic pathways involved in tumour development, it is hypothesized that poorly differentiated HCC could be better evaluated with [¹⁸F]FDG and well-differentiated HCC with [¹¹C]CH [19, 20]. This led to the consideration of dual-tracer imaging in HCC as tumour differentiation of lesions within a given patient may vary, some taking up only one tracer. Variability of uptake can also be seen between separate portions of a single HCC lesion. Due to high background uptake of liver parenchyma of [¹⁸F]FDG, intrahepatic lesions might more difficult to detect, further complicated by the display of significant treatment effects seen on PET/CT in post-therapy HCC cases [21]. Therefore, this study aims to evaluate the role of dual-tracer PET imaging with [¹⁸F]FDG and [¹¹C]CH PET/CT in diagnosed HCC patients with suspicion of recurrent or progressive HCC in clinical decision making.

This case series study included HCC patients who underwent both [¹¹C]CH and [¹⁸F]FDG PET/CT in our institute from February 2018 to December 2021. No exclusion criteria were applied. HCC patients were included if they received both [¹¹C]CH and [¹⁸F]FDG PET/CT within 4 weeks of each other and had at least 6 months of follow-up. Patients underwent both radiotracer PET/CT scanning because intra- and extrahepatic recurrence of HCC was suspected on conventional imaging or patients had elevated serum tumour markers when conventional imaging was inconclusive. Tumour markers were considered elevated it they significantly increased compared with previous tumour marker measurements as part of normal follow-up. As the study was performed retrospectively and both radiotracers are part of clinical routine in our institution, individual informed consent was not required, according to the Dutch Act on Medical Scientific Research involving Human Beings (WMO). All research was conducted in accordance with both the Declarations of Helsinki and Istanbul.

A multidisciplinary team consisting of a hepato-pancreato-biliary surgeon, radiologist, and nuclear medicine physician discussed follow-up status of HCC patients, often with a history of treatment, such as microwave ablation (MWA), selective internal radiotherapy (SIRT) or transarterial chemoembolization (TACE), and surgical resection. These patients underwent dual-tracer PET/CT scanning in cases of unexplained and suspicious CT/MRI anatomical imaging or sudden rise of serum tumour markers without anatomical imaging evidence. General patient characteristics, medical imaging, relevant histopathology, alfa-fetoprotein (AFP) and des-gamma-carboxyprothrombin (DGCP), BCLC stage, follow-up therapy, and biopsy data for primary tumours, not metastases, were extracted from patient medical records.

PET/CT images were examined by author EBV and related to radiologist’s report of both tracer studies and noted whether a lesion was detected by any of the tracers. No case discrepancies were encountered. A lesion was considered malignant if there were non-physiological foci of high uptake, unless the imaging context concluded benign origin. A malignant lesion was considered critical when the finding had prognostic consequences leading to treatment changes. On the contrary, non-critical findings were defined as findings not altering treatment strategy, for example in case of a novel extrahepatic metastasis in patients with already known extrahepatic disease.

Nineteen patients who underwent [¹¹C]CH and [¹⁸F]FDG PET/CT examinations were identified (Table 1). Patients were previously diagnosed with HCC by imaging (n = 10) or by pathology (n = 9). Of these nine patients, histopathologic results of the primary HCC lesion revealed four well-, two moderately-, and three poorly differentiated tumours. All but seven patients were previously treated for HCC, including MWA (n = 6), SIRT (n = 1), surgical resection (n = 1), TACE (n = 1), MWA and resection (n = 2), and MWA and SIRT (n = 1). Six patients had not yet been treated, as they received dual-tracer PET/CT while awaiting planned therapy (four SIRT, two MWA). Other patients were under active surveillance, with no planned treatment. All but two patients had underlying diffuse liver disease: cirrhosis (n = 9) with one case caused by chronic hepatitis B virus and one by hereditary hemochromatosis, fibrosis (n = 1), non-alcoholic steatohepatitis (NASH, n = 6), and steatosis (n = 1). Nine patients were considered BCLC stage A, seven stage B, and three stage C. The minimum length of follow-up was 12 months, with a maximum of 24.

This study found detection rates of 68% for [¹⁸F]FDG and 84% for [¹¹C]CH for lesions that were pivotal in subsequent clinical decision making in patients with HCC. In only one patient both PET tracers failed to detect the anatomical substrate for suspected disease progression, resulting in a combined radiotracer detection rate of 95%. We contend that dual-tracer PET scanning may not have been necessary for this particular patient, given that the lesion was detected through an MRI conducted later, after the dual-tracer PET had already been performed. These promising results underline the potential critical role that dual-tracer approach can have in patients with suspected progressive disease. In the current study, we identified more clinically relevant lesions when combining [¹¹C]CH and [¹⁸F]FDG PET/CT in HCC, compared to [¹⁸F]FDG or [¹¹C]CH alone.

[¹⁸F]FDG appears to have low sensitivity (27–70%) for intrahepatic HCC, probably due to high background uptake of [¹⁸F]FDG in liver parenchyma [2, 10]. One study reported a sensitivity rate for the detection of intrahepatic HCC by choline-radiotracer of 88% [22]. Our data shows moderate intrahepatic detection rates for both tracers (65%). Differences in patient selection could explain this: seven out of 19 patients in our study had a direct therapy-related critical finding, such as needle tract metastases and ablation site recurrences. [¹⁸F]FDG found most ablation site recurrences (80%) and none of the needle tract metastases, with contrasted results for [¹¹C]CH (40% and 100%, respectively). HCC tumour aggressiveness, besides technical failure of tumour ablation procedure, is linked to lesions that require multiple bouts of locoregional therapy, of which ablation site recurrence is indicative as well [23]. Our study did not have enough histopathological data to match these results to tumour differentiation.

While both tracers detected a bilateral lung carcinoma, only [¹¹C]CH revealed one case of prostate carcinoma (Gleason 8). Detection of extrahepatic metastasis is known to be an independent predictor of poor survival and thus critical in deciding optimal treatment, especially for OLT and liver resections work-up [24, 25]. In four cases out of 16, [¹⁸F]FDG missed crucial extrahepatic metastases, of which three were identified after addition of [¹¹C]CH. Therefore, [¹¹C]CH PET/CT scanning in patients with negative extrahepatic [¹⁸F]FDG scans may increase sensitivity in extrahepatic HCC and suspected needle-tract or local recurrence after local–regional treatment.

Lack of [¹⁸F]FDG PET/CT uptake in well-differentiated tumours may be due to low amounts of FDG-6-phosphatase activity in these tumour cells [11]. Choline-tracers have shown increased uptake in well-differentiated tumours as these have increased cell membrane metabolism, of which choline is a substrate [18]. It is assumed that both tracers reveal different stages of tumour differentiation, advocating for dual-tracer diagnostics [11]. Although supported by recent meta-analysis, it recommends caution in accepting this rationale, as both methodological differences and lack of standardized histological grading practices were noted [26]. [¹⁸F]FDG failed to identify one well- and one moderately differentiated tumour in our study: growth of known HCC lesion and a needle tract recurrence. [¹¹C]CH identified all lesions with known differentiation. A link between elevated AFP and vascular invasion with [¹⁸F]FDG avidity has been proposed [26, 27]. Our study found eight cases with elevated AFP or DGCP as reason for dual-tracer PET/CT. One of these lesions was only found by [¹⁸F]FDG, three only by [¹¹C]CH, and four lesions by both tracers. With our limited number of patients, our study does not support a link between differentiation grade or serum tumour marker elevation to [¹⁸F]FDG or [¹¹C]CH detection rates.

Multimodal dual-tracer PET-based imaging can play a significant role in revealing changes in within-tumour metabolism, although the link between oncological changes and metabolic effects are not yet fully understood. Whereas [¹⁸F]FDG is sensitive for unexplained rise of serum levels of AFP, choline tracers appear to be not [20, 22, 28]. Studies on unexplained serum AFP elevation following locoregional HCC treatment found that [¹⁸F]FDG PET/CT had detection rates of 64% to 98% [2, 29]. Of the eight patients with elevated tumour markers, seven critical lesions were detected by [¹¹C]CH and five by [¹⁸F]FDG. In all but one the finding was intrahepatic, often growth of a known tumour. Due to heterogeneity in metabolic and/or genetic traits within a single tumour and between lesions within the same patient, overlap between both tracers in their detection rates is plausible [28].

To our knowledge, the presented study is the first to look at intra- and extrahepatic distribution of HCC with a dual-tracer approach with an emphasis on post-therapy follow-up. The retrospective nature, small patient sample, and the limited amount of available histopathology are limiting our conclusions. Further studies combining histopathology, serum tumour markers, four-phase CT/MR imaging, and genetic analysis are essential to further unravel the link between HCC pathogenesis and appropriate detection techniques, such as dual-tracer PET/CT solutions.

Although PET imaging combining both [¹⁸F]FDG and Choline-radiotracers has shown its benefit in staging and therapeutic management of patients with HCC, its use is still not commonplace. Due to the intrinsic characteristics of HCC pathogenesis and the resulting metabolic differences, a negative [¹⁸F]FDG PET/CT does not exclude recurrent HCC. Our study favours a dual-tracer approach in HCC staging when conventional imaging is non-conclusive.