The biology and treatment of oligometastatic cancer.

Clinical reports of limited and treatable cancer metastases, a disease state that exists in a transitional zone between localized and widespread systemic disease, were noted on occasion historically and are now termed oligometastasis. The ramification of a diagnosis of oligometastasis is a change in treatment paradigm, i.e. if the primary cancer site (if still present) is controlled, or resected, and the metastatic sites are ablated (surgically or with radiation), a prolonged disease-free interval, and perhaps even cure, may be achieved. Contemporary molecular diagnostics are edging closer to being able to determine where an individual metastatic deposit is within the continuum of malignancy. Preclinical models are on the outset of laying the groundwork for understanding the oligometastatic state. Meanwhile, in the clinic, patients are increasingly being designated as having oligometastatic disease and being treated owing to improved diagnostic imaging, novel treatment options with the potential to provide either direct or bridging therapy, and progressively broad definitions of oligometastasis.


INTRODUCTION
Hellman first proposed the theory of oligometastases in 1995 as a sequel to the spectrum theory of cancer metastasis. Hellman hypothesized that the process of cancer metastases occurred along a continuum, from locally confined cancer to widely metastatic disease. Although the phenomenon of limited and treatable cancer metastases had been noted historically, Hellman and Weichselbaum proposed the term oligometastases, suggesting that in some patients with a limited number of clinically detectable metastatic tumors, the extent of disease exists in a transitional state between localized and widespread systemic disease. In this model local control (LC) of oligometastases would have the potential to yield improved systemic control, going against the dogma that control of oligometastatic disease would not have a therapeutic benefit since it represents a clinical manifestation of a few detectable lesions in the setting of widespread occult disease. Oligometastasis continues to be defined as a state of metastatic disease that is limited in total disease burden, usually by number of clinically evident or radiographic sites (either 1-3 or 1-5), and that is not rapidly spreading to more sites. The clinical implication of oligometastasis suggests that if the primary site (if still present) is controlled, or resected, and the metastatic sites are ablated (surgically or with radiation), there will be a prolonged disease-free interval, and perhaps even cure. As the understanding of the mechanisms underlying cancer metastasis have evolved, possible mechanisms for the oligometastatic state must be explained and examined within that context.

Biological basis for oligometastasis Theories of metastasis
Metastasis is the cause of most cancer-related deaths [1]. In 1889, Stephen Paget [2] theorized that circulating tumor cells would "seed" to an amenable "soil", suggesting that metastasis was not a matter of chance. Five years later, [3,4], in 1894, Halstead theorized that cancer was an orderly disease that progressed in a contiguous manner, by direct extension from the primary tumor through the lymphatics, to the lymph nodes, and then to distant sites. Halstead proposed that breast cancer metastasis was a progressive, anatomical process of contiguous seeding; his hypothesis supported the use of radical surgery and radiotherapy. Radical en bloc surgery, radical hysterectomy, and primary and regional irradiation for several tumor sites were all based on Halstead's theory of tumor spread [4]. James Ewing, in 1928, complemented the Paget and Halstead theories to propose that cancer cells grow at a particular site because they are directed by the direction of blood flow and lymphatics. [5] The 'systemic' theory of metastasis, first suggested by Keynes [6] and further developed by Fisher [7], held that clinically apparent cancer was a systemic disease, and that small tumors were an early manifestation of systemic disease [8]. In this theory, nodal involvement was not part of an orderly contiguous extension but rather a marker of distant metastases. According to this theory, local control would not impact survival.
In contrast to the 'Halsted' theory and the 'systemic' theory, the 'spectrum' theory of cancer metastases, first described for breast cancer metastases in 1994, held that disease stage at the time of initial disease presentation fell into a spectrum ranging from indolent disease to widely metastatic, with the degree of clonal evolution determining the ability of the tumor to metastasize [9]. The spectrum theory was refined just one year later to describe the limited metastasis of any solid tumor and the term 'oligometastasis' was coined.
The spectrum theory conceptualized the entire range of metastatic competence, analogous to a diapason, which is the entire range of an instrument. To that end, the social sciences concept of a diaspora has recently been utilized to inform biologic understanding and therapeutic paradigms of cancer metastasis [10]. A diaspora refers to the scattering or movement of a population from its original homeland. In the case of systemic metastases, the diaspora resembles an imperial colonization in which the populations spread widely and eventually conquer the new host lands (aggressive cancer clones to multiple organs). Oligometastases resemble trading post diasporas, representing a limited number of outposts with limited growth potential (less aggressive cancer clones to few organs). (Table 1, Figure 1). Systemic versus oligometastatic diasporas may be dependent on the types of mutations present in the cancer cells (quality of the diaspora migrants), the quality of the original tumor site (factors in the homeland that cause the population to migrate), and the quality of the new hostland (factors that allow immigrants to establish and flourish) ( Table 2).

Biology of systemic and oligometastases
It is now widely accepted that there are discrete steps in tumor metastasis. Initially there is a loss in cellular adhesion, followed by increased motility > invasiveness of the primary tumor > entry into and survival in the circulation > entry into new organs > eventual colonization of these organs [11,12]. Shortfalls at any stage of this metastatic progression could result in phenotypes of limited metastatic potential [13]. Gupta et al. described specific tumorigenic genes -initiation, progression, and virulence genes -that fulfilled specific roles in the metastatic cascade. Initiation genes afford a selective advantage to primary tumor cells to enter circulation. Progression genes fulfill rate-limiting steps for colonization. Virulence genes provide a selective advantage in cells to colonize a secondary site(s) [14]. As Weichselbaum and Hellman noted, this paradigm suggests "that there may be primary tumor cells with a limited capability in one or more of the necessary biological requirements for metastasis; thus proposing a possible biologic explanation of the origin of oligometastases" [15].
In 2000 and updated in 2011, Hanahan and Weinberg proposed the, widely accepted, 'hallmarks of cancer' [16]. The original cancer hallmarks consisted of six transformations in cellular physiology that allow cells to survive, proliferate and disseminate, and together support carcinogenesis. The update posited that underlying the original hallmarks were two 'enabling hallmarks': 'genome instability and mutation' and 'tumor-promoting inflammation'. Additionally, the update proposed two new 'emerging hallmarks': 'deregulation of cellular energetics' and 'avoidance of immune destruction' [17]. The exact sequence in which the transformations occur, and thus the appearance of the 'hallmarks' (selfsufficiency in growth signals ↔ insensitivity to antigrowth signals ↔ tissue evasion and metastasis ↔ limitless replicative potential ↔ sustained angiogenesis ↔ evading apoptosis) can vary throughout the course of progression. Despite the ordering, collectively, the hallmarks can terminate into a cancer. However, the ordering and the degree of specific hallmarks, may potentially allow for a subsequent oligometastatic state. For example, cancer cells that lack the hallmarks to actively metastasize still may be able to slough into the circulation and inefficiently establish metastases. Other cells may be less efficient at proliferation, establishing slow growing metastases. Warburg (1956) suggested that "We may have cells which indeed look like cancer cells but are still energetically insufficient… such cells which are clinically not cancer cells, have lately been found not only in the prostate, but also in the lungs, kidney, and stomach of elderly persons such cells have been referred to as "sleeping cancer cells" [18].

Preclinical models of oligometastasis
Traditional clinicopathologic factors are inadequate when attempting to define the potential underlying biology of oligometastases. Several investigations have demonstrated the marked genetic and epigenetic This could be (a) secondary to due to environmental conditions in the primary tumor forestalling evolutionary clonal pressure, (b) cancer cells that slough out of the primary tumor that do not have the properties necessary to survive the circulation and invade into target organ sites, and/or (c) the cancer cells land in inhospitable target organs. (right) Systemic disease. Widespread metastatic growth potential is unlimited. This could be (a) secondary to due to environmental conditions in the primary tumor creating many undifferentiated, aggressive clones, (b) cancer cells that actively migrate out of the primary tumor that have the properties necessary to survive the circulation and invade into target organ sites, and/or (c) the cancer cells land in hospitable target organs.  [10] heterogeneity present in metastatic cancer sites within the same patient [19][20][21][22]. These studies demonstrate that cancer cells at different sites within a patient can have varied malignant potential [23][24][25]. Preclinical models of tumors with varying degrees of metastatic potential, including low metastatic potential exist. Using a cell line derived from B16F1 melanoma, Fidler et al. [26] found that variant metastatic cells pre-exist in a heterogeneous primary tumor as opposed to originating through adaption during metastasis from an otherwise homogenous primary tumor. This finding was advanced in later work that showed the KHT sarcoma line demonstrated similar heterogeneity whether grown in vitro or in vivo, which suggested that clonal variation seen in vitro derived from heterogeneity present in the primary tumor [27]. As an extension of this work in the KHT sarcoma line, it was demonstrated that effective metastatic variants developed at a high rate with low frequency, as opposed to the more frequent and stable subpopulations of metastatic variants [28]. In comparing B16 cell lines, Cillo et al. showed that the more highly metastatic, and less genetically stable cell line, generated increased metastatic variants corresponding to increased chemotherapy resistance [29]. Numerous in vitro studies and analyses of animal models have indicated that cells isolated from metastases differ greatly-both genetically and phenotypically-from cells isolated from their parental primary tumors [30]. The preclinical models point toward variation in individual tumor cells' metastatic potential, which supports the concept of oligometastases [15].
Given that stochastic models have been used to predict biologic phenomena, a Bayesian model has been proposed to predict the chance of occult metastases in the presence of detectable oligometastases [31]. Using the size and number of metastases, the proposed model inferred, (1) that the probability of occult metastases may increase substantially with minor increases in metastatic potential and (2) that extended disease-free periods were predictive of a substantial decrease in additional occult disease. Although compelling, such models are in their infancy and as yet remain in pre-clinical testing where the host, tumor and experimental factors are controlled. [31] Clinical evidence of oligometastasis Evidence for the evolution of the oligometastatic phenotype comes from various clinical and pre-clinical sources [26,[32][33][34]. Recent studies of the molecular biology of renal cell cancer metastasis have implied biologic differences between less and more aggressive metastases, as well as between fewer and multiple metastases. In order to better determine which patients presenting with localized RCC harbor an aggressive tumor and may not benefit from surgery, Kosari et al. using gene expression profiling, found gene expression alterations associated with an aggressive tumor and metastatic potential in the primary tumor [35]. From a cohort of 20 resected pulmonary metastases taken from 18 patients, Wuttig et al demonstrated the predictive potential of identified gene signatures, when comparing

Widespread metastasis Δ Q O = decreasing [(E FnΔt ) < (M FnΔt ) (S FnΔt )] [TR] [(H1 Qn) ) + (H2 Qn) ) = high
Primary microenvironment quality: Δ Q O = the changing quality ( Δ Q) of the primary cancer site over time ( Δ t). The quality of the primary microenvironment is dependent on multiple factors, including pH, oxygenation, amount of nutrients, interaction with supporting host cells, and the quality of the immune response. Primary nonlethal, epithelial cancer cells in a highly vascularized environment with rich nutrients are presumed to be less likely to evolve to a more aggressive clone by undergoing an epithelial to mesenchymal transition (EMT) and leave the primary over time. As the quality of the environment decreases, it is likely that the generation of lethal, mesenchymal clones increases.
Fitness of migrant cancer cells: E FnΔt = the number (n) and fitness (F) of passively shed epithelial cancer cells (E) over time (Δt). This represents the likelihood that a cancer cell passively shed into the circulation will survive transport to a target organ. M FnΔt = the number (n) and fitness (F) of actively emigrant cancer cells (M) over time (Δt). S FnΔt = number (n) and fitness (F) of cancer stem cells (S) over time. It is likely that the fitness of a passively shed cell is less than a migrating mesenchymal cell or stem cell that actively exits the primary tumor through the lymphatics, nerves, or circulation. Fitness depends on many variables, including EMT state, stemness, ability to secrete MMPs, ability to avoid anoikis, etc.
Quality of metastatic sites: H Qn = the quality (Q) of the target organ or host-land sites (H1, H2…). Migrating cancer cells will land in multiple sites (n) within different target organs in order to immigrate. Success depends on the quality of the soil of each of these microenvironments as well as the host immune response. disease-free intervals (DFI) and number of metastases, both of which are predictive of prognosis in metastatic RCC (mRCC). There were 306 differentially expressed genes in comparing DFI ≥5 years and DFI ≤9 months, and 135 differentially expressed genes in comparing multiple metastases (≥16) and few metastases (≤8). [36] In colorectal cancer, there is growing evidence that liver-limited disease is a distinct biological cohort that may benefit from aggressive management. While only a minority of patients are technically resectable, approximately 40% of patients with resected liver limited disease are alive 5 years after diagnosis compared with less than 1% for those with disseminated disease. [37] There is genetic evidence that patients undergoing hepatic resection for metastatic cancer had a different disease than those who did not [37]. For example, it was noted that BRAF V600E mutant tumors, which were typically associated with aggressive biology, rarely came to liver resection [38,39]. In addition, novel chromosomal aberrations have been identified that are associated with intra-and extra-hepatic recurrence after liver resection [40].
MicroRNAs, small non-coding RNA known to regulate tumor proliferation and apoptosis, are frequently dysregulated in cancer and metastasis [41][42][43]. MicroRNA profiling has shown a possible method to distinguish patients with oligometastases from those with polymetastatic disease. Examples of pro-metastatic microRNAs include microRNA-10b (upregulated in primary breast tumors that had metastasized), microRNA-21 (correlated with advanced stage, incidence of metastases, and poor outcomes in breast and pancreatic tumors), and microRNA-373/520c (increased expression in breast metastases) [43]. MicroRNA-210, a known transcriptional target of the HIF-1α signaling pathway, was elevated in sera from patients with metastatic castrate-resistant prostate cancer, as compared to controls, and was correlated with treatment response as assessed by change in PSA [41]. Lussiter et al. found microRNA-200c was associated with polymetastatic progression in a oligometastatic cell line, derived from patients treated with high-dose radiotherapy, tested in a xenograft model [44]. The investigators then stratified patients with resected pulmonary oligometastases into subgroups, based on high-risk versus low-risk of further metastatic progression. Differential microRNA expression patterns were identified between these two groups (high rate of progression (n = 16 prioritized microRNAs) and low rate of progression (n = 32 prioritized microRNAs) and, in an independent dataset, the expression patterns were associated with risk of progression and decreased overall survival. [8] Most recently, Uppal et al. identified three microRNAs overexpressed in clinical metastasis samples from patients with limited metastatic disease. MicroRNA-127-5p, microRNA-544a, and microRNA-655-3p were shown to limit, but not fully inhibit, metastasis in a model of breast cancer lung colonization [45].

Controversies surrounding the treatment of oligometastatic disease
Hellman theorized that, 1) whereas some tumors were destined to remain localized, 2) other tumors, as they increased in size, acquired an increasingly greater metastatic phenotype, suggesting at an early stage these tumors seeded distant sites with clones that had not reached full metastatic potential, and finally, 3) that some tumors already had occult distant dissemination at the time they are diagnosed. He also proposed that metastatic potential was not only directed by the tumor phenotype, but that it was also influenced by the tumor's location, venous drainage, and host factors [4,46]. Based on Hellman's theory, later scientists further theorized that tumors in the oligometastatic disease state were tumors early in their evolution of metastatic progression; therefore they produced metastases that were limited in number and location. These data support the presumption of a temporal evolution with an intermediate stage of limited metastatic capacity, where oligometastatic tumors may not have acquired the broad array of genetic changes required to develop widespread metastases [14,47,48].
The clinical implication of the oligometastatic state is that locally ablative therapies, given with the intent of targeting sites of clinically evident metastatic disease could result in long-term survival or cure [15,49]. Treatment of oligometastatic disease may also result in decreased overall tumor burden, decreasing morbidity and increasing survival. These arguments are opposed by the concept that clinical metastases are evidence of systemic disease and locally directed treatment will not alter the natural history of the disease course within a patient. In this scenario, only systemic therapy may be beneficial. Indeed, oligometastatic treatment paradigms are controversial (due to the limited data available) [4,34,50,51]. Without randomized studies, it is impossible to know if treatment of oligometastatic disease helps the patient. In addition, oligometastatic disease may represent indolent disease that does not require potentially toxic treatments [4,34,50,51].
Patients are increasingly being diagnosed with oligometastatic disease due to the advent of sensitive imaging technologies as well as effective therapies that are allowing patients to live longer with the diagnosis of cancer [34,52]. In addition, the fact that novel treatment options with acceptable safety profiles, such as stereotactic radiation, cryoablation, and minimally invasive surgery, are available to treat limited metastases, has led to a renewed interest in treating oligometastatic disease. Treatment of oligometastatic disease not only has the potential to prevent further evolution of genetically unstable clones and metastatic spread, it may improve overall disease control and delay more toxic systemic treatment [13,34,53,54]. Finally, the definition of oligometastases had gradually evolved (Table 3), which further inflates the increasing population of patients diagnosed with oligometastasis. In the absence of data to guide decisions, treatment of oligometastatic disease may be seen as a quality-of-life oriented approach, choosing personalized treatments with a reasonable risk to benefit ratio and taking into account the patient's own attitude in guiding them toward more or less, intensive therapy [50].

Therapeutic options for oligometastases Radiation therapy
While much of the literature supporting the oligometastatic states is within the surgical literature, there is an increasing body of literature describing the use of stereotactic body radiotherapy (SBRT) and stereotactic radiosurgery (SRS), in addition to conventional fractionated radiotherapy techniques [51]. SBRT is a noninvasive method of delivering high doses of radiation to ablate a target lesion while sparing the neighboring normal tissue, thus reducing long-term effects of radiation on the non-malignant tissues. The radiation is delivered from many beams originating from multiple directions that converge on the target site. [55] Through improved targeting and management of tumor motion, SBRT may improve tumor control and reduce treatment-related toxicity, as compared to conventional fractionated RT. Improved radiation targeting allows for higher-dose, hypofractionated, more efficient treatment regimens that can be delivered within narrow margins sparring adjacent organs. [56] 'Hypofractionation' is the delivery of large doses of radiation over a shorter time period as compared Oligometastasis "…metastases (from tumors early in the chain of progression) limited in number and location because the facility for metastatic growth has not been fully developed and the site for growth is restricted…" [46] Oligometastatic disease Solitary or few detectable metastatic lesions that are usually confined to a single organ [50] Oligometastases Due to limited metastatic competence and does not occur following otherwise successful systemic treatment. New metastases in this situation, albeit even limited, is likely to have more extensive malignant capabilities that were somehow spared from eradication by therapeutic means, or from the development of resistant clones [15] Induced oligometastases Occurs when widespread micrometastatic disease is mostly eradicated by systemic chemotherapy but drug resistant clones are left behind, or tumor foci is located in a site not accessed by chemotherapy [4] Oligorecurrence Limited metastases in the presence of a controlled primary lesion [195] Sync-oligometastases ≤5 metastatic or recurrent lesions in the presence of active primary lesions [196] Synchronous oligometastasis Oligometastatic disease is detected at the time of diagnosis of the primary tumor, therefore there is an active primary tumor [196] Metachronous oligometastasis Development of oligometastatic disease after treatment of the primary tumor; interval for classification of metachronous versus synchronous is not standardized; between Controlled primary lesion except for concomitant primary and distant recurrence [196] Oligoprogression Progression of a limited number of metastatic deposits, while remaining metastases are controlled with systemic therapy [197] Oligometastasis (specific to prostate cancer) Rising PSA following primary therapy, with oligometastasis on imaging, in whom local treatment (surgical metastasectomy (usually LN dissection), or SBRT for bony mets or LN recurrence) is required to defer initiation of ADT [54] Oligometastasis (specific to prostate cancer) Castrate resistant prostate cancer with a rising PSA and oligometastasis on imaging, in whom local treatment (surgical metastasectomy (usually LN dissection), or SBRT for bony mets or LN recurrence) may allow deferral of ADT [54] Abbreviations: LN=lymph node; SBRT=stereotactic body radiation therapy; mets=metastases; ADT=androgen deprivation therapy; PSA=prostate-specific antigen to conventional radiation fraction sizes. Therapy can generally be completed in 1-5 sessions, as compared to conventional radiation therapy that is delivered in smaller doses 5 days/week over ≥ 6 weeks. [55] SBRT can be used to manage oligometastatic disease presentations that would be associated with added morbidity if managed by surgery, such as deep-seated or osseous lesions. Patients who are poor surgical candidates may often be treated with SBRT, given that it is noninvasive and has a modest morbidity profile. Whereas surgery, when used to manage oligometastatic disease, tends to be seen as the gold standard as it allows for pathologic evaluation and assessment of the surgical margins. In addition, lesions greater than 7-8 cm and those difficult to target with SBRT, are better left for surgical management. [55] Although traditionally radiation therapy was thought to be immunosuppressive, there is increasing pre-clinical and clinical evidence that high dose, hypofractionated radiation -SBRT-may reverse antitumor immunity via CD8+ T-cells and cellular stress signals. [55] Although uncommon, the abscopal effect (regression in tumors distant to the targeted field of radiation) is an example of recovery of anti-tumor immunity following RT [57]. While the exact mechanism is unknown, it has been proposed that the radiation effect may result in the release of anti-tumor proliferative antigens and cytokines. Matzinger [58] theorized a 'danger model', which suggested that the immune system is stimulated by injured tissues rather than 'non-self', therefore damaged tissues in irradiated sites may stimulate the immune system. Lee et al. [59] demonstrated in a murine model that hypofractionated, high dose RT may possibly reverse T-cell unresponsiveness to primary and metastatic tumors. More specifically, considering that local RT to a tumor may modify its microenvironment by producing inflammatory cytokines that may increase its immune responsiveness, Lugade et al. demonstrated that IFN-γ (a cytokine), following RT, promotes T cell function [60]. It is thought that a tumor has ongoing crosstalk with the immune system-from one end of a spectrum where the immune system eliminates the tumor, to tumor-immune system equilibrium (a subclinical tumor), until selective pressure from the immune system stimulates the evolution of tumor cells resistant to the immune system (clinically detectable). A way to think of how RT may promote immune recovery is that the irradiated tumor is converted into a 'vaccine' that promotes tumor specific T cells to bestow immune memory against non-treated tumors. [57]. Thus, an emerging advantage of SBRT is the possibility that it may be exploited to benefit the immune system.

Surgery
Historically, most of the surgical data in regards to oligometastatic disease is centered on hepatic resection. [37]. Perioperative mortality related to hepatic resection has decreased from 20% (before 1980) to 1%. Recent improvement in overall survival following hepatectomy, are likely from improvements in patient selection (shift in definition of resectability to new criteria based on whether a macroscopic and microscopic complete resection of the liver lesion as well as complete resection of any extrahepatic disease), surgical technique, and more effective adjuvant therapy. The use of portal vein embolization and of neoadjuvant chemotherapy have also expanded the population of patients who are eligible for resection. [61] In a recent series of patients having undergone hepatic resection, despite progression on chemotherapy, the 5-yr survival was 53% [51].
Secondary resection remains a worthwhile therapeutic goal; patients brought to resection by systemic therapy enjoy comparable long-term survival to patients who had resectable disease at the time of presentation, and far superior to those receiving palliative systemic chemotherapy [62]. 10-20% of patients that develop colorectal liver metastases present with, or are converted by systemic treatment to, an oligometastatic state defined as metastatic lesions that are limited in number and involving only a single organ. This type of disease is potentially amenable to local therapeutic modalities, of which hepatic resection is the most effective. [51] Cryotherapy Cryotherapy has been utilized in multiple settings for the ablation of metastatic disease [63,64]. In the largest series published to date, Littrup and colleagues treated a total of 251 oligometastatic tumors from multiple primary cancers in 126 patients [65]. Sites of treatment included retroperitoneal, superficial, intraperitoneal, bone, and head and neck; average diameter of tumors was 3.4 cm. At 11 months average follow-up (range, 0-82 months), a 10% total recurrence rate (26 of 251) was noted; three occurred within the ablation zone, for a local progression rate of 1.2%. The average time to recurrence was 4.9 months, and, at 21 months, the initial ablation zone had reduced in volume by 93% [65].

Treatment of oligometastases in different cancers
Anecdotal reports of tumors with limited metastases, having undergone treatment to the metastases and had long-term response date back to the 1930's [66]. Several studies, with a variety of endpoints and meeting different levels of evidence criteria, have been performed in a variety of cancers to treat oligometastatic disease. For each clinical study reviewed, we considered the definition of oligometastasis that was used, the therapy given, and sample size. In order to provide an ordinal categorization to assess the strength of the study designs and of the study endpoints, we used the National Cancer Institute Levels of Evidence for Adult and Pediatric Cancer Treatment Studies ranking system to assess each study [67]. In this system, the strength of the studies in descending order is: randomized controlled trials (double-blinded): Level 1i > randomized controlled trials (non-blinded): Level 1ii > non-randomized controlled clinical trials: Level 2 > case series (population-based, consecutive series): Level 3i > case series (consecutive cases (not population-based)): Level 3ii > case series (nonconsecutive cases): Level 3iii. The system ranks the study endpoints in descending order of strength as follows: total mortality: Level A > cause-specific mortality: Level B > quality of life: Level C > event-free survival: Level Di > disease-free survival: Level Dii > progression-free survival: Level Diii > tumor response rate: Level Div.

Breast cancer
Metastatic breast cancer (MBC) at diagnosis constitutes 3.5-7% of all new breast cancers [68], while oligometastases comprises 1-3% of the total MCB population [50]. Traditionally MBC patients were managed primarily with systemic modalities, with limited local treatments given only for palliative intent.
We searched PubMed using the terms 'breast oligometastases', 'breast oligometastasis', and 'oligometastatic breast cancer'; results were as follows, n = 20, n = 5, and n = 51, respectively. We omitted reviews (including special features), studies including mixed primary tumors, case reports, studies not in the English language, studies not focused on oligometastatic breast cancer, and a survey. There were seven clinical studies remaining; oligometastases held six definitions and each study used a different therapy ( Table 4).

Lung cancer
Non-small cell lung cancer (NSCLC) is the leading cause of death worldwide with > 50% of patients having metastatic disease at diagnosis [76,77]. The primary treatment for most patients with metastatic NSCLC is palliative chemotherapy (CT), which results in median survivals of 8-11 months [78]. However, multiple studies have demonstrated a subset of long-term survivors with OMLC [76,77]. Additionally, curative outcomes were documented in patients with treated adrenal metastases [79].
We searched PubMed using the terms 'lung oligometastasis', 'lung oligometastases', and 'oligo metastatic lung cancer'; results were as follows, n = 14, n = 90, n = 123, respectively. We omitted reviews, mixed primary tumors, studies not focused on oligometastatic lung cancer, case reports, editorials, studies not in the English language, surveys, commentaries, cost analyses, studies including less than 10 patients, and studies we were unable to retrieve. There were twenty clinical studies remaining (Table 5); oligometastasis held 17 different definitions and each study had a different treatment paradigm.

Melanoma
Approximately 30% of patients with melanoma will develop metastases. The 5-yr survival of stage IV melanoma is about 5%.
We searched PubMed using the terms 'melanoma oligometastasis', 'melanoma oligometastases' and 'oligometastatic melanoma'; results were as follows, n = 5, n = 5, and n = 15, respectively. We omitted reviews, studies including mixed primary tumors, and studies not focused on oligometastatic melanoma. An additional search through the bibliographies of the review papers allowed us to retrieve one additional clinical study for review, therefore we reviewed two clinical studies ( Table 6). The two studies differed in their therapies, which included resection [99], and SRS followed by +/− immunotherapy [100], and definitions of oligometastases.
The evidence for treatment of oligometastases in melanoma is weak based the two, retrospective

Colorectal cancer
Colorectal cancer is the fourth most common cancer diagnosis in the world (around 1.2 million diagnoses each year), and accounts for the second highest number of deaths [37]. Nearly one-fourth of patients with newly diagnosed colorectal cancer (CRC) will present with synchronous liver metastases [101,102]. Hepatic resection is considered a standard treatment option for metastatic CRC and can result in 10-year survival rates of 20-26%, and potential cure [103][104][105][106][107][108][109]. We searched PubMed using the terms 'colo rectal cancer oligometastasis', 'colorectal cancer oligo metastases', and 'oligometastatic colorectal cancer'; results were as follows, n = 5, n = 24, and n = 28, respectively. We omitted reviews, studies including mixed primary tumors, studies not focused on oligometastatic colorectal cancer, case reports, perspectives, preclinical reports, and earlier reports if later reports for the same study were available. There were nine studies remaining (Table 7); oligometastasis held 9 different definitions and each study had a different treatment paradigm.

Sarcoma
Soft tissue sarcomas are relatively rare with about 10, 000 cases diagnosed each year in the United States. If diagnosed early, the prognosis is excellent, however, in some subtypes (of which there are greater than 50), up to 50% of patients will develop distant metastases and will have a 5-year survival of below 10%. The median overall survival for patients with advanced and metastatic soft tissue sarcoma is about 12 months [118].
We searched PubMed using the terms 'sarcoma oligo metastasis', 'sarcoma oligometastases', and 'oligometastatic sarcoma'; results were as follows, n = 3, n = 3, and n = 13, respectively. We omitted reviews, studies including mixed primary tumors, studies not focused on oligometastatic sarcoma, and case reports. There were two clinical studies remaining (Table 8); both studies used different definitions of oligometastasis and a different treatment paradigm.
The evidence for oligometastases in sarcoma was weak based on the two retrospective studies available for review. Both studies were nonconsecutive case series (Level 3iii). Falk et al. included a moderate sample size (n = 281) and MOS was reported (45.3%) (Level A) [119]. The second study was small (n = 16) and the endpoint was time to progression (17-months) (Level Diii) [120].

Renal cell carcinoma
Approximately 25-30% of patients diagnosed with renal cell carcinoma have metastatic disease at initial presentation. Approximately 1/3 with clinically localized primary tumor at diagnosis will eventually develop metastatic disease [121]. Historically patients with metastatic RCC (mRCC) have a poor prognosis, with 5-yr survival rates of ≤ 10%, however prolonged survival has been noted in those with solitary or oligometastatic disease amenable to resection. RCC is often considered resistant to cytotoxic chemoconventional RT and cytokinebased immunotherapy. The lung is the most common site of metastases in mRCC; the second most common is bone. SABR is being suggested as a potential new therapeutic option.
The standard of care in mRCC is systemic therapy; however, in patients with solitary or limited metastases, aggressive local therapies may potentially prolong survival. The literature suggests a survival benefit with surgical metastasectomy, with a reported 5-year survival as high as 45% in those who achieve complete resection. [55] We searched PubMed using the terms 'renal cell carcinoma oligometastasis', 'renal cell carcinoma oligometastases', and 'oligometastatic renal cell carcinoma'; results were as follows, n = 1, n = 9, and n = 3, respectively. We omitted reviews, studies including mixed primary tumors, and studies not focused on oligometastatic renal cell carcinoma. There were three clinical studies remaining (Table 9); each study used a different definition of oligometastases and a different treatment paradigm.
Additional searching through the bibliographies of other selected papers allowed us to retrieve two clinical studies that investigated the effects of removal of the primary tumor in patients with metastatic renal cell carcinoma (mRCC) ( Table 9) [125,126]. The evidence for removal of the primary tumor in mRCC was strong; both studies were nonblinded randomized controlled clinical trials (Level 1ii), had moderate sample sizes, and reported median OS (17 versus 7 months, [125] and 11.1 versus 8.1 months [126]) (Level A). The therapies given in both studies were similar: nephrectomy + interferon versus interferon only [125] and nephrectomy was followed by interferon versus interferon only [126]. The evidence for oligometastases in RCC was not assessed in these studies. However, they were important studies owing to strong design, with comparable treatments and endpoints, and most importantly they demonstrated that removal of the primary tumor in metastatic disease increased survival.

Prostate cancer
There is a body of evidence, contrary to historical clinical practice, that treating men with metastatic prostate cancer to the lymph nodes at the time of diagnosis with surgery or radiation therapy results in increased longterm survival (Table 10), [127][128][129][130][131] and Table 11 [101,132]. Historically, the ideal patient to be cured by radical prostatectomy (RP) was one with organ-confined cancer. At that time, the morbidity of RP was substantial; therefore the surgery was generally only offered to those where the probability of cure was high. More recent data, however, show that RP may provide a survival benefit-albeit not a cure -to men with metastatic prostate cancer. Engel et al. [101] conducted a retrospective study (n = 938) and found that in men with prostate cancer and positive lymph nodes, treated with +/− RP, that the 5-and 10-yr survival rates and the prostate cancer specific survival rates, were better in men who had undergone RP. Cadeddu et al [132] conducted a retrospective study (n = 38) and found that in men with prostate cancer and positive lymph nodes, treated with LND +/− RP, that the 5-and 10-yr prostate cancer specific survival was better in men who had undergone LND + RP.  conducted a systemic review of the literature of metastasis-directed therapy of regional and distant recurrences after curative treatment for PCa (prostate cancer). They found that  salvage LND and RT appear to be safe in treatments for OM PCa recurrence. [133] Culp et al. [134] studied the impact of survival of definitive treatment of the prostate in men diagnosed with metastatic PCa. Using the SEER database, he reviewed 8185 men treated with NSR (no surgery no radiation), brachytherapy, or RT. The 5-year OS and disease specific survival (DSS) was significantly higher in men with metastases having undergone RT. Zapatero et al., reported on a study of men with intermediate and high risk localized prostate cancer (n = 362), treated with RT + long term ADT versus RT + short term ADT, and found that long term ADT was superior in 5-year biochemical disease free survival, metastasis free survival, and OS [135].
The standard of care treatment for metastatic prostate cancer is androgen deprivation therapy (ADT) [136]. Men with newly diagnosed metastatic prostate cancer, entered into the control arm of the Systemic Therapy Multi-Arm Randomized Controlled Trial (STAMPEDE) and treated with ≥ 2 years of ADT alone, had a 2-year OS of 72% [137]. The response rate for primary hormonal therapy for men with metastatic prostate cancer exceeds 80% and the median duration of response is approximately 18-24 months [138]. In men with prostate cancer with limited metastases, radical prostatectomy may be advantageous in that the primary tumor and its ability to continuously metastasize, to secrete tumor promoting growth factors and immunosuppressive cytokines, and to generate bulk-related morbidity, is removed. Heidenreich et al. recently evaluated survival outcomes following radical prostatectomy (RP) in men with low volume metastatic prostate cancer. RP led to improved progression-free survival, time to castrate resistance and overall survival, as compared to a cohort treated with androgen deprivation therapy alone. [139] Moreover, Abdollah et al. found that in men with pN1 prostate cancer, treated with RP and extended lymph node dissection, adding adjuvant radiotherapy improved cancer-specific mortality [140].
We searched PubMed using the terms 'prostate cancer oligometastasis', 'prostate cancer oligometastases', and 'oligometastatic prostate cancer'; results were as follows, n = 3, n = 19, and n = 22, respectively. We omitted reviews, studies including mixed primary tumors, studies not focused on oligometastatic prostate cancer, case reports, editorials/commentaries, studies that we were not able retrieve, and studies with no results reported. There were ten clinical studies remaining (Table 11); oligometastasis was defined differently in each study and there were 7 different treatment paradigms.
Currently there are two studies, not yet completed, for which we are anxiously awaiting the results. Decaestecker et al. are conducting the first randomized, phase 2 trial on men with an oligometastatic recurrence to assess the impact of metastases-directed therapy -versus active surveillance -on the start of palliative ADT [150]. Attard et al., as part of the ongoing STAMPEDE trial, describes the addition of a new treatment arm (enzalutamide, abiraterone and prednisone with ADT) for men with newly diagnosed M1 disease [151].

Current trials of oligometastatic disease
We reviewed clinicaltrials.gov to gauge the ongoing studies for oligometastatic disease. Multiple trials are underway to determine if the treatment of oligometastatic disease is beneficial in cancer as reported on clinicaltrials. gov [152] (see Table 12).

CONCLUSIONS
Hellman proposed that tumors early in the chain of progression may give rise to metastases limited in number and location due to an undeveloped metastatic competence of the cancer "seeds" as well as the restrictive nature of the host tissue at some metastatic sites. The rationales of discrete steps of metastases [11,12] and the hallmarks of cancer [16,17] have both provided a biologic framework to support the oligometastatic state. Molecular diagnostics have not yet arrived to the point to determine where an individual metastatic deposit is within the continuum of malignancy [32- 44,55], preclinical models are beginning to lay the groundwork of the oligometastatic state [19][20][21][22][23][24][25][26][27][28][29][30][31].
While Hellman posited that oligometastasis may be underrecognized due to limitations in imaging diagnostics, imaging has advanced to such a degree that we are now in a yin-yang position such that, 1) imaging often likely demonstrates very early findings not likely to manifest into disease and, 2) imaging demonstrates very early disease that is destined to become widespread, which cannot be distinguished from oligometastases.
Hellman described potential cure as an attractive upshot of treating an oligometastatic state, but prolonged survival has been added as a potential benefit. There is an increasing shift toward individualized, multidisciplinary management of OM disease [34] because it is difficult to conduct randomized controlled trials in oligometastatic patients due to the variety of presentations. While other trial designs may be more realistic, such as N of 1 studies, tissue-based approaches, propensity matched analyses and adaptive design methods [55], Hellman originally endorsed that the ability to identify the oligometastatic state would profit from careful review of past experience, i.e. re-analysis of clinical data found in archival materials