mia (all types considered together), and good for thyroid cancer.5 Survival for pancreatic cancer is so poor that the estimated number of new cases per year can be higher than the estimated number of people surviving at a given time during the year.6

The committee found no broad compilation of data on the prevalence or incidence of rare diseases in the United States. It did, however, locate a recent report from Orphanet that lists estimated European prevalence for almost 2,000 rare diseases (out of an estimated 5,000 to 8,000 such conditions) (Orphanet, 2009). The list has much in common with the NIH list of rare conditions cited in Chapter 1. The demography, living conditions, and other characteristics of Europe and the United States likewise have much in common. Thus, despite the limitations discussed below, the committee believes that the overall portrait of rare diseases prevalence in the Orphanet report is likely to approximate that in this country.

Figures 2-1A-D show the distribution of rare conditions according to prevalence as presented in the Orphanet report. They reveal an overall distribution that is highly skewed to very rare conditions. In fact, data for approximately 1,400 of the approximately 2,000 conditions (about 70 percent) consist only of case reports for individuals or families. For the conditions not included in the study, the distribution may be even more skewed given that the project began with what were thought to be the more common rare conditions (Eurodis, 2005).

In general, the limitations of the data in the Orphanet report include the use of single numbers for conditions with widely varying estimates of prevalence in the literature7 and the lack of bibliographic citations and explanatory details.8 The committee did not systematically check the data presented in the report, but it did note that a few of the listed conditions


SEER identifies four primary types of leukemia: acute lymphocytic leukemia (ALL), acute myeloid leukemia (AML), chronic lymphocytic leukemia (CLL), and chronic myeloid leukemia (CML) but also reports data on several other types (Horner and Ries, 2007).


An NCI working group has defined rare cancers as having an incidence of 40,000 or fewer cases rather than in terms of prevalence (Mikhail, 2005). This specification apparently relates to the specific challenges of clinical research involving populations that include many individuals who have undergone therapies, sometimes multiple therapies.


For example, the prevalence report lists malignant hyperthermia (a rare, life-threatening reaction to certain anesthetics and other agents) with an estimated prevalence of 33 per 100,000 population, but a 2007 study published in the Orphanet journal cites a highly variable incidence from 1/5,000 to 1/50,000–100,000 anesthesia episodes (Rosenberg et al., 2007). For Prader-Willi syndrome, the prevalence report cites a figure of 10.7/100,000, whereas an article in an Orphanet-associated journal cited a range of 1/15,000 to 1/30,000 (Cassidy and Driscoll, 2009).


The report does not include citations of source data but generally cites EMEA (European Medicines Agency), new scientific publications, gray literature, and expert opinion (Orphanet, 2009). Short overview discussions of individual conditions in the Orphanet database vary in the specificity of their citations of sources.

The National Academies of Sciences, Engineering, and Medicine
500 Fifth St. N.W. | Washington, D.C. 20001

Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement