Introduction

Solidarity is a fundamental social value in many European countries, though the precise practical and theoretical meaning of this term is open to dispute.Footnote 1 Our concern is with the application of the concept of solidarity to health care. Hence, our first task will be to critically assess various meanings attached to solidarity. Solidarity is often used in a descriptive mode. However, a dominant feature of health care today, whether in the US or the EU, is that it is too costly. No society, whether for political or economic reasons, can afford to provide to all its members anything and everything medicine might offer, no matter the cost, no matter how meagre the benefit (Wilking et al. 2017; Aggarwal and Sullivan 2014).Footnote 2 This implies the need for health care rationing and priority-setting. The emergence of precision medicine, primarily for the benefit of metastatic cancer patients, significantly exacerbates this problem of health care cost control (Vokinger et al. 2020).Footnote 3 In addition, it threatens to fragment solidarity in practice, not simply because of associated costs, but because of the way in which precision medicine fragments the population of metastatic cancer patients and gives pre-eminence to the treatment of cancer, as opposed to many other life-threatening medical conditions How can health care solidarity be sustained in these circumstances?

Answering this last question requires a normative understanding of solidarity. However, the prevailing understandings of solidarity are too abstract and too vague to be adequately responsive to this very complex challenge. Consequently, I argue that a normatively useful understanding of solidarity must include a conception of health care justice. However, multiple conceptions of justice are part of our ethical armamentarium. They too can contribute to the fragmentation of solidarity. Consequently, I ultimately argue that carefully constructed, fair, inclusive processes of rational democratic deliberation are necessary both to articulate just limits on health care spending (fair and affordable precision medicine) and to specify the shared commitments that will constitute solidarity in practice so far as access to the fruits of precision medicine is concerned.

Health care solidarity: what might it mean?

Appeals to solidarity have been used in a number of social and political contexts over the past two centuries to build, for example, coalitions of workers to fight for their rights and better wages. The message to workers was, “We are all in this together; we are all oppressed by the capitalistic owners of these factories.” This is a descriptive and practical use of the concept of solidarity intended to build political strength through the creation of a political identity. However, our focus will be on the concept of solidarity in a health care context. This is much more of a challenge because health care itself, and what we call health care needs, are extremely complex and heterogeneous. Marx could appeal to “workers of the world” to unite for their common good. It is much harder to concoct a convincing appeal to “patients of the world” in the hopes of uniting for a common good.Footnote 4 The good of cardiac patients is seen in practical terms as being very different from the good of cancer patients or renal patients or patients with a variety of disabilities, and so on. If health care resources were unlimited, these differences would make no practical difference. But health care resources are limited by the willingness of taxpayers to raise taxes or to pay higher health insurance premiums “for the benefit of others” (as perceived by payers in very good health).

In a health care context, the concept of solidarity will capture a number of related notions. Saltman (2015), citing Bayertz (1999), includes in the notion of solidarity “reciprocity (asymmetrical for the needy); social cohesion; altruism and fellowship; citizenship duties; universal brotherhood; political and/or social justice.” All these notions have a virtuous quality about them. They are readily endorsed by any reasonable person living in a community where all are vulnerable to the various misfortunes of life. If my house is struck by lightening and burns to the ground, I can count on others in my community to help me rebuild. And I would do the same for them. This is a simple example of reciprocity. However, things are more complicated in the world of health care, in part because of the complexity of health care itself, in part because of the insurance mechanism, whether public or private.

What does a commitment to reciprocity mean when millions in the EU are afflicted with chronic degenerative conditions (sometimes lifelong) that effectively undermine their ability to reciprocate? Prior to 1990 (to pick a somewhat arbitrary date), health care costs were largely socially affordable in the EU. It did not matter that some individuals had minimal ability to reciprocate by hel** to meet the health care needs (and related costs) of others. What mattered was that everyone was vulnerable to serious health needs, sometimes life-threatening, but treatable at significant expense. All were in this together. Beyond 1990 many more expensive, emerging medical technologies began to proliferate. They were a potential threat to reciprocity (and solidarity) because it was not as if these technologies were emerging more or less uniformly across a broad range of diseases and other health infirmities. Some would benefit; others would not. Further, it was not as if these technologies were emerging randomly; these technologies came to be as a direct result of deliberate public and private investment in specific health needs and research (not with a commitment to reciprocity and solidarity). Still, the effects over the past thirty years might have been a relatively minor erosion in overall commitment to reciprocity and solidarity with regard to health care if it were not for the emergence of these very expensive, marginally beneficial targeted therapies at the heart of precision medicine.

Rapid increases in health care costs have generated demands for controlling those costs. It is easy to imagine everyone endorsing greater efficiency and eliminating wasteful care. That can be seen as reinforcing solidarity. We are all better off getting health care more efficiently. More problematic has been the emerging rhetoric around personal responsibility for health, especially when failure in that regard results in very high health care costs for the rest of society. The argument is that individuals who fail to take responsibility for their health (smoking, poor diets, lack of exercise, medical non-compliance) are exhibiting indifference to the reciprocal aspects of solidarity (Davies and Savulescu 2019).Footnote 5 They are consuming health care resources which they would not have needed if they had been more attentive to protecting their own health. This, it is argued, is irresponsible and incompatible with a genuine commitment to solidarity.Footnote 6 I will not get into an argument about the ethical legitimacy of these judgments. Instead, my point is that individuals who are judged to have been irresponsible with their health are pushed out of the health care community of solidarity, even if only metaphorically, perhaps through some form of subtle political shaming as opposed to any outright denial of needed health care.Footnote 7 Still, this does represent some erosion of an all-embracing conception of health care solidarity that includes some understanding of reciprocity.

Going back to Saltman (2015) and Bayertz (1999) and their attributions to understandings of solidarity, they speak of social cohesion. However, my comments above would suggest a loosening of that social cohesion in health care as a result of efforts to control health care costs, selective investments in emerging technologies that benefit only some patient groups, and attributions of costly personal irresponsibility regarding health. Saltman and Bayertz also mention altruism as integral to some understandings of solidarity, though this is open to dispute. Altruism is selective regarding its commitments and largely voluntary. This might not be especially problematic outside the field of health care, but it would loosen the bonds of solidarity within health care. Could a society claim to be committed to solidarity and choose to ignore the health care needs of HIV + patients, or patients who struggle with substance abuse, with the explanation that it simply chose not to be altruistic in that regard? That seems intuitively incoherent and arbitrary.

Saltman and Bayertz also link citizenship duties and universal brotherhood to some understandings of solidarity. However, those linkages fail to address the question of what the scope of those citizenship duties might be so far as assured access to needed health care is concerned. The same is true with regard to the notion of universal brotherhood. If extremely expensive life-saving or life-prolonging medical technologies consistently yielded cures or greatly improved length and quality of life, it would be clear that a society committed to solidarity owed that to all its citizens who had those needs, even if there were overall concerns about controlling health care costs. However, all of that is contrary to fact. There are persistent and growing concerns about controlling health care costs throughout the EU, and those costs are largely driven by very costly emerging life-prolonging medical technologies that often yield only marginal gains in life expectancy and quality of life, far out of proportion to those costs. Precision medicine is a prime example of this challenge.Footnote 8 Precision medicine not only reinforces this trend, but it exacerbates that trend, thereby undermining a normative commitment to solidarity.

The vast majority of metastatic cancer patients achieve only marginal gains in life expectancy at very great social expense. That would suggest that such patients should be denied access to these therapies at social expense, though (as we discuss below) some number of patients generally not identifiable before the fact do gain extra years of life from these drugs. These patients are referred to as “super responders.” However, if such patients could be identified reliably before the fact, and if they alone had socially funded access to these drugs, that would entail an obvious risk of fragmentation to solidarity. Alternatively, an unlimited commitment could be made to funding any and all somewhat effective therapies for these metastatic cancer patients. However, that too would risk the fragmentation of social solidarity, unless social funding were extended to all patients faced with chronic degenerative illness for anything medicine could offer that would forestall death. That would mean giving up to a large extent the goal of health care cost control, which would yield yet another threat to social solidarity. At this point, Saltman and Bayertz call attention to the role of social justice as integral to social solidarity.

Saltman (2015, at 1–2) writes, “To some, solidarity means that every individual regardless of income or social standing has the same services delivered by the same health care providers and with the same clinical outcome.” For others, Saltman observes that solidarity is about shared financial sacrifice that included some understanding of fairness regarding degree of sacrifice. However, what those understandings of solidarity do not specify is the precise content of health services that will be assured to all as part of that commitment to solidarity. These latter uses of the concept are intended to be taken normatively. Consequently, Saltman and Bayertz agree with other writers who will see some conception of health care justice as being intimately linked to solidarity. Habermas (1990), for example, sees justice and solidarity as two sides of the same coin, a notion endorsed by Meulen (2015) as well. The thought is that a commitment to solidarity by itself is too vague to have significant normative force. Solidarity implies a commitment to a common purpose. In the case of health care, the common purpose would be to prevent premature death and to restore health when individuals have suffered as a result of disease or accident. That can mean roughly equal access to effective health care for all. In other words, if a particular costly drug for multiple sclerosis is an approved treatment for that disease, then all patients with that disease will have access to that drug at social expense (assuming no medical contraindications). This undergirds the expectation of universal coverage. In addition, it can also include commitment to the idea that higher income groups will pay more health care costs than lower income groups (using a progressive tax mechanism), and lower health risk groups will help pay the costs of higher health risk groups. However, I will still argue that precision medicine represents a threat to this societal commitment to this last normative understanding of solidarity.Footnote 9

Precision medicine and the need to control costs

Precision medicine today (2021) includes more than one hundred and fifty targeted cancer therapies or immunotherapies (almost exclusively for advanced and metastatic cancer). The “target” of these therapies are certain genetic features of a cancer, mutations that are responsible for “driving” that cancer and causing it to spread and multiply. These targeted therapies generally have prices of €50,000 to €75,000 annually or for a course of treatment.Footnote 10 These prices might be justified if they saved many lives, but the sad fact is that for the vast majority of patients gains in life expectancy will be measurable in months, not years. Cure is not a reasonable expectation. One form of immunotherapy, CAR T-cell therapy (for various leukemias), has front-end costs of $475,000, and at least 30% of these patients will not survive a year. There are 1.3 million cancer deaths each year in the EU.Footnote 11 Here is our critical question: Must a commitment to solidarity mean that all these targeted cancer therapies and immunotherapies are included as part of a benefit package guaranteed to all in each EU country, no matter what the cost, no matter what degree of effectiveness? If not, how is solidarity preserved when some metastatic cancer patients are denied access to the only drugs that might yield some gain in life expectancy?

I argue that health care cost control is essential to preserving solidarity. We can see the reason for this if we simply look at practice in the United States. In 2020 total health care expenditures in the US were about $4.0 trillion, or 18% of our Gross Domestic Product [GDP], compared to 1960 when total expenditures were $26 billion [roughly $224 billion in 2020$], or 5.2% of GDP (Keehan et al. 2020). European countries today are generally spending 8–12% of GDP on health care. Roughly half of that $4.0 trillion in the US represents public spending, i.e., Medicare for the elderly, state Medicaid programs for the poor. The other 50% is largely paid through private, employer-based insurance or out of pocket.

Employers see their profits at risk because of rapidly escalating health care costs, largely related to costly advancing medical technologies, such as these targeted cancer therapies. Consequently, employers use a variety of techniques to control the costs of health insurance for which they have assumed responsibility.Footnote 12 Most of those techniques involve shifting more health care costs and health insurance costs to their employees, either by requiring employees to pay a larger portion of the insurance premium, or by reducing the scope of covered services. Employers also achieve some degree of cost control in relation to these targeted cancer therapies (and other very high cost medical interventions) through (1) very high co-pays required to access these therapies, or (2) these therapies are simply removed from a basic comprehensive benefit package and left to the purchase of supplemental private insurance paid by individuals, or (3) cancers related to poor health behaviors (smoking and lung cancer) may be excluded from coverage.Footnote 13

What should be clear is that all of these mechanisms used in the US would certainly be destructive of solidarity if adopted in a European context and if “solidarity” is understood to include something like “equal access to needed and effective health care without financial barriers.” What is common to all these mechanisms is that access to needed and effective health care is dependent upon the financial status of individuals. Individuals in the lower half of the economic spectrum who are faced with very high health care costs will simply forego that care and accept the consequences for their health of doing so, including premature death. This can be accurately described as “self-rationing,” which has the desirable effect from an employer’s perspective of freeing them from blame or responsibility for the “unfortunate” health outcomes that then befall their employees.

European readers might believe that patients who are covered by public programs in the United States would be spared these economically motivated risks to health that are common in the private sector. However, taxpayers (mostly healthy individuals) pressure state and federal legislators to control costs in the Medicare and Medicaid programs. Consequently, coverage of health services for the poor uninsured (variously defined by each state) will vary considerably from state to state, depending upon the willingness of legislators to raise taxes to fund the Medicaid program. The Medicare program used to be described as “national health insurance for the elderly”. However, over the past decade Medicare has become increasingly privatized through individual ability to purchase a variety of Medicare Advantage policies with various mixes of coverage and costs to individuals at the time of medical need (Miller 2020). This was a deliberate choice by Congress with the intent of reducing health care costs to Medicare and shifting more of the responsibility for absorbing those costs to Medicare recipients or the private insurers promoting Medicare Advantage policies. European countries could choose to make comparable choices in order to control the costs associated with these targeted cancer therapies (and other very high cost life-prolonging medical technologies) but such choices would surely represent the fragmentation of solidarity, as I argue below.

Solidarity: the challenge of limited resources

The basic ethical problem we address can be captured in these two questions: (1) What does it mean to be a just and caring society when that society has limited resources to meet unlimited health care needs? (2) What does it mean to be a society committed to solidarity as a fundamental social value when that society has limited resources to meet virtually unlimited health care needs? Resources are limited by what taxpayers are willing to pay to assure everyone with specific health needs to have those needs met effectively. Resources for health care are also limited by the legitimate demands for resources required by a broad range of other social needs. Health care needs are virtually unlimited because, as Daniel Callahan (1990) has argued, what we regard as morally compelling health needs are tied directly to emerging medical technologies. In other words, there was no “need” (in an ethically relevant sense) for these targeted cancer therapies until these targeted cancer therapies were invented. Finally, we must ask whether cancer is somehow morally special, such that it justifiably commands unlimited resources while all other health care needs are subject to cost constraints. This has recently been referred to as “onco-exceptionalism,” and is regarded as being ethically problematic (Salcher-Konrad and Naci 2020; Fu et al. 2021).

Precision medicine as a threat to solidarity

Why does precision medicine represent a threat to solidarity? The first reason would be the extraordinarily high cost of the interventions that are covered by the rubric of precision medicine. The second reason would be the very large number of patients with metastatic cancer who would be eligible for these interventions annually.Footnote 14 The third reason would be the very marginal gains in life expectancy that are the result for the vast majority of these patients.Footnote 15 These are the very broad reasons; more specific reasons are discussed below. Before turning to those reasons, some medical and scientific background regarding metastatic cancer and the distinctive features of precision medicine will be useful.

If a cancer is diagnosed at stage one or two, 80% of those patients will survive at least ten years. Ordinarily, we might think of those patients as being ‘cured’. However, most medical researchers are reluctant to use that term, preferring instead ‘long term remission’. This reflects the fact that so much is not understood regarding the origins, evolution, complexity, and recurrence of cancer. For many diseases, medicine can identify a specific initiating event in an individual, i.e., infection with a virus of some sort. In other cases, such as heart disease, monitoring cholesterol levels and blood pressure, allow careful disease management and tracking. Neither of these options generally apply to cancer.Footnote 16

Precision medicine and the wiliness of cancer

Recent research suggests that a cancer might get established with the dysregulation of one or more tumor suppressor genes. TP53 is the most common of those tumor suppressor genes. TP53 is responsible for repairing genes that might have become mutated, or, alternatively, initiating cell death in those cells with mutated genes to prevent the development of a cancer. TP53 itself is highly mutated in ~ 50% of all cancers and minimally functional in the rest of cancers. Certain mutations of TP53 result in its acquiring oncogenic properties, i.e., actually promoting the spread of a cancer instead of destroying cancer cells.Footnote 17 However, unlike viral infections whose initiation can usually be identified quickly, it might be fifteen years after TP53 mutation before there is detectable clinical evidence of a cancer. This is one of many insidious aspects of cancer that makes early identification, control and cure difficult to impossible. Another insidious element is the genetic heterogeneity of cancer, especially metastatic cancer. Cancer cells tend to be genetically unstable, which means that successive generations of cancer cells become more and more mutated. What this means in practice is that a particular tumor might have multiple drivers of that cancer within that tumor.Footnote 18 Also, tumors throughout the body will often have multiple drivers, some of which will be genetically different from the driver of the original tumor. The consequence of this genetic diversity is cancer drug ‘resistance’.

As noted above, targeted cancer drugs are aimed at stop** the functioning of specific drivers of tumors. For example, tyrosine kinase inhibitors [TKIs] block specific enzymes that send growth signals within cells. TKIs will be used to treat several types of cancer. Non-small cell lung cancer would be one good example. Erlotinib is a TKI that inhibits the epidermal growth factor receptor [EGFR] when mutated and driving the growth of a cancer. This drug works well in preventing disease progression for as long as ten to fourteen months. However, because these tumors are constantly acquiring new mutations (sometimes in response to the drug itself), resistance develops, typically in the form of a T790M mutation. Osimertinib, a next generation TKI, can then be given, though resistance to it will develop as well, primarily as a result of a C797S mutation in exon 20. Very recently (August, 2021), research was reported on the drug quercetin, another TKI, which will attack non-small-cell lung cancers with that C797S mutation (Huang et al. 2021). Why should this medical scenario be a threat to solidarity?

Solidarity and targeted cancer therapies

As noted already, roughly 1.3 million patients in the EU die of cancer each year. These targeted cancer therapies have costs in excess of €75,000 or more per patient per year. If any of these patients have a solidarity right to one of these drugs, then they would all seem to have that right (assuming no medical contraindications).Footnote 19 To give those 1.3 million metastatic cancer patients an extra year of life would cost €98 billion. At present, we have a targeted cancer drug or immunotherapy for the genetically specific cancer of 50% of all those patients, though the ultimate goal of medical research is 100%. The goal is actually much more ambitious than that, as our discussion above of TKIs makes clear. If one drug fails after a year to halt cancer progression because of resistance, then we want to provide another drug that is responsive to the new genetic driver of that cancer, thereby gaining another year, and a year after that, and another year after that. In other words, cancer may not be curable, but maybe it can become a managed chronic disease, as has happened with HIV. HIV + patients have gained as much as thirty extra years of life from the three- and four-drug combinations used to manage the virus as it mutates.

In the case of metastatic cancer, let us consider a more modest scenario of five-year survival with drugs costing €75,000 per year. Five cohorts of future metastatic cancer patients would add almost €500 billion annually to health care costs in the EU, which would not be the total cost of cancer care. This huge figure just represents the cost of these drugs, not the cost of genetically analyzing tumors or managing side effects, and so on. A cost increase of this size for one disease either squeezes out whatever is judged to be lower priority care for some other diseases, or it requires a very substantial increase in taxes, or it requires some form of rationing access to these drugs, i.e., trying to articulate and discriminate among high-value cancer care and low-value cancer care. For example, would individuals with early stage dementia or moderate dementia and metastatic cancer not be offered these targeted cancer drugs? Would offering these drugs in these circumstances be regarded as low-value (solidarity optional) care? Roughly 20–30% of patients with non-small cell lung cancer will have mutations in KRAS genes that are driving that cancer. In a review article, Parums (2014) notes that “KRAS mutations are associated with a decreased response to EGFR TKIs and with a worse prognosis” (at 515). It is not that the TKIs are completely ineffective. The median gain in overall survival might be four weeks. Is that gain insufficient to justify the cost of these TKIs? May these drugs then be denied to these patients with KRAS mutations? All these considerations suggest potential cracks in solidarity.

Why precisely would there be cracks in solidarity? Patients with metastatic cancer can accept the fact that at this time there is no targeted therapy for their cancer with this specific driver of their cancer. This is just bad luck and does not undermine solidarity. In the KRAS mutation example, the four weeks is a median gain in overall survival. Where is the tail on that curve? That tail might be six months out, or one year out. The number of patients represented by that tail will be very small. It is easy to imagine a metastatic cancer patient with that KRAS mutation wanting that small chance of being in the tail. Denying that patient that option will feel like a denial of solidarity. Saying to that patient that everyone who has that cancer with that driver mutation will be denied that TKI at social expense, thereby honoring some form of fairness, will not speak to the solidarity issue. From the perspective of that patient, everyone in that diagnostic category will have been removed from solidarity with other cancer patients. Further, our hypothetical patient might argue that patients with his cancer but a different driver mutation for which a drug promises a nine-month median gain in overall survival will have that drug paid for socially. However, patients at the very low end of that median figure might only gain two months of additional life, or even less. They are embraced in solidarity, which will seem unjust to our hypothetical patient. The obvious response from a societal representative is to say we simply do not know before the fact who will be at either end of that tail. However, as discussed below, much research is ongoing, seeking prognostic and predictive biomarkers. If future research uncovered biomarkers predicting poor response to a targeted therapy, i.e., less than two months gain in overall survival for specific patients (though predicted median gain in overall survival for the entire cohort would be nine months), and if that knowledge were used to deny these poor responding patients that targeted therapy, that would look like another crack in solidarity. Allowing such patients to purchase access to these drugs at their own expense, or even with a 50% social subsidy, would only make the cracks worse. Further, why would the “poor responder line” be drawn at two months overall survival gain rather than three months or four months? There are many more potential cracks.

Cracks in solidarity

Another new class of drugs is called checkpoint inhibitors. They are a form of immunotherapy. Ordinarily we would expect the immune system to attack cancer cells because they are ‘not us.’ That is, they should be recognized as something foreign in our bodies. When our bodies are subject to a serious viral infection, the immune system attacks in force. That attack would itself cause damage if it were not ‘checked’ once the alien microbes had been destroyed. Cancer cells use that checkpoint strategy to trick the immune system into believing nothing was awry. Checkpoint inhibitors, such as pembrolizumab and nivolumab, block the checkpoints [PD-1 or PD-L1] and allow the immune system to recognize and kill the cancer cells. Unfortunately, only a small percentage of patients [~ 20%] will derive significant benefit from these drugs (Grisham 2017), though half of these cancer patients would be medically eligible for these drugs. Cost is again relevant. Generally, these drugs will cost €80,000 [$156,000] per year or for a course of treatment. Melanoma and non-small-cell lung cancer are the cancers most often treated with these drugs, though at least 15 major cancers would be treated with these drugs (National Cancer Institute 2019).

Recent research suggests that patients who derive the most benefit, more than an extra year of life, are patients whose tumors have the most mutations, which seems to make the tumors more visible to the immune system and these drugs (Klempner et al. 2020). Here we have another challenge to solidarity. Is it acceptable, from a solidarity perspective, to deny patients with melanoma or lung cancer access to these drugs at social expense because their tumors do not have ‘enough’ mutations? In other words, these drugs would be unlikely to yield much in the way of benefit. This is what Callahan has characterized as the ‘ragged edge’ problem (Callahan 1990, chap. 2). Where should the line be drawn that identifies enough mutations in a tumor that these drugs are more likely to yield substantial benefit? It would be easy to draw that line if below it there were no chance of benefit (and only the risk of toxic side effects), while above the line would be substantial benefit. Instead, the reality is a continuum, which means wherever the line is drawn some individuals would be denied some degree of medical benefit that they may have wanted. Is that outcome something congruent with a certain understanding of solidarity? The very costly alternative is to provide these drugs to everyone with melanoma or lung cancer (or any other cancer for which a checkpoint inhibitor would be appropriate), though this would be largely wasteful in as much as 80% of those cases. This would be a very high price to pay, ethically and economically, to preserve a specific sense of solidarity (perhaps a very strict egalitarian sense of solidarity).

More recently, much research has focused on using combinations of checkpoint inhibitors, sometimes with one of the targeted cancer therapies. This would be as first-line treatment instead of the current chemotherapy regimen. The first thing to be noted is that this substantially increases cost, roughly €125,000 ($256,000) per year or per course of treatment. There are corresponding gains in medical benefit. For example, in the case of metastatic non-small cell lung cancer, a combination of nivolumab and ipilimumab will yield progression free survival for one year of 42%, if there are more than ten mutations per megabase, though median progression free survival will be 7.2 months. If total mutational burden is less than 10 mutations per megabase with this drug combination, then median progression free survival will be 3.2 months. The point to emphasize is that this latter figure is not one week. It might well be significant to patients in this latter category who have no other life-prolonging options. Also, we are talking about a median, which means half the patients will do better than 3.2 months, some very much better. For the sake of overall cost control, does a commitment to solidarity permit not treating patients with this very expensive drug combination if their mutational burden is less than 10 mutations per megabase? This looks like another potential fracture in solidarity.

The trial that generated the data above was the Checkmate 227 trial. Another statistic worth noting from that trial is the following: Among all the patients in the trial median survival was 17.1 months with the nivolumab plus ipilimumab combination and 14.9 months with chemotherapy (Hellman et al. 2019). The median net gain in survival was 2.2 months, which would come at a very high price for that limited gain (~ $400,000 vs. ~ $30,000). Should a commitment to solidarity require paying that price from limited social resources? Alternatively, should a 30% co-pay be required from patients who wanted the nivolumab-ipilimumab combination while the chemotherapeutic regimen would be wholly paid from social resources? This would mean the relatively wealthy could afford the checkpoint combination. This would address the need for cost control, though this could readily be seen as eroding solidarity (in the moderate egalitarian sense endorsed by Saltman and Bayertz). The problem is this: It is one thing for two patients to have metastatic melanoma, only one of whom has a genetic mutation in their cancer that can be attacked by some specific targeted therapy. The other patient might feel badly but they would not feel as if they had been treated unjustly. From a precision medicine perspective, it is as if these patients had two different diseases, only one of which had a relevant therapy. This does not corrode solidarity. In the scenario above, however, all those patients have the “same disease,” and we are treating the financially well off with a somewhat more effective therapy while relegating the financially less well off to a less effective therapy. This will have a corrosive effect on solidarity, especially if we are mindful of the fact that the financially less well off will be hel** to finance through their taxes that 70% of the cost of the combination therapy.

Is solidarity affordable?

Probably the most expensive cancer therapy in practice today is CAR T-cell immunotherapy. This therapy involves extracting a patient’s T-cells, re-engineering them so that they can more effectively attack one or another blood cancer through targeting specific proteins on the surface of cancer cells, then injecting them back into a patient. In the United States, the front-end cost of this therapy is $475,000. In addition, more than half of these patients will experience a side effect known as cytokine release syndrome (with a range from somewhat serious to very serious), which will require a lengthy ICU stay for recovery and add more than $200,000 to the overall cost of the procedure if more serious (Santomasso et al. 2019). In the US ~ 70,000 individuals with leukemia or lymphoma would be potential recipients of CAR T-cell therapy (if all were fully insured and if the technical capacity to meet that volume were available). In the EU that number would be ~ 120,000 individuals. Depending upon specific medical circumstances, 30% + of adult patients offered CAR T-cell therapy will not survive a year. The remainder may gain two to four years of life expectancy, at which point the T-cells will often be exhausted. Research is going on to develop a version of this therapy that can be used to treat tumors in solid organs (Grens 2019). If such research were successful, that would dramatically increase the number of individuals who would be candidates for this therapy. Consider now the solidarity-rupturing questions that are associated with this therapy.

Considerable research is occurring with regard to all these immunotherapies and targeted therapies to identify predictive biomarkers aimed at making ‘more appropriate’ use of these therapies (Bairi et al. 2019). First question: Should such research be funded when the goal is to find biomarkers that will identify patients who have a 90% chance of not surviving a year after CAR T-cell therapy? Second question: If such research is successful, should it be used to deny patients access to CAR T-cell therapy at social expense, assuming a 90% likelihood they will not survive a year? What needs to be kept in mind is that this is the proverbial ‘last chance’ therapy; there are no other therapeutic options for such patients. What also needs to be kept in mind is the question of whether cost-effectiveness is a relevant consideration. Do such considerations necessarily corrupt solidarity? The implication of cost-effectiveness is that two patients might have the same cancer requiring the same targeted therapy with different predictable outcomes, such that from a social point of view one of those patients might be twice as expensive from a cost-effectiveness perspective as the other, even though the same volume of dollars or euros might be spent in both cases. If a patient survives (and is predicted to survive) only six months with CAR T-cell therapy, then from a cost-effectiveness perspective, those six months cost $950,000 per life-year gained (at least), while another patient receiving that same therapy and surviving three years would have a cost of $150,000 per life-year gained. Alternatively, is there some limit to what should be spent reasonably to protect solidarity?

Next question: Researchers are now exploring whether patients who fail CAR T-cell therapy the first time should be offered the option of trying it again. The conclusion from one small study is that the results of a second infusion are not as good as the first infusion (Keown 2019). The solidarity implications of this conclusion can be taken in a couple different ways. If the judgment was that individuals who were predicted to fail CAR T-cell therapy in less than a year still had to be offered the therapy at social expense (because appeals to cost-effectiveness were a threat to solidarity), then that logic would require offering a second chance at CAR T-cell therapy. In addition, if individuals failed CAR T-cell therapy after three or four years, then this same logic would require offering all of them another round of CAR T-cell therapy. It would be unlikely that they would gain another three or four years, but a two-year gain would be significant. If these conclusions are accepted, then it would seem the implicit message would be that cost does not matter when metastatic cancer patients have no other life-prolonging options. This would demonstrate an unwavering commitment to protecting solidarity……among cancer patients. However, this is really a very awkward conclusion. It represents tribal solidarity. This is hardly the sense of solidarity that most advocates for solidarity have in mind.

Is tribal solidarity good enough?

If cost is irrelevant for metastatic cancer patients when only one life-prolonging therapy is available to them, then non-tribal solidarity will require providing comparable life-prolonging therapies to patients with end-stage heart disease, or liver disease, or lung disease, or multiple sclerosis, and so on. We can take the logic of this position one step further. What we have described in this last sentence is one very large, very inclusive tribe of individuals who are terminally ill but for whom something can be done to prolong their lives. This is a tribe that few of us are enthusiastic to join, especially if we have a chronic degenerative condition that will ultimately result in death. What we would much prefer, if cure of our condition is just not medically possible in a predictable future, is that the slope of decline be ameliorated and prolonged as much as possible by anything medicine might offer us. Also, if we are paying taxes to pay for extremely costly life-prolonging technologies for individuals with very late-stage illness, in effect saying cost does not matter, then we might well expect that cost should not matter with regard to our medical needs as well. What, we might ask rhetorically, is the moral logic that would dictate penuriousness with respect to the use of therapeutic resources in the earlier stages of a chronic degenerative illness, so that we could speed along to those final stages when unrestrained spending would be the rule? For example, the current cost of fourteen drugs that would slow the decline of multiple sclerosis will range from $67,000 per year to $92,000 per year (Huff 2018). This is a lot less than the front-end costs of CAR T-cell therapy. However, patients can be on these drugs for two or three decades, yielding an aggregated cost per patient of two to three million dollars over the course of a lifetime. In the United States there are about one million people living with multiple sclerosis. How should this be thought about from the perspective of solidarity? Can there be a cost-conscious form of solidarity that does not yield fragmented or tribal solidarity?

Another recent example is a drug called tafamadis. It would be used to treat a very specific form of heart failure called transthyretin amyloidosis. Roughly 100,000 to 150,000 patients in the US have this version of heart failure. The cost of this drug is $225,000 per patient per year (Court 2019). This is a drug, like the MS drugs, that could be taken for several years. One of the medical oddities associated with this drug is that only 4% of patients with this condition are correctly identified as having this condition in current medical circumstances because their symptoms are not distinctive enough to be recognized by most cardiologists. Consequently, they are simply identified as heart failure patients, and they are treated with relatively inexpensive drugs. This raises a provocative question for countries committed to solidarity. Should all patients now identified generically as heart failure patients undergo more sophisticated diagnostic testing to identify correctly those with this very specific version of heart failure, potentially adding $20 billion per year to the cost of treating these patients? Powerful economic considerations for health care cost control would likely speak against such an effort. Would solidarity considerations make such an effort ethically obligatory? Recall that a small percentage of these patients would already be correctly diagnosed by a few astute clinicians who now would know of this drug. Would solidarity be best preserved if this drug were not offered to any of these patients? This sounds like a plausible option. However, in the arena of targeted cancer therapies and immunotherapies, all manner of genetic testing is being done to identify patients whose tumors have specific genetic mutations likely to elicit a beneficial response with some novel drug. This information is not being hidden from these patients. Why should such information be withheld from cardiac patients who would stand to benefit from tafamadis?

The challenges to solidarity, cost control, and precision medicine might be manageable if these last two examples were the only two examples of concern. However, the list of emerging, very costly, life-prolonging medical technologies is growing rapidly. For patients with cystic fibrosis there is now Trikafta®, a drug that partially corrects for the genetic deficiency that results in cystic fibrosis. The cost of that drug is $310,000 per patient per year. This is another drug that could be taken for several decades with per patient lifetime costs in the millions of dollars or euros. There is also Zolgensma® for spinal muscular atrophy at a one-time cost of $2.12 million per patient. There are enzyme replacement therapies for patients with either Pompe disease or Gaucher’s disease, each of which has costs of $300,000 per patient per year. The cost of dialysis in the United States is $91,000 per patient per year, currently for 570,000 patients, who can gain seven to ten extra years of life, sometimes more. The implied argument from these statistics is that if we are willing as a society to pay these costs now for all these other patients, how could we justify refusing to pay the costs associated with the cancer therapies under the rubric of precision medicine, especially in countries that embrace solidarity as a fundamental social value?

Someone might argue that all the therapies listed above are very effective, resulting in additional years or decades of life. In contrast, for a large majority of metastatic cancer patients the targeted therapies are marginally effective at best, and only somewhat more effective for most of the rest of those patients. However, defenders of these targeted therapies will point out that the social costs for many of these other patients are in the millions of dollars for lifetime costs, whereas the costs for these cancer patients might be $100,000. Also, it might cost $150,000 to give a metastatic cancer patient an extra year of life while these other patients are getting many years of life for $300,000 or more per extra year of life. That, the argument goes, does not appear congruent with a commitment to solidarity. However, someone might then argue that perfect solidarity is a utopian illusion; we just have to accept some moderate deviations from solidarity, given the complexity of medical needs and costs.

This last point generates the obvious follow-up question: How should we (a society committed to solidarity) decide which deviations from solidarity are acceptable? What should be the relevant criteria? For example, women who are HER-2 + for breast cancer are typically treated with the targeted therapy, trastuzumab. If they are judged to be at low risk of recurrence, then that therapy will go on for a year. However, some recent research has shown that a nine-week course of this therapy is almost as good as a year-long course of therapy. To be precise, at the five-year mark disease-free survival was 87.5% for the one-year group and 85.4% for the nine-week group (Nixon et al. 2018). The advantage of the shorter course is reduced risks of toxicity and a savings to a public funder of about $60,000 per patient. As a cost control measure, should all women in these circumstances simply receive the nine-week course of therapy? Is this only a minor deviation from solidarity that all should accept as reasonable? After all, the absolute difference is only 2%. However, we should note that it was a five-year study. What is unknown is whether at the ten-year mark a much wider survival gap might emerge, perhaps a difference of thirty or forty percentage points. The likelihood is that outcome will never be known, especially if a policy decision needs to be made at five years based on this research. How should that be assessed from the perspective of a commitment to solidarity?

Precision medicine or precision prevention: whither solidarity?

Another challenge to solidarity comes from advocates for a much stronger effort to prevent cancer in the first place, or else to identify and treat it as early as possible. On the face of it, such efforts seem eminently reasonable and affordable. One such advocate for this strategy is Azra Raza, who is a physician/ cancer researcher who recently authored a book, The First Cell and the Human Costs of Pursuing Cancer to the Last. Her research for the past twenty years has been about myelodysplastic syndromes and acute myeloid leukemia. The sub-title of her book makes clear her point of view. She regards all the resources spent on metastatic cancer, all the targeted therapies and immunotherapies, to be a huge waste of resources. She wants all those resources redirected to intense preventive efforts. She writes, “The prevention I am talking about is through identification and eradication of transformed cancerous cells at their inception, before they have had a chance to organize into a bona fide malignant, incurable disease” (Raza 2019).

Raza sees blood tests for circulating cell-free DNA, so-called liquid biopsies, as the key to identifying cancer in its earliest stages. It was after the publication of her book that a company called GRAIL announced they had developed such a test that could detect 20 different cancers and their likely tissue of origin with 99% specificity. Sensitivity, however, varied with the specific type of cancer and the stage at which that cancer was detected. More precisely, sensitivity for Stage I was 32%; 76% for Stage II; 85% for Stage III; 93% for Stage IV (Baldwin 2019). In early 2021 Grail announced a new liquid biopsy test capable of detecting fifty different cancers from a single blood draw (Reno 2021). The cost of this test is $950.

A first question for consideration would be who should have access to these tests. A second question would be who would pay for these tests. Relative to drugs that cost $100,000 or more for a course of treatment, this sounds like an astounding bargain at $950. However, given the lifetime risk for cancer for anyone in the US or the EU is 40% (American Cancer Society 2020), that would suggest that 200 million adults in the US or 370 million adults in the EU would be potential users of this test. That translates into an annual cost in the US of $190 billion and $350 billion in the EU. Why would all these people want to have such a test every year? The short answer is that it is non-invasive; it could catch a cancer very early when theoretically treatable; the cost would be socially absorbed; it would eliminate cancer anxiety. Furthermore, some cancers, such as ovarian and pancreatic cancers, tend to present with symptoms only in very late stages when a terminal prognosis is nearly certain. Where does a commitment to solidarity fit regarding this discussion of cancer prevention?

Recall that Raza judges that the money we are spending on treating late stage metastatic cancer is a waste. She wants that money to be used instead to pay for something like the GRAIL liquid biopsy test (known as Galleri).Footnote 20 As noted already, in the US, about 1.7 million individuals are diagnosed with some form of cancer annually. That represents less than 1% of the adult population who would be candidates for this liquid biopsy. That means that 99% + of these individuals would derive no benefit (other than psychological assurance) from the test each year, and more than half of those individuals will never have cancer. Raza’s critics will see that as a waste of resources, likely arguing that a much more targeted approach is necessary. Furthermore, if all the resources needed for this preventive effort come from withdrawing of funds from these targeted cancer therapies and immunotherapies, that means these metastatic cancer patients will die ‘prematurely’. For a majority of these cases, individuals would be denied extra months of life, perhaps a year of life, which they might well desperately want. Others, however, the strong responders, would lose extra years of life for the sake of saving the ‘statistical lives’ of unnamed and unnamable others. This will strike many as being incongruent with solidarity, certainly if we accept Bayertz’, Saltman’s and Habermas’ understanding of solidarity. The core of the problem is that patients with late stage metastatic cancer are identifiable individuals. They have names and faces; they are visibly suffering and facing death, knowing today that these therapies exist that might postpone death for them.

A substantial debate has been ongoing regarding the question of whether statistical lives in this sort of situation should be weighed just as heavily, morally speaking, as the identifiable lives of patients burdened with disease (Faust and Menzel 2012; Cohen et al. 2015). This is not a debate that can be addressed in this essay. However, it is a debate that is integral to our understanding of solidarity. We might imagine seeking a compromise of some sort. More specifically, liquid biopsies should only be available at social expense for individuals in various high-risk categories. Individuals with a strong family history of cancer would be one obvious category. More challenging would be individuals who had an elevated risk of cancer associated with their DNA, as known through whole genome sequencing. This is another place where the ragged edge problem would emerge with a vengeance, again as a significant threat to solidarity.

I cannot think of any cancer related to our DNA that is 100% penetrant. Instead, the much more common reality is that an individual could be told that whole genome sequencing indicated that their lifetime risk for cancer ‘X’ was 20% (using a polygenic risk score), while the lifetime risk for any random individual for that same cancer might be 10%. Should we conclude from those statistics that the individual with that 20% lifetime risk should have access at social expense to the GRAIL liquid biopsy? However, what do we say, as solidarity supporters, to the individual with that 10% lifetime risk? Are they not eligible for an annual liquid biopsy at social expense? They will point out that it will still be the case that one out of every ten individuals with that 10% risk for cancer ‘X’ will in fact be diagnosed with that cancer. Further, some of those individuals will be diagnosed too late; their cancer will be advanced and incurable. How can such an outcome be congruent with a commitment to solidarity? What if that lifetime risk of cancer ‘X’ is 5% or 3%? Is that clearly the point at which denial of access to that liquid biopsy annually at social expense is justly denied? That is the ragged edge. What is a reasonable response to that challenge from the perspective of solidarity?

The scenario suggested here is in reality too simplistic. For the vast majority of us, the likelihood is that we might be vulnerable to several different cancers. For almost all of them the lifetime risk might be in the low single digits. However, we can imagine there might be one that poses a lifetime risk of 15%. Should we then conclude that solidarity would require annual access at social expense to that liquid biopsy test? The vast majority of cancers (90%) are not a simple result of mutations somewhere in our DNA; environmental factors are often key contributors to the initiation of a cancer, such as radiation exposure (natural or artificial), chemical exposure, smoking, excess alcohol consumption, and many other known or suspected factors. Individuals might have some control over some of these factors but very little control over others. How should such factors be considered in determining whether any particular individual, or class of individuals, should have access to this liquid biopsy test at social expense?

Solidarity, justice, and rational democratic deliberation

Solidarity is a fundamental social value that ought to be nourished and cultivated. It undergirds a commitment to universal health care access in many European countries.Footnote 21 However, as I have argued in this essay, precision medicine and the growing dissemination of very costly, marginally beneficial targeted cancer therapies, threaten to fragment solidarity. The rapid rise in the cost of cancer care threatens the adequacy of resources needed to address many other serious health care problems. Within cancer itself we can see the potential emergence of competing cancer tribes, defined by genetic features of their cancers and the drugs used to treat those cancers. The basic problem is that the concept of solidarity does not have the conceptual resources within itself to address the fracturing tendencies inherent in the medical and economic aspects of precision medicine. The concept of solidarity is simply too abstract a notion.

In another essay I have argued that the concept of solidarity must be qualified by a pluralistic conception of health care justice (Fleck 2015). We need to articulate a conception of just solidarity. To be clear, a well-articulated conception of just solidarity will still not be able to completely address the ethical challenges being created by the emergence of precision medicine. The contemporary world of medicine, economics and health care politics is simply too complex. Relevant considerations of justice (egalitarian, utilitarian, libertarian, prioritarian, sufficientarian, luck egalitarian) are likewise too complex for pure rational resolution. The philosopher John Rawls (1993, at 54–58) refers to this as the ‘burdens of judgment.’ What this notion means is that at the practical level of resource allocation within health care we may be able to do no better than hope to achieve ‘rough justice’ and ‘supple solidarity.’ The need for tradeoffs is inescapable. We cannot afford to do everything that is medically possible, most especially when life is at stake and a patient is faced with a terminal prognosis (as in the case of metastatic cancer).

The ‘we’ I have been referring to is all of us. Multiple fundamental social values are relevant to making allocation and rationing decisions within any health care system.Footnote 22 A reasonable balance must be achieved among them that is integral to protecting a commitment to solidarity, what Rawls (1993, at 8–11) refers to as a ‘reflective equilibrium’. There is no perfect balancing among these values. Most often several possible balancings will be ‘just enough’ and ‘protective enough’ of a commitment to just solidarity. However, we cannot simply stop there. We need to make specific choices that are reasonably fair and affordable. We all need to be participants in the processes of rational democratic deliberation through which such balancing can be articulated and legitimated.

Let me offer a very succinct picture of what a rational democratic deliberative process might look like. We will use the example of metastatic colorectal cancer treated with the checkpoint inhibitor, pembrolizumab. Recall that tumor mutational burden (TMB) has proven to be a good predictive biomarker for prolonged therapeutic effectiveness (greater overall survival) (Klempner et al. 2020). It seems the more mutations in the tumors, the stronger the response from pembrolizumab (or other checkpoint inhibitors). This will be especially true with a TMB greater than 35 mutations/Mb. If that number drops to 25 mutations/Mb, the degree of effectiveness will decrease significantly, but not to zero. This is the ragged edge problem. The deliberative process could endorse a rationing protocol as being “just enough” that would not fund from public resources pembrolizumab for a colorectal cancer below that TMB of 35 mutations/ Mb because cost-control judgments had to be made somewhere. That would reduce to some degree the survival benefit for patients whose colorectal cancer was below that TMB mark. Some deliberators might object to sacrificing those extra months of life “for the sake of money.” Some deliberators might argue that this checkpoint inhibitor ought to be provided to any patient who might benefit to any degree with a TMB that would predict some likelihood of some benefit. In other words, they would bypass the “ragged edge” problem by refusing to draw any bright line “in this case.”

Another deliberator could ask what it was about “this case” that would justify not drawing a bright line that would save resources. Unless such a rationale were forthcoming, the implication of endorsing this view is that no bright line would be drawn in the case of any targeted therapy or immunotherapy that promised any degree of benefit with regard to prolonging survival. To be clear, there would be nothing intrinsically unjust about paying for everything under the rubric of precision medicine that yielded any degree of medical benefit. However, another deliberator could reasonably ask whether the same would be true in every other area of medicine when individuals were faced with a terminal prognosis and some very expensive intervention could yield some marginal benefit. If that were the practice, this would have drastic economic and ethical consequences either for other non-life-threatening medical needs or for some range of non-medical social needs. From the perspective of Rawls’ notion of wide reflective equilibrium, the outcome would be unjust and disruptive of solidarity. No justice-relevant compelling reasons can be given for providing unlimited resources either to cancer care or medical therapies for patients with a terminal prognosis. Moreover, it would be socially imprudent to endorse such targeted largesse, given the scope and diversity of health care needs in the population as a whole. It would also be imprudent for each and every individual to endorse such limited largesse, given that in the real world the vast majority of us are behind a health care veil of ignorance.Footnote 23

The practical, justice-relevant implication of this attenuated democratic dialogue is that bright lines (even with a little wiggle) must be drawn through the deliberative process rather than allowing very arbitrary and very wide ragged edges to determine resource allocation decisions in cases such as our TMB case above. That sort of arbitrariness would certainly undercut solidarity. Agreeing to create these bright lines does result in “rough justice” because we know some marginal gains in life expectancy will be sacrificed, and sometimes those foregone gains will be more than marginal. Further, those sacrificed gains are attached to patients who command our empathy because they are very visible. However, those empathic sentiments must not be allowed to distort the just allocation of resources for the needs of other patients who will necessarily be invisible.

Rawls invokes the notion of a ‘veil of ignorance’ to assure impartiality in choosing principles of justice. Critics often deride the unreality of that notion. However, the veil of ignorance is perfectly real and relevant when it comes to the deliberative process that I have in mind for addressing the problem of health care rationing, whether narrowly in the context of precision medicine or more broadly. For the vast majority of us at any point in our lives we are entirely ignorant of the health-related disorders with which we might be afflicted. That provides a suitable environment for rational, objective, impartial thinking about specific health care rationing problems. Further, even if I have an early-stage serious cancer (or know that I am at greater risk for some cancer than the average person), and might be initially biased toward more funding for cancer-related health care interventions that most would regard as not being very cost-effective, I can be reminded through the deliberative process that I am still vulnerable to many other threats to my life and my health other than cancer problems. I would also be reminded through the deliberative process that there are many other people to whom I am emotionally connected who are at risk for a broad array of health problems, some of which would be life-threatening. Consequently, I would want to make with my fellow deliberators the fairest and most prudent choices possible, given that we have only limited resources and virtually unlimited health care needs. Put another way, the deliberative process need not degenerate into competing interest groups or health care tribes. Instead, we are all capable of recognizing that we are largely incapable of taking complete responsibility for meeting our health care needs; we all need to take responsibility for each other, given the uncertainty and complexity of future possible health needs.Footnote 24

In conclusion, given a society committed to accepting a reasonable pluralism, it will rarely be the case for any particular rationing problem that there will be one “most just” choice that all reasonable persons would recognize. The more common situation will involve several possible choices that might all be regarded as being ‘just enough’ or non-ideally just (along with other possibilities that would not be ‘just enough’). The goal of the deliberative process would be to legitimate a choice that would be ‘just enough’ and ‘not reasonably rejectable’ by those who might have favored a different option. This is how a just solidarity is constructed in a diverse society for purposes of addressing the ethical ambiguity created by precision medicine.

To be clear, precision medicine is not a unique threat to solidarity and health care justice. The broader threat is represented by various life-prolonging technologies that are extraordinarily expensive and generally yield only marginal benefits for a substantial majority of patients. Unchecked expenditures for these technologies and these patients threaten the just allocation of health care resources elsewhere in the health care system. If solidarity and health care justice are nothing more than abstract idealizations, that threat cannot be abated. To counter that threat, the normative force of solidarity and health care justice must be institutionalized and concretized through fair and inclusive processes of rational democratic deliberation, as briefly illustrated above. Precision medicine is in many respects the scientific quintessence of the healing arts, but it is also the worst instantiation of the technological imperative run amok, generating “wicked” ethical problems that threaten the integrity and justness of the healing arts.Footnote 25