Oligometastatic cancer: an entity, a useful concept, or a therapeutic opportunity?

The word ‘oligometastases’ was coined in 1995¹ by Hellman and Weichselbaum who proposed ‘the existence of a clinical significant state of oligometastases’. It seemed to them to be a useable term by which to identify a subset of patients who have disseminated cancer, but in whom metastases were few in number. The diagnostic term oligometastatic disease is now being used with increasing frequency in the argot of multidisciplinary team meetings. It provided the title of a session at the 14th World Conference on Lung Cancer (IASLC July 2011).² How has this come about? In this analysis I consider the usage of the term from four view points.

1. First the historical perspective known as ‘framing’ disease.

2. Then, because oligo means few, is there a numerical basis for this categorization?

3. Next usage: where and how has the term been used in medical writing?

4. And finally the oligometastatic state as a therapeutic opportunity.

From dropsy to ESRD; from consumption to oligometastatic disease

Social historians of medicine have used the term ‘framing disease’ to describe how medical diagnosis is arrived at and how it changes through history.³ While diseases exist in nature, diagnoses are man-made. The categorization which is entailed in giving a diagnosis is a medical and societal convention. In this essay I argue that for oligometastases, it is the opportunity to administer a surgical or ablative treatment that frames the disease.

Prior to the 19th century diseases were framed by clinical features hence consumption and dropsy. The study of ‘morbid anatomy’ then refined disease descriptions. Oedema due to heart disease, and that due to kidney disease, were differentiated by Bright and others.⁴ What had been a unifying clinical diagnosis of dropsy as seen by the physician William Withering, and treated with foxglove, had to be reframed to differentiate between cardiac and renal causes of fluid retention following Bright's observations in the postmortem room. Later, during the 20th century, diagnosis increasingly relied on the biochemical laboratory. Measurements of creatinine, blood sugar, enzymes, cholesterol, troponin and so on are used to define – or frame – disease. The diagnosis of end-stage renal disease (ESRD) which since 1972 entitles an American patient to centrally funded renal replacement therapy,^4,5 is not clinical, aetiological, or pathological, and although underpinned by laboratory measurement, is framed by the need for renal replacement therapy to sustain life.

The other great unify diagnosis was consumption – but 19th-century consumption and present day tuberculosis are by no means synonyms. When the body thinned and shrank it was considered to consumed by disease. If this state were to be accompanied by symptoms related to the lung (cough, sputum, haemoptysis) the diagnosis of pulmonary consumption was made. In the 18th century there was no distinction that could possibly be made between the consumption due to tuberculosis or that due to cancer in the lungs. Koch was yet to identify the tubercle bacillus and recognition of cancer as disease that arose locally and spread elsewhere, came from the work of Campbell de Morgan in the 1870s.⁶ Stephen Paget, author of an early text on surgery of the chest,⁷ studied the distribution of metastases in cancer. It was he who first used the seed and soil analogy.⁸

Weight loss is not required now to make a diagnosis of disseminated cancer, maybe well before weight loss becomes a feature, due to remarkable advances in imaging. Modern imaging allows us also to identify metastases, previously undetectable in the living patient⁹ but perhaps more important in framing oligometastatic disease is that imaging allows the absence of metastases to be inferred. It is the belief that we can know when metastases are not there that enables us to put a number on them, and make the diagnosis of an oligometastatic state.

Counting and other numerical considerations

It is important to use numbers to convey the necessary information on which clinical management can be rationally offered and its implications understood by patients.¹⁰

It is the ‘fewness’ that ‘frames’ oligometastatic disease so we should consider how well the term stands up from a numerical perspective and, I will argue, not well. To understand the point, consider the age distribution of patients with the very familiar condition of spontaneous pneumothorax (Figure 1). Towards the left is the population of young patients whom we recognize as having primary spontaneous pneumothorax, and toward the right, the older patients in whom the pneumothorax is more likely to be secondary to degenerative lung disease. A numerate and scientifically minded individual, knowing nothing of the clinical picture, the morbid anatomy, or pathophysiology of pneumothorax, might question, on observation of the data alone, whether patients with pneumothorax share one diagnosis. Contemporary clinical thinking allows two diagnostic frames, primary and secondary pneumothorax, which differ in their aetiology, clinical picture and clinical implications.

Figure 1.

Age and sex distribution of all 127 patients operated on in Guy's Hospital Thoracic Unit in 2004 with the diagnosis of pneumothorax

In the absence of a bimodal distribution isn't oligometastatic cancer simply the tail end of a continuous distribution? There are no data that refute this suggestion. A study unusual in that it provides a denominator as well as numerators, shows solitary metastasis to be a very rare finding.⁹ Among 1509 patients with lung cancer there were only 10 with a single metastasis, fewer than 1 in 150 patients. Six were in the brain and two each in bones and adrenal glands. A solitary metastasis in the adrenal gland from lung cancer is a well-recognized clinical entity, but it was seen only once in 750 patients.

We know from many clinical reports that patients with a single known metastasis dominate the practice of pulmonary metastasectomy. Reports are very highly selected by therapeutic intent with the consequence that 60% of 3504 patients with colorectal cancer in a systematic review¹¹ and 46% of 5206 patients in the International Registry of Lung Metastases¹² had a solitary lung metastasis. How should their disease be termed? It should be noted in passing that these patients cannot strictly speaking be deemed to have an ‘oligometastasis’ because the word is internally contradictory: they have a monometastasis as has been observed by the reviewer, Spyros Retsas. The question remains, is oligo/mono/metastatic cancer a discrete entity? There is no hint among these data that there is a bimodal distribution in the number of metastases. An exploration of the mathematics¹³ concluded that ‘the rate at which a primary tumor sheds new metastases increases exponentially, in parallel with its exponential growth’ and that this ‘creates a log-log linear relationship between the volume distribution of metastases and number of metastases’. Why do some patients apparently pass through a so-called oligometastatic state while others run an explosive ‘polymetastatic’ course? As far as can be determined from the evidence they are at different points in a spectrum of biological aggressiveness. There is nothing found so far to indicate anything other than a continuous distribution.

The emergence and dissemination of the concept of an oligometastatic state

A search in PubMed:MEDLINE returned 46 titles containing <oligometast*> up to early 2011 (see Appendix – http://jrsm.rsmjournals.com/lookup/suppl/doi:10.1258/jrsm.2011.110279/-/DC1). This search captures the singular, plural, and the derived adjective ‘oligometastatic’. No titles were retrieved from before 1995 so, as far as can be determined, Hellman and Weichselbaum coined the term.¹ As far as PubMed titles are concerned the term was not used again until 2001¹⁴ when Downey, a thoracic surgeon, uses the word in the context of a primary lung cancer with metastases, which are few in number, but may, in any given patient, exist in more than one organ.

As judged from published titles, the term oligometastases has been hesitantly adopted as years passed; its usage over 15 years can be followed in Figure 2. Including Downey's report, 8/44 reports concern metastases from lung cancer. There are five confined to breast cancer. In 25/44 publications, the origin of the metastases is from a variety of organs. The publications predominantly describe outcomes with radiotherapy and in papers from 2008 to 2011 the reports tend to specifically concern the use and results of stereotactic body radiotherapy (SBRT).

Figure 2.

A count of publications from 1995 to 2010 in PubMed containing oligometast-asis/ases/atic in the title

It is generally true that the more metastases the worse the outlook – it would be surprising if it were otherwise, but authors attempting to define oligometastatic disease by number do not agree on what number defines the state. Five or fewer is the inclusion criterion in several reports. In prostate cancer people with <5 fared better than those who have more. In individual reports four or fewer were included for brain metastases and three or fewer for vertebrae. One might reasonably observe that it is self-evidently futile to put a number on it but if you cannot put a number on it how else can the diagnostic frame be defined?

There is separate consideration, given by some of the authors, to the number of organs involved. This makes complete sense in practical terms. Three metastases in one lobe of the lung provide a practical proposition for surgery in contrast to three metastases from a primary lung cancer: one in an adrenal gland, one in the brain and one in a bone. The first is a therapeutic opportunity, the second a considerable technical challenge, whatever interventions are being considered.

Oligometastatic disease as a therapeutic opportunity

The authors of the original description predicted that ‘The importance of the oligometastatic state will be dependent on the size of the group of patients for whom it offers curative prospects’. This brings us back to framing disease and the note that change in societal perception or the influence of the availability of a treatment modality are among the factors that can be used to define a group of patients as having a ‘diagnosis’ in common.^3,4,15,16 ESRD entered common use as a diagnosis following a USA Congress decision in to provide federal financial support for all Americans whose future survival would depend on dialysis.^4,5 For these people ESRD becomes the only diagnosis that matters. The underlying pathology of their kidney disease ceases to be important to them; their continued survival depends on being classified as being eligible for state funded renal replacement therapy.

Following the same principle, the availability of an effective intervention may be as rational a way to frame disease as any other. An operational definition of the oligometastatic state is when metastases are ‘limited in number and location and are amenable to regional treatment’.¹⁷ Hellman and Weichselbaum wrote ‘An attractive consequence of the presence of a clinically significant oligometastatic state is that some patients so affected should be amenable to a curative therapeutic strategy’.¹ But that pulmonary metastasectomy cures disseminated cancers with, by definition blood born spread, has never been substantiated¹⁸ although it is common practice. Pulmonary metastasectomy for oligometastatic colorectal cancer is now the subject of a randomized trial.¹⁹

Recalling the aphorism attributed to Mark Twain (to a man with a hammer everything looks like a nail) this definition of the oligometastatic state would be like defining a nail as a thing you can drive in. It was therefore interesting to encounter in the course of this review a title ‘Using a bigger hammer’ in which the author concludes ‘although SBRT seems to have given us a bigger hammer, we still have much to learn about how and when to strike the nails’.²⁰ While arguably causing less harm to the patient than thoracotomy, SBRT or radiofrequency ablation are no more likely to be beneficial but it is in promoting these forms of treatment that the oligometastatic concept is most often invoked.

The word oligometastatic has an authoritative sound. The Greek composite is also convenient since any equivalent in English requires a relatively clumsy phrase. But the language of Hippocrates gives the ‘oligometastatic state’ a certain caché that it might not deserve. It appears, from reading the subsequent literature, that rather than having a sound biological foundation oligometastases is a diagnostic term adopted to frame a therapeutic opportunity. The bottom line could be that oligometastatic defines a state in which there are few enough metastases to ‘zap’!

DECLARATIONS

Competing interests

None declared

Funding

TT is Chief Investigator of the PulMiCC (Pulmonary Metastasectomy in Colorectal Cancer) Trial which receives funding from Cancer Research UK; TT receives no funding

Ethical approval

Not applicable

Guarantor

TT

Contributorship

TT is the sole contributor

Acknowledgements

The author is very grateful to Chris Lawrence for many discussions concerning ‘framing disease’ during and since his sabbatical at the Wellcome Institute for the History of Medicine in 1995 and to Wilfrid Treasure, author of Diagnosis and Risk Management in Primary Care: Words that Count, Numbers that Speak, who made detailed comments on an earlier draft. The author is also very grateful to Lesley Fallowfield, Martin Utley and the reviewer, Spyros Retsas, for helpful and constructive comments