Schrödinger's Cat Paradox: Bartonella Serology Cannot Be Used to Speciate Bartonella Endocarditis

To the Editor—We read with great interest Ordaya and colleagues' recent article titled “Let the Cat Out of the Heart,” describing the clinical characteristics of 16 cases of Bartonella endocarditis [1]. We congratulate the authors on an intriguing case series that highlights Bartonella endocarditis's association with renal failure, embolization, and PR3-ANCA positivity. Publishing case-level data on Bartonella endocarditis is increasingly relevant in light of the 2023 updates to the modified Duke criteria that include Bartonella diagnostics as a major endocarditis criterion [2].

However, we have concerns that the article, including the witty title, may mislead readers to consider Bartonella henselae as the predominant etiology of Bartonella endocarditis without providing species-level evidence to support this claim. The 16 cases described were diagnosed by Bartonella serologic positivity, with 7 patients confirmed to Bartonella genus level by Bartonella polymerase chain reaction (PCR) on explanted cardiac tissue. Species-specific PCR targets such as 16S rRNA, ribC, rpoB, and gltA genes were not performed [3]. While the authors mention Bartonella species other than B henselae in the introduction, they proceed to focus on B henselae, emphasizing cat exposure and Ctenocephalides flea vectors. “Cat exposure” was reported in 62.5% of the described cases, but the authors failed to define what this exposure meant. Were all “cat exposures” scratches or simply a recollection of a feline in the vicinity? The authors also failed to include important risk factors for Bartonella quintana, such as homelessness, pediculosis, or immigration from a low-income country [4, 5]. This may reflect heuristic, confirmation, and premature closure biases [6]; the authors imply that any “cat exposure” is a mental shortcut for B henselae infection without sufficient information to exclude alternative Bartonella species.

The authors appropriately recognize the limitations of serology, recognizing that Bartonella endocarditis may cause reactive Coxiella burnetii serology. Eight of the 16 patients in this study resulted in false-positive C burnetii serology. Analogously, Bartonella serology lacks specificity and cross-reacts between Bartonella species. Bartonella quintana and B henselae commonly cause serologic positivity for the other species, with similar titers [7]. In Manitoba, a Canadian province that borders Minnesota (jurisdiction of the authors), B quintana was a more common cause of Bartonella serologic positivity than B henselae [7]. One cannot assume Bartonella species based on serology.

In the introduction, Ordaya et al appropriately describe homelessness, human immunodeficiency virus (HIV), and alcohol use as risk factors for B quintana infection. However, only HIV and alcohol use were reported in their Table 1; current homelessness, previous homelessness, and shelter exposure were not described. Moreover, it is increasingly recognized that B quintana is endemic to many low-income countries [5, 8]. Thus, immigration history is an important and often neglected risk factor for B quintana infection that was not described here. Last, B quintana has previously been isolated from cats, and human cases of B quintana have occurred after a feline bite [9, 10]. Cat exposure alone is not sufficient to confirm B henselae among endocarditis cases with Bartonella serologic positivity.

In quantum mechanics, Schrödinger's cat is a thought experiment to demonstrate quantum superposition [11]. In this parable, a cat, a flask of poison, and a radioactive source are sealed in a box. If radioactivity is detected, the poison is released from the flask, eventually killing the cat. After a period of time, until one looks into the box, the cat can be both dead and alive. Only after opening the box can one identify either an alive or dead cat. Using this analogy, if we do not open the box and use molecular testing to identify the Bartonella species, Bartonella endocarditis may be caused by either B henselae or B quintana (or another species) and may or may not be linked to a cat.

Note

Potential conflicts of interest. All authors: No reported conflicts.