The utility of FDG-PET for assessing outcomes in oligometastatic cancer patients treated with stereotactic body radiotherapy: a cohort study

Metastases are the most common cause of cancer related mortality. The standard treatment for most metastatic cancer patients is usually chemotherapy or hormonal deprivation. Although chemotherapy has improved median survival duration in patients with metastatic cancer, response rates are low and chemotherapy is rarely curative, except in cases of hematologic malignancies and germ cell tumors [1‐6]. Previously, it has been proposed that within the spectrum of metastatic disease, a group of patients exists with metastases limited in number and location (i.e. oligometastatic patients), who may benefit from metastasis-directed therapy [7]. This hypothetical disease state is supported by long-term survival rates of 20-50% in patients undergoing limited pulmonary and hepatic metastasectomy [8, 9]. Recent technical developments in radiotherapy planning and delivery have allowed metastatic patients who are not medically fit for surgery, or those who are technically unresectable, to be treated with a few fractions of high doses of radiation to each lesion. Termed stereotactic body radiotherapy (SBRT), these techniques have been shown in prospective series to render some patients with limited metastases disease-free [10, 11] .

Response assessment following SBRT is typically performed via computed tomography (CT) scans with each individual lesion measured using RECIST criteria [12]. However, due to architectural changes of normal tissues surrounding treated metastases, particularly post-SBRT scarring, RECIST criteria are often difficult to use [13]. Furthermore, osseous metastases are not measureable by standard RECIST criteria, complicating assessment of treatment response in these regions commonly involved with oligometastases.

2-[18F]-fluoro-2-deoxy-D-glucose positron emission tomography (PET) imaging is a common metabolic imaging technique used to stage many cancer patients as well as assess outcomes following treatment. PET is particularly helpful in cases where anatomy has been altered, such as due to radiotherapy fibrosis or necrosis, where magnetic resonance imaging (MRI) and computed tomography (CT) may not be able to differentiate malignant pathology from post treatment changes [14, 15]. To our knowledge, there are no data on the use of PET for assessing SBRT response in patients with oligometastatic disease. Therefore, we evaluated the utility of PET as an indicator of treatment response in oligometastatic patients undergoing SBRT.

Patients included in this analysis were selected from participants in an IRB approved dose escalation study assessing the role of SBRT to all known sites of metastases in oligometastatic patients. From these patients, we identified those with PET imaging both prior to and after radiotherapy to form our study population.

We have previously reported trial methods and treatment outcomes [10]. Briefly, patients were eligible if they had pathologically confirmed stage IV metastatic cancer, were ≥ 18 years of age, had life expectancy of > 3 months, with metastases amenable to radiation therapy as seen on standard imaging (CT, MRI, PET, or bone scan). Patients were excluded if they had coexisting malignancies, uncontrolled medical comorbidity, active infectious processes, or exudative, bloody, or cytologically malignant effusions. Concurrent systemic chemotherapy was not allowed during radiotherapy. Each metastatic lesion was grouped into one of the following five disease sites: liver, lung, abdominal, head and neck, or extremity.

This analysis was performed after the first 50 patients participating in the dose escalation trial had been enrolled. Of these, 31 patients had PET and CT data available before and after treatment for analysis in this study. Patient and tumor characteristics are presented in Table 1. Median follow-up was 14 months (range: 3–41 months). Median time from cancer diagnosis to the development of metastastic disease was 11.1 months (range: 0–299 months). Median time from development of metastatic disease to SBRT was 6.9 months (range: 0–59.1 months). Median number of chemotherapy regimens received prior to SBRT was 1 (range: 0–4).

The majority of patients were women (52%). The most common primary histology treated was non-small cell lung cancer (9 patients), followed by sarcoma (5 patients), breast (4 patients), colon (4 patients), and head and neck cancer (4 patients). In total 58 metastatic lesions were treated, including 19 lung, 11 osseous, 11 nodal, 9 liver, 6 adrenal and 2 soft tissue. Treatment dose was selected per protocol and listed in Table 1. Two patients were treated with 50 Gy in 10 daily fractions of 5 Gy.

Following treatment, the median time to the first post-therapy PET was 1.2 months (range: 0.5-4.1). All patients had one post-therapy PET, 24 of 31 (77%) had two post therapy PET scans and 9 of 31 (29%) had three post-therapy PET scans. The median time to the second PET scan was 4.4 months (range: 1.4-26.8 months) and the median time to third PET scan was 7.1 months (range: 6.4-20.4 months).

Clinical decision making for metastatic patients undergoing treatment hinges on the clinician’s ability to accurately assess the initial therapy response, as well as the duration of response. CT and MRI are routinely used to evaluate response to therapy, with RECIST criteria being used to assess response for patients on study and similar criteria used for off study patients. However, as used in routine clinical practice, these technologies and measurement standards have limitations. Postsurgical and postradiotherapy changes alter fascial planes and create fibrosis and necrosis of both malignant tissue as well as surrounding normal tissue, often making it difficult to differentiate between malignant and non-malignant tissues in surveillance CT and MRI imaging [14, 15, 17, 18]. This is especially the case when the size of tumor does not change even though pathologically the tissue is no longer malignant.

PET technology is reliant upon the altered metabolic biology of malignant cells compared to normal cells [19]. 2-[18F]-fluoro-2-deoxy-D-glucose is the most frequently utilized radiotracer in clinical practice. As availability and technology improve, PET is increasingly being used in diagnosis, staging, radiotherapy planning, early response assessment, and restaging after completion of therapy for several malignancies [19]. In the detection of recurrent head and neck tumors, PET has been shown to have a sensitivity and specificity approaching 90-100% [20‐22], and in primary lung cancer, PET is increasingly used to differentiate tumor from atelectasis and consolidation.

In our study, we found that FDG-PET appeared to be a useful adjunct for response assessment for patients with oligometastatic disease treated with SBRT. Not only did most patients with a CT response have a PET response, but patients who were not evaluable with CT were able to be assessed with PET scan effectively. Furthermore, many of the patients with partial response on PET proceeded to have a complete PET response with further follow-up. This can be seen in Figure 1, which depicts the sequential PET imaging from a patient in our study.

In our patient population, PET was helpful in assessing a response in patients with stable disease by anatomic evaluation on CT. This has significant implications on clinical decision making. The concern for treatment resistant disease makes it crucial for the clinician to rely on imaging to monitor disease response. However, disease may not morphologically change on CT scan. In this situation, PET allows for an understanding of tumor physiology, and hence a decrease in metabolic uptake (PR or CR) would indicate decreased tumor activity and possible response to treatment. In the setting of a stable CT examination, this could decrease unnecessary or untimely changes of treatment for suspected resistance to treatment. A stable CT scan can mean a number of things in this situation. The patient could have fibrosis or necrosis secondary to treatment, or conditions such as atelectasis in the lung [14, 22, 23]. On CT it could be difficult to identify viable tumor tissue compared to these post-treatment changes, but on PET it would be easier to distinguish tumor from normal tissue. PET may have a role in assessing response after SBRT in primary and metastatic lung tumors [24, 25, 26]. PET imaging’s strength is in this ability to differentiate for both planning purposes as well as long term analysis of patterns of progression.

Interestingly, patients in our study with osseous metastases had very encouraging early and late response to SBRT. As osseous metastases cannot be evaluated with standard RECIST criteria, the use of PET in response of assessment is particularly important [12]. As is demonstrated by Figure 2, PET is an effective tool to use in evaluating for response in these patients.

This is the first study to evaluate the utility of PET for assessing response following SBRT for patients with limited metastatic disease. PET has previously been evaluated in patients treated with SBRT for Stage 1 non-small cell lung cancer. Ishimori et al. studied the use of FDG-PET in assessing the response of patients treated with SBRT [18]. PET studies were performed 1 week before and 1–8 weeks after treatment. Of nine patients with imaging, two had a PET CR and seven had a PET PR. Hoopes et al. found that in 28 patients with PET follow-up after SBRT, 4 patients had prolonged elevation of SUV at 22–26 months but no clinical or radiographic evidence of local, nodal, or distant disease [11, 23].

It is interesting to note that following treatment, residual PET activity may be present. This could indicate a protracted inflammatory response to higher dose per fraction radiotherapy. While protracted low level PET activity has not been seen to predict for recurrence or late toxicity, this may warrant further investigation.

It is important to acknowledge that a limitation of our study is the relatively short median follow-up of our cohort, mostly due to death from malignancy. Additionally, it is difficult to make conclusions from our data regarding the cumulative PET and CT response to SBRT for time periods longer than our follow-up. Our results must also be interpreted in light of the small and heterogeneous patient population as well as the non-standardized follow-up PET schedule amongst our cohort.

PET was effective at detecting responses in oligometastatic patients undergoing curative intent SBRT. CR by CT correlated with PET CR. PET was able to detect response in patients with stable disease by RECIST criteria. Long-term PET response was related to radiation dose delivered. PET scan should be part of the routine follow-up for patients undergoing curative intent SBRT to determine response to therapy.

Supported by the Ludwig Center for Metastasis Research and the University of Chicago Cancer Research Center Grant 5–30073