This is a classic case control study in that it demonstrates both the strengths and weaknesses of this study design. I was also intrigued with how prenatal factors could possibly influence a cancer that tends to strike adolescents.
Source: Endicott AA, Morimoto LM, Kline CN, et al. Perinatal factors associated with clinical presentation of osteosarcoma in children and adolescents. Pediatr Blood Cancer. 2017;64(6): e26349; doi:10.1002/pbc.26349. See AAP Grand Rounds commentary by Dr. Mary-Jane Staba Hogan (subscription required).
Researchers used California Department of Public Health (CDPH) vital statistics to find 670 patients with osteosarcoma and 2860 control children matched for sex, birth year, and race/ethnicity. Their final analysis showed that higher birthweight was associated with more advanced stage of osteosarcoma as well as greater tumor extension into surrounding tissues and likelihood of distant metastases at the time of diagnosis. Further quantification showed a 1.11-fold increase in the odds of having metastatic disease with each 200 g increase in birthweight.
This is a perfect scenario for a case control study. Osteosarcoma is a fairly uncommon event, only 400 children are diagnosed per year in the United States, so it would be difficult to perform prospective population-based studies. The investigators chose to have 4 control children for each osteosarcoma case, a fairly hefty number. Their analysis is accurate, but of course the data are shaky because of the problems inherent in any case control study. The number of variables available for analysis was limited by what was collected by the CDPH, and the authors also commented that this data could have problems with accuracy.
Another limitation is unique to the biology of this question, and the authors did a good job explaining this. Key to any hypothesis for an association of a disease with particular risk factors is whether the association is biologically plausible. Stated differently, does it make sense based on current understanding of disease pathogenesis that the risk factor (birthweight) could have some bearing on development or severity of disease (osteosarcoma)? In this example, it has been proposed that the connection is insulin-like growth factors (IGFs). Higher birthweight infants may have increased IGF exposure in utero, and IGF also has been shown to act as an in vivo osteoblast chemotactic factor and in vitro mitogen for osteosarcoma growth. The authors note, however, that birthweight is affected by a number of other factors, such as maternal conditions and parental height and weight, and as such is a crude measurement that may mask effects of more significant risk factors. Unfortunately, the CDPH data did not track those other conditions.
So, this case control study mostly points to potential mechanisms for osteosarcoma presentations later in life, which in turn could lead to investigations of pathogenesis that ultimately could be targeted for prevention and treatment of this serious disease.
