A critical review on oligometastatic disease: a radiation oncologist’s perspective

The lack of consensus regarding the definition of OMD does not favor the prognostic characterization of these patients. In the last 20 years, numerous studies have begun to detect factors that could be helpful in daily clinical practice for a better patients’ selection [13, 14]. We herein report the most important series.

Number

Fode et al. [13] showed borderline significance (p = 0.049) for solitary metastasis in terms of OS, in a series of 321 OM patients treated with SBRT. In contrast, Franceschini et al. [14] reported no correlation between the number of metastasis (more than 1) and OS (p = 0.792) in a large cohort of 358 patients. Moreover, in various recent studies [14,15,16,17], the total number of metastases treated in OM patients does not seem to affect OS but only Progression-Free Survival (PFS). Phillips et al. [18] in a prospective II trial of SBRT in oligorecurrent prostate cancer showed that total consolidation of all Positron Emission Tomography (PET)-avid lesions resulted in a significant threefold increase in 6-month PFS and Distant Metastasis Free Survival (DMFS) with no grade 3 toxicity compared to those whose lesions were left untreated. These findings were mirrored in a phase II randomized trial of ART in patients with OM Castration Resistant Prostate Cancer (CRPC), where those who benefited most of the intensification of Abiraterone and SBRT in terms of complete biochemical response at 6 months were patients who underwent Prostate-Specific Membrane Antigen (PSMA) PET staging compared to non-PSMA PET staging [Odds Ratio 8.34 vs 1.32; p = 0.05)] [19]. Taken together these observations are consistent with the hypothesis that lesion consolidation by SBRT might alter the natural history of OMD by interfering with signals that promote further development of metastatic disease. Therefore, a numerical-based decision to withhold a local therapy may reduce the benefit of MDT in the treatment of OMD. Indeed, a recent ESTRO- American Society for Radiation Oncology consensus on OMD definition [20] claimed that “there is no biological evidence supporting the maximal number of metastases, or the maximal lesion size, that can be treated to provide clinical benefit”.

Site

Location of metastases is a controversial prognostic factor. Franceschini et al. [14] reported a strong correlation between the presence of lung or nodal metastases and longer OS (p = 0.001), whereas patients with liver or brain localizations were found independent predictors of any progression and poorer OS. In a large cohort of 270 patients with OM-ColoRectal Cancer (OM-CRC), Franzese et al. [17] showed a longer OS of lung metastasis compared with non-lung sites. The incidence of brain and liver metastases has increased with the advent of Magnetic Resonance Imaging, as advanced imaging has definitely improved our ability to detect small lesions earlier in the course of disease. This observation, along with more shared and homogeneous treatment regimens in the RO community, could allow us to achieve better results also in this setting. The ESTRO-EORTC consensus statement [12] about the definition of OMD did not distinguish extracranial from intracranial metastases, and recommended that patients with intracranial metastases should not be excluded from trials on OM patients. A recent review published by Suh JH et al. [21] highlighted that a subset of brain metastasis patients may live for years after diagnosis, especially those with “limited” intracranial disease (up to 4 metastasis), from specific primary histologies (e.g., breast cancer) and with targetable molecular alterations [22]. These findings emphasize the prognostic relevance of metastases location, and namely that lung lesions are associated with a better prognosis. However, the site of the metastasis as prognostic factor must be analysed in close relation with the primary histology, the number of lesions, and their size.

Size

The prognostic relevance of metastases’ size is still unclear. In a recent multicenter retrospective study with 1378 patients published by Yamamoto et al. [23], a maximum OM tumor diameter (per 1-cm increase) has showed a strong correlation (p < 0.001) with OS. However, Girard et al. [24] compared tumor size and detectability by number of tumors Doubling Times (DT). The authors hypothesized that the probability that even undetectable lesions are present increases as small as the metastases are found. More specifically, they suggest that only a single metastasis that has reached a diameter of 32 mm (34 of DT) would have an 80% probability of being truly solitary. Moreover, three large cohort studies [13, 17, 25] reported that a metastasis cut-off of 30 mm significantly correlated with better OS. Franzese et al. [17] in patients with OM-CRC showed that a Clinical Target Volume > 30 mm was associated with worse prognosis (p = 0.03). In a recently published multicenter large retrospective database on the personalization of Stereotactic ABlative Radiotherapy (SABR) use in lung metastases from CRC (LaIT-SABR study) [26], a correlation between tumor size and the development of the polymetastatic disease was demonstrated, other than metastases number. Specifically, patients with metastases diameter exceeding 20 mm and with > 3 metastases had a significantly short time to polymetastatic conversion. These findings support the presence of a survival prognostic size cute-off.

Tumor markers

Although no biomarkers that can differentiate between the oligometastatic and the polymetastatic state have been so far validated, some are routinely used as prognostic factors especially in CRC and prostate cancer. A short Prostate-Specific Antigen (PSA) DT is known to predict both the development of metastasis and prostate cancer-specific mortality in patients who underwent primary treatment [27]. PSA kinetics may be an important predictor of mortality in recurrent prostate cancer [28] and PSA DT is also a strong predictor of metastasis and survival in non-metastatic CRPC [29].

However, in the OM setting, PSA DT was not predictive of OS and PFS but only of Androgen Deprivation Therapy Free Survival (ADTFS) [17, 30, 31]. In metastatic CRC Thompson et al. [32], reported that pre-SBRT Carcinoembryonic Antigen (CEA) was a significant predictor of better OS with a predictive cut-off of 100 ng/ml.

In recent years, novel biomarkers are emerging through the use of liquid biopsy. The blood test of Circulating Tumor Cells (CTCs) and Circulating Tumor DNA (ctDNA) showed a prognostic value in OMD. In a prospective analysis of 43 patients with various primitive histologies, a lack of CTC clearance to ≤ 15/ml after 100 days by the end of SBRT was associated with progression of the irradiated lesion [33]. Lebow et al. [34] analyzed 820 patients with advanced Non-Small Cell Lung cancer (NSCLC) who underwent liquid biopsy with plasma next-generation sequencing of ctDNA. OMD was associated with a lower rate of ctDNA detection compared to polymetastatic disease, identifying a strong correlation between number of disease site and ctDNA. In the near future, these novel biomarkers will likely be integrated into a more comprehensive algorithm for defining OMD other than the number of metastases, and will represent a helpful tool to optimize the treatment strategy.

Timing

Disease-Free Interval (DFI), defined as the time between primary diagnosis and the detection of the first metastasis, is a deep-analyzed prognostic factor. Alongi et al. [35] reported a longer OS for DFI < 30 months for lung oligometastases. Similar data were found in a recursive partitioning-based analysis by Franzese et al. [17], who observed a DFI cut-off of 34 months in OM-Prostate Cancer. These data suggest that DFI is a promising parameter to detect which OM patients would benefit most from SBRT.

In a multi-institutional database recursive partitioning analysis reported by Chen et al. [36], patients with extracranial OM disease and metachronous presentation over 24 months showed a better OS (36.5 vs 17.1 months) compared with metachronous presentation ≤ 24 months. Interestingly, the time factor influenced OS only in case of extrapulmonary disease and specific histologies (NSCLC, Head and Neck, Breast triple negative, Melanoma, Sarcoma).

The EORTC-ESTRO OligoCare consensus recommendation differentiated into synchronous versus metachronous states, according to the interval between primary cancer diagnosis and development of OMD. As such, OMD is defined as synchronous if metastases are detected within 6 months from the initial diagnosis, and metachronous in case of a later appearance (at least 6 months) from the initial diagnosis [12]. There’s no shared consensus on defining the time point for synchronous/metachronous and the prognostic implications still remain unclear. Moreover, there are a number of studies with controversial data about the real prognostic impact on OS of metastatic disease timing. Three large retrospective studies reported no correlation between OM presentation (synchronous/metachronous) and survival outcomes (OS and PFS) [14, 15, 37]. A retrospective analysis of 194 patients with synchronous OM-NSCLC treated with MDT (radiotherapy, surgical therapy or other local ablative therapy) [38] showed a 5-years OS rate of 27–32% with a median follow-up of 52 months. In contrast, Fode et al. [13] show a favorable prognosis for metachronous metastases (p = 0.02) in a population represented for 98% of a location confined to a single organ (lung or liver). Obviously, the time factor is closely related to the ability of the imaging modalities to detect even the smallest lesions for a correct definition of a synchronous vs metachronous disease. Modern and more accurate staging systems (e.g., PET PSMA or liquid biopsy) are needed to consolidate DFI and OM presentation as prognostic factors. A recent prospective phase II trial testing the OM hypothesis in patients with positive PET PSMA for prostate cancer recurrence treated with MDT found a biochemical complete response rate of 22% without the use of Androgen Deprivation Therapy (ADT) or other therapies [39].

Prior systemic therapy

The numbers of systemic therapy lines prior to MDT could heavily influence the local response and survival outcomes in OM-CRC patients. Three large cohort studies found that numbers of systemic lines administered prior to SBRT have a negative influence on OS and PFS, suggesting that tumor cells surviving after ChemoTherapy (CT) acquire an improved DNA repair capacity, switching to a more radioresistant phenotype [17, 32, 40]. In particular, Klement et al. [40] showed a consistent dose–response relationship among pre-SBRT chemotherapy and local control. Unlike chemotherapy-naïve metastases, SBRT treatment with prior systemic therapy required a Biologically Effective Dose (BED) of more than 209 Gy10 to achieve 90% local control at 2 years. Furthermore, patients who received more than 3 lines of chemotherapy had the worst outcomes in terms of PFS rates (26% vs 55%). Conversely, immunotherapy seems to enhance the effect of radiotherapy, in particular at high doses per fraction. A recent retrospective analysis by Kroeze et al. [41] of 108 patients with oligoprogressive or polyprogressive disease treated with multiple line of Target Therapy (TT) or ImmunoTherapy (IT) and MDT shows a significant correlation between previous lines of systemic therapy and PFS (p = 0.033) at multivariate analysis. This data suggest that the reduced efficacy of subsequent lines of systemic therapy (CT or TT/IT) could drive distant progression. In this scenario, MDT could delay the systemic line therapy-switch and therefore play an increasingly important role.

Histology

The primary histology seems to be less correlated with local control, maybe due to the current wide use of ablative doses. Franceschini et al. [14] reported no significant association between any survival outcomes and tumor histology with a median BED administered of 105 Gy. However, some studies suggest a sort of radioresistance for CRC metastases [35, 40]. The presence of radioresistant histologies is largely due to their biological characteristics (e.g., high proportion of hypoxic cells) but also to a clonal selection caused by several lines of pre-RT chemotherapies [40]. Nevertheless, a strong evidence of correlation with OS exists for breast or prostate primary tumors, both showing better OS compared to the other tumors [42, 43]. Milano et al. [42] showed a significant discrepancy in terms of OS (p < 0.00001) for OM-breast cancer treated with a BED of 100 Gy10, with a 6-year OS rate of 47% compared to 9% for non-breast-histology. Franzese et al. [43] reported a 3-years OS rate of 88% in a population of OM-prostate cancer treated with a median BED of 157.5 Gy. In a recent update of a multi-institutional database consisting of 1033 patients with OM (≤ 5 metastases) treated with SBRT between 2006 and 2017 [44], conditional PFS stratified by primary site significantly increased over time for patients with CRC, breast and kidney cancer, remained stable for NSCLC and kidney cancer and significantly decreased for prostate, breast and CRC. Although primary histology remains a strong prognostic factor in terms of OS, regardless of the dose used, a routine use of ART doses is required to ensure a high rate of local control of the disease.

Performance status

The prognostic value of Performance Status (PS) is often underestimated. Most studies have demonstrated a strong correlation between OS and PS [13, 23]. In contrast, others large cohort studies [14, 17, 32] did not showed a significant statistical correlation at multivariate analysis, but only at univariate analysis. PS score incorporates a series of independents biological parameters of the patient (e.g., age, cognitive impairment, sarcopenia, malnutrition, advanced disease, comorbidities, pain cancer etc.) that heavily influence OS. Each of these parameters must be assessed individually based on the location of the disease and the primary histology of the tumor (e.g., cognitive impairment for brain metastasis, age and comorbidities for prostate cancer, pain for bone locations, sarcopenia and malnutrition for head and neck tumors, etc.). PS score should always be evaluated before offering treatments with a high risk of toxicity or less established therapeutic approaches. More personalized prognostic score based on activities of daily life and metastases site of tumors are required.

留言 (0)

沒有登入
gif