December 2, 2025
This GoldCare Zelenko Memorial Grand Rounds opens with a tribute to Dr. Vladimir “Zev” Zelenko—honored for his intelligence, courage, and refusal to bend during the early COVID-19 years, even while facing terminal cancer. In that spirit, the evening turns to another outspoken dissident voice: Canadian pathologist Dr. Roger Hodkinson.
Introduced as a “dissident” who spoke plainly against COVID policies in Canada, Dr. Hodkinson describes himself as a general pathologist—someone who does not claim “massive expertise” in one niche, but sees the whole landscape. That generalist view, he argues, is exactly what is needed to understand how microbiology went off track and how DNA diagnostics might finally fix it.
To frame the problem, Dr. Hodkinson asks us to picture a modern diagnostic lab divided by a figurative dotted line.
On one side: chemistry and hematology—fast, automated, highly accurate, and relatively inexpensive. Emergency departments rely on these tests because they deliver rapid, reliable numbers that can guide urgent decisions.
On the other side: microbiology.
Here, he says, Louis Pasteur and Robert Koch “would feel quite comfortable.” Many methods remain essentially unchanged: urine cultures are still streaked onto seaweed agar; fungal and parasitic work depends on a technologist’s eyes; some tests may take days or even weeks.
According to Dr. Hodkinson, this side of the lab is:
He stresses that the term “microbiology lab” is misleading. In practice, it is “limited bacteriology” with only “lip service” paid to virology, mycology, and parasitology.
Meanwhile, real life is syndromic. Patients do not walk in labeled “bacterial vs viral vs fungal.” They present with diarrhea, sore throat, meningitis, respiratory symptoms. Clinicians make an educated guess and start treatment while waiting days for results that may not even cover all likely pathogens. A few modern tests are faster, some take even longer, but overall, he says, the system is overdue for a complete overhaul.
To move microbiology into the present, Dr. Hodkinson turns to molecular diagnostics. Every pathogen—virus, bacterium, fungus, or parasite—contains DNA or RNA. Within those genetic sequences are small stretches unique to that organism. These “signature sequences,” often as short as 12–20 nucleotides, can serve as unmistakable identifiers.
The challenge is not the concept of looking for these signatures, but the way current technologies try to do it.
In everyday language, almost everyone talks about “PCR tests.” Dr. Hodkinson is careful to point out that PCR itself is not an identification method. Polymerase chain reaction simply multiplies whatever genetic material is present; it “has nothing to do at all with identifying what you’ve multiplied.”
Identification usually comes from probes. He explains them using a simple shape analogy.
Imagine the target genetic segment as one shape and the probe as its complementary partner. When they fit perfectly and bind, a signal is produced—conceptually, a lightbulb turning on. The test reads that signal as a positive result.
The problem arises when the fit is only partial. A probe may still stick to a similar but not identical sequence. The light still goes on. The system records a positive, but now it is a false positive. And crucially, the user has no way to know it is false—they only see that the “lightbulb” lit up.
He also notes a structural limitation: most routine probe-based PCR panels are practically limited to about three, at most four, targets. For complex syndromes with many possible causes, that is a very thin slice of reality. When labs push beyond that number, sensitivity and reproducibility drop.
Another major issue is cross-contamination. Opening tubes in a molecular lab creates aerosols. Material from one patient’s sample can drift into another’s tube. If that happens, the test can produce a positive result for the wrong person, and there is no automatic way to detect it. Cross-contamination, he says, can become a “nightmare” for labs.
Next-generation sequencing (NGS) is often promoted as the cutting edge. Dr. Hodkinson explains that NGS essentially miniaturizes Sanger sequencing and then uses algorithms to piece together countless short fragments.
He acknowledges that NGS is “fantastic” for drug discovery and academic research, but he sees serious problems for routine clinical use:
His conclusion is blunt: NGS is “a solution in search of a problem” when it comes to everyday infectious disease diagnostics. It is powerful technology but not well matched to the needs of ordinary clinical practice.
By contrast, Sanger sequencing has been around for decades and remains the “gold standard” for definitive nucleotide determination. It is the method that reads the actual sequence—the A, G, C, and T letters—of DNA.
The reason it has never become routine in clinical microbiology is simple: traditional Sanger cannot be multiplexed. If multiple targets are sequenced together, the output becomes an indistinguishable “soup” of letters. You cannot tell which letter belongs to which organism. As a result, Sanger sequencing has been limited to one target at a time—too slow and too costly for standard diagnostic panels.
Dr. Hodkinson then describes the technology developed by his company, MultiSeq (multiseq.bio), which aims to solve exactly that problem: multiplexing Sanger sequencing.
The key idea is to separate the different targets by their molecular weight. After the first target is sequenced, the heaviest final nucleotide in that run might have, for example, a molecular weight of 100. The primers for the second target are designed so that the lightest final nucleotide in its run begins at a higher weight—say 110.
By carefully spacing these ranges, the machine sees distinct “cohorts” of sequence peaks, one set after another, each belonging to a different target. In this way, multiple sequences can be read in a single run without blending into a genetic “soup.”
According to Dr. Hodkinson, this approach makes it possible to:
He notes that they believe they can reach around twenty targets in one run, close to the physical capacity of current sequencing instruments. Their first panel, launching from a lab in Greensboro, North Carolina, is designed for upper respiratory tract infections and includes influenza A and B, RSV, COVID-19, and a fifth target.
One of the most striking features he describes is how the system checks for cross-contamination.
Each patient sample is spiked with a unique artificial nucleotide marker. If “Joe’s” sample contains a particular synthetic nucleotide and “Susie’s” sample contains a different one, the lab knows exactly which marker belongs where.
When sequencing Susie’s sample, the system expects to see only her four known nucleotides plus her specific artificial marker. If Joe’s artificial marker appears in Susie’s sequence, that is direct evidence that cross-contamination occurred.
Dr. Hodkinson emphasizes that this allows the lab to prove that a given sequence truly comes from the patient’s own sample, something no other platform has built into the process in this way.
By combining multiplexed Sanger sequencing with liquid-handling robots and modern sequencers, he argues that the whole workflow can be fully automated. Samples can move from arrival in the lab to final result within the typical eight-hour window of a shift—dramatically faster than many current microbiology cultures that may take three days or more.
Cost, he adds, also drops when manual handling is minimized. And because DNA and RNA are universal, the same platform can be applied to bacteria, viruses, fungi, and parasites, as long as there are known signature sequences in public databases like GenBank.
In the second half of the lecture, Dr. Hodkinson turns to another frontier: using circulating DNA in the blood to detect cancer.
He explains how some major centers in the United States already monitor patients after surgery and treatment by looking for known tumor mutations in the bloodstream. Even when imaging is clear and the patient feels well, tiny metastases may be shedding mutated DNA into circulation. Detecting those specific mutations can signal a recurrence long before traditional tests.
Conceptually, he notes, that same principle could apply to people with no prior cancer history—screening apparently healthy individuals for minute amounts of tumor-derived DNA. Companies like GRAIL, with its Galleri test, are moving in that direction using NGS-based methods.
Here, his central concern is accuracy. Labeling a healthy person as having cancer based on a molecular test demands extreme certainty. False positives in this context would be unacceptable. He emphasizes that sequencing-based, verifiable results are crucial before anyone embarks on early treatment based solely on circulating tumor DNA.
He also describes a potential advantage of very early intervention: targeting smaller, less heterogeneous tumors before they become “polyclonal” collections of many different cancer cell types. The goal, he suggests, would be to reduce the tumor burden enough that the immune system can complete the job, rather than relying on chemotherapy alone.
Looking ahead, Dr. Hodkinson describes a scenario in which molecular diagnostics move directly into the physician’s office.
In his example, a patient with a suspected sexually transmitted infection has a sample collected, but instead of shipping it to a distant reference lab, the clinic staff load it into a disease-specific cassette—labeled, for instance, “diarrhea” or “upper respiratory infections”—and insert that cassette into an in-office machine.
About thirty minutes later, the result is ready. The patient, still in the waiting room, is called back and receives a diagnosis and targeted treatment on the spot.
For clinicians, he notes, this could be both a major advance in care and a new revenue stream. Rather than being paid only a small fee to collect and send out samples, practices could perform FDA-approved testing themselves. He quotes a familiar saying from pathology: “He who controls the sample controls the game.”
During the question-and-answer period, concerns are raised about over-reliance on lab data and the dangers of building protocols around single test results. Both Dr. Hodkinson and the moderator, Dr. Richard Amerling, acknowledge that laboratory testing should never stand alone; it must be integrated with clinical evaluation.
Dr. Hodkinson recalls his experience in Canada pushing back against excessive thyroid screening that consumed a large portion of the provincial lab budget without clear benefit. Dr. Amerling highlights how diagnoses like “hypercholesterolemia” have been treated as disease entities based primarily on lab values, feeding massive long-term use of drugs such as statins.
They also reflect on how COVID-era practices—like counting “cases” based purely on PCR results in asymptomatic people—distorted medicine and public policy.
From there, the discussion widens to corruption in medicine: the influence of pharmaceutical companies, the role of journals as marketing vehicles rather than neutral academic platforms, and the way negative trials have often disappeared from view. COVID, they agree, exposed just how deeply these problems run.
Finally, Dr. Hodkinson touches on what he sees as the future: the merging of diagnostics and therapeutics into “theranostics.”
Today, he says, much of prescribing still amounts to a guessing game captured in two words: “try this.” Different people respond differently because their receptors—encoded by their DNA—are not the same.
He envisions a time when, for chronic non-acute conditions, patients would first have their DNA analyzed to clarify which receptor variants they carry. On that basis, clinicians could choose among multiple drugs for the same condition, selecting the one most likely to work for that specific person.
This shift would likely make drugs more expensive to develop and produce, but potentially more effective, with fewer side effects. In his view, understanding the patient’s genome and receptor profiles is central to moving beyond the trial-and-error model that dominates much of current pharmacology.
Throughout the lecture, one theme remains constant: microbiology, as currently practiced, lags far behind the rest of diagnostic medicine.
By enabling accurate, multiplexed Sanger sequencing with built-in contamination controls, Dr. Hodkinson believes it is possible to drag microbiology “kicking and screaming” into the twenty-first century—across infectious disease, oncology, and eventually point-of-care practice.
In the spirit of the Zelenko series, the class does more than show a new technology. It asks a larger question: what happens when a long-trusted corner of medicine finally has to face how outdated, limited, and distorted it has become—and what might change when that is no longer acceptable?
