AML's Seasonal Peak Suggests Viral or Environmental Etiology

Javier Cotelo, MD

August 16, 2022

MADRID, Spain — Most diagnoses of acute myeloid leukemia (AML) are made during January. This finding strongly implies that seasonal factors, such as infectious agents or environmental triggers, influence the development or proliferation of the disease, which points to prevention opportunities. This was the conclusion of an international study led by a team from the Jiménez Díaz Foundation University Hospital Health Research Institute (IIS-FJD) in Madrid, Spain, in collaboration with colleagues from the University of Bristol, United Kingdom. Their work was published in the British Journal of Haematology.

The study's aim was to investigate the potential seasonal and long-term trends in AML diagnosis in an overall population and in subgroups according to sex and age. To do so, the researchers examined 26,472 cases of AML diagnosed in Spain between 2004 and 2015. They found seasonality in the diagnosis of this type of leukemia. This "could point to there being an underlying seasonal etiology at play," noted one of the main authors of the study, Juan Manuel Alonso, MD, a physician in the IIS-FJD's department of hematology and hemotherapy.

"The environmental triggers involved could be radiation, pollution, allergens, or infectious agents like viruses. We're leaning toward viruses, because there are already distinct solid tumor and hematologic cancers that are caused by them and because, in the winter months, there's an increased incidence of cancers due to viral infections," Alonso told Medscape Spanish edition. "The etiological mechanism should be different from that exerted by chronic viral pressure, because here we're dealing with an acute and aggressive disease that probably needs a short incubation period."

Various Hypotheses

Medscape Spanish edition also spoke with David Martínez, MD, a hematologist at La Fe University Hospital in Valencia, Spain, to get his opinion as a specialist who did not participate in the study. "An extremely well done and much-discussed study on AML, a disease that appears to be diagnosed more frequently at a certain time of year — namely, January.

"There's no clear explanation for this finding," Martínez said. "Several possible reasons have been put forward and are being talked about. The one that seems to hold the most water is the hypothesis that infectious agents and environmental factors may have a greater influence. This is because the idea that they're involved in neoplastic diseases is nothing new. In fact, there are a lot of publications and a good amount of scientific evidence that link viral infections and environmental factors with the development of oncologic diseases."

AML is a rare disease yet is responsible for many cancer-related deaths. Mutations that cause AML can occur due to an inherited mutant gene or exposure to certain carcinogens, such as chemotherapy, radiotherapy, ionizing radiation, tobacco, and benzene. These findings are broadly similar to those of a large US-based study by Calip et al., who found a peak of adult AML diagnoses during December and January from 1992 to 2008. Previous smaller studies have provided conflicting evidence, likely due to lower power or to the use of less advanced statistical approaches.

Seasonal Factors Involved?

Demonstration of seasonal variation in the occurrence of AML would, firstly, provide supportive evidence of etiology by seasonal factors, such as infectious agents or environmental factors, and, secondly, focus research onto the etiologic role of such factors.

The current study used population-based data on cases of AML occurring in Spain from a nationwide hospital discharge registry for the years 2004 to 2015. "This is, to our knowledge, the largest study aimed at investigating the potential seasonal and long-term trends in AML incidence in an overall population and in subgroups according to sex and age while employing novel statistical models with serial dependence for discrete-valued time series," wrote the researchers.

They extracted information from the register of each case about the date of admission, discharge date, the anonymous identifier for each patient, International Classification of Diseases (ICD)-9 codes, sex, and date of birth, from which they derived age groups as described for the at-risk population. For patients hospitalized on more than one occasion, only the record corresponding to their first diagnosis of AML was selected.

AML cases per month were standardized to months of equal length.

Age/sex-standardized monthly incidence rates of AML were calculated using the census of Spanish population in 2010 as a “standard” population. Age-standardized and sex-standardized monthly incidence rates of AML were calculated.

Nine separate time-series decompositions were performed as an initial exploratory analysis on the monthly incidence rates of AML using data for all cases and data for each sex and age group. Nine separate Poisson generalized linear autoregressive moving average (GLARMA) models were fitted to evaluate the temporal dynamics in AML incidence using data for all cases, and data for each sex and age group.

Long-Term Trend

A total of 26,472 patients with a first diagnosis of active AML were hospitalized in Spain and registered at the country's Minimum Basic Data Set (CMBD) during 2004-2015. In the end, there were a total of 26,475 patients in the study population; a greater proportion of cases were male (56.0%), and the median age at diagnosis was 67 years.

Seasonal and trend decomposition using Loess decomposition of the incidence rates observed in the overall population exhibited seasonal fluctuation with a peak in January. A slight upward trend was apparent from visual inspection with an upturn in early 2005 and a downturn at the end of 2013. As for the differences by sex groups and age groups, Alonso said, "For both sexes and in age groups 5-19, 20-49, and 50-64 years, we found that the results were identical to those found in the overall population."

The final model included an upward linear long-term trend, as well as the variables monthly seasonality and December 2015. The estimated monthly long-term trend implies that the monthly incidence rates of AML diagnoses annually increased by 0.4% (95% CI, 0.2%-0.6%; P = .0011), given that the other covariates are held constant.

January displayed the highest incidence rate of AML, with a minimum average difference of 7%, when compared with February (95% CI, 2%-12%; P = .0143) and a maximum average difference of 16%, compared with November (95% CI, 11%-21%; P < .0001) and August (95% CI, 10%-21%; P < .0001).

The incidence rate of AML for December 2015 was 0.43 (95% CI, 0.34-0.54; P < .0001) times the average incidence rate for the rest of the study period.

Potential Role of Viruses

"We have to keep in mind that infectious agents (viral infections) and environmental factors (allergens) don't disappear in the warmer months," Martínez added. "There are just other viruses and different factors. We don't know the role or the weight that each one of the factors has, either individually or specifically, in the development of AML. In addition, we know that AML is a very heterogeneous disease and that various factors, including genetic ones, can be involved in its etiopathogenesis."

With respect to the stem cell theory in this leukemia, Alonso emphasized that, "in theory, the virus could fit into it with no problem. That said, any other environmental agent could also produce the described phenomenon where the rapid proliferation of quiescent leukemic stem cells is stimulated, thereby hastening the diagnosis."

"Should the etiological factor be found," Martínez noted, "we can try to reduce exposure and thereby decrease the incidence of AML. On the other hand, discovering how the environmental factor stimulates the proliferation of quiescent leukemic [stem] cells could enhance our knowledge about the regulation of that."

As to whether there is evidence for the involvement of infections in other hematologic malignancies, Martínez reported, "This has already been seen. And this study shows other examples (Epstein-Barr virus and human T-cell lymphotropic virus type 1 with lymphomas), and there could also be  Helicobacter pylori  and lymphomas."

Outside of hematology, human papillomavirus has been associated with cervical cancer, tobacco with lung cancer, sun with skin cancer, and diet with the development of some solid neoplasms.

"The study speaks about the concept of a latency period. To accept the idea that a factor or virus that's more prevalent in winter produces, on its own, AML in a few weeks or months means accepting the idea of a very short latency period — something that's not usually the case. For that, another explanation is given: an abnormal immune response or that a seasonal infectious agent can be capable of promoting leukemogenesis. These are also hypotheses to be explored in the future," suggested Martínez.

New Research Network

Several potential limitations of this study should be considered. One limitation is that AML cases were obtained from the CMBD registry as defined by ICD-9, and no other AML classifications were available. Another limitation is that information on the date of onset of clinical symptoms was not available for analysis. In addition, a further limitation related to the source of their data may have led the researchers to underestimate the incidence rates of AML in older patients, as only hospitalized patients were captured in their study.

As for continuing the research, the results make it necessary to carry out complementary epidemiologic studies that will examine the association between seasonal risk factors and the increased diagnosis of AML during winter months.

To go forward, the first step would be to secure funding. For this purpose, a network is being put together featuring collaborators from other world-renowned research groups that are at the top of their respective disciplines. Through this network, they hope to be able to apply together for public research grants from countries in Europe and elsewhere as well as to establish collaborations with various companies in the private sector.

"This could open up new therapeutic avenues in the future, as we could try to force leukemic stem cells to divide, thereby reducing the resistance that the standard treatments usually demonstrate," concluded Alonso.

Alonso received research funding from Incyte, Pfizer International, and Astellas Pharma outside the present work. Martínez disclosed no relevant financial relationships.

This article was translated from the Medscape Spanish edition.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.