A DLT Isn’t Just an Adverse Event—It’s One of the Most Important Decisions a Medical Monitor Makes

Medical monitor standing at a crossroads representing a dose-limiting toxicity decision in an oncology clinical trial

One of the biggest misconceptions I see about early-phase oncology trials is that dose-limiting toxicities (DLTs) are simply identified by reading the protocol. If only it were that straightforward.

I’ve sat through countless safety review meetings over the years where the discussion wasn’t about whether an adverse event occurred. Instead, the real discussion centered on something much more nuanced: Does this actually represent a dose-limiting toxicity?

This distinction holds significant importance, as a DLT is not simply a checkbox in a database. It has the potential to influence dose escalation, alter cohort expansion plans, change the recommended Phase 2 dose (RP2D), and even shape the future of an entire development program. That’s a tremendous amount of responsibility resting on what often appears to be a single line in a protocol. 

It’s also why I believe DLT adjudication is one of the most intellectually demanding parts of medical monitoring. Over the years, I’ve learned that identifying DLTs goes beyond memorizing protocol criteria. It requires a deep understanding of the clinical story unfolding before you.

The protocol is the foundation, not the answer

Every protocol includes a DLT definition. Grade 3 laboratory abnormalities, specific hematologic criteria, defined observation windows, recovery requirements, and exceptions are also included. These definitions are important because they ensure consistency across different sites and studies. 

However, they rarely provide all the answers. Two patients can experience the exact same CTCAE grade but arrive at that grade through completely different clinical pathways. For instance, I’ve seen patients with Grade 3 transaminase elevations that resolved almost as quickly as they appeared, with no significant clinical consequences. On the other hand, I’ve also seen patients with seemingly similar laboratory values that were the first indication of evolving immune-mediated hepatitis requiring aggressive intervention. 

Despite the numbers appearing almost identical on paper, the clinical presentations of these patients were very different. That’s where medical judgment comes into play. Medical monitoring has never been about replacing clinical thinking with algorithms. If anything, early-phase oncology demands even more thoughtful interpretation because uncertainty is an inherent part of the job.

The events that deserve your attention often begin quietly

One valuable lesson from experience is that potential DLTs rarely make themselves known. They often appear subtly at first. For instance, a laboratory value that consistently increases over several visits, an investigator mentioning that a patient “just doesn’t seem quite right,” an unplanned extra clinic visit, a dose held for what initially seems like a minor issue, and so on. 

Individually, none of these events necessarily qualify as a DLT. However, when considered collectively, they may indicate the onset of an important safety signal. This is why I strongly encourage medical monitors to continuously review safety data rather than waiting until the end of the DLT observation window. 

Delaying the review often means losing valuable time. Effective medical monitoring is proactive. The goal isn’t simply to determine whether an event met DLT criteria after the fact. Instead, it's to recognize evolving toxicity early enough to support investigators, guide additional evaluations, and ensure that the participant receives appropriate care. This shift in mindset changes everything.

Context matters more than people realize

One of the questions I ask myself whenever a potential DLT emerges is surprisingly straightforward: What else is happening with this patient? The adverse event is just one part of the story. Has the patient recently developed an infection? Were corticosteroids started? Did concomitant medications change? Has disease progression complicated the picture? Is this toxicity biologically plausible based on the mechanism of action? Could another explanation be more fitting? 

These questions can’t be answered in a spreadsheet. They require stepping back and viewing the patient as a whole. I believe this is where experienced medical monitors bring immense value to development teams. We’re not simply reviewing isolated data points. Instead, we integrate laboratory trends, imaging, medical history, pharmacology, investigator observations, and clinical judgment into a coherent narrative. This is very different from simply assigning a CTCAE grade.

Tracking isn’t administrative, it’s strategic

I must admit that not many people are enthusiastic about DLT trackers. However, they should be. Some of the most effective early-phase teams I’ve worked with maintained incredibly disciplined DLT tracking processes. 

Every potential DLT was meticulously documented, every pending adjudication was visible, every observation window was monitored, and every unresolved question had an assigned owner. This level of organization doesn’t just keep a study running smoothly, but it changes the quality of decision-making. 

When a Safety Review Committee (SRC) meets to discuss dose escalation, nobody should be scrambling to reconstruct timelines or searching through email threads to understand whether an event resolved before Day 28. Instead, the focus should be on scientific discussions, not administrative matters. Effective tracking allows the medical discussion to remain precisely where it belongs.

Communication may be the most overlooked skill in medical monitoring

Here’s something we don’t discuss enough: being right isn’t enough. The medical monitor must also communicate effectively. 

A potential DLT may involve conversations with investigators, sponsor physicians, pharmacovigilance teams, clinical operations, data management, and the Safety Review Committee. Each group requires slightly different information. Investigators need practical guidance for patient management. Sponsors want to understand the broader implications for the development program. Clinical operations needs to know if enrollment timelines may change. The Safety Review Committee needs a balanced interpretation supported by data. One poorly communicated DLT can create confusion across an entire study team. 

Conversely, clear communication builds confidence. I’ve found that the most effective medical monitors don’t just present data. They explain their reasoning. They describe why they reached a particular conclusion, what uncertainties remain, and what additional information could change their assessment. That’s what transforms a medical review into a medical opinion.

Experience teaches humility

Over the years, oncology drug development has consistently reinforced the notion that certainty is often an illusion. Not every DLT is obvious, and not every challenging case has a definitive solution. Sometimes, additional information can alter your interpretation even days later. 

Moreover, two experienced physicians may reasonably disagree on a particular decision. This is perfectly acceptable. Medical monitoring is not about presenting every decision as a clear-cut black-and-white scenario. Instead, it involves applying sound clinical judgment, maintaining intellectual honesty, documenting your reasoning, and being open to revisiting conclusions as new information becomes available. 

Ironically, I believe that the willingness to acknowledge uncertainty is one of the most reliable indicators of a physician’s experience.

Final thoughts

People sometimes describe medical monitoring as reviewing safety data. However, I believe our role is to assist development teams in making some of the most important decisions they’ll make throughout a drug’s life cycle. These decisions encompass dose escalation, participant safety, protocol interpretation, and risk assessment. These decisions go beyond a mere literal interpretation of protocol language; they require thoughtful clinical interpretation. 

Ultimately, a DLT is not simply an adverse event; it’s a decision, and decisions of this nature require something that no protocol can fully capture: experience, judgment, and the willingness to see the patient behind the data. This is precisely where medical monitoring makes its most significant impact.

-The Medical Monitor’s Desk 

Previous
Previous

AI Won’t Replace Medical Monitors, But It Will Redefine How We Work.

Next
Next

Medical Monitoring Isn’t Reactive and Treating It That Way Is Risky