Should the current diagnostic criteria for CLL be reconsidered? A reply to Johnstone et al. “De‐diagnosing chronic lymphocytic leukaemia: An ethical and scientific case for changing diagnostic criteria”

I read with interest the recent article by Johnstone et al., and in particular the sincere and courageous account provided by one of its authors, Peter Allen, Vice‐President of the UK CLL Support Association, who was diagnosed with chronic lymphocytic leukemia (CLL) in 2001 and placed on a “watch and wait” approach. ¹ Allen's account is especially important because it certainly represents only the tip of the iceberg, being one of many unreported cases in which a diagnosis of CLL—at a stage that does not require immediate treatment—nevertheless entails a range of psychological difficulties, not to mention lifestyle changes, professional challenges, health insurance concerns, and related issues. ²

As a physician, I recall encountering a similar situation in 2005. A healthy 46‐year‐old woman with a completely normal complete blood count, who belonged to a family with a history of CLL, was included in my PhD thesis. ³ She had just received a diagnosis of monoclonal B‐cell lymphocytosis (MBL). At that time, there was little information regarding the probabilities of progression to CLL, and my supervisor and I spent approximately two weeks deliberating whether we should notify her of the diagnosis. Ultimately, because research participants should be informed of research results, including newly discovered risks and benefits, ⁴ we decided to disclose the diagnosis to her. In the subsequent years, she was followed at our center, and although she was repeatedly informed that no treatment was required—since she was considered not to have a disease but rather a condition that might develop into one—we observed that she was almost invariably distressed and apprehensive as a result of the diagnosis she had received.

Based on situations such as these, Johnstone et al. argue that the current diagnostic paradigm for CLL (and, a fortiori, for MBL, as I would add) constitutes a form of overdiagnosis, insofar as it imposes a diagnosis that may cause significant distress to patients without conferring corresponding clinical benefit. They therefore propose revising the diagnostic threshold for CLL to a clonal B‐cell count of >10 × 10⁹/L, on the grounds that this size of the B‐cell population is a better predictor of treatment‐free survival (TFS) and overall survival (OS). ⁵ While I agree that the scenario described by the authors calls for serious and careful reflection on the current diagnostic criteria for CLL, I would like to offer some considerations that weigh in favor of the other side of the balance, suggesting that, at present, the existing criteria—that is, a clonal B‐cell count of >5 × 10⁹/L—should not be modified. ⁶

The diagnosis of CLL, whether in what might be called its qualitative aspects (immunophenotypic profile) or in its quantitative aspects (B‐cell count), is not free from difficulties. ⁷ , ⁸ , ⁹ With regard to this last issue, the landmark study by Shanafelt et al. revealed that the B‐cell threshold that best predicted TFS and OS, in subjects with a clonal population of cells of CLL phenotype, was 11 × 10⁹/L. Notably, the current diagnostic criteria (>5 × 10⁹/L) was also able to predict TFS (although it did not predict OS), but its statistic for TFS was slightly inferior to that observed with the 11 × 10⁹/L threshold. ⁵ This is undoubtedly an argument in favor of changing the current B‐cell count for the diagnosis of CLL. Yet, I would like to point out that it remains unclear to what extent reclassifying a patient with CLL as having MBL would effectively reduce emotional distress, which is the main reason advanced by Johnstone and collaborators. ¹ Indeed, individuals diagnosed with MBL also experience a substantial psychological impact upon learning that their condition carries a potential risk of progression to leukemia. ¹⁰ In short, it is open to question whether the label MBL is psychologically less distressing than that of CLL.

Another aspect that warrants discussion is the suggestion of “intentional non‐investigation for some individuals with asymptomatic lymphocytosis.” ¹ Although the decision not to investigate certain clinical situations may at times be justified, I am not sure whether this is the appropriate strategy in cases of asymptomatic lymphocytosis. I have some reasons for this view. First, it cannot be assumed that all cases of asymptomatic lymphocytosis will display a CLL immunophenotype and therefore receive a diagnosis of CLL or CLL‐like MBL. Some cases may instead correspond to non‐CLL/SLL‐type MBL, which requires thorough evaluation to exclude, for example, the possibility of peripheral blood involvement by non‐Hodgkin lymphomas. ¹⁰ , ¹¹ Second, individuals with MBL and CLL have an increased risk of infection and secondary malignancies, such as lung cancer and melanoma. ¹⁰ , ¹¹ , ¹² , ¹³ Failure to establish the diagnosis implies that these subjects will not be enrolled in appropriate surveillance and screening programs. Third, in an era in which medical information is readily accessible on the internet—a trend further intensified by the advent of generative artificial intelligence (AI)—it is difficult to imagine that an apparently healthy subject who receives a “routine” blood test showing lymphocytosis would not immediately seek guidance from the many freely available AI tools. ¹⁴ It is even more difficult to conceive that, when among the various AI‐generated responses (such as influenza, COVID‐19, or tuberculosis) the possibility of MBL or CLL appears, any medical professional could persuade that person not to pursue further investigation into the cause of the lymphocytosis.

All this said, I would not claim that my view represents the best possible approach in such scenarios. However, I do believe that there are serious considerations that support it, and I offer it in that spirit. Ultimately, issues of clinical validity (that is, how well a diagnosis of MBL or CLL predicts a clinically meaningful outcome) and clinical utility (that is, the likelihood that knowledge of an MBL or CLL diagnosis will lead to an improvement in the patient's well‐being—for example, through prevention of other diseases) constitute the core bioethical considerations at stake. ¹⁵

More importantly, the outstanding article by Johnstone et al. clearly opens a broad range of questions that, I believe, remain unanswered. In this article, I merely address other aspects of the problem.

AUTHOR CONTRIBUTIONS

Daniel Mazza Matos: Conceptualization; writing—original draft.

CONFLICT OF INTEREST STATEMENT

The author declares no conflict of interest.

FUNDING

This research received no funding.

DATA AVAILABILITY STATEMENT

Data sharing not applicable to this article as no datasets were generated or analysed during the current study.