“Screening” for Prostate Cancer in New York’s Skid Row: History and Implications

Abstract

The Bowery series, open perineal biopsies performed on more than 1200 alcoholic men recruited from homeless shelters in New York City’s Bowery section, began in 1951 and persisted for more than a decade. If frozen sections revealed prostate cancer, men typically underwent radical perineal prostatectomy, orchiectomy, and diethylstilbestrol treatment. This poorly informed, vulnerable population was subjected to health risks that investigators knew others would not accept. Although the knowledge produced had little impact on practice, the Bowery practices foreshadowed and have troubling continuities with later developments. Currently, more than a million American men each year undergo prostatic biopsies. But the efficacy of prostate-specific antigen screening and the treatment that typically follows has never been established. The Bowery series and subsequent developments are part of one continuous story of how medical and lay people came to believe in the efficacy of population screening followed by aggressive treatment without solid supporting scientific evidence.

THE RECENT US PREVENTIVE Services Task Force recommendation against prostate-specific antigen (PSA) screening for prostate cancer is the latest salvo in a controversy that has existed since it was introduced in the late 1980s.¹ But looking for prostate cancer in asymptomatic men and treating detected cases with radical interventions has a longer—if forgotten—history.

Starting in 1951 and continuing for more than a decade, urologist Perry Hudson and his colleagues recruited more than 1200 homeless, alcoholic men from New York City’s skid row, the Bowery, to come to a new municipal cancer hospital for invasive tests and procedures, including open perineal biopsy (OPB) of the prostate gland. If frozen biopsy sections revealed cancer, the patients typically underwent radical prostatectomy and orchiectomy, followed by a course of diethylstilbestrol.

To present-day observers, the unethical aspects of these practices—the failure to provide full informed consent and exposing a vulnerable population to undue risk—are disturbing. Yet the Bowery series, as Hudson called these practices, were published in leading medical journals and frequently cited, joining the long list of unethical studies performed in full public view, such as those revealed in Henry Beecher’s influential 1966 exposé.²

I have detailed the history of the Bowery series, providing the medical and social context for why and how these practices were begun, persisted, and eventually faded away. Examining this history does more than explain—without excusing—an important but hitherto forgotten large-scale ethical misadventure. It also reveals continuities with prostate screening beliefs and practices today. Recounting and recovering this history is a window into what is at stake—ethically and scientifically—in the long-running controversy over the early detection and treatment of prostate cancer. The juxtaposition of past and present also opens up for ethical analysis new developments, such as the mass diffusion of interconnected clinical practices, each understood to work on the basis of limited endpoints by well-intentioned actors, before—and potentially subverting—robust evidence of their overall efficacy.

THE BOWERY SERIES

Prostate cancer circa 1950 resembled pancreatic cancer today. Patients almost always had an incurable disease at the time they sought medical attention. Despite Johns Hopkins surgeon Hugh Young’s demonstration at the turn of the century that prostate cancer could be cured with radical prostatectomy, the technically challenging operation was rarely done, largely because of patients’ late stage at the time they sought medical attention.³ There was also little push to mobilize the public for early detection. Autopsy studies and pathological examination of tissue removed during procedures to relieve benign urinary obstruction showed cancer at rates far in excess of what was clinically detected.⁴ Urologists believed these cancers, labeled latent or incidental, caused no harm, and thus there was no need to detect and treat them.

Perry Hudson, who in 1951 was an ambitious, 33-year-old, newly minted, Hopkins-trained, Columbia University urologist, wanted to change his fellow urologists’ pessimism and clinical inaction. He had just been appointed head of urology at the Francis Delafield Hospital, a public cancer hospital built and paid for by New York City where many Columbia physicians with clinical and research interests in cancer had appointments. Hudson was in charge of 45 beds and a full floor of laboratories plus another five urological laboratories at nearby Columbia-Presbyterian Hospital.⁵

Newly in charge of these extensive resources, Hudson decided to scale up and experiment with existing clinical practices that would advance objectives that were diverse and in flux. First, he hoped to determine the prevalence of prostate cancer and get a better understanding of the disease’s natural history. Second, he hoped to prove that OPB could be used to diagnose prostate cancer at earlier stages.⁶ Third, by performing radical prostatectomy on men whose intraoperative frozen sections demonstrated cancer, he hoped to treat men before they developed metastatic disease and thus persuade urologists that prostate cancer detected incidentally in asymptomatic men was neither latent nor harmless and should be treated aggressively. Overall, he hoped to contribute to the ultimate development of an early detection and treatment paradigm in prostate cancer similar to what was beginning to take hold in breast cancer and cervical cancer.⁷

Participant in Bowery series talking to doctors at Francis Delafield Hospital, 1957. Photo by Walter Sanders. Printed with permission of Getty Images.

Hudson routinely performed OPBs on men before they underwent surgery to relieve benign obstruction. If biopsy showed cancer, he instead did a radical perineal prostatectomy. But he recalled wondering about people “wandering about the street who don’t have a benign enlargement. What about them? … How do you screen for it?”⁵ The idea of more systematically screening men occurred to him while caring for a former Princeton University history professor who had become a homeless alcoholic living in a Bowery flophouse. Hudson visited his flophouse to see whether he could recruit men to come uptown to Delafield for OPB and other procedures. His initial attempts were failures, but his luck turned when he befriended two welfare department workers at the municipal shelter (“the Muni”), a large city-run facility for homeless men. Hudson convinced the workers of the importance of his investigation, and they were able in turn to convince Muni residents to participate in Hudson’s studies “so that just about everybody volunteered.”⁵

A physician resident who worked under Hudson and helped recruit “subjects” in the 1960s and participated in their later clinical care (and who I interviewed on the condition he not be named) recalled that the Bowery men were only dimly aware they were participating in research but that “research was a very nebulous term”⁸ at that time. The Bowery men were also a source of clinical material, allowing Hudson and his trainees to perform more perineal biopsies and prostatectomies than anywhere else in the world.

Hudson recalled that recruitment was easy. Bowery men knew they would receive “total care” at Delafield. According to a contemporary popular account of Bowery life that included a chapter on the Bowery series, once persuaded to volunteer, “the authorities lifted the man’s meal ticket and returned it only when he had kept his appointment.”⁹ Hudson’s resident recalled, “We would meet these men as they came out of the shower … and say ‘we’re doing examinations on your prostate.’”¹⁰ This physician limited recruits to men with a prostatic finding or problem, but he recalled that other trainees would recruit whoever they could, usually saying something like, “Oops, you have a prostate; let’s go up to the hospital.” They explained to the Bowery men that at the hospital they would “get an open biopsy to test your prostate.”¹⁰ The men who went along were attracted to staying “four or five days in a hospital [with] clean sheets, and three meals a day.”¹⁰

These recruiting practices were part of a larger moral, political, and monetary economy of Bowery life. Men (and a few women) provided souls and remained temporarily sober for church missions, which in turn provided meals, lodging, and a respite from street dangers; men hustled each other and outsiders for cash, which they exchanged for food and cheap drink; low-rent flophouses and cheap bars turned profits from the large flow of down-and-out men; men provided cheap and expendable day labor for small amounts of cash. The Bowery concentrated homeless, alcoholic, and physically and mentally distressed people who were often the waste material of the work economy and made them invisible and scarce elsewhere in a part of the city long impoverished and violent, literally under the shadows of a soon-to-be-removed elevated train.¹¹ The exchange of bodies for research and practice and training was only a small part of this larger, historically and geographically specific system of exploitation.

After agreeing to participate, Bowery men were taken to Delafield, where they underwent a complete physical examination, including prostatic massage for cytology; urine and blood analysis; x-rays of the abdomen; intravenous pyelograms; and cystoscopy, often accompanied by transbladder biopsy of the prostate. Then, under general or local anesthesia, Hudson or his residents removed a 2.5 × 1.0 × 0.5 centimeter core of the posterior prostate.¹² One half was sent to the pathology laboratory for frozen section, the other half retained for permanent preparation. If the frozen section showed cancer, the patient typically underwent a perineal prostatectomy and orchiectomy, followed by diethylstilbestrol treatment, a nonstandard and aggressive practice that Hudson rationalized on the grounds that the three interventions together might benefit those patients with apparent localized disease who were not cured by surgery alone.¹³ If the biopsy was negative, the patient’s perineum was stitched back together.

Initially, Hudson and his colleagues selected for biopsy 141 patients who apparently had some symptoms or physical signs suggestive of urinary obstruction. But subsequent recruits were unselected.¹⁴

Although the published studies did not systematically report complications from biopsies, Hudson claimed they caused no harm, including no impotence.¹⁵ Partially in response to this claim, another independent group described a case series of 24 patients who had undergone OPB, two thirds of whom had serious complications, such as rectal perforations, cardiac events, and impotence. These dramatic differences from the Bowery series, the authors said, were “difficult to explain.”¹⁶ Hudson’s public report notwithstanding, Hudson’s resident later recalled that impotence was common and that the most feared danger of perineal biopsy was rectal perforation.¹⁷

Most cancers detected by OPB were said to have occurred in men with normal physical, laboratory, and radiological results.¹⁸ In other words, these other tests all lacked even the minimal sensitivity to detect cancer at a curable stage. The most complete prevalence data were from a 1966 report in which cancer was detected in 10% of 891 “survey patients,” presumably those Bowery men recruited without symptoms or signs.¹⁹

In 1957 Hudson published the only report on longer-term outcomes, citing data on 686 men who had undergone OPB and had generally been treated with radical prostatectomy, orchiectomy, and diethylstilbestrol therapy if the biopsy was positive for cancer. Mortality was high, both among men with negative biopsies (20%) and those diagnosed and treated for cancer (30%). Hudson noted that only 1 patient diagnosed by OPB had died of prostate cancer and that this man had refused treatment, dying, in Hudson’s view, of untreated “latent” cancer.²⁰ More follow-up data were apparently collected but never published because some data were lost when Hudson left Delafield in 1960 (for reasons unrelated to his Bowery work) and because Hudson had been persuaded to stop the studies and forego publishing his results.

Although Hudson continued to recruit Bowery men after he moved to Montefiore Hospital in 1960,²¹ he stopped publishing when an editor of the journal Cancer wrote him

a very careful letter and asked me what protection I had from the university’s legal department. And what measures I had taken to guard against lawsuits. And I had to tell him, “Nothing, I had done nothing.” That sort of scared me. So when I did the last of the studies I never published them.

The editor who warned him was a friend of mine. He didn’t send that out as anything but a friendly warning. That meant that he had heard things. Other people did too. I was aware of that.²²

Except for a passing remark about the studies’ “ethical problems” in one study,²³ no explicit criticism of their ethics was made in scores of citations in the contemporary medical literature or in National Institutes of Health study section reviews. The study was heroically depicted in a 1957 Life Magazine photoessay and a 1961 popular account of Bowery life.²⁴

Years later, Hudson’s resident expressed regrets over the absence of adequate informed consent, especially that the dangers of rectal perforation had not been explained and documented on written forms. “Today we wouldn’t do that in a million years, but we did that.”²⁵ He recalled, however, that “the goal was a noble one, to find cancer at an early stage”²⁵; that the procedure was generally safe; and that Hudson’s residents were well trained as a result of the experience.

Hudson today has no regrets about the ethics of the Bowery series. Acknowledging that many men were mentally ill and alcoholic, he nevertheless believed that they were not “simpleminded” and “understood everything that was going on.”²⁶ He pointedly noted that he never gave money to participants to participate, which in his view constituted an ethical lapse. He was aware that others objected: “Among urologists, I got called a collection of really fancy names … I didn’t pay any attention to them.”²⁶ The people Hudson admired “understood what I was doing. So I paid no attention to it.”²⁶

Hudson also remains an advocate for perineal approaches to surgery and biopsy. He is the author of two surgical atlases that highlight perineal surgery. However, the incomplete detail in publications and the absence of adequate controls meant that the Bowery series would never ignite use of OPB for screening or as a way to diagnose cancer before obstruction-relieving surgery.

CONTINUITIES FROM THE BOWERY TO MAIN STREET

Although there were few objections raised at the time, the many unethical aspects of the Bowery series are largely evident to us now. The risks of screening through hospitalization and invasive testing, especially OPB, were large; the efficacy of treating “early” cancers with radical surgery, castration, and diethylstilbestrol was unknown; and there is no evidence that anything but rudimentary consent was obtained. Hudson used Bowery men because only desperate, poor, and unknowing men would participate; it was unimaginable that the average American man would volunteer. In other words, Hudson recruited men from Bowery shelters because they were either not fully competent to understand the risk and benefit calculation or they were desperate enough to make a calculation that someone less vulnerable was unlikely to make. The former aspect ran against the ethical principle of informed consent, which had been explicitly codified as a result of the Nuremberg trials in 1947, whereas the protection of vulnerable populations in research might arguably be understood as a less established norm before Beecher’s 1966 article and the multitude of regulations and codes that followed the public airing of the Tuskegee experiments starting in the 1970s.

In addition to these ethical failures of the Bowery experiments, the men were also used as clinical material for teaching and practice. The Bowery practices occurred in the nebulous boundary zone between human experimentation and medical therapeutics. Although this boundary zone is not unique for its period or the present, it is fertile soil for ethical mishaps.²⁷

Although many 20th-century bioethical tragedies have involved racial minorities, institutionalized juveniles, and prison and other captive populations, the Bowery men were almost uniformly White. Their vulnerability resulted from their poverty, mental and physical illness, alcoholism, and the resulting homelessness. These demographic factors may partially explain why the Bowery series has not previously attracted any bioethical attention—in addition to possible medical embarrassment over the parallels with present practices.

What was unimaginable in the 1950s has become routine: more than a million American men each year undergo prostatic biopsies for cancer, and the greatest risk, then and now, is from the active treatments triggered by a cancer diagnosis. Although men undergoing biopsy today are typically selected by screening blood tests and biopsy practices that are safer than OPB, the efficacy of this entire screen-and-treat paradigm remains unclear.^28,29

In the decades following the Bowery series, many incremental technological and practice innovations were introduced. Starting in the 1960s, radiologists experimented with newly developed ultrasonography to diagnose prostate cancer, but it was not until the 1980s that technological improvements in transrectal ultrasound made it clinically useful.³⁰ By the late 1980s, urologists were using relatively safe spring-loaded biopsy guns, along with transrectal ultrasounds, to biopsy palpable lesions.³¹ Walsh’s modifications to the radical prostatectomy in 1982 reduced its sexual and urinary complications. He noted that before this time, “only 7% of men with localized prostate cancer underwent surgery,”³² demonstrating the persistence of the inertia and pessimism that Hudson had hoped to change 30 years earlier.

What was crucial to overcoming this inertia and pessimism was the search for something less invasive and risky than screening OPB. In the 1980s, there was concerted testing and innovation with prostatic acid phosphatase, first developed by Hudson’s co-workers at Columbia 50 years earlier³³ and marketed as a male Papanicolaou test. However, the male Papanicolaou test³⁴ was eclipsed by the development and implementation of PSA screening and diagnosis in the 1980s and early 1990s.³⁵

These technological changes were important improvements over OPB and the radical perineal prostatectomy used on asymptomatic men in the Bowery series. They helped lower the threshold for mobilizing asymptomatic men and made more palatable the screen-and-treat paradigm. As prostate biopsies were safer and done on men believed to be at high risk, the potential for rapidly increasing the scale of surveillance on the general population grew. But because the efficacy of the entire screen-and-treat paradigm was never established by a controlled trial during the period in which it was widely diffused, there was also much greater potential for large-scale harm. Whether efficacious or not, these technological and practice innovations resulted in much more testing, more prostate cancer diagnoses, and greater numbers of men who survived for long periods of time with a prostate cancer diagnosis. This greatly improved survival led to a perception that the different elements of the screen-and-treat program worked, leading to more compliance and more incorporation of these practices.

In the decades following the Bowery series, these incremental developments traveled along the fault lines that Hudson blazed: put needles into asymptomatic men, find some prostate cancer, treat maximally, and kindle a cascade of action and belief that would transform prostate cancer from a deadly to a survivable disease. The first step, as Hudson urged and foresaw, was to make many more cancer diagnoses confined to the prostate gland. Urologists were gradually persuaded that these cancers were dangerous and subjected many more men to biopsies and radical interventions. With little change in prostate cancer mortality (it rose in the decades before the early 1990s and then declined), these interventions were believed effective because survival rates increased and case fatality plummeted, although these changes could have resulted solely from increased numbers of diagnoses (i.e., overdiagnosis).³⁶

Men detected “early” and urologists could plausibly perceive that that they were participating in important medical progress, although there was no demonstrated net mortality benefit and many men were potentially harmed. Only a few years after PSA screening was introduced, the American Urological Association and American Cancer Society had endorsed its use in all men older than 50 years.³⁷ Some critics nevertheless noted at the time that there was “no controlled study of health outcomes after PSA screening”³⁸ and that results of trials already started would not be available for more than a decade.

Starting in the 1990s, the ever-increasing numbers of “cured” prostate cancer survivors have been mobilized for advocacy and disease awareness campaigns and constitute for many lay and medical people evidence in themselves of efficacy. Much had changed, including mass incorporation of screening and new medical and lay faith in the efficacy of modern practices, but these changes did not necessarily mean that screening per se had reduced overall mortality.

In these ways, the practices of Hudson’s successors have succeeded where Hudson’s crude practices failed. Throughout this history, the goals were reasonable and the practices could be perceived as effective as long as there was no disconfirming evidence from a clinical experiment with appropriate controls. Urologists, radiotherapists, and radiologists in the decades after the Bowery series diffused more acceptable diagnostic and treatment practices and then used the resulting transformed disease picture (apparent greater survival rates, a new curable stage of disease) as evidence of the efficacy of these very same practices.

Since the mid-1980s, increasing numbers of men received a prostate cancer diagnosis via PSA screening and biopsy, submitted to active intervention, and lived long cancer-free lives. Many such men as well as outside observers interpreted this sequence—just as Hudson believed about his Bowery patients—as diagnosis, treatment, and cure, and they touted it as such, leading to more faith in the paradigm, more screening, more decisions for active treatment, and more recruits into the mass of cured cancer patients. This logic has often resisted skeptical challenge. Not only is progress undeniable but the selling of fear and uncertainty also makes the technologies that promise to banish them irresistible. In sum, a self-reinforcing cycle of practice and attitudinal change took root. This cycle has been stabilized by deeply held assumptions about prostate cancer’s one-way, destructive natural history and orderly classification (a cancer is a cancer).

These continuities are not simply guilt by association. The Bowery series and subsequent developments are part of one continuous story of how medical and lay people have come to believe that a massive screen-and-treat paradigm works and of the interconnected transformation of prostate cancer as a disease. The many incremental practice changes and technological developments that catalyzed this transformation had plausible rationales. Urologists evaluated and justified their use on the basis of limited endpoints (e.g., improved case fatality, removal of cancer, early stage of screened cancers, posttreatment PSA determinations). By 1995, the earlier urological skepticism about the early detection and treatment of prostate cancer had been largely reversed. An editorialist noted that for many physicians “withholding screening while men die of prostate cancer is unethical,”³⁹ whereas for others, “encouraging screening without this evidence is unethical (primum non nocere).”³⁹

Hudson and subsequent actors had good reason to pursue the strategy of kindling a self-reinforcing pattern of behavior and attitudinal change. Prostate cancer circa 1950 was, and in 2013 remains, a serious, deadly problem. As we have long had the means to remove or destroy cancer, many people then and now have understood the biggest obstacles to be the stage at which patients seek treatment and the pessimistic and fatalist mood of doctors and the lay public.

The transformation of prostate cancer occurred before the results of any large-scale clinical trials of screening using mortality as an endpoint were available. Furthermore, the successful diffusion of the screen-and-treat paradigm may have made a meaningful trial impossible to interpret. The “usual care” control groups of any “intention to X” trial contain too many men screened and treated to determine with any confidence whether there is or is not a small net health benefit if trial results were negative.

CONCLUSIONS

To an observer today and to some at the time, the ethical failures of the Bowery series are apparent. But we generally do not see it as an ethical problem that more than 1 million mostly asymptomatic American men each year have prostatic biopsies. The screen-and-treat paradigm leads to predictable psychological and bodily risks and was widely diffused in the absence of solid evidence that such risks were balanced by health benefits. This continuity between the Bowery practices and those of subsequent decades should lead us to question our very different reactions to the past and present.

There were differences of course. The Bowery men were poorly informed and vulnerable research subjects exposed to risky interventions of uncertain benefit. American men who have prostatic biopsies after elevated PSA levels are often fully informed patients, although many are poorly informed about the PSA test that triggered the biopsy. But as the efficacy of the screen-and-treat paradigm was never established, the practices of the past few decades can be understood as a mass experiment conducted on ill-informed men.

My answer to why we view the Bowery series and current practices so differently is that a half century of incremental technological tinkering has normalized and made more palatable practices that could only be done on ill-informed, vulnerable participants in the past. This tinkering created the conditions for a self-reinforcing interaction among unsubstantiated assumptions about cancer’s natural history and overdiagnosis and a transformed disease picture. Witnessed but limited evidence of progress—prostate cancer became a treatable disease and the number of survivors rapidly grew—led to the rapid diffusion of the screen and radical treatment paradigm. There have been rational motivations for the many incremental steps taken by different individuals (urologists, technology innovators, policymakers), which together constituted the cascade of changes that transformed our approaches to prostate cancer.

But this rationality at the individual level does not mean there is not an ethical or moral problem. We take a myopic view of ethics when we limit moral judgments to individual actors at one point in time. This individual focus prevents us from ethical reflection on the overall operations of a system of practices that evolved incrementally and with seeming rationality and whose individual innovators were not clearly acting unethically. Although the conceptual and practical challenges of evaluating and regulating how clinical practices unfold are large (e.g., for the autonomy of clinician–patient decision making and access to unproven treatments), so are the consequences of the evidence-free emergence and reification of population-wide practices of unproven efficacy.

The unethical aspect of the Bowery series was not only exposing a poorly informed, vulnerable population to undue risk but also the well-intentioned strategy to transform a disease and jumpstart an aggressive screen-and-treat paradigm in ways that bypassed and eventually subverted more robust clinical evaluation of its efficacy. “Randomize the first patient” may be an unrealistic ideal, but it might be less so if we recognized it as a moral, not simply scientific, dictum. This is especially true of interventions that can affect large populations and transform the target condition. If American clinicians and lay persons had recognized the mass diffusion of the screen-and-treat paradigm before high-quality clinical evidence of its efficacy as an ethical challenge, there might have been at least a partial brake on the runaway, self-reinforcing processes that I have described and that have also occurred in breast cancer and other diseases.⁴⁰ This is a challenge to the way research, technological innovation, and clinical practices are understood and promoted, not to doctors’ and patients’ individual decisions once the transformation of prostate cancer was in full swing, as it was by the early 1990s.

Ethical tragedies are difficult to recognize in the present. They do not necessarily appear in the same guise as past ones. Future observers may view the massive evidence-challenged expansion of our screen-and-treat paradigm in prostate cancer in the critical way we now view the Bowery series practices.