Les diagnostics odorants diagnostiquent les infections en analysant l’haleine expirée.

Using the Nose and its Equivalent as a Diagnostic Tool

In this full-length doctor's interview, William Hanson, M.D., explains how doctors are using computerized noses to diagnose infections by analyzing exhaled breath.

Ivanhoe Broadcast News Transcript with William Hanson, M.D., Anesthesiologist/Intensivist, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania, 
TOPIC: Smelling out Sinusitis

How did you get involved with the electronic nose?

Dr. Hanson: The way I got involved, interestingly enough, and I hesitate to admit this, I was reading the Wall Street Journal, and about seven or eight years ago, they had an article on electronic noses, sort of generically, and one of the devices was a device from a company in England. I approached the company, and I said I would be very interested in using your electronic nose as a potential diagnostic tool in medicine.

Physicians going way, way back to the dawn of time used their noses to smell people’s breath, urine and bodies to diagnose diseases, so I thought that there might be something that we could use medically in today’s very high-tech environment. As it turns out, we’ve been able to use this handheld device, which is very suited to the medical environment.

How long have you been working with it now?

Dr. Hanson: Seven or eight years, I think.

How, in that amount of time, have you seen the allocation of it change or what you know about what it is capable of doing?

Dr. Hanson: We originally wanted to test this in patients on ventilators, and the reason to do that is when you have patients who are breathing into a tube, you don’t have to worry about bacteria in the teeth or bad breath or whatever.

It’s all bypassed with a tube in the lungs for patients who are unfortunately on ventilators. Our original problem with the devices was that we started with personal computer-sized devices and getting the sample to the machine was difficult.

We had to collect the sample in a bag, walk it to a machine, and in the meantime, the sample tended to condense on the side of the bag and we lost a lot of the odor in the process. The newer devices are smaller -- miniaturized, as a lot of things are -- and they’re handheld. In this case, they’re perfect for the kinds of applications we’re looking for. So, that’s one way in which it’s changed.

There’s a whole, huge wave of new advances coming out of laboratories that involve sensing with things like DNA sensors, nanosensors or combinations of electrical and biosensors. This first-generation device that we’re working with right now is a sort of a non-specific sensor that is very good at handling a wide variety of odors, and we’re looking to see more specific devices for the medical market.

You were involved in a study with this recently, right?

Dr. Hanson: We have an ongoing series of studies. The studies I’m working with right now involve an ongoing series of pneumonia patients. My colleague is working on patients with sinusitis, which is another big outpatient disease, whereas mine is an inpatient disease.

We’ve shown that we can distinguish different bacteria from one another just by their smell. These are plated bacteria, bacteria that are in petri dishes. You can smell them and say, “This is one organism,” and smell a different organism and identify it.

We’ve also, interestingly enough, been able to smell different body fluids, one example of which is cerebral spinal fluid, which can leak, in patients with certain diseases, out of your nose, and it’s hard to know whether it’s a very dangerous situation where you have spinal fluid leaking out of your nose or you have something just cold secretions of whatever.

Distinguishing between those two things is very difficult currently in the lab. So, that’s one of the many studies that we have that’s shown some interesting result.

What would make spinal fluid leak out of your nose?

Dr. Hanson: In patients who’ve had a head injury, their dura, which is the leathery material that surrounds the brain and encloses it, can tear and they can have cerebral spinal fluid leaking through the dura into the nose. They can also have the same problem as sinus surgery. In both cases, there’s a potential for bacteria to get into the lining of the brain, which is obviously a bad thing.

What is the benefit a patient gets out of a doctor using an electronic nose as a diagnostic tool?

Dr. Hanson: The way we currently diagnose a number of diseases involve getting a sample from the area of interest. It might be the sinuses, it might be the lungs, it could be a wound, and then sending that sample off to the laboratory for results, which would likely come back two or three days later.

In the meantime, the clinician, the doctor or nurse has to make a decision as to whether or not to treat the patient with antibiotics, and if so, what kind of antibiotics. Either they’re broad-spectrum or their narrow-spectrum antibiotics. Our goal in the first instance is to get a diagnostic method that really is essentially non-invasive and painless that would allow us to determine whether or not the patient has an infection that we need to treat.

That would be a double. A home run would be we would be able to determine what the infection comes from and then use the appropriate antibiotic for that kind of an organism, and in both cases, you’d start therapies either earlier, which would be good, or you wouldn’t start therapies that weren’t warranted, which would also be good.

I imagine that would help with the overuse of antibiotics if you were not prescribing it when they’re not needed.

Dr. Hanson: One of the problems that we’re all concerned about is the inappropriate use of antibiotics in medicine. Patients are oftentimes treated with antibiotics when they have symptoms that could be either viral or bacterial. A sore throat is an obvious example of that, but so are situations with the sinuses or with the lungs where you have a cough and you have the low-grade fever.

It could be a virus or it could be pneumonia. If we can avoid the inappropriate use of broad-spectrum antibiotics, we’re way ahead of the game in terms of limiting the spread of antibiotic-resistant organisms, which as you know is a big problem for all of us.

What do you think will become the most common use of the electronic nose?

Dr. Hanson: One of the most likely applications of electronic noses is not so much as the specific diagnostic device but as a screening device. For example, if a patient came in with symptoms you were concerned about, you might sample the patient’s breath or sinus exhalations, and if they were worrisome, you would then send a culture off.

If they were absolutely not worrisome, you wouldn’t send the culture, which has the obvious effect of saving money for inappropriately cultured patients that don’t need to be cultured.

In a basic sense, when would you use this?

Dr. Hanson: With the device that we’re talking about right now, I think the kinds of applications you can imagine are in the outpatient setting, using this in any patient that comes in prior to being seen by the doctor with a sinus complaint. So, they would presumably sit down in some holding area while they’re getting their vital signs done and get an anode sample.

Using that information as part of the information that you have available to you, including the patient’s temperature and blood pressure and whatever, you would make your decisions more precisely than you would otherwise.

In the intensive care unit, you could imagine having an electronic nose of one sort or another attached to every respirator that’s continuously monitoring the patient and would then alert the physician or nurse when something was happening that was worrisome in terms of the trend of the exhaled gasses, suggesting that there was a pneumonia brewing.

How widespread is the use of this?

Dr. Hanson: This should be looked at right now as a tool that’s being investigated in several areas including cancer, sinus disease, pneumonia, some of the inflammatory diseases of the lungs like SARS, which we’ve heard a lot about recently. These are all areas that are very fruitful for investigation, and that’s sort of the stage at which these tools are being used right now.

What we saw when the patient breathed in and you saw the lines move, what is the significance of the line that went way up? What does that tell you?

Dr. Hanson: The way in which these sensors work are there are differences amongst different company’s sensors, but in effect, what happens is that the molecules that are in the odor bind to the sensor and change the resistance of the sensor to electricity. So, the resistance goes up, which is what that upward deflection on the screen was that you would see on a specific track.

Then for a sensor that was very responsive to the components in your sample of interest, there would be a very significant deflection of that particular trace, the red line, for example, whereas if you had a sensor that was essentially unresponsive to that particular odor, it would be very little deflection, and when you take the aggregate movement or the bulk movement of all 32 sensors, you have what can imagine is a point in 32-dimensional space, which you obviously can’t work with, but there are mathematical ways to take that point and then compare it to other points and say are these odors the same, like or different.

So you would basically need to know what level someone with pneumonia registers on the sensor?

Dr. Hanson: The way in which you would probably do this is you would start with a group of known with no pneumonia. You’d also have a group of water samples to calibrate the device. Then you would train the nose to recognize people with pneumonia and you would say pneumonia, yes or no, which group do you fall in.

If you got better and more specific with your device, you would say this patient has no pneumonia or they fall into the group of patients that have pneumococcal pneumonia or they fall into the group of patients that have mycoplasma pneumonia, which is walking pneumonia. So, ultimately what you’d like to have is a nose that has been trained by known samples to recognize unknown samples.

END OF INTERVIEW

This information is intended for additional research purposes only. It is not to be used as a prescription or advice from Ivanhoe Broadcast News, Inc., or any medical professional interviewed.

Ivanhoe Broadcast News, Inc., assumes no responsibility for the depth or accuracy of physician statements. Procedures or medicines apply to different people and medical factors; always consult your physician on medical matters.

If you would like to get more information, please contact:

The Perelman Center for Advanced Medicine

3400 Civic Center Boulevard

Suite A-5

Philadelphia, PA 19104

Office: 215-662-4856

Fax: 215-662-7625