Author:

Sophie Fessl

Introduction

Choriocarcinoma, a rare and once fatal solid tumour that develops from trophoblasts during pregnancy, is today one of the most curable of cancers, with a 100% survival rate among women with low-risk disease and 98% survival rate in high-risk disease. Much of the credit for this progress goes to Kenneth Bagshawe, a young English doctor with no particular interest at that time in oncology, whose powers of observation, curiosity, persistence and innovative thinking saved the life of a patient referred to his ward with chest pain. The process by which he diagnosed and managed that patient led to a breakthrough in the treatment of gestational trophoblastic tumours. It also established two important milestones in the treatment of solid tumours – the use of biomarkers in diagnosis and management, and the use of combination chemotherapy to guard against resistance.

Bagshawe was not the first to cure a choriocarcinoma. That distinction belongs to Roy Hertz and Min Chiu Li, two oncologists at the US National Cancer Institute, who had been working in parallel with Bagshawe without either side being aware of the other (Li et al, 1956). Bagshawe had already begun treating his patient with mercaptopurine when Hertz and Li published their first case of treating choriocarcinoma with methotrexate. But it was Bagshawe’s discovery of the human chorionic gonadotropin (hCG) as a diagnostic biomarker that set the standard for the diagnosis and monitoring of the disease. His role in setting up a national registration and data gathering system, and later a national specialist gestational trophoblastic disease service, laid the basis for rapid progress in developing the techniques for risk stratification, and risk-adapted treatment protocols in choriocarcinoma and other gestational trophoblastic diseases.

Learning from a rare cause of pulmonary hypertension

Kenneth Bagshawe took up a position as senior registrar in medicine at St. Mary’s Hospital, London, in 1956. At the time, he hadn’t decided which speciality to pursue – cardiology being one of the options – until three puzzling cancer cases set him on the course of becoming an oncologist.

In a short Lancet article published in 2004 under the title ‘I never did become a cardiologist,’ Bagshawe describes his own first encounter with a choriocarcinoma:

“The admission card from casualty read ‘Hysterical hyperventilation! Good teaching case.’ She was 19 and was so dyspnoeic that she crawled into the hospital. As a hopeful would-be cardiologist, I found a thrusting right ventricle and normal breath sounds. A radiograph and electrocardiogram were consistent with right ventricular strain and the lung fields were clear. It was concluded she had pulmonary hypertension. She died a few days later.”

At autopsy, the first impression was of pelvic sepsis which had spread to the lungs, causing thrombosis and narrowing of the pulmonary arteries. However, another pathologist present at the autopsy, A. M. Joakes, commented that he had seen a similar case roughly 10 years earlier, which had been due to choriocarcinoma (Begent & Seckl, 2024). That case, he said, had been reported in a cardiology textbook. As Bagshawe writes (2004):

“A subsequent search confirmed that it was the only such recorded case. Histology in both cases showed the pelvic vessels and pulmonary arteries were occluded by choriocarcinoma and secondary thrombus and there was no extravascular tumour.”

Typically, choriocarcinoma presents weeks, months or even years after the pregnancy has ended; the tumour tends to metastasise rapidly, and it can present in different ways. One such presentation – as Bagshawe saw in his own patient and the case reported 10 years previously – is as pulmonary embolism, where the pulmonary arteries get blocked with tumour tissue, causing severe breathlessness and death.

Putting the learning to use - a novel biomarker

Choriocarcinoma is a rare cancer, so the odds that Bagshawe would encounter another patient with choriocarcinoma were extremely low. Yet, just a few months later, a patient with highly similar symptoms presented. This was probably in 1958, though there are discrepancies in the literature.

The 34-year-old woman showed signs of pulmonary embolism 10 days after hysterectomy for menorrhagia, and was started on anticoagulants. When the patient later complained to her general practitioner about shortness of breath, she was sent to the hospital for breathing exercises. While at the hospital, she experienced chest pain and was admitted to the ward Bagshawe was working on (Bagshawe, 2004). A re-examination of her uterus confirmed that there was no malignant disease. As Bagshawe later described in an interview with Cancer Research UK, “I thought the chances of it being another case of this incredibly rare form of choriocarcinoma were millions to one, but I kept it in the back of my mind,” (Howlett, 2017).

As the patient’s condition continued to deteriorate, Bagshawe tried to work out how to rule out the possibility of a choriocarcinoma. What was needed, he decided, was to do a pregnancy test.

The reason why Bagshawe wanted to carry out a pregnancy test on a patient who had undergone hysterectomy, and could not possibly be pregnant, was the conclusion he had drawn about choriocarcinoma, says Professor Michael Seckl, consultant oncologist at Charing Cross Hospital, and director of the national Gestational Trophoblastic Disease Service, who had trained under Bagshawe.

“Ken Bagshawe had been busily reading the literature, trying to understand how a diagnosis could be made here, other than waiting for the patient to die and then get a post-mortem done. He reasoned that this is placental-type tissue, so it should be making placental hormone – the pregnancy hormone hCG [human chorionic gonadotropin]. As [the patient] wasn’t pregnant, the hCG should be normal, unless it is this particular cancer diagnosis of choriocarcinoma.”

The pregnancy tests in use back then measured the level of hCG, as they do now. However, unlike modern kits, the tests in 1956 involved rabbits. A urine sample from the woman would be injected into a female rabbit, and the rabbit’s ovaries would be examined a few days later. If there was hCG in the urine, the rabbit’s ovaries would change detectably, and the test would deliver a positive result.

Despite still being a junior doctor, Bagshawe was determined to have the test done for his patient – though as he would later recall, he didn’t give enough thought to how he would convince the hospital lab to carry it out.

“I took a urine specimen to the laboratory and foolishly fumbled my reasons for the request to the laboratory head. He went through his ritual response to a fatuous request from a junior doctor, which was to remove his glass eye and polish it on a red and white handkerchief before replacing it to deliver a paralysing stare. He said he refused to sacrifice a rabbit (the Friedman test) for a woman who could not be pregnant and whose uterus showed no malignant disease,” (Bagshawe, 2004).

One way or another, that test did get done. Legend has it that Bagshawe asked a young nursing colleague to submit the sample under her name, says Seckl. “Of course, the laboratory head couldn’t refuse a member of staff asking for the test.”

The following Friday, the result came back, strongly positive. In the meantime, the patient’s condition had deteriorated significantly. She had been placed in an oxygen tent and couldn’t eat anymore, as she was too breathless. At the time, no cure for choriocarcinoma was available. As Bagshawe would later recall (2004), he extrapolated from the knowledge that mercaptopurine – then the only available chemotherapy drug at his hospital – produced responses in some leukaemia patients.

“Choriocarcinoma was reputedly the most rapidly growing of all solid carcinomas so it seemed possible that the newly introduced antimetabolites, which were beginning to show responses in childhood leukaemia, might do something useful. In the pharmacy I found 6-mercaptopurine tablets which she started taking that night. During the weekend she seemed ever closer to death, but on the Monday morning she was out of the oxygen tent, having breakfast.”

Pre-empting resistance and monitoring impact

The treatment with mercaptopurine had worked. But, with remarkable foresight for that era, Bagshawe was wary of a miracle cure. During Bagshawe’s training as a junior doctor, he had witnessed how the single antibiotic treatment that once seemed so promising for TB had failed. As the bacteria evolved to resist the antibiotic, patients were left with ever fewer treatment options. Everything changed with the discovery that a combination of drugs could attack the bacteria more effectively than a single drug. This breakthrough marked the beginning of combination therapy for TB, restoring its status as a curable disease.

So, when his hospital received a second chemotherapy drug, methotrexate, Bagshawe decided to combine methotrexate with the 6-mercaptopurine that he had already prescribed. The patient’s response to treatment, which was monitored using hCG assays, dropped to undetectable levels within three months (Begent & Seckl, 2024). After treatment for six months, the patient recovered completely, and survived for more than 45 years. Her case is thought to be the first example of a solid cancer being eradicated using combination chemotherapy.

Bagshawe’s first treatment of a case of choriocarcinoma involving the pulmonary artery established the principles for managing choriocarcinoma as a disease. “Ken built a team around him that helped him to develop the chemotherapy protocols that we largely use today,” says Seckl.

In the US, the first attempts at finding a cure for choriocarcinoma/gestational trophoblastic disease had been made at a around the same time. In 1956 Hertz and Li had published their experience of treating a patient with trophoblastic disease with methotrexate. However, the two teams were not aware of each other’s efforts, says Seckl. “Subsequently, they met each other and started having meetings on trophoblastic disease. But it was different people stumbling upon the answer separately.”

Developing the first highly sensitive cancer biomarker

One of Bagshawe’s most innovative contributions was to employ hCG levels as a proxy for the number of choriocarcinoma cells present. On moving to London’s Charing Cross Hospital in 1961, Bagshawe’s first priority in the laboratory was to establish a reliable bioassay for hCG. The team had the idea that varying hCG levels in the blood were due to changes in the number of cancer cells. If this were the case, hCG levels could serve as a proxy to determine whether tumours were shrinking in response to treatment.

Speaking to Cancer Research UK, Professor Stan Kaye, who worked in Bagshawe’s lab at the time, described how hCG tests unfolded (Howlett, 2017): “Tuesdays and Fridays were crucial times in the lab. This was when the results of the HCG tests came in and it was my job to write the results up on the board, which Ken would be waiting anxiously for.

“The fascinating thing was that we kept seeing the HCG levels go down in patients who were having treatment. This meant the cancer was disappearing, and we could track the response as it was happening in real-time. It was the first time we’d seen a marker that was so exquisitely linked to a cancer.”

Relying on rabbits’ responses to injected urine samples, however, was unduly cumbersome, so Bagshawe focused on finding other ways for measuring hCG levels. At the time, radioimmunoassays were being developed for other hormones, says Seckl. “Ken wondered whether he couldn’t do the same thing for hCG, and that’s exactly what he did. He set up to develop a radioimmunoassay specific for hCG, and having succeeded in doing so, he built a company to automate the process. This was a major step forward.”

Today, hCG remains a commonly used biomarker in managing choriocarcinoma, adds Seckl. “We use hCG levels for the surveillance of patients with pre-cancerous disease, we use it for monitoring patients on treatment, we use it to guide, in part, the choice of which type of treatment to use and, once treatment is finished, we use the hCG level to monitor patients for relapse… Measuring hCG levels plays an absolutely central role in managing patients with choriocarcinoma and I can’t overemphasise the importance of the hCG assay that Ken developed with his team.”

Driving clinical progress

Bagshawe also deserves credit for setting up the structures needed to make rapid progress in this rare cancer. By 1973, Bagshawe and his team had set up a national registration scheme and centralised monitoring service for patients with hydatidiform mole. In 1975, a scoring system was developed to define the probability of drug resistance emerging. From the data available through the national registry, Bagshawe and his team developed a risk stratification system, allowing the treatment to be tailored more specifically to each patient. The EMA-CO, a combination of five chemotherapy drugs (etoposide, methotrexate, and dactinomycin alternating weekly with cyclophosphamide and vincristine), which was developed by Edward Newlands, part of the Bagshawe team, remains standard of care for high-risk gestational trophoblastic neoplasia today. The combination protocol of methotrexate with folinic acid that is still used worldwide as a minimally toxic treatment for low risk disease was originally developed by Bagshawe and his team back in 1962 (Begent & Seckl, 2024). In an early example of specialist centres for a rare cancer, in 1984, three hospitals – in Scotland, north England and London – were designated centres for hCG monitoring and registration of new patients with gestational trophoblastic disease (GTD), as part of a national GDT service spearheaded by Bagshawe.

“Bagshawe was a man with very far reaching vision, and built the systems to help us to register patients and ensure we have a robust database,” says Seckl. “None of the progress made would have been possible without Ken’s national centralised GTD service, which has served as a model for the rest of the world to emulate.”

An extraordinary achievement

While Bagshawe may not have been the first to cure a choriocarcinoma, his impact was much more far-reaching than that of the two US oncologists who published the first case of a choriocarcinoma cure, argues Seckl. “Hertz and Li didn’t work out how to develop an assay for hCG, they didn’t set up a centralised service, they didn’t work out the other chemotherapy regimens and also not the scoring system to work out which patients might benefit from single agents as opposed to multi-agent treatment.”

The true extent to which Bagshawe pushed the boundaries in treating choriocarcinoma is easily underestimated if not seen in the context of the knowledge available at the time, argue Seckl and Begent in a contribution to the series of Biographical Memoirs of Fellows of the Royal Society (Begent & Seckl, 2024).

“Pulmonary hypertension caused by choriocarcinoma in the pulmonary artery had never been diagnosed in life before. Using a biochemical marker with the clinical signs of pulmonary hypertension to make the diagnosis was a conceptual advance. Using the knowledge that choriocarcinoma, a solid tumour, had rapidly dividing cells which hypothetically might be susceptible to antimetabolite drugs was novel. Having the skill to keep a patient alive in extremis for the critical first days without modern intensive care facilities was exceptional. Monitoring the treatment to its successful conclusion with a biochemical marker was original. Succeeding with the essential elements of this between Friday afternoon and Monday morning was almost beyond belief. … It is easy to look back and underestimate the achievement, but that would be wrong because none of the components was based on direct proven prior knowledge. Rather, they depended on a hypothetical understanding of mechanisms and clinical skill assembled into a diagnostic and therapeutic plan.”

The impact on the prognosis of patients could hardly have been greater, they argue. “Survival rates were close to zero prior to his work, but today, through his efforts and those of his successors, nearly all affected women can expect to be cured in the UK and globally.

“Ken transformed the field from a 100% death rate at the outset to – by the time Ken retired in the 1990s – cure rates that were approaching about 85% survival for high-risk patients,” says Seckl. “For low risk-disease, this approached 99% survival.” However, treatments have been even further refined since Bagshawe has retired. “For low-risk disease, I haven’t lost a single woman in the last 25 years, so it’s a 100% cure rate. In high risk-disease, we are approaching 98% survival. It is fair to say that, without having had Ken set up our national service, the developments that we have been able to make since then would not have been possible. He is really fundamental to a lot that has been achieved.”

References

Bagshawe, K. D. (2004). I never did become a cardiologist. Lancet, 363, 1502.

Begent, R. & Seckl, M. (2024). Kenneth Dawson Bagshawe. 17 August 1925–27 December 2022. Biogr Mems Fell R Soc, 76, 49-67.

Howlett, E. (2017). Our milestones: How a combination of events led to a cure for choriocarcinoma. https://news.cancerresearchuk.org/2017/08/15/our-milestones-how-a-combination-of-events-led-to-a-cure-for-a-rare-cancer/.

Li, M. C., Hertz, R., Spencer, D. B. (1956). Effect of Methotrexate Therapy upon Choriocarcinoma and Chorioadenoma. Proc Soc Experimental Biol Med. 93(2), 361-366.