A dialysis drug softens the shield around pancreatic cancer in patients’ biopsies

Detective: a lead from the lab turned up again in patients' biopsies

Pancreatic cancer has a terrible reputation among oncologists. It has earned it. The tumour rarely causes early symptoms, so many people are diagnosed only once it has already spread. Chemotherapy can slow it down, but rarely for long. In Poland this cancer kills several thousand people every year. In its metastatic form, five-year survival is measured in single-digit percentages.

One reason is the tumour's surroundings. Pancreatic cancer cells grow a dense, hard layer of connective tissue around themselves. It is built by fibroblasts, the cells that heal wounds in a healthy body. Next to a tumour they behave like a building crew that never finishes the job. The result is a kind of shield. A drug from an IV drip has a harder time getting through. Lymphocytes that could attack the cancer stay outside.

A team from Dana-Farber Cancer Institute in Boston and the Salk Institute in California tested whether this shield can be loosened with a drug pharmacies have stocked for years. The results appear in Nature Cancer. The drug is paricalcitol, a synthetic analogue of active vitamin D. The US FDA approved it long ago, but for a completely different purpose. It is given to people with chronic kidney disease whose parathyroid glands release too much hormone.

The first lead: a receptor that calms fibroblasts

The investigation began in Ronald Evans's lab at Salk. Evans has studied nuclear receptors for decades. These are proteins in the cell nucleus that respond to hormones, vitamins and fats, then switch genes on or off. According to Salk, roughly one in eight FDA-approved drugs works through such a receptor.

One of them is the vitamin D receptor. Evans's team noticed that fibroblasts in the liver and pancreas carry a lot of it. When the researchers stimulated this receptor with synthetic forms of vitamin D, the fibroblasts calmed down. In animals this held back liver fibrosis and pancreatitis. In models of pancreatic cancer something more happened. Tumour-associated fibroblasts stopped being activated, and chemotherapy worked better.

Synthetic forms have an advantage here over ordinary vitamin D. The body does not break them down as quickly. That is why the researchers chose paricalcitol rather than a supplement from the drugstore. This is an important caveat. The study does not show that vitamin D tablets treat cancer.

The second lead: biopsies before and after

A theory from mice is not yet proof in people. So Brian Wolpin and Kimberly Perez of Dana-Farber designed a trial meant to answer a simple question first: is it safe?

It enrolled 36 people with metastatic pancreatic cancer who had not been treated before. All received standard first-line chemotherapy, gemcitabine with nab-paclitaxel. They were randomly split into three groups of 12. One group also received a placebo. The second got paricalcitol intravenously. The third got the same drug as tablets.

The most interesting part of the trial lies in the biopsies. Doctors took tumour tissue at the start and again after four to six weeks of treatment. They then examined it with two modern methods. One stains many proteins at once in the same section. The other shows which genes are active in a specific spot of the tissue.

The picture matched what had been seen in animals. In patients on paricalcitol, the share of activated fibroblasts fell. The fibroblasts did not disappear; their state changed. More CD8 T cells, the cells that can kill cancer cells, also entered the tumour. Importantly, they sat closer to the cancer cells than before treatment.

The third lead: who responded

The trial was not designed to settle whether paricalcitol extends life. It was too small. Still, the authors report signals that are hard to ignore.

Among the 24 people who received paricalcitol, the tumour clearly shrank in 10. In the placebo group this happened in one person out of 12. A year after treatment began, the disease had not progressed in five people on paricalcitol. There was no such person in the placebo group.

The next lead says the most. The amount of vitamin D receptor in the tumour varied widely between patients. Some had a lot, others little. That amount predicted the response to treatment with paricalcitol. Patients with high receptor levels who received the drug lived the longest.

This may be the most valuable result of the whole study. If it holds up, a doctor could test a tumour sample before treatment and judge who should get paricalcitol added and who should not.

How to read these numbers

In oncology, a "partial response" has a strict meaning. It usually means the tumour lesions shrank on imaging by at least 30 percent. The tumour does not vanish. It shrinks enough for the radiologist to see a clear difference.

The type of trial matters too. The authors call it a randomized run-in trial. Such a design checks whether a new combination does harm and collects material for tissue analyses. The placebo group serves mainly as a background here. It helps separate changes caused by chemotherapy alone from changes after adding paricalcitol.

That is why the biopsies matter more in this paper than the response rate. They show the mechanism in people, not just in mice. The efficacy numbers are, for now, a hint for whoever designs the next trial.

The price of the lead: calcium in the blood

Safety was the main goal of the trial, and the result here is good, with an asterisk. The combination could be given. However, five of the 12 people taking the tablets had raised blood calcium. Doctors had to reduce their dose. This is a known effect of drugs that act like vitamin D. It can be managed with standard blood tests, but a larger trial will have to watch it from day one.

The asterisk also says something about the dose. The IV form and the tablets did not behave identically. The next trial will have to pick one route and one dose.

What is solid here, and what is not yet

Three things are solid. The combination can be given to patients with metastatic disease. Biopsies show a change in the tumour's surroundings, in line with what animal studies predicted. The vitamin D receptor varies widely between tumours and is linked to response.

The efficacy numbers are soft. Thirty-six people is a very small group. Ten responses out of 24 sounds good, but with so few patients chance has plenty of room. The trial was also not designed to compare survival. The authors themselves write that larger trials are needed.

It is also worth remembering that pancreatic cancer has already seen many promising ideas. Various drugs have tried to break down its shield. Some failed in large trials. The difference here is the approach. Paricalcitol does not destroy fibroblasts; it calms them. Earlier work by other teams suggested that removing this tissue entirely might even help the tumour.

The business layer: a cheap drug, expensive proof

From a market point of view, this is an unusual story. Paricalcitol is an old molecule, also available as a generic. No one will earn as much on it as on a new antibody. That is good news for payers and bad news for research funding. A large phase 3 trial costs so much that a drug company usually pays for it only when it holds a protected patent.

This work was funded mainly by foundations. They include Stand Up to Cancer, Cancer Research UK and the Lustgarten Foundation, which focuses solely on pancreatic cancer. That model may repeat at the next stage. For off-patent drugs, the burden of proof often falls on foundations, universities and public agencies.

There is also a second, less obvious market. If the level of vitamin D receptor in the tumour truly predicts response, a test will be needed. Such a test has to be checked in large groups of patients, standardised and written into guidelines. At that point, diagnostics rather than the molecule itself may be where the money is.

Poland: a drug known from dialysis, a problem known from oncology

For Polish readers, it is interesting that paricalcitol is not exotic. Nephrologists and dialysis units have known it for years. If a large trial confirmed a benefit, adoption would not need a new factory or a new delivery technology. The barriers would rather be registration of a new indication and reimbursement.

Pancreatic cancer is among the cancers with the worst prognosis in Poland. Polish cancer centres run large multicentre clinical trials, so Polish patients could join the next stage if a sponsor opens recruitment in Europe. For now that is a condition, not a plan.

There is also a lesson for Polish research teams. This story began with basic research on a nuclear receptor. Many years passed between the first observations in animals and patient biopsies. Basic science and clinical oncology had to work together the whole time.

Why this lead is worth following

Most new cancer drugs target the cancer cell itself. This work takes another route. It changes the surroundings the tumour built for its own protection, and checks whether chemotherapy and the immune system can then break through more easily.

For now this is an early stage. We have a safe combination, a change visible under the microscope, and a promising but statistically weak efficacy signal. We also have a candidate marker that could point to the patients with the best chance of benefit. The next step is clear: a larger randomized trial that tests survival and confirms the role of the receptor.

If it works, oncologists will get something rare in pancreatic cancer. A cheap, familiar drug that does not kill the tumour, but takes away part of its defence.

Sources

1. Perez K.J. et al. Gemcitabine and nab-paclitaxel with or without the VDR agonist paricalcitol for metastatic pancreatic cancer: a randomized, multiarm, run-in phase trial. Nature Cancer, vol. 7, p. 928 (2026). DOI: 10.1038/s43018-026-01165-8. Trial registry: ClinicalTrials.gov NCT03520790.
2. Salk Institute / ScienceDaily, report on the paricalcitol trial in pancreatic cancer (3 October 2026): https://www.sciencedaily.com/releases/2026/09/260930225501.htm