Evaluation of response using FDG-PET/CT and diffusion weighted MRI after radiochemotherapy of pancreatic cancer: a non-randomized, monocentric phase II clinical trial—PaCa-DD-041 (Eudra-CT 2009-011968-11)

Pancreatic ductal adenocarcinoma (PDAC) is associated with a poor prognosis. When detected early and treated adequately, the 1‑ and 5‑year median overall survival can be up to 20–25%. In Germany, PDAC is the third leading cancer death and the incidence corresponds to the mortality rate [1]. One of the reasons for this is that relatively few and late symptoms occur due to the organ’s localization in the retroperitoneum. Only about 20% of patients can be treated with a curatively intended resection at the time of diagnosis [2]. For patients with non-metastatic but locally advanced tumors, neoadjuvant chemo- or radiochemotherapy followed by surgery in responding patients is an option to improve outcome [3]. Ferrone et al. [4] showed a significant increase of overall survival by the neoadjuvant use of FOLFIRINOX chemotherapy.

The main potential advantages of neoadjuvant therapy include the downsizing of the tumor, increase of the resection rate, decrease of margin-positive resections, and decrease of lymph node metastases, all of which are expected to reduce the likelihood of locoregional recurrence [5]. Recent studies in patients with PDAC showed promising results regarding neoadjuvant chemotherapy or radiochemotherapy. However, as neoadjuvant treatment plus surgery increases toxicity it appears necessary to establish biomarkers for response monitoring. These may help to improve monitoring of the individual treatment of patients.

In this study on pancreatic adenocarcinoma, the value of [¹⁸F]2-fluoro-2-deoxy-d-glucose-positron emission tomography/computed tomography (FDG-PET/CT) and diffusion-weighted magnetic resonance imaging (DW-MRI) for the prediction of histopathological response following neoadjuvant radiochemotherapy was investigated. In previous publications, only the correlation between FDG-PET/CT findings and histopathological response had been reported for various other gastrointestinal tumors [6‐8]. In particular studies on early response evaluation for cancers of the gastro-oesophageal junction have been published. Lordick et al. [6] conducted the MUNICON trial and found FDG-PET to predict early metabolic response after platinum- and fluorouracil-based induction chemotherapy. Weber et al. [8] also found FDG-PET to discriminate responders from non-responders early after cisplatin-based chemotherapy. For pancreatic cancer only a small number of studies using PET for diagnosis, staging and response can be found, thereby leaving room for this prospective study [9‐11].

DW-MRI detects changes in intra- and extracellular water mobility [12, 13]. Tumors with a high cell density show a higher DW signal compared to signals arising from inflammatory processes [12]. For quantification, tissue specific properties, namely the apparent diffusion coefficient, can be calculated (ADC, [mm²/s]). If tumors exhibit a low ADC in initial imaging, DW-MRI may be a valuable tool for response evaluation. For pancreatic cancer, only a few reports for the use of the ADC for predicting response to chemotherapy have been published. The main reason for this is that quantitative DW-MRI is technically challenging with respect to image acquisition and analysis and to a certain extent is dependent on patient cooperation. Niwa et al. [14] reported a lower high b-value ADC to be predictive for early progression in patients with advanced pancreatic cancer treated with a gemcitabine-based regimen.

Based on the scarcity of data, the authors initiated a non-randomized, monocentric phase II clinical study of combined neoadjuvant chemotherapy followed by neoadjuvant radiochemotherapy with subsequent curative resection in patients with locally advanced adenocarcinoma of the pancreas. The aim was to evaluate the prognostic value of FDG-PET and DW-MRI obtained prior to and twice during neoadjuvant treatment.

A total of 50 patients suspected of having PDAC were screened between 06/2011 and 12/2015. The median patient age was 68 years, 14 (61%) were female, the Karnofsky Index was 90 and in 83% (19 patients) the tumors were located in the head of the pancreas. Of the screened patients, 13 had metastatic disease during the staging procedures, four had other pathological findings, one had a second synchronous tumor, in four patients adenocarcinoma could not be pathologically confirmed, and four were excluded for other reasons. Finally, 25 patients were enrolled in the study. Of these, two patients were excluded due to early development of metastatic disease and cardiac decompensation during the first cycle of chemotherapy, leaving 23 patients available for analysis. Since the study was poorly recruiting, it was terminated prior to accrual of the calculated number of patients. All enrolled patients received their therapy according to protocol.

In routine staging for PDAC, FDG-PET plays only a minor role due to its low specificity and sensitivity of 61–94% and 85–97%, respectively [22, 23]. The main advantage of FDG-PET is its sensitive detection of distant metastases; furthermore, FDG-PET is a good method for the evaluation of metabolic response. This has been shown in numerous other tumor entities [24, 25]. In this trial using the decrease of SUV_max, the majority of patients were detected as responders and only a minority as non-responders. However, a clear correlation between responders in FDG-PET and histopathological response could not be found. Based on these findings, it is not possible to predict histopathological response on the basis of metabolic FDG-PET response. A number of other studies had established a correlation between the metabolic and histopathological response [6‐8].

It can be concluded that, by using FDG-PET/CT, the selection of patients with non-responding or progressive disease is possible. Thus, patients can be offered alternative therapeutic approaches, such as intensified neoadjuvant chemotherapy or a palliative treatment regimen early on.

Due to a distinct desmoplastic reaction induced by neoadjuvant radiotherapy no significant decrease in tumor size on the CT scan is expected [4, 26]. Therefore, the response can be poorly estimated and lead to misinterpretation. As expected, the authors were not able to show size changes of the tumors after neoadjuvant radiotherapy as measured by morphologic MRI. Due to this, as well as changes in tumor cellularity, the use of DW-MRI is increasingly considered [27]. Based on the data from this trial, however, the evaluation of the response with DW-MRI was not possible, with the ADC values showing no significant correlation. This could stem from fibrosis induced by radiochemotherapy. So far, only small cohort studies have been conducted and, for the future, studies with larger cohorts are required [28]. One major limitation of quantitative DW-MRI of the pancreas is its high sensitivity to motion artifacts, both to movement of the organ and to phase encoding artifacts due to bowel motion. When performing DW-MRI, high costs need to be taken into account as major downsides of this imaging modality. Nevertheless, DW-MRI seems to be an interesting non-invasive option for short-term follow-up during oncologic therapy of various primary tumors [29‐31]. The authors believe that further developments in this area will increase the diagnostic performance of DW-MRI.

Regarding the neoadjuvant therapy in the PaCa-DD-041 trial, the feasibility of the induction chemotherapy with gemcitabine and oxaliplatin, followed by radiochemotherapy with gemcitabine, was shown. Generally, the therapy was well tolerated and no grade 5 toxicities were observed. As determined using the RECIST criteria, a response (partial and complete) rate of 26% was recorded. In all, 61% of the study population showed stable disease. The high rates of stable disease can be explained by the moderate cytotoxicity of gemcitabine and oxaliplatin. By using more effective regimes like FOLFIRINOX, response rates can possibly be increased [17].

In summary, the PaCa-DD-041 trial was terminated early due to low accrual. A significant metabolic response based on FDG-PET was found in 50% of patients; however, this did not correlate with pathologic response following neoadjuvant treatment. Nevertheless, by using FDG-PET, monitoring of neoadjuvant therapy for trend prediction and to avoid ineffective and potentially harmful treatment is possible. The authors did not find a potential role for DW-MRI in early response evaluation. Further research needs to show whether FDG-PET response evaluation saves patients from undergoing ineffective, potentially toxic treatment.