Stanford researcher challenges assumptions about cystic fibrosis

In Herzenberg Lab, Rabin Tirouvanziam makes revolutionary discovery, explores new treatments using JMP^®

The conventional wisdom used to be that neutrophils – the most abundant white cells in our blood – played a passive secondary role in the destruction of lung tissue in cystic fibrosis. Rabin Tirouvanziam, a researcher and instructor at the Stanford University School of Medicine, has turned that wisdom on its head.

Using JMP statistical software to analyze and interpret his data, Tirouvanziam has learned that live neutrophils are central to both the onset and the progression of cystic fibrosis lung disease. Ongoing clinical studies suggest that his findings are likely to help extend lives.

In general terms, Tirouvanziam’s work involves looking at the relationships among chronic disease, inflammation and metabolism within the human system.

“We know that our bodies invest a lot in our immune systems,” says Tirouvanziam. “In conditions where the immune system is being challenged, it’s a constant cost on our metabolic balance.”

Tirouvanziam studies those interactions. “That’s the overarching theoretical umbrella of my research,” he says.

Beneath that umbrella, Tirouvanziam and his colleagues have been engaged in the study of cystic fibrosis, a genetic disease that strikes the lungs and digestive system. In the US, about one in 30 Caucasians, one in 45 Hispanic-Americans and one in 65 African-Americans are carriers of cystic fibrosis gene mutations. While medication and dietary regimens help alleviate the effects of the disease – and life expectancies have improved – cystic fibrosis continues to cause serious damage to the lungs. A thick mucus builds, lung tissue eventually breaks down and those afflicted essentially choke to death.

“Lung disease is what’s making those with cystic fibrosis sickest and ultimately killing them,” says Tirouvanziam. “The end game is transplantation. Fifty percent of lungs transplanted in the US are transplanted because of cystic fibrosis. So not only is it a huge drain on the quality of life of these patients, but it’s also a huge burden on our health care system.”

But Tirouvanziam and his colleagues have made a revolutionary discovery.

“The paradigm that’s been dominant for a number of years is that neutrophils have a bystander effect in cystic fibrosis,” Tirouvanziam explains. “But we’ve turned that paradigm around with our data.”

Researchers have long believed that cystic fibrosis developed when bacteria became trapped in built-up mucus in the lungs. At this point, neutrophils enter. Neutrophils attach to bacteria and, in healthy lungs, destroy the bacteria, serving as innate protectors against disease. But in the lungs of those with cystic fibrosis, the neutrophils weren’t doing their job – and were very quickly dying.

Or so researchers believed. Tirouvanziam and his colleagues discovered something quite different – that, in fact, neutrophils were hanging around, ultimately breaking down lung tissue.

“We analyzed cells from the lungs and the blood of patients, and we demonstrated that these cells are in fact not dying quickly in the lungs,” Tirouvanziam says. “They’re very much alive, they’re being reprogrammed, and they are genuine targets for therapies.”

This work was carried out through the use of flow cytometry, using a fluorescence-activated cell sorter (FACS), as developed in Stanford’s Herzenberg Laboratory, with which Tirouvanziam has long been associated (see sidebar).

“We collected a lot of multiparametric data from the FACS,” Tirouvanziam continues, “and obviously we needed a robust statistical package to look at all that data, and look at it longitudinally and run multivariate correlations and see how we would associate it with different severities of disease, different genotypes, etc.

“Our work with JMP started very early on, with our basic studies, and those basic studies have led to a couple of pathways that are treatable,” he says.

Tirouvanziam’s research has indicated that the level of inflammation in those with cystic fibrosis can be reduced with an oral medication called N-acetylcysteine (sometimes called NAC). He and his team are now engaged in an FDA-approved placebo-controlled trial involving 80 patients enrolled in 10 centers across the US.

“The framework for analyzing the data has broadened,” Tirouvanziam says, “but the platform that we use is still the same – we’re still using JMP to tabulate all the laboratory data as well as the clinical data that’s coming in from patients across the six months of this trial. JMP is our go-to platform.”

Tirouvanziam is pleased that JMP helps him perform analyses properly. “What I like about JMP,” he says, “is that it has a lot of underlying statistical assumptions that you don’t get in other tabulating software – which means that you can’t go wrong with the way you enter your data and then the way you analyze it.

“For example, you can’t fool the program into thinking that a numerical column that is continuous can have a character. Or you can right away define boundaries for your data; so, inline, as you enter it, your data is being checked for appropriateness with regards to the columns into which it’s being placed.”

Tirouvanziam regularly designs experiments with JMP. “Most biologists don’t get to do much with designed experiments,” he says. “But in the realm of clinical studies, this is really significant. Basically, you have a first set of data, and you project yourself into the future. You want to see how many patients you will need to enroll to get sufficient statistical power to make a statement as to whether, for example, something is being modulated by a drug. This is built very well into JMP, and it’s really useful.”

Tirouvanziam says he likes how JMP has evolved over time. “Each version of the software has continued to bring new things to the table,” he says, citing, in particular, the graphics.

He also appreciates the ease with which JMP works with third-party software. “A lot of people still work with Microsoft Excel,” Tirouvanziam says, “and I think the ability to go from JMP into Excel is very helpful. And from Excel into JMP, there’s that data check that occurs inline. If you have errors in Excel, Excel doesn’t prompt you. It doesn’t really care whether you make an error or not. So when you have huge data sets, tabulating your data in Excel is a recipe for making mistakes – while JMP doesn’t allow that.”

Tirouvanziam has now begun research into autism, a disease for which the underlying causes are currently unknown.

“We have very good data suggesting that at least a subset of patients has chronic inflammatory pathways that are being activated,” Tirouvanziam says. “So we’re pursuing a couple of studies – one of them observational, the other interventional – again, using JMP.

“We’re using JMP for all of our other exploratory studies,” Tirouvanziam says, “looking for new pathways and new targets. It’s what we’ve been relying on. JMP is really the bag into which we throw all our marbles.”