Keywords

1 The Oligometastatic Hypothesis

Once a solid cancer is found to have metastasized to a distant organ, discussions between doctors and their patients change (Aitini and Aleotti 2006). These difficult conversations focus on palliative treatments, as opposed to curative measures. This common approach in oncology implies tumor cells are present in both macro- and micrometastases as soon as the malignancy has spread distantly and therefore cannot be completely eradicated. As such, systemic therapy is the mainstay of treatment for patients with metastatic disease.

Breast cancer was the original model for the metastatic sequence of solid tumors. In the early 1900s, William Stewart Halsted pioneered the paradigm of radical treatment for localized breast cancer (Halsted 1907). He contended cancer spreads sequentially, from a single location to regional lymph nodes, before eventually spreading to distant organs. Expanding on the work of Keynes (1954), Bernard Fisher presented an “alternative hypothesis.” suggesting breast cancer is a systemic disease at the time of diagnosis and local therapy is unlikely to impact the chance of overall survival (Fisher 1980). Fisher postulated cancer disseminated at the onset, not in a contiguous progression as Halsted had suggested. Samuel Hellman offered a third model for breast cancer spread, the “spectrum theory” (Hellman 1994) implying cancer presents on a spectrum of localized disease to wide spread distant metastases. In his theory, Hellman indicated metastatic sites, either nodal or distant, could be a source of further disease spread. Shortly after proposing the spectrum theory, Hellman and Weichselbaum described an intermediate state between local and widespread disease which they coined “oligometastases” (Hellman and Weichselbaum 1995).

2 Biology of Oligometastases

In their original publication, Hellman and Weichselbaum stated “… in the light of the emerging information on the multistep nature of cancer progression, we propose the existence of a clinical significant state of oligometastases. For certain tumors, the anatomy and physiology may limit or concentrate these metastases to a single or a limited number of organs.” Since the original “seed and soil” hypothesis by Stephan Paget in 1889 (Paget 1889), the biological progression of localized malignancy to distant spread has been further elucidated (Fidler 2003). This process includes local proliferation and angiogenesis with subsequent loss of cellular adhesion and increased motility. This leads to the interaction of malignant cells with platelets and other intravascular cells, which are transported throughout the circulatory system. This cell cluster will arrest in organs with adherence to the vessel wall followed by extravasation into tissue. Tumor cells will evade the host defense to establish a microenvironment, proliferate, and undergo angiogenesis in order to develop a marcometastasis.

It has been suggested patients with oligometastatic disease consist of deposits grown from sloughed cancer cells from the primary site, but have limited further metastatic and proliferation potential (Reyes and Pienta 2015). A plethora of preclinical in vitro and in vivo studies have shown a wide variation in the phenotypes of cells isolated from different primary and distant malignant tumor sites. Biologic basis for the clinical discrepancy between widespread and oligometastatic disease may include different primary tumor microenvironments, fitness of the migrant cancer cells, and the hospitability of host sites (Pienta et al. 2013).

Patients with a limited number of indolent metastatic deposits in different anatomic locations may represent the only trace of malignancy that remains. The natural history of cancer in this limited state may behave differently than the clinical course of a patient with diffuse metastatic disease. This crucial point of the oligometastatic theory suggests metastases-directed therapy through surgery and/or radiation combined with systemic therapy offers hope for patients previously deemed “incurable.” The theoretical curative potential of treating oligometastases makes aggressive treatment for these patients enticing.

3 Defining “Oligo” Metastases

There is not a consensus definition of what constitutes “oligo” with respect to counting the number of metastases (Treasure 2012). Most studies have defined the oligometastatic state to be a limited distant hematogenous spread of disease, generally involving 1–5 metastatic lesions. Furthering the oligometastic hypothesis, Niibe et al. described oligorecurrence to distinguish patients with controlled primary tumors who experienced improved outcomes compared to patients with uncontrolled primary disease (Niibe and Hayakawa 2010). The process of counting metastases to define oligometastatic disease is predicated upon the reliability of imaging studies used for staging. Novel imaging modalities have become readily available, including PET-CT and MRI, which allow for increased ability to evaluate patients for the presence of metastatic disease. For example, pretreatment work up with PET-CT for lung cancer leads to increased detection of occult metastases in 19% of patients (MacManus et al. 2001). Beyond medical imaging, there is a developing body of literature demonstrating the utility of circulating tumor cells to evaluate metastatic disease (Krebs et al. 2011). This creates a clinical predicament, how you look for metastases may ultimately determine the presence or absence of the oligometastatic state. Mathematical modeling to predict presence of additional occult metastases has been proposed (Kendal 2014), but has not been widely adopted. Additionally, biological prognostic tools are currently being studied and are discussed in more detail later on in this chapter.

With the aforementioned caveats, cancers presenting with presenting with fewer metastases have a distinct clinical behavior relative to patients with increased burden of disease. In prostate cancer, patients with five or fewer metastatic deposits have similar survival to patients with no evidence of metastatic disease at 5 years (73% vs. 75%) and 10 years (36% vs. 45%) (Singh et al. 2004). Furthermore, patients with more than five lesions exhibited a 5-year survival of 45% with only 18% of patients alive at 10 years. Early stage breast cancer patients who experience oligorecurrent disease, with less than five sites of disease, have improved median survival (108 vs. 22 months) compared to patients with greater than five sites of disease (Dorn et al. 2011).

A limited number of reports inform the incidence of oligometastatic presentation or recurrence. A retrospective series determined breast cancer relapses were isolated to the liver and/or one other organ in 59% of patients (Pentheroudakis et al. 2006). Data from prospective trials performed for the first line treatment of metastatic breast cancer indicate up to 91% of patients enrolled had ≤4 metastases at time of enrollment (Albain et al. 2008). Memorial Sloan Kettering Cancer Center published a series of patients with sarcoma and found 19% of patients presented with isolated pulmonary metastases as their first site of failure (Gadd et al. 1993).

4 Surgical Resection of Limited Metastases

In the mid 20th century, anecdotal evidence demonstrated metastatic renal adenocarcinoma to lung could be controlled long term with surgical resection of metastatic deposits (Barney 1945). There is a strong body of evidence supporting local treatment for limited metastatic disease in the setting of intracranial metastases. Randomized trials have demonstrated improvements in disease control and overall survival for patients treated with surgical resection or stereotactic radiosurgery (SRS) in addition to whole brain radiation therapy (WBRT) (Patchell et al. 1990; Andrews et al. 2004). Outside of the brain, there is surgical data demonstrating long-term disease control and survival in patients treated with metastectomy from sarcoma (van Geel et al. 1996) and breast cancer (Hanrahan et al. 2005) amongst other primary tumors. Patients presenting with spinal cord compression from solid tumors who undergo surgical decompression in addition to radiation have improved ambulatory function, continence, and survival compared to radiation monotherapy (Patchell et al. 2005).

Fong et al. published their experience with metastectomy of hepatic oliogmetastases for 456 patients with colorectal cancer treated between 1985 and 1991 (Fong et al. 1997). The treatment was well tolerated with low mortality and a post resection median survival of 46 months and 38% 5-year survival. A later publication showed 22% of these patients achieved 10-year survival and were effectively cured of their disease (Fong et al. 1999). Subsequent studies (Simmonds et al. 2006) lead to hepatic resection for oligometastases from colorectal cancer becoming the standard of care in the absence of a prospective clinical trial in an era prior to oxaliplatin and ironotecan chemotherapy backbones. Long-term survival post lung metastectomy has also been published. The International Registry of Lung Metastases reported outcomes of surgical resection of lung metastases on 5206 patients with metastases from a variety of primary tumor histologies. The series demonstrated 15-year survival rates of 22% (Pastorino et al. 1997). Intriguingly, patients with fewer metastases and a longer disease-free interval fared even better. There is preliminary evidence to suggest a subset of patients with limited metastatic disease may be curable with localized treatment beyond chemotherapy.

5 Using SBRT for Extracranial Oligometastases

In general, SBRT is less invasive than surgical resection and can be used to treat anatomic locations that may not be surgically accessible. SBRT is an attractive treatment modality for oliogmetastases since it is rapidly deployable, allowing limited interruptions in systemic therapy. Advancements in radiation treatment planning and delivery platforms have improved the quality and reliability of delivering ablative doses of radiation. However, there is a scarcity of high quality prospective randomized trials evaluating the use of SBRT in this setting. Multiple groups have analyzed retrospective case series or performed single arm dose escalation studies in an effort to better understand the clinical history of oligometastases which have been treated with ablative radiation therapy.

Investigators at the University of Chicago recently updated their series of 61 patients with five or less extracranial metastases who were treated on a dose escalation trial in which all known sites of metastasis were treated with ablative RT (Wong et al. 2016). At a median follow up of 2.3 years (6.8 years for survivors), Kaplan-Meier estimates of treated metastases control were 51% at 2 years and 44% at 5 years. 13 patients (21.3%) were alive at last follow up and 11.5% of patients never progressed after protocol therapy. Treatment was well tolerated with only 2 patients experiencing acute grade 3 toxicity and 6 patients with late grade 3 toxicity. There were no grade 4 or higher toxicities. The University of Rochester prospectively analyzed the role of SBRT in the treatment of one to five oligometastases, present in one to three organs. Patients with breast cancer showed a 2-year overall survival of 74% with 52% of patients free from widespread distant metastasis, and local control rate at 2 years of 87%. Long term (6 year) overall survival was 47% in this subset of patients 87% local control achieved. These values were all significantly higher than rates of disease control achieved for patients with metastases from non-breast primary tumors. On multivariate analysis, patients with bone metastases or single metastatic lesion experienced significantly improved survival. This study offers insight into selecting patients who may experience therapeutic benefit from the utilization of SBRT for oligometastases. It appears the patients most likely to garner benefit from SBRT include individuals with breast cancer primary tumors, single bony metastases, and stable to responsive disease prior to SBRT.

The largest published series evaluating outcomes after SBRT for oligometastases comes from Vrije University in Brussels, Belgium (de Vin et al. 2014). Their study included 309 patients with ≤5 metastases, 209 of whom were treated with SBRT to 430 extracranial lesions. 82.6% of extracranial lesions were treated with 10 fractions of 4 or 5 Gy. The majority (74%) of patients had a single anatomic site of disease with 46.3% of patients having only one metastatic lesion and 29.8% of patients having two lesions. The most common sites of disease were brain (34.6%), lymph node (28.5%), liver (24.9%), or lung (18.1%). Patients with a solitary extracranial metastasis had a median survival time of 40 months, whereas patients with two to five sites of disease achieved a median survival of 26 months. In an attempt to build a prognostic model for patient selection, de Vin et al. determined male sex, nonadenocarcinoma histology, presence of intracranial metastases, and synchronous presentation of metastases were associated with inferior outcomes. Stratifying patients by number of risk factors showed patients with two or fewer risk factors had a median overall survival of 23 months compared to 9 months for patients with three risk factors and 4 months if all four risk factors were present. Table 1 outlines the studies with the longest follow-up and highest patient numbers.

Table 1 Select studies of SBRT for multisite oligometastases
Full size table

6 First, Do No Harm?

The safety of SBRT to a distinct anatomic site of oligometastatic disease has been explored. A multi-institutional phase I/II study investigated the use of SBRT for oligometastatic cancer to lung (Rusthoven et al. 2009). Thirty-eight patients with an assortment of primary cancers were treated with SBRT on a dose escalation trial of 48–60 Gy in 3 fractions. The majority of patients (82%) were treated to 1 or 2 lesions with no extrathoracic metastases in 87% of patients. Local progression was only observed in 1 patient conferring a local control rate of 96% at 2 years. Two year overall survival was 39 and 63% of patients had distant progression. Treatment was well tolerated with no grade 4–5 toxicity. Only three of the 38 patients experienced grade 3 toxicity.

Berber et al. published the largest series exploring the use of SBRT for liver metastases (Berber et al. 2013). 153 patients with 363 metastases were treated to a dose between 31.3 and 46.5 Gy in 3 or 5 fractions. With a mean follow-up of 25.2 months, the overall local control rate was 62% and 1 year overall survival was 62%. Treatment was well tolerated with no grade 4–5 toxicity and only 3.2% of patients experiencing grade 3 toxicity. Other series exploring treatment of liver metastases with SBRT have shown grade 3–5 toxicity rates up to 18% (Carey Sampson et al. 2006). In one published experienced, three of 31 patients experienced grade 5 toxicity (Blomgren et al. 1995).

The use of SBRT for spinal metastases was studied in a multi-institutional phase II/III trial, RTOG 0631. Phase II results included 44 patients with 4 cervical, 21 thoracic and 19 lumbar sites treated with a single fraction of 16 Gy (Ryu et al. 2014). There was high quality treatment delivery with on 26% of patients with minor deviations in target coverage and spinal cord dose constraint met in 100% of patients. Treatment was well tolerated with only one patient experiencing grade 3 neck pain and no grade 4–5 events. The phase III component is randomizing patients to receive single fraction high dose SBRT (16 or 18 Gy) compared to standard palliation with a single fraction of 8 Gy with a primary end point of pain control at 3 months post treatment. A recent multi-institutional series of 541 patients (594 tumors) treated with spine SBRT showed a total of 34 patients (5.7%) had a new or progressive vertebral compression fracture following SBRT, with a median time to fracture of 3 months (Jawad et al. 2016). Preexisting fracture, solitary metastasis, and higher prescription dose (≥38.4 Gy) were associated with increased risk of fracture.

In summary, these limited data suggest for some patients with limited metastatic disease, local treatment of macroscopic tumor sites is generally well tolerated and may improve disease free intervals, and potentially, overall survival for select patients.

7 SBRT Treatment Planning

There is no absolute definition for high dose ablative radiation for extracranial disease. Stereotactic body radiotherapy (SBRT) and stereotactic ablative radiation (SABR) are used interchangeably. AAPM TG 101 suggested SBRT is typically comprised of 1–5 fractions of 6–30 Gy doses per fraction (Benedict et al. 2010). As summarized above, early studies evaluating the use of radiation therapy consisted of a more prolonged treatment course of hypofractionated radiation. The optimal radiation dose is influenced by several factors including the number and location of target lesions. Desired local disease control must be balanced with respecting surrounding normal tissue tolerance. In early stage lung cancer, there are data showing improved local control when the biologically effective dose (BED) is greater or equal to 105 Gy (Grills et al. 2012; Kestin et al. 2014). Excellent rates of local disease control with use of high BED SBRT has been shown in the oligometastatic setting (Salama et al. 2012). NRG-BR001 outlines a location-adapted approach for multi-organ site ablative radiation therapy (MOSART) SBRT (Table 2).

Table 2 MOSART prescription doses used in NRG-BR001
Full size table

In order to provide high precision SBRT, accurate patient positioning and immobilization is required. Respiratory motion analysis and management is imperative, particularly for lesions in the lung or liver, which exhibit significant movement with respiration (Benedict et al. 2010). GTV, CTV, ITV, and PTV volumes should be delineated depending on the anatomic location of the tumor and clinical scenario. There are many commercially available treatment delivery systems to enable reliable, high fidelity SBRT. The prescription isodose surface is chosen such that 95% of the target volume (PTV) is conformally covered by the prescription isodose surface. When evaluating target coverage, doses less than 95% of the prescription dose are restricted to the outside edges of the PTV. The prescription isodose surface selected used should typically be ≥60% and ≤90% of the dose maximum within the PTV. Treatment plans must be optimized to limit the high dose spillage to surrounding tissue. To assess the dose fall off, the ratio of prescription isodose volume to the PTV volume should be kept below 1.5 with a goal of 1.2. Moreover, the ratio of the 50% prescription isodose volume to the PTV (R50%) and the maximum dose a 2 cm (D2 cm) should be minimized. Suggested guidelines are outlined in Table 3. Priority should be placed on limiting radiation exposure to surrounding organs at risk, particularly for organs with grave potential toxicities (e.g. spinal cord). One, three, and five fraction SBRT OAR dose limits proposed in NRG BR002 (Table 4) are tabulated below. Circumferential irradiation of gastrointestinal tract structures (esophagus, duodenum, bowel, and rectum) should be avoided.

Table 3 Recommended treatment plan evaluation parameters
Full size table
Table 4 Organ-at-risk (OAR) dose limits used in NRG-BR002
Full size table

8 Future Directions

8.1 Combining SBRT with PD-1 Blockade

An intact immune system is important for controlling the neoplastic process. To enhance their proliferative transformation, tumors garner the ability to evade this immune regulation (Vinay et al. 2015). After decades of interest, but limited clinical relevance in solid tumors, the use of cancer immunotherapy has entered the mainstream over the past decade. With the identification of regulatory immune receptor to ligand interactions which influence immunity, “checkpoint” blocking monoclonal antibodies have become standard of care in multiple tumors (Pardoll 2012). The first of these approaches to enter clinical usage was the inhibition of cytotoxic T lymphocyte antigen 4 (CTLA4) being approved for the treatment of metastatic melanoma (Hodi et al. 2010). Since then, CTLA4 blockade has been studied in several other primary tumors including non-small cell lung cancer (Lynch et al. 2012).

Subsequent to the development of CTLA4 blocking antibodies, cancer immunotherapy has gained a broader usage with the production of monoclonal antibodies against the Programmed Death 1 (PD1): Programmed Death Ligand (PDL1) axis. The PD1:PDL1 interaction appears to be a major immune-evasion pathway up-regulated by some tumors to suppress anti-tumor immunity. Preliminary data suggests a potential synergistic effect on tumor response using PD-1 blockade in combination with radiotherapy (Drake 2012; Deng et al. 2014). This effect was observed in both tumors within the radiation field as well as distant tumors, suggesting the beneficial effects of radiation on the immune response have systemic impact. Clinical case reports have shown this abscopal response in sites distant from radiation while patients are receiving CTLA4 blocking immunotherapy (Postow et al. 2012).

Tumor cell death after high dose per fraction SBRT appears to be mediated through pathways beyond DNA damage and may enhance immune surveillance of tumors (Liang et al. 2013). The mechanism for this enhanced effect seems to include, at least in part, increased tumor antigen exposure, improved antigen presentation, and T cell function as well as modulation of immunosuppressive cell populations such as T regulatory cells and myeloid derived suppressor cells (Gaipl et al. 2014).

Beyond synergistic mechanisms of modulating the immune response, direct tumor debulking by radiation may also be particularly well suited as an adjunct to immunotherapy. Radiation to sites of bulk tumor would be presumed to improve the overall response rate of combination therapy. Additionally, reports of SBRT combinations with systemic therapies have suggested time to progression is delayed (Milano et al. 2012; Iyengar et al. 2014). Anti-PD1 antibody treatment may particularly be boosted by this approach. Clinical reports of the drug pembrolizumab suggest lower disease burden at the time of treatment initiation has been associated with higher response rate and one year survival in advanced melanoma (Joseph et al. 2014). Several phase I studies are ongoing to evaluate treatment with SBRT to various metastatic sites in patients with advanced solid tumors in conjunction with immune modulators (NCT02608385) (Bernstein et al. 2016).

8.2 Biological Prognostic and Predictive Tools

There have been recent advancements in the use of biologic markers to forecast disease behavior in oligometastatic patients. One such technology, microRNA classifiers, may help assess tumor biology and predict clinical outcomes. Significant differences in expression of microRNA200c occur between polymetastatic and oligometastatic phenotypes, with polymetastatic phenotypes expressing significantly higher levels of microRNA200c (Lussier et al. 2011). Using an oligometastatic-polymetastatic xenograft model, the group demonstrated oligometastatic cell lines could be induced to progress in a polymetastatic manner via the enhancement of microRNA200c. In the clinical setting, microRNA expression analysis in patients treated with pulmonary metastastectomy identified patterns that predicted higher rates of progression and lower rates of survival (Lussier et al. 2012). Wong et al. performed a microRNA expression analysis on 17 patients treated on their institutional protocol showing differential survival for patients exhibiting high and low classifier scores. Overexpression of a subset of microRNAs, miR-517a, miR-519c, and miR-521 directly correlated with poor long-term outcomes and increased cell proliferation. These data suggest certain tumors may exhibit an indolent nature, supporting Hellman and Weichselbaum’s original hypothesis. A priori selection of patients with indolent tumors may justify local therapy to interrupt further metastatic potentiation. These developments emphasize the importance of prospectively collecting biological and clinical outcomes in the treatment of oligometastases on a randomized controlled clinical trial.

9 Ongoing Clinical Trials

Several ongoing studies are accruing patients to assess the use of SBRT for oligometastases (Reyes and Pienta 2015). SABR-COMET is an international randomized phase II trial enrolling patients with up to 5 metastases (NCT01446744). All patients will be treated with standard of care chemotherapy and randomized to SBRT directed to all known oligometastases or no SBRT with the primary endpoint designed to detect a difference in in overall survival (Palma et al. 2012). The UK and Australia are conducting CORE trial (conventional care or radioablation in the treatment of extracranial metastases) (Aitken et al. 2014). This is a phase II trial enrolling patients with three or less extracranial metastases with metastatic non-small cell lung cancer, breast cancer, or prostate cancer with a primary endpoint of progression free survival. Patients are randomized to either standard of care with systemic therapy or standard of care systemic therapy combined with SBRT. Also in the UK, the Stereotactic Ablative Radiotherapy for Oligometastatic Non-small Cell Lung Cancer (SARON) trial is evaluating the use of systemic chemotherapy with or without radial RT to primary disease and up to 3 metastatic sites (NCT02417662). In prostate cancer, the Surveillance or metastasis-directed Therapy for OligoMetastatic Prostate cancer recurrence (STOMP) trial is currently ongoing. With a primary endpoint of androgen deprivation therapy free survival, the investigators are randomizing patients with metastatic disease to local therapy (surgery or radiation) or active clinical surveillance (NCT01558427). NRG Oncology has sponsored NRG-BR001 “A Phase 1 Study of Stereotactic Body Radiotherapy (SBRT) for the Treatment of Multiple Metastases” (NCT02206334). To parlay off the results of the phase I study, NRG-BR002 is a phase II/III trial comparing standard of care treatment to standard of care in addition to SBRT for women with 1–2 breast cancer metastases (NCT02364557). The trial is powered to address progression-free survival in the phase II study, and the study will automatically expand to a phase III design in the event a benefit in progression-free survival is observed in the phase II component.

10 A Cautionary Tale

With no randomized data to show the therapeutic benefit of SBRT for the treatment of extracranial oligometastases, the field may be in danger of putting the “cart before the horse.” Despite the lack of high quality evidence, local treatment for oligometastases has become the de facto standard of care (Bartlett et al. 2015; Lewis et al. 2015). This sets the stage for a phenomenon known as a “medical reversal,” when a widely adopted and accepted intervention is later found to have no clinically significant benefit (Prasad and Cifu 2013; Prasad et al. 2013). In oncology we are keenly aware of widespread implementation of an unproven therapy. Based on promising observational studies in the late 1980s and 1990s, it became commonplace to treat locally advanced and metastatic breast cancer with high dose chemotherapy followed by autologous hematopoietic stem-cell transplantation (Belanger et al. 1991). The proliferation of transplant clinics was sparked by a 1995 randomized study showing improvements in DFS and OS, which was later retracted (Bezwoda et al. 1995; Vickers and Christos 2000, 2001). Clinical trials published in the 2000s showed contrary findings, which prompted a steep decline in the use of transplant in breast cancer (Antman et al. 1997; Tallman et al. 2003; Berry et al. 2011). There has been an exponential rise in publications referencing oligometastases (Fig. 1) since the original publication by Hellman and Weichselbaum in 1995. To prevent another medical reversal, we encourage the prospective collection of data, preferably on a clinical trial. These data will allow us to conduct high quality analyses to answer clinical questions in order to best serve our patients now, and for years to come.

Fig. 1
figure 1

Number of publications, by year, with “oligometastasis”, “oligometases”, or “oligometastatic” in title

Full size image