Just as the Twig Is Bent, the Tree’s Inclined

A time-honored remedy for respiratory symptoms has been to expose the cough sufferer to increased airborne moisture. A simple homemade approach has been the readily available hot shower, an option when one’s child begins to cough and wheeze unexpectedly in the middle of the night. But concerned and conscientious parents of children with chronic or recurrent cough will often invest in small commercial humidifiers meant to be run at a child’s bedside through the night. By the late 20th century, epidemiologists expressed concern when they identified strong associations between bedroom humidifier use and the presence of childhood asthma. Could such an association imply a cause-and-effect relationship? Other associations reported by epidemiologists were almost certainly causative; exposure to secondhand tobacco smoke or the use of household natural gas for cooking have plausible mechanistic roles in producing airway inflammation and injury. Those who espoused a cause-and-effect relationship between humidifier use and childhood asthma observed that dust mites thrive in conditions of high relative humidity as do a variety of molds (1). Indeed, at least one epidemiologic study has reported an association between humidifier use and subsequent new asthma diagnoses (2). However, a precise causal pathway between humidifier use and asthma has been difficult to confirm. Although the theory that humidifier use increases asthma risk by encouraging dust mite exposure remains controversial, all parents and physicians agree that if humidifiers are used in a child’s bedroom, the humidifier must be kept clean. And so once again, we are reminded that “the road to hell is paved with good intentions.”

Between 1994 and 2011, when the products were withdrawn from the market, up to 20 humidifier disinfectants were available in South Korea, and it is believed that up to 8 million people were exposed to these agents, the chief culprit among them being polyhexamethylene guanidine hydrochloride (PHMG) (3). Various pulmonary injuries have been reported with the official death toll set at 130 people as of 2016. It has since been suggested that the actual death toll may be far higher.

In this issue of AnnalsATS, Lee and colleagues (pp. 1523–1532) report their careful examination of children exposed to these toxic humidifier cleaners and subsequently treated for asthma (4). As compared with children treated for asthma without a history of exposure to PHMG, children with a history of low-dose exposure had asthma that was different in terms of airway physiology and inflammatory markers. Children with asthma unrelated to such toxic exposures had typical responses to methacholine, greater airflow limitation on spirometry, and a familiar pattern of type 2 inflammatory markers such as elevated serum periostin levels. Children receiving asthma therapy after prolonged low-dose exposure to PHMG were less likely to exhibit these features. The study by Lee and colleagues takes advantage of an unfortunate but discrete and readily defined childhood exposure to examine the genesis of a distinct asthma phenotype. Their study has broad implications for asthma care and research.

Although the phrase “personalized medicine” is often spoken, it is still uncommon to find evidence of a personalized approach to common respiratory diseases in clinical settings. In the case of asthma, we have begun to characterize phenotypes of severe asthma, an approach driven by highly specific monoclonal interventions (5). But this approach seems limited to the 5 or 10% of asthma sufferers considered to have severe disease; the majority receive a “one-size-fits-all” approach. Post-PHMG asthma may be rare and seldom seen outside South Korea but a more general phenotype of postinjury asthma without type 2 inflammatory biomarkers is likely to be readily identifiable in clinics. A history of early life pulmonary injury and less variable airflow limitation might be the first clue that a patient suffers from such an asthma variant. Could exhaled nitric oxide be used to distinguish postinjury asthma from more commonplace type 2 inflammatory airway disease (6)? This seems plausible but this clinically available measurement was not part of the post-PHMG study.

The authors remind us that early childhood can influence adult lung health. Indeed, prenatal exposure to tobacco smoke, low birth weight, and childhood asthma are well known risk factors for chronic obstructive pulmonary disease in adulthood, a fact largely unknown to the general public and overlooked by most clinicians (7). The lungs have little or no ability to regenerate or return to normal after significant injury whether acute or chronic, yet pulmonologists tackling adult dyspnea seldom look beyond the current and obvious factors to explain the abnormalities they see. Although none of us would minimize the impact of current smoking or pet dander exposure, we are unwise to ignore a lifetime of potentially significant events. The bronchopulmonary dysplasia of prematurity looks like emphysema in the young and breathless adult (8). An episode of respiratory syncytial virus bronchiolitis may make it seem that adult asthma has been accompanied by surprisingly premature “remodeling” (9). Lee and colleagues have described a phenotype of asthma present in a relatively small number of children exposed carelessly to a substance now withdrawn from the marketplace and widespread household use. But the children they describe will almost certainly have different responses to therapy and different outcomes as compared with children with more commonplace asthma phenotypes. Sadly, we have done little to develop this notion of preserving lung health throughout a lifetime. Dentists have been more successful preserving adult dentition than pulmonary doctors have in ensuring adults can continue to breathe comfortably into old age. Except for cystic fibrosis, few pulmonary diseases cross the artificial pediatric/adult divide in terms of our awareness and shared approaches to care.

As thought-provoking as we may find the study by Lee and colleagues, it is not without its limitations. Some might argue with the term asthma as applied to the PHMG-exposed group given that they were less likely to be hyperresponsive to methacholine, had less pronounced obstruction, and had little or no inflammation of the sort associated with early onset allergic asthma in children. In this regard, the investigators were pragmatic and used a working definition of physician-diagnosed and treated asthma. A more important limitation was the small number of subjects studied and the limited number of patient characteristics measured. The conclusions of the study rest heavily on the examination of just 10 PHMG-exposed children. As the authors note, the different and refractory nature of their PHMG-exposed cohort may reflect a self-referral bias by concerned parents. In addition, the physiologic studies were limited to spirometry, there were no reported radiologic studies, and there was no assessment of airway inflammation by quantification of exhaled nitric oxide or sputum eosinophil counts. As an opportunistic study aiming to be minimally intrusive in the pediatric setting, these limitations are understandable. One hopes that these intriguing data will be followed by larger and more detailed studies and that eventually the adult long-term consequences of the finding will be reported.