Mr. Michael Jarvis: Welcome, everybody. Thank you very much for tuning into this presentation, where I'll be introducing the new AB SCIEX Triple Quadrupole 6500 LC-MS/MS System, a new tool for achieving the next level of performance for steroid research

So, first, let me introduce myself. My name is Michael Jarvis. I'm the Technical Marketing Manager for Clinical Research at AB SCIEX

And I'll be starting off with a hardware overview of the new components of the AB SCIEX Triple Quadrupole 6500 System, featuring IonDrive technology

Then I'll hand off to Dan Holmes, who is the Clinical Associate Professor of Pathology and Laboratory Medicine at the University of British Columbia and the Division Head of Chemistry at St. Paul's Hospital in Vancouver, Canada, who will talk about the analysis of aldosterone by LC-MS/MS

The new AB SCIEX 6500 series is the newest introduction to the family of modern AB SCIEX instrument and joins the 4500 and 5500 systems

Like the 4500 and 5500 systems, it shares a compact footprint with dimensions of 32 by 32 by 24 inches. And it comes equipped with convenient integrated features, such as the six-position diverter valve and the syringe pump

Major advantages in LC-MS/MS performance on this system are powered by IonDrive technology, which consists of three principle components. The IonDrive Turbo V Source, the IonDrive QJet Ion Guide, and the IonDrive HE Detector

Furthermore, advanced dual-RF electronics makes the system a no-compromise triple quadrupole mass spectrometer with capability to operate in two different modes, either high-sensitivity mode with a mass range from 5 to 1,250 Daltons, or extended mass range mode, with mass range from 5 to 2,000 Daltons

This system is also compatible with SelexION Ion Mobility Technology for enhanced selectivity

So, this system was designed with the philosophy of creating more ions, capturing more ions, and detecting more ions. The system-wide improvements include modifications in the Turbo V Ion Source, the QJet Ion Guide, and the detector

All of these improvements serve to drive LC-MS/MS performance to the next level

Let me start by describing the IonDrive Turbo V Source, designed with the idea of creating more ions. The larger-diameter heaters are 11 millimeters in diameter as opposed to 4 millimeters on the original Turbo V Source

The optimized geometry ensures more efficient heat transfer, which effectively desolvates and ionizes analyte ions more efficiently

These heaters cover a larger cross section of the spray zone, resulting in a much wider sweet spot for optimization of probe position

Furthermore, these changes for the source design ensure more robust performance against fluctuations in gas flow dynamics and source-to-source differences

As we can see on this slide, this figure highlights the differences between the original Turbo V Source and the new IonDrive Turbo V Source

As we can see, the much larger heaters provide much larger sweet spot for ionization and much more efficient probe optimization

The second component of the multi-component IonDrive technology is the new IonDrive QJet Ion Guide, designed to capture more ions

This RF ion guide is a two-stage ion guide. The first stage serves to better capture ions from the larger orifice. The second stage serves to better transfer ions to the Q0 region

This design provides high robustness and a better barrier against neutrals and microdroplets entering the mass spec system

Compared to the original QJet Ion Guide on previous systems, the new IonDrive QJet Ion Guide provides ion transmission efficiencies which are two times better

In this slide, we can see a computed gas flow model showing flow through the ion guide. The larger diameter of the first stage ensures better ion capture, while the smaller diameter of the second stage ensures better ion focusing into Q0. The end result, more ions transmitted and fewer neutrals reaching Q0

The third and final component of the IonDrive technology is the IonDrive High Energy Detector, designed with the idea of detecting more ions

This detector insures that sensitivity gains are not at the expense of dynamic range. Ultra-fast pulse counting allows ten to the eight counts per second. And this is a 20 times improvement compared to conventional detectors and previous mass spectrometry systems

The higher saturation point combined with the pulse counting to ensure a low-end sensitivity is not compromised results in a total dynamic range of six orders of magnitude

On this slide, we get a sense of the sensitivity gains on the Triple Quad 6500 System relative to the Triple Quadrupole 5500 System

On the left side, we see a compound verapamil with a sixfold gain in signal and fourfold gain in signal to noise. On the right side, we see a compound, ginsenoside RB2 with a gain of 20 times in signal and a gain of four times in signal to noise

Overall, we expect that the average gain in sensitivity on new Triple Quadrupole 6500 System will be up to 10 times

As I've already mentioned, it's expected that the new HE detector provides an extended linear dynamic range of up to six orders of magnitude. As we can see here, we have a calibration curve ranging from one pictogram per mL to one microgram per mL of aldosterone

The inset in the bottom right-hand corner demonstrates that, even at low concentrations, the accuracy and linearity of the calibration curve is very good

So, in conclusion, the IonDrive technology on the new AB SCIEX Triple Quadrupole 6500 System drives LC-MS/MS performance to the next level

It does so by improving analyte ionization and desolvation with the new IonDrive Turbo V Source. It also improves ion capture with the new dual-stage IonDrive QJet Ion Guide

And finally, the new IonDrive HE Detector improves sensitivity and extends the dynamic range

The overall performance of this system provides up to 10X more sensitivity compared to the 5500 series and up to six orders of linear dynamic range

Furthermore, this system is compatible with the SelexION Ion Mobility technology, thus providing enhanced selectivity capabilities for challenging assays

And finally, this no-compromise system employs advanced dual-RF electronics to allow for operation in two different modes for either highest sensitivity analysis or for extended mass range analysis up to M-over-Z 2000

So, I'd now like to turn over to Dr. Daniel T. Holmes from British Columbia in Canada, who will be speaking about the analysis of aldosterone by LC-MS/MS

Mr. Dan Holmes: Thanks, Michael, for that very interesting presentation about the 6500. Those look like really positive improvements

We're going to talk about some initial work that we've done on a difficult analyte on the 6500. But, we'll start with our background of what we've done on the 5000

So, start with an outline. First, we'll talk about the research motivation behind a desire to measure aldosterone, looking at the structures, the sample prep, the chromatographic challenges that we had. And we'll be comparing the sensitivity of aldosterone determination on the API 5000 to the new Triple Quad 6500 System

We're going to talk about our research motivation behind measuring aldosterone. Aldosterone and related mineralocorticoids are of interest for those who are doing research in secondary forms of hypertension

And when we say secondary forms of hypertension, we mean curable forms that have a definite cause. Most hypertension doesn't have a well-defined treatable cause

Primary aldosteronism as the most common form of secondary hypertension is thought to comprise up to 10 percent of hypertension in all comers

There's also increasing interest in the physiological effects of aldosterone and other mineralocorticoids outside the classical renin angiotensin aldosterone system paradigm. That is we might call those pleiotropic effects

In other words, what effects does aldosterone have other than the effective retaining sodium and wasting potassium at the level of the renal tubule and therefore maintaining blood pressure

There's also an interest in measuring aldosterone in tissue provided a couple of references here from a colleague Carlos Gomez Sanchez and another researcher Kirin Selma [sp] from the University of British Columbia

This is a diagram of the renin angiotensin aldosterone system. And just as an overview, starting on the left looking at the liver, the liver makes a protein, a globular protein called angiotensinogen, which is released into the plasma

And under the action of an enzyme produced by the juxtaglomerular apparatus of the kidney, the enzyme is renin, a decapeptide called angiotensin I is cleaved off

Angiotensin I is not physiologically active as far as anyone knows. But, as it passes through the endothelium of the lung and the renal endothelium, it is subsequently cleaved to another peptide called angiotensin II, which has a number of direct physiological activities

Relevant to us, angiotensin II has action at the adrenal gland to increase the production of aldosterone. Aldosterone, as I alluded to earlier, at the level of the renal tubule causes the retention of sodium and in exchange for potassium

And with sodium comes the water that surrounds that ion. And that maintains the plasma blood pressure by increasing the plasma volume

The concentration of aldosterone's very low in the plasma. It's in the picomolar range. And that makes it a difficult steroid to analyze, unlike cortisol, which is about 1,000-fold higher in concentration measured in the nanomole per liter range

Why would we want to measure aldosterone by LC-tandem MS? Well, it's a well-known problem that there are significant method-dependent biases in aldosterone radioimmunoassays

And so, when you're doing research and you happen to change vendors for your assay, you're going to see a big shift in the numbers. And we've seen that in our own research work

We've been through two incarnations of radioimmunoassay. And we've moved onto LC-tandem MS. And we have to adjust all of the methods and normalize them into one method in order to do statistical analysis

So, LC-MS offers us a way of standardizing to something that's going to be more predictable and not have interferences from aldosterone-related compounds

The problem with RIAs really became significant when there was no longer an extraction step. They were the so-called homogeneous RIAs. And those are the ones that are used typically today

This has been a well-known problem that the homogeneous RAs have some cross-reactivity with non-aldosterone steroids, probably metabolites of aldosterone, which is a particular problem in people with chronic kidney disease because the metabolite half-life is much longer

This is an example from Pizzolo et al. in Clinical Chemistry 2006, showing--comparing different radioimmunoassay with a chemiluminescent assay

And you can see from this comparison that the two RIAs compare quite well. But, with the chemiluminescent assay, there's a large bias scene

I personally have tried two different RIAs in my own laboratory. And the slope between the two of them was about 0.65

This is another slide from Sherpenbock [sp] et al. in 2006. And what this is showing is a mean aldosterone concentration measured with arrow bars by four different methods, Adaltis, Nichols, DSL, and an in-house radioimmunoassay for aldosterone

And you can see at the time zero point for the same sample the mean aldosterone concentration can be as much as twofold different

So, when you have biases like that and you're trying to adhere to clinical practice guidelines or you're trying to do meaningful comparisons between study specimens at different sites, you have a significant problem with normalization of your results

What are available methods for aldosterone by LC-tandem MS? Well, there's a method from Taylors [sp] on the water symbiosis system involving a protein precipitation and analysis in electrospray mode and electrospray negative mode on the Quattro Premier

There's also a well-known publication from Dr. Solden [sp] requiring online SPE and use of atmospheric pressure photo ionization on the AB SCIEX API 5000

From Ursula Turpeinen in Finland, there's a method for aldosterone determination on the AB SCIEX API 3000 using extraction with dichloromethane diethylether and reconstitution with 50-50 methanol water

Finally, Quest Diagnostics reported an APCI method in positive ion mode in a book called Steroid Analysis by Makin and Gower. And finally, there's a derivatization method by Yoshimata et al

What are our limitations? Many of the listeners will be from countries other than Canada and won't know some of the barriers we face in our public healthcare system

But, we have very significant cost concerns in how we manage healthcare dollars here. So, we didn't want the high disposables cost of solid-phase extraction, whether online or offline

At the time of initiation of the method, we actually didn't have a vacuum system for a Hamilton stylus. And so, offline SPE that would be automated was not going to be a possibility

We did not have an APPI source and the extra pump needed for the toluene dopant. We wanted to have a liquid-liquid extraction approach because that's simple. But, we didn't want to use the chloromethane ether because of the cancer risks associated with DCM and the high vapor pressure of both solvents

So, equipment was always a battle for us. We wanted to do this with the equipment we had, which was an out-of-the-box API 5000 system

So, initially, we tried methyl tert-butyl ether in mixture with methanol in nine to one. But, the presence of the methanol caused the columns to clog. And it was a fairly consistent pattern we were seeing of clogging columns every second run or so

So, we moved to pure MTBE 2,500 microliters with 500 microliters of sample. Now, that's a bit large sample. But, trying to use 250, the LOQ was too high for our work

Using 500 microliters of sample our LOD is about 20, and our LOQ based on a 10 to one signal-to-noise ratio is about 50 picomolar. We vortexed in a 5-millimeter Eppendorf spin and dried down in a 96-well format

We recon with 125 microliters of 20-80 methanol water, which gives us much tighter peaks than 50-50 methanol water

All of our pipetting is automated. And the injection volume is 50 microliters in the LC-MS system

In terms of IS and calibrators, we use d7-aldosterone from IsoSciences, and we made weighed-in calibrators using aldosterone from Sigma spiked into steroid-free serum from Golden West Biologicals

For reference, here are our mass spectrometric parameters so the viewer can copy those down if they want to try this

And here's our solvent gradient. A is basically water with some ammonium acetate. And B is essentially methanol with ammonium and acetate

We're using a Gemini-NX 3-micron C18 column that is 10 centimeters long, and we're using a C18 guard pack to go with that. HPLC is a Shimatsu system

Here, a typical chromatogram from a real patient sample, this patient is about 334 picomolar, which is smack in the middle of the normal range. And you can see that there are a number of interfering shoulders that are chromatographically separated

That shoulder we've determined is probably 18-hydroxycorticosterone. That's what it's looking like when we do spiking experiments

But, chromatography's quite good, even below the limit of quantitation. This is 31 picomolar. And you can see we have a decent peak there. You can see more clearly the interfering chromatographically separated peaks

We have good linearity up to 3,500 picomolar

And we have good precision between 3 and 5 percent in the main portion of the analytical range and somewhere around 10 percent down at the limit of quantitation

In terms of recovery, we did direct mixing experiments with high pools and low pools and got good recoveries

We've done extensive interference studies in coordination with AB SCIEX. And we've only found that 18-hydroxycorticosterone has the potential to interfere. That's that shouldering peak would have to be very high in concentration in order to accomplish that. And we have never observed that in any samples that we've actually run

The qualifier and quantifier peaks compare very well, as do EDTA plasma versus serum

Naturally, we did a comparison with radioimmunoassay, in this case the Siemens Coat-A-Count. And we got a slope of 1.15 with a correlation coefficient of 0.94

So, there is a little bit of bias, both proportional and constant bias, not really surprising, and perhaps more scattered than people would expect. But, there should be no concern

When we did a comparison versus LC-MS at another site using a different method, supported liquid extraction in this case, and their own LC-MS method for aldosterone, also on an API 5000 system, we get a much tighter slope. And we get a coefficient of determination of 0.96

Why is more sensitivity desirable? Well, it'll improve your limits of quantitation. You can use lower sample volumes. And you can use smaller injections. And if you need to, you can do a reinjection

So, there's a number of reasons why we would want to have better sensitivity when we're doing low-concentration steroids like this one

So, we've done some initial investigation on the 6500 system in the determination of aldosterone. You can see on this slide that we can see the aldosterone peak spiked into steroid-free serum down to about 3 picomole

In terms of measurement in peripheral blood, that would handle just about any research interest that anyone could have

We have good linearity up to 1,000 picogram per mL, which is 2,800 picomole per liter and a good correlation coefficient on the calibration curve

But, this is what we are really interested in at level of 40 picomole per liter, 39.6 picomole per liter, which is below our LOQ

You can see that we have about 500 counts per second on the 5000 system. But, we have greater than 2,000 counts per second on the 6500 system. So, that's an improvement of about fourfold

So, in conclusion, LC-MS is an appropriate method of analysis for mineralocorticoid since it avoids all the problems associated with RIA

And in fact, we've done a number of other mineralocorticoids in addition to aldosterone, the most difficult of which in terms of its concentration would be 11- deoxycorticosterone. But, we've also done some of the 18 steroids

We've developed a method of analysis for the 5000 system based on LLE with an LOQ of about 50 picomole per liter

The new AB SCIEX 6500 system should provide a sensitivity of about fourfold and perhaps an LOQ down to the 10- to 20-picomole-per-liter range. That would be in real specimens, not in spiked specimens

The extended linear and dynamic range is up to potentially six orders of magnitude. And that'll afford analysis of very high concentration samples without the need for dilution, which can be a convenient thing

What context, you see very high concentration samples. Neonates on occasion, we've seen up to 50 or 100,000 picomole per liter; samples from the adrenal vein, where we've seen concentrations of up to around 3 or 400,000 picomole per liter; and of course, urine samples, which are in the nanomole per liter range

So, thank you for your attention. And I'm going to hand the talk back over to Michael Jarvis to finish up

Mr. Michael Jarvis: Okay. Well, I'd like to say, first of all, thank you very much to Dr. Holmes for a very interesting presentation on the analysis of mineralocorticoids

And thank you, also, to everybody who has listened into this presentation today. I hope that you found it very informative

And lastly, I just want to put up a slide showing the speaker contact information. If you have any questions about anything that you've heard here today, you can either go to the Website at www.absciex.com, or you can get in touch with the authors of this presentation, either myself, [email protected], or Dr. Holmes at [email protected]

Thank you very much