AIMS vs. LC-MS for Cancer Diagnostics: ccRCC Metabolomics with Rachel Wood

What if you could assess a tumor's metastatic potential — without extensive sample prep — directly from tissue? Rachel Wood is working on exactly that.

Rachel Wood is a PhD candidate at Queen's University (Kingston, Ontario), 
working jointly in the labs of Dr. Richard Oleschuk (Chemistry) and Dr. Chris 
Nicoll (Pathology & Molecular Medicine). Her research applies Ambient Ionization Mass Spectrometry — specifically the Liquid Microjunction Surface Sampling Probe  (LMJ-SSP) and Desorption Electrospray Ionization (DESI) — to profile the metabolomics of non-metastatic and metastatic clear cell Renal Cell Carcinoma (ccRCC) cell lines. The long-term goal: a rapid, minimally invasive MS-based prognostic tool for kidney cancer.

This is the first reported use of LMJ-SSP to assess metastatic status in ccRCC — 
and early PCA results are already showing distinct separation between cell line 
profiles.

In this episode, David and Rachel cover:

• Why ccRCC is such a resistant biomarker problem despite active genomic, proteomic, and metabolomic research
• The clinical stakes of "diagnosed by accident" — and what a prognostic biomarker actually needs to do for a clinician
• How AIMS complements (not replaces) LC-MS in a discovery-phase cancer workflow
• LMJ-SSP vs. DESI: what each technique sees that the other misses, including spatial resolution differences (~1mm vs. ~50µm)
• Why Rachel runs LMJ-SSP on both a triple quad and a Q-TOF — and what each instrument contributes to the same research question
• How her lab repurposed a 3D printer as a custom automated XYZ sampling stage with 50µm movement precision
• Managing high-dimensional metabolomics data across two polarities, two cell lines, and two mass analyzers — the pipeline explained
• What unsupervised machine learning (PCA) showed in early results
• The path from cell lines to excised tumor tissue to, ultimately, liquid biopsies
• Advice for analytical chemists considering work at the chemistry–oncology interface

Video Transcription

Kidney cancer remains a difficult biomarker problem, particularly when researchers try to distinguish tumors that are likely to behave aggressively from those that are not. In a recent episode of Concentrating on Chromatography, PhD researcher Rachel Wood discussed how she is combining rapid ambient ionization mass spectrometry, LMJ-SSP, DESI, multivariate data analysis, and conventional LC-MS workflows to investigate the molecular differences associated with clear cell renal cell carcinoma and its metastatic potential.

Returning to the research question that started it all

Wood first encountered kidney cancer research during her undergraduate studies. Although she later moved into a more instrumentation-focused master’s project involving breast cancer, the subject of kidney cancer remained unfinished business.

When the opportunity arose to choose a PhD project, she returned to the field that had originally sparked her interest in research. She is now conducting her doctoral work at Queen’s University in Kingston, Ontario, in an interdisciplinary environment spanning chemistry, pathology, molecular medicine, computer science, and clinical research.

Why clear cell renal cell carcinoma is such a difficult biomarker problem

Wood’s research focuses on clear cell renal cell carcinoma (ccRCC), which represents the majority of kidney cancer cases discussed in the interview.

The difficulty is not a lack of research effort. Genomics, proteomics, and metabolomics studies have all produced promising biomarker candidates. The challenge is that different studies often identify different sets of markers, with relatively little overlap between them.

Patient-to-patient biological variability further complicates the picture. A candidate biomarker panel that performs well in one cohort may not reproduce equally well in another. Selecting which features deserve further investigation therefore becomes a major analytical and biological challenge.

From incidental diagnosis to prognostic information

Another difficulty is the way kidney cancer is often discovered.

Rather than being detected through a dedicated screening program, tumors may be found incidentally during imaging performed for another reason. According to Wood, the ideal solution would be a diagnostic biomarker capable of detecting kidney cancer early. Such a marker is not currently available within the scope of the work she described.

That places additional importance on prognostic biomarkers. Once a tumor has been detected, clinicians need as much information as possible about how aggressive it may be and how likely it is to progress.

Wood’s current work is therefore not attempting to declare directly that a particular patient has metastatic disease. Instead, the goal is to understand what changes biologically as a tumor moves from a less aggressive to a more aggressive state.

That means asking which molecular pathways change, which mechanisms become altered, and which analytes may eventually provide useful indicators of tumor behavior.

AIMS and LC-MS answer different questions

A central theme of the interview was the relationship between rapid ambient ionization mass spectrometry and conventional LC-MS.

Wood is clear that AIMS is not intended to replace LC-MS. Instead, she sees the two approaches as complementary.

AIMS provides a rapid way to obtain a broad molecular overview of a sample with minimal preparation. This makes it especially useful during the discovery phase, when the goal is to gather as much molecular information as possible and identify features that distinguish biological groups.

LC-MS then becomes valuable when those features need to be separated, characterized more carefully, and quantified with greater confidence.

In Wood’s workflow, rapid MS screening can therefore help identify which analytes deserve closer attention before more targeted LC-MS investigations are performed.

What “minimal sample preparation” actually means

The amount of sample preparation required depends strongly on the type of material being analyzed.

Wood currently works mainly with cultured human cell lines, which means much of the preparation begins with growing the cells themselves. Depending on the experiment, the cells may then be chemically lysed or subjected to fixation, paraffin embedding, and sectioning before analysis by LMJ-SSP or DESI.

She also notes that the possibilities would be different for freshly excised tumor material. In an appropriate perioperative or biosafety laboratory setting, LMJ-SSP could potentially be used to analyze tumor margins more directly.

The practical meaning of “minimal preparation” is therefore highly dependent on the sample and the analytical question being asked.

Why combine LMJ-SSP and DESI?

Rather than choosing between LMJ-SSP and DESI, Wood uses the two approaches because they provide different types of information.

LMJ-SSP offers rapid, localized analysis and can accommodate samples with uneven surface topography, which is particularly useful for excised tissue.

DESI, by contrast, provides substantially higher spatial resolution. In the workflow described in the interview, DESI can achieve spatial resolution on the order of tens of micrometers, whereas LMJ-SSP operates at a much larger sampling scale depending on probe design.

That higher spatial resolution is important when moving from simplified cell-line models toward actual tumor tissue, where the molecular organization of different regions may reveal biologically relevant differences.

Combining both techniques therefore enables researchers to balance speed, surface compatibility, and spatial information.

Two mass analyzers for two parts of the same problem

Wood also uses two different mass spectrometers in her LMJ-SSP work.

Most of the discovery work is carried out using a triple quadrupole mass spectrometer, particularly because her project focuses strongly on lower-mass analytes and small molecules.

Once statistically important analytes have been identified, a Q-TOF instrument can be used for MS/MS experiments and tentative structural identification.

The two instruments therefore address different stages of the workflow: one supports broad discovery and comparison, while the other provides additional structural information needed to connect selected features with possible metabolic pathways.

Turning a 3D printer into an automated sampling platform

One of the most unusual elements of the project is the automated sampling system used with LMJ-SSP.

Historically, LMJ-SSP measurements could be performed manually or with custom-built XYZ positioning stages. Wood’s laboratory instead adapted a 3D printer into an automated positioning system.

Custom software controls the printer and moves the sampling probe with precise control over its X, Y, and Z positions. According to Wood, the system enables movements down to approximately 50 µm and also controls parameters such as the contact time between the probe and sample and the delay between consecutive sampling events.

This level of automation improves reproducibility because variables that would otherwise depend on manual operation can be controlled systematically.

Even the Z-position can be programmed, allowing the probe to operate at constant or varying heights depending on the sample surface.

Managing high-dimensional MS data

The experimental design creates a large data-analysis challenge.

Wood is working with multiple cell lines, both ionization polarities, and more than one mass analyzer. To maintain traceability, she analyzes the data in clearly defined subsets.

A typical workflow might begin with one mass analyzer and one polarity. Within that condition, the lower-aggressiveness and higher-aggressiveness cell models can then be compared directly.

The data processing pipeline includes steps such as:

• binning,
• normalization,
• transformation,
• dimensionality reduction,
• identification of discriminating molecular features.

Once relevant features are identified, MS/MS analysis can be used to investigate them further. The same process is then repeated across other polarities and instruments.

Unsupervised machine learning reveals distinct molecular profiles

Machine learning is particularly valuable because the AIMS datasets are highly dimensional, containing very large numbers of molecular features across multiple samples.

Wood’s initial analysis has included principal component analysis (PCA), an unsupervised dimensionality-reduction technique.

The early results are encouraging: the non-metastatic and metastatic cell-line models separate clearly in PCA space.

That does not itself identify a definitive biomarker panel, but it indicates that the two biological states have distinguishable molecular profiles even though both belong to the same kidney cancer subtype.

This provides an important foundation for identifying which specific metabolites or molecular pathways may be responsible for the separation.

Testing LMJ-SSP for metastatic status

A particularly novel aspect of Wood’s project is the use of LMJ-SSP to investigate differences associated with metastatic status.

She had previously worked on distinguishing healthy and cancerous tissues, where the biological differences are relatively large. Comparing two states within the same cancer type is more demanding because the molecular changes may be much subtler.

The scientific motivation, however, is strong. Tumors discovered at later stages carry greater concern for metastatic progression, making it important to understand the molecular changes associated with increasingly aggressive behavior.

For Wood, this made the methodological risk worthwhile: the analytical challenge directly addresses an area where better prognostic information could ultimately have clinical value.

A long-term goal: rapid MS-based prognostic testing

The long-term vision extends beyond cell-line research.

Wood is interested in whether rapid, minimally invasive mass spectrometry could eventually contribute to clinical decision-making.

One possibility would be identifying a diagnostic biomarker in a liquid biopsy, potentially using blood or urine. A urine-based marker capable of contributing to earlier detection would be particularly attractive for kidney cancer.

For prognostic applications, the goal would be slightly different. Once a tumor is discovered, molecular information could help characterize its likely behavior and provide clinicians with additional evidence when deciding between immediate treatment, surgery, or other management strategies.

Wood emphasizes that such a tool would not replace standard clinical workflows. Instead, it would provide additional molecular information to complement existing diagnostic and pathological approaches.

Why interdisciplinary work matters

Throughout the interview, Wood repeatedly returns to the importance of collaboration.

Her research sits at the intersection of analytical chemistry, pathology, oncology, computation, and clinical medicine. No single researcher can remain equally current in all of these fields.

Working closely with pathologists, computer scientists, surgeons, medical students, and other specialists therefore provides both scientific expertise and a way to stay connected with developments outside one’s immediate discipline.

For Wood, the biological context has also changed how she thinks about analytical chemistry. Developing a new instrument or method is important, but understanding who may ultimately benefit from that analytical capability makes the research more meaningful.

That patient-centered perspective is now an integral part of how she approaches the project.

An intersection full of opportunities

Wood describes the interface between analytical chemistry and oncology as an especially exciting area for young researchers.

The work can range from troubleshooting instrumentation to interpreting pathology, developing statistical methods, and collaborating across disciplines. That diversity creates constant opportunities to learn and to apply analytical chemistry to clinically relevant questions.

Her project illustrates that modern mass spectrometry research is increasingly about more than generating spectra. It involves connecting instrumentation, spatial sampling, multivariate statistics, machine learning, biological interpretation, and clinical needs into a single analytical strategy.

And in kidney cancer, where reliable prognostic biomarkers remain difficult to establish, that integrated approach may be particularly valuable.

This text has been automatically transcribed from a video presentation using AI technology. It may contain inaccuracies and is not guaranteed to be 100% correct.

Concentrating on Chromatography Podcast

Dive into the frontiers of chromatography, mass spectrometry, and sample preparation with host David Oliva. Each episode features candid conversations with leading researchers, industry innovators, and passionate scientists who are shaping the future of analytical chemistry. From decoding PFAS detection challenges to exploring the latest in AI-assisted liquid chromatography, this show uncovers practical workflows, sustainability breakthroughs, and the real-world impact of separation science. Whether you’re a chromatographer, lab professional, or researcher you'll discover inspiring content!

You can find Concentrating on Chromatography Podcast in podcast apps:

• Spotify
• Apple podcast
• podscan

and on YouTube channel