Abstract
Purpose
Patients with oligo-metastatic disease (OMD) can be safely treated with Stereotactic Radiation Therapy (SRT). Further disease progression is common in these patients. In most cases, patients relapse again with oligo-metastases, however some can experience a poly-progression after a local ablative treatment (LAT). The purpose of this study was to retrospectively identify factors associated with poly-progression in patients receiving SRT for OMD.
Methods
Data from a monocentric database were retrospectively analyzed. Patients treated with SRT for OMD and who developed progression after LAT were selected. Patients were categorized as oligo- or poly-progressive according to the number of new/progressing metastases (≤ or > 5). Herein, we analyzed data about patients’ characteristics, oligo-metastatic presentation and radiation treatment characteristics to evaluate their relationship with progression type.
Results
From 2013 to 2021, data on 700 patients progressing after LAT were analyzed. Among them, 227 patients (32.4%) experienced a poly-progression; the median time to poly-progression was 7.72 months (range 1–79.6). Five variables associated with poly-progression were found to be statistically significant in the univariate analysis: performance status (p < 0.001), site of the primary tumor (p = 0.016), ablative dose (p = 0.002), treated site (p = 0.002), single or double organ (p = 0.03). Of those, all but the number of involved organs retained their significant predictive value on the multivariate analysis.
Conclusion
Our study identified four independent factors associated with poly-progression in patients with OMD receiving SRT. Our data may support comprehensive characterization of OMD, better understanding of factors associated with progression.
Keywords: Oligo-metastases, Stereotactic radiation therapy, Predictive factors, Pattern of failure, Poly-progression
Introduction
Despite increasing evidence and studies on the existence of an oligo-metastatic state (OMD), its definition has remained generic and almost unmodified since 1995 (Hellman and Weichselbaum 1995). Recent efforts conducted by ESTRO and ASTRO (Lievens et al. 2020) did not end up with a more accurate definition, so OMD was still defined as 1–5 metastatic lesions, a primary controlled tumor being optional, but where all metastatic sites must be safely treatable.
A parallel initiative by ESTRO and EORTC (Guckenberger et al. 2020) contributed to creating a common language for OMD and identified nine possible subcategories of this state. This commendable work highlighted a common clinical observation: OMD is a very heterogeneous scenario, and the absence of reliable biomarkers for identifying it makes clinical results extremely variable.
One of the clinical correlates of this variability is the different failure patterns of OMD after local ablative treatments (LAT), including stereotactic radiotherapy (SRT). Other studies observed that a significant proportion of OM patients at the time of progression remains “oligo”. For instance, Poon et al. (2020) found in an extensive series of 1033 OM patients that 33.1% of progressing patients had a limited disease dissemination, still satisfying the criteria for OMD classification. In a specific clinical scenario like prostate cancer, Decaestecker K. et al. (2014) found that 75% of patients experienced further OM relapse after a first course of SRT for oligo-metastases. As shown in this study, there is a clinical relevance to this particular recurrence pattern since these patients can be again treated with a LAT, with possible positive prognostic implications. Repeated LAT and multiple courses of RT have been linked to long-term survival by Sebastian et al. (Sebastian et al. 2021). In this study, patients who received more than 5 RT courses had a numerically more prolonged overall survival (OS) (median 4.5 vs 2.8), although this difference was not statistically significant (p value = 0.073).
Changing the point of view, this progression pattern also means that many OM patients, unfortunately, experience a poly-metastatic diffusion as the first event after LAT. Considering that one of the cornerstones of the OMD definition is the indolent behavior, a rapid poly-metastatic progression could be regarded as a signal of a misleading, although unpredictable with available knowledge, classification of that patient as oligo-metastatic. At the same time, treatment intensification could be pursued for these patients if we could distinguish them from clinically similar “true” oligo-metastatic patients.
On this background, we designed this analysis on progressing patients derived from our institution's larger database of OM patients treated with SRT. We analyzed progression patterns, distinguishing oligo and poly-progression, to identify factors associated with poly-progression.
Materials and methods
Study population
Data from an institutional database of OM patients treated with SRT were retrospectively analyzed to select patients who had a further distant disease progression after the treatment. Database inclusion criteria have already been described in previous publications (Franceschini et al. 2019, 2022). For the present analysis, patients were excluded if they did not have a distant progression after SRT, if they had exclusively a local progression of the previously irradiated lesion(s) and if, at the time of the first SRT, they already had other sites of disease not treated because apparently under control (extra target disease). Patients with oligo-progression were included in the present analysis only if they were treated for all visible diseases.
The study was conducted with the approval of institutional review boards, and each patient signed an informed consent at SRT time to use data for future research.
The endpoint of the analysis was the type of distant progression, categorized as oligo or poly according to the number of new appearing lesions (≤ or > 5). Data about patients’ characteristics, oligo-metastatic presentation and radiation treatment characteristics were used to evaluate their relationship with progression type.
Statistical analysis
Kaplan–Meier actuarial analysis was used to generate the poly-progression-free time evolution data. Time was scored from the first day of SRT.
Univariate analysis was performed on several potential clinical and technical predictors to identify the subset of factors influencing the poly-progression history of the patients.
These included: sex, age (dichotomized at the median), site of the primary tumor (categorized as colon, lung, breast, prostate and other sites), the histology of the primary tumor (adenocarcinoma, squamous cellular carcinoma, other), disease-free interval defined as the time from primary diagnosis to first occurrence of metastatic diseases (categorized at one-year threshold), presence of comorbidities, ECOG performance status (PS) (categorized at 0 or ≥ 1), presence of synchronous or metachronous metastases, oligo-metastases type [de novo, repeated or induced according to ESTRO EORTC classification (Guckenberger et al. 2020)], the execution of ablative treatments (with a biological equivalent dose [BED] threshold at 100 Gy), the number of treated lesions (1,2, more than two), the treatment site (lung, brain, liver, adrenal gland and lymph nodes), the treatment of a single or double organ, the delivery of previous chemotherapy cycles, the presence of concomitant systemic treatments delivered immediately before, during or after SRT, local progression.
Factors identified as statistically significant (log-rank p value ≤ 0.05), all factorial, were included in a multivariate Cox regression model. Hazard Ratios (HR) were computed with regard to the last category for each factor.
The analysis was carried out using the SPSS Statistics software v.22 (IBM, Armonk, USA).
Results
From the original database, including more than 1300 patients, 700 were selected for the analysis. The median age at SRT was 68.6 years (mean: 67.2 ± 11.9, range: 20.1–91.9 years). Table 1 summarizes some descriptive statistics for the patient's cohort.
Table 1.
Descriptive statistics for the whole patient’s cohort
| Factors | Groups | Number of patients (%) |
|---|---|---|
| Sex | Males | 281 (40.1) |
| Females | 419 (59.9) | |
| Age | Median | 68.6 years |
| Range | 32–89 years | |
| Site of primary tumor | Colon-Rectum | 174 (24.9) |
| Lung | 157 (22.4) | |
| Breast | 55 (7.9) | |
| Prostate | 60 (8.6) | |
| Melanoma | 26 (3.7) | |
| Gastroesophageal | 36 (5.1) | |
| Gynecological (mix) | 33 (4.7) | |
| Pancreas | 48 (6.9) | |
| Renal | 30 (4.3) | |
| Sarcoma | 17 (2.4) | |
| Other | 64 (9.1) | |
| Primary tumor histology | Adenocarcinoma | 460 (65.7) |
| Squamous Cell Carcinoma | 47 (6.7) | |
| Other | 193 (27.6) | |
| Disease-free interval | ≤ 1 year | 296 (42.3) |
| > 1 year | 404 (57.7) | |
| Comorbidities | No | 185 (26.4) |
| Yes | 515 (73.6) | |
| Performance status (ECOG) | 0 | 422 (60.3) |
| 1 | 236 (33.7) | |
| 2 | 41 (5.9) | |
| 3 | 1 (0.1) | |
| Metastases timing | Synchronous | 188 (26.9) |
| Metachronous | 512 (73.1) | |
| Oligo-metastases type | De novo | 300 (42.9) |
| Repeat | 102 (14.6) | |
| Induced | 298 (42.6) | |
| Previous chemotherapy | No | 231 (33.0) |
| Yes | 469 (67.0) | |
| Treated lesions | 1 | 403 (57.60 |
| 2 | 181 (25.9) | |
| ≥ 3 | 116 (16.6) | |
| Site of treated lesions | Lung | 221 (31.6) |
| Brain | 78 (11.1) | |
| Liver | 129 (18.4) | |
| Adrenal glands | 18 (2.6) | |
| Lymph-nodes | 175 (25.0) | |
| Other | 79 (11.3) | |
| Multiple organs | Single | 628 (89.7) |
| Double | 72 (10.3) | |
| Ablative dose | < 100 Gy | 274 (39.1) |
| ≥ 100 Gy | 426 (60.9) | |
| Concomitant systemic therapy | No | 571 (81.6) |
| Yes | 129 (18.4) | |
| Local Progression | No | 569 (81.3) |
| Yes | 131 (18.7) |
Four hundred seventy three patients (67.5%) progressed in an “oligo” way. More than half of them (53.6%) had a further relapse in the same organ, 37.8% in a different organ and the remnant 40 patients both in the same and other organs. Among the 227 poly-progressive patients (32.5%), in 62 (27%) cases, progression was limited to the same organ of the SRT target lesion(s). Only different organs were involved in 81 patients (35.7%). At the same time, progression in the same and other body sites was experienced by 84 patients (37%). Median time to distant progression was 7.8 months (range 1–80.76), for oligo progressive patients 8.2 months (range 1–80.76) and for poly-progressive patients 7.7 months (range 1–79.6).
No further treatment was performed in 94 (13.4%) patients due to advanced age, poor performance status or early death. Further LAT was administered to 298 patients (42.5%), all but eight having an oligo-progression, without changing or initiating a new systemic therapy. A combination of LAT and new systemic therapy was chosen for 28 cases (4%), while a pharmacologic approach only was adopted in 236 patients (33.7%), more than half of them having a poly-progression. The remnant cases were candidates only for palliative RT.
Table 2 summarizes the univariate analysis for the factors that resulted as statistically significant. The results are reported as the mean and the median time to poly-progression in years, together with the corresponding 95% confidence intervals and the p values. Results are provided for all sub-groups for each factor. Figure 1 shows the corresponding actuarial curves. Five factors resulted significant: the delivery of an ablative dose, the initial performance status, the treatment of single or double organ, the primary tumor site and the treated site.
Table 2.
Univariate analysis summary for statistically significant predictors. Mean and median time (in years) to poli-progression
| Predictor | Mean | 95% C.I | Median | 95% C.I | p | |
|---|---|---|---|---|---|---|
| Ablative dose | < 100 Gy | 2.1 ± 0.2 | 1.8–2.5 | 1.3 ± 0.2 | 0.9–1.6 | 0.002 |
| ≥ 100 Gy | 3.1 ± 0.2 | 2.7–3.6 | 2.3 ± 0.3 | 1.8–2.8 | ||
| Performance status | = 0 | 3.2 ± 0.3 | 2.7–3.7 | 2.3 ± 0.2 | 1.8–2.8 | 0.001 |
| ≥ 1 | 2.3 ± 0.2 | 1.8–2.8 | 1.4 ± 0.2 | 1.1–1.8 | ||
| Organ single/double | Single | 3.0 ± 0.2 | 2.6–3.4 | 1.9 ± 0.2 | 1.4–2.4 | 0.03 |
| Double | 1.4 ± 0.2 | 1.1–1.7 | 1.0 ± 0.1 | 0.8–1.3 | ||
| Primary tumor | Colon | 3.0 ± 0.3 | 2.5–3.5 | Not reach | – | 0.016 |
| Lung | 2.5 ± 0.3 | 1.9–3.1 | 1.9 ± 0.3 | 1.2–2.6 | ||
| Breast | 2.0 ± 0.5 | 1.0–3.0 | 0.7 ± 0.2 | 0.3–1.2 | ||
| Prostate | 2.5 ± 0.2 | 2.1–2.9 | Not reach | – | ||
| Other | 2.7 ± 0.3 | 2.2–3.2 | 1.5 ± 0.3 | 1.0–2.0 | ||
| Treated site | Lung | 3.2 ± 0.3 | 2.6–3.7 | 2.4 ± 0.6 | 1.2–3.6 | 0.002 |
| Brain | 1.6 ± 0.2 | 1.2–1.9 | 1.5 ± 0.1 | 1.3–1.7 | ||
| Liver | 2.4 ± 0.4 | 1.6–3.2 | 1.0 ± 0.1 | 0.7–1.3 | ||
| Adrenal gl | 2.4 ± 0.6 | 1.3–3.6 | 1.6 ± 0.4 | 1.4–1.9 | ||
| L.N | 2.8 ± 0.2 | 2.4–3.2 | 3.0 ± 0.3 | 2.2–3.5 |
Uncertainties are expressed as standard error and 95% confidence interval
Fig. 1.
Poli-progression-free survival rate according to significant factors at multivariate analysis
Table 3 summarizes the results of the multivariate analysis with the factors retained in the model, the corresponding p value and the HR). From the list of significant factors of the univariate analysis, only the single/double organ treatment was excluded from the final model. Patients treated with ablative dose, with a better PS, treated for lung metastases or lymph nodes and with prostate cancer as primary site resulted in better performance as derived from the univariate and multivariate analysis. In detail, among patients treated with a BED < 100 Gy, 38% had a poly-progression compared to 28.9% of patients treated with a higher dose. Patients with a PS of 0 had oligo-progression in 70.6% of cases and poly-progression in 29.4%, compared with 62.9 and 37.1% respectively in patients with a PS of 1 or higher. Patients irradiated on lungs or nodes had a similarly low rate of poly-progression (27.1 and 29.2%, respectively), while the brain and primarily liver metastases correlated with a higher risk (34.6% and 41.1%, respectively). Regarding the primary tumor site, the highest rate of poly-progression was found in breast cancer (43.6%), followed by lung cancer (31.8%). A lower risk was linked to the prostate or colorectal cancer (28.3 and 25.9%).
Table 3.
Summary of multivariate Cox regression (backward conditional)
| Predictor | p | Hazard ratio |
|---|---|---|
| Ablative dose | 0.02 | 1.54 |
| Performance Status | < 0.001 | 0.59 |
| Treated site | 0.02 | |
| Lung | 1.0 | |
| Brain | 1.2 | |
| Liver | 2.0 | |
| Adrenal glands | 1.1 | |
| Primary tumor site | 0.01 | |
| Colon | 0.72 | |
| Lung | 0.86 | |
| Breast | 1.49 | |
| Prostate | 0.7 |
All predictors are categorical and hazard ratio is computed with respect to the last category
Discussion
In the present study, we report factors associated with poly-progression after the first course of SRT in OM patients. To the best of our knowledge, this is the first analysis focusing on this specific research question. We found that four parameters influenced the type of progression: BED, PS, metastatic site and primary histology.
The prognostic role of RT dose has already emerged in other studies on OM patients. Hong et al. (2018) generated a model for OS prediction in OM patients. In their experience, a BED higher than 75 Gy was significantly correlated with outcome, with a 3 year OS of 61 vs 43% (p < 0.01). BED maintained its predictive role also at multivariate analysis. Nicosia et al. (2022) analyzed many colorectal lung metastases treated with SRT. They found BED ≥ 125 Gy significantly reduced the risk of local progression (HR 0.24, 95% CI 0.11–0.51; p = 0.000), poly-metastatic (HR 0.45, 95% CI 0.29–0.69; p = 0.000) and prolonged OS. (median OS with a BED < 100 Gy, 100–124 Gy, and ≥ 125 Gy were 35.5, 42, and 58.5 months, respectively, p = 0.0045). In our study, we used a cut-off of 100 Gy, historically used to discriminate between ablative and sub-ablative doses in SRT. In line with this data, we found that a BED > 100 Gy also reduced the risk of poly-metastatic dissemination at first recurrence after SRT. This again highlights the importance of delivering a radical treatment to OM patients to have a real impact on the disease trajectory.
Performance status is a well-known prognostic factor in oncology. Even in the OM setting, PS is often included in prognostic models, for instance, in the Metabank (Begin et al. 2019). This tool created by Van den Begin et al. included PS, sex, the timing of oligo-metastases, metastatic site and histology in a score to predict OS in OM patients. Since PS represents an indirect sign of general patient status, its predictive role is intuitive. However, the correlation between PS and pattern of progression as in our study is less immediate. We think that a higher PS could be a manifestation of a higher, although occult, disease burden that becomes manifest at the time of poly-progression.
In our study, we also found that metastatic sites could predict the pattern of failure. Indeed we found a protective role against poly-progression of lung and nodal metastases, with a higher risk of disease dissemination in case of brain and liver irradiation. The fact that brain metastases on one side and nodal metastases on the other side represent different types of metastatic disease was quite predictable. Instead, the large difference we found between lung and liver metastases was less expected. However, looking at the literature, a protective role of lung metastases concerning disease progression and risk of death has also been reported in other experiences. Poon et al. (Poon et al. 2020) found a statistically significant correlation at multivariate analysis between OS and lung only (HR, 0.58; 95% CI 0.48–0.72; p < 0.001) or nodal/soft-tissue metastases only (HR, 0.49; 95% CI 0.26–0.90; p = 0.02). Therefore, they included lung localization of metastases in their survival model, consisting of 3 risk classes (Chen et al. 2021).
Lastly, in our study, the primary tumor site also influenced the pattern of progression, with the worse outcome for patients affected by breast cancer. This result could seem in contrast with traditional predictors of survival in the OM scenario. Indeed, breast cancer primary is often related to a longer OS in the large database analysis aiming at prognosis estimation (Hong et al. 2018; Vin et al. 2014; Wong et al. 2016). However, other studies showed a higher HR of death for breast cancer patients compared to prostate cancer (Poon et al. 2020). Anyway, the endpoint of our analysis is not survival but a pattern of progression. With available effective systemic therapies, it is likely that, despite this higher risk of widespread dissemination, survival outcomes of breast cancer patients will end up being similar or better than those obtained with other histologies.
The high number of patients collected and the numerous parameters analyzed are the strength of the present study. However, the heterogeneity of the series and the retrospective nature still limit the validity of our results. Moreover, the exclusion of patients who did not progress after SRT could have introduced a selection bias. However, our aim was to focus only on progressing patients, to better understand if we could predict the unfavorable event of early poly-progression.
We are convinced that clinical parameters, like the one analyzed in this study, have intrinsic limitations to their predictive power in the OM scenario. Until we will not be able to study and understand the underlying disease biology and until we will continue to define OM disease based on a generic clinical definition, all the models and the prediction will be severely limited. As shown by Lussier et al. (2012) almost ten years ago, the key is linked to the genetic signature of the disease. They analyzed microRNA expression patterns from lung metastasis samples of 63 OM patients resected with curative intent. With this approach, they could discriminate between two classes of patients with a high and low rate of progression.
There are logical barriers to this approach for patients treated with SRT since an invasive diagnosis is often not performed. However, the recent developments in liquid biopsy and next-generation sequencing (NGS) could give us a unique opportunity to understand OM disease better.
Conclusion
We found a correlation between BED, PS, primary tumor site and irradiated metastatic site with the risk of poly-progression in OM patients treated with SRT. A better understanding of the risk of progression, particularly of the risk of rapid, widespread dissemination, could be helpful for an individualized treatment intensification or de-intensification. Translational research is required to reveal the intrinsic biology of OM disease and increase prognostic classifications' accuracy.
Author contributions
DF: Conceptualization; Data Curation; Supervision; Writing original draft. LC: Data Curation; Formal Analysis; Methodology; Writing–review and editing. VV: Data Curation; Writing original draft. AMM: Data curation; Writing–review and editing. BM: Data curation; Writing–review and editing. PN: Writing–review and editing. MS: Supervision; Writing–review and editing.
Funding
The authors declare that no funds, grants, or other support were received during the preparation of this manuscript.
Data availability
Research data are stored in an institutional repository and will be shared upon request to the corresponding author.
Declarations
Competing interests
Luca Cozzi: Stock or stock options and Other financial or non-financial interests with Varian Medical Systems All other authors declare no conflict of interest
Ethical approval
This study was performed in line with the principles of the Declaration of Helsinki. This is a retrospective observational study.
Consent to participate
All patients signed an informed consent allowing the use of their data for scientific purposes
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- Chen H, Poon I, Atenafu EG et al (2021) Development of a prognostic model for overall survival in patients with extracranial oligometastatic disease treated with stereotactic body radiation therapy. Int J Radiation Oncol Biol Phys 000(00):1–10 [DOI] [PubMed] [Google Scholar]
- de Vin T, Engels B, Gevaert T, Storme G, De Ridder M (2014) Stereotactic radiotherapy for oligometastatic cancer: a prognostic model for survival. Ann Oncol 25:467–471 [DOI] [PubMed] [Google Scholar]
- Decaestecker K, De Meerleer G, Lambert B et al (2014) Repeated stereotactic body radiotherapy for oligometastatic prostate cancer recurrence. Radiat Oncol 9:135. 10.1186/1748-717X-9-135 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Franceschini D, De Rose F, Franzese C et al (2019) Predictive factors for response and survival in a cohort of oligometastatic patients treated with stereotactic body radiation therapy. Int J Radiat Oncol Biol Phys 104(1):111–121 [DOI] [PubMed] [Google Scholar]
- Franceschini D, Polenghi V, Franzese C et al (2022) Oligoscore: a clinical score to predict overall survival in patients with oligometastatic disease treated with stereotactic body radiotherapy. Acta Oncol 61(5):553–559 [DOI] [PubMed] [Google Scholar]
- Guckenberger M, Lievens Y, Bouma AB et al (2020) Characterization and classification of oligometastatic disease: a European society for radiotherapy and oncology and European organisation for research and treatment of cancer consensus recommendation. Lancet Oncol 21:e18-28 [DOI] [PubMed] [Google Scholar]
- Hellman S, Weichselbaum RR (1995) Oligometastases. J Clin Oncol 13:8e10 [DOI] [PubMed] [Google Scholar]
- Hong JC, Ayala-Peacock DN, Lee J et al (2018) classification for long-term survival in oligometastatic patients treated with ablative radiotherapy: a multi-institutional pooled analysis. PLoS ONE 13:e0195149 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lievens Y, Guckenberger M, Gomez D et al (2020) Defining oligometastatic disease from a radiation oncology perspective: an ESTRO-ASTRO consensus document. Radiother Oncol 148:157–166 [DOI] [PubMed] [Google Scholar]
- Lussier YA, Khodarev NN, Regan K et al (2012) Oligo- and polymetastatic progression in lung metastasis(es) patients is associated with specific MicroRNAs. PLoS ONE 7(12):e50141 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Nicosia L, Franceschini D, Perrone-Congedi F et al (2022) A multicenter large retrospectIve database on the personalization of stereotactic ablative radiotherapy use in lung metastases from colon-rectal cancer: the lait-SABR study. Radiother Oncol 166:92–99 [DOI] [PubMed] [Google Scholar]
- Poon I, Erler D, Dagan R et al (2020) Evaluation of definitive stereotactic body radiotherapy and outcomes in adults with extracranial oligometastasis. JAMA Netw Open 3(11):e2026312 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Sebastian M, Christ SM, Ahmadsei M, Wilke L et al (2021) Long-term cancer survivors treated with multiple courses of repeat radiation therapy. Radiat Oncol 16:208 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Van den Begin R, Engels B, Collen C et al (2019) The METABANK score: a clinical tool to predict survival after stereotactic radiotherapy for oligometastatic disease. Radiother Oncol 133:113–119 [DOI] [PubMed] [Google Scholar]
- Wong AC, Watson SP, Pitroda SP et al (2016) Clinical and molecular markers of long-term survival after oligometastasis-directed stereotactic body radiotherapy (SBRT). Cancer 122:2242–2250 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
Research data are stored in an institutional repository and will be shared upon request to the corresponding author.