Modeling timelines for translational science in cancer; the impact of technological maturation

Abstract

This work examines translational science in cancer based on theories of innovation that posit a relationship between the maturation of technologies and their capacity to generate successful products. We examined the growth of technologies associated with 138 anticancer drugs using an analytical model that identifies the point of initiation of exponential growth and the point at which growth slows as the technology becomes established. Approval of targeted and biological products corresponded with technological maturation, with first approval averaging 14 years after the established point and 44 years after initiation of associated technologies. The lag in cancer drug approvals after the increases in cancer funding and dramatic scientific advances of the 1970s thus reflects predictable timelines of technology maturation. Analytical models of technological maturation may be used for technological forecasting to guide more efficient translation of scientific discoveries into cures.

Fig 1. Quantitative model of the technology growth cycle S-curve.

Growth of research in a specific area is modeled by a best-fit logistic equation (black line) fit to the log of the number of publications identified by a PubMed search. New areas of research emerge into a nascent stage and mature through a near-exponential growing stage before becoming established as limits are encountered. The rate of acceleration of the best fit model (red line) (i.e., d²y/dx²) is used to identify the initiation (Ti), representing the point of maximum exponential acceleration, and when the technology is established (Te), representing the point of maximum exponential deceleration of cumulative publications.

Fig 2

(A) Annual appropriations for the National Cancer Institute (NCI) and publications related to cancer, 1950–2012. NCI appropriations are shown in constant 2012 dollars and exclude supplemental funding from the American Recovery and Reinvestment Act in 2009–2010. Publications on cancer were identified in PubMed.gov using the search term “neoplasms”[MeSH]. (B) Annual approvals of anticancer drugs. Approved drugs are classified as biologic, phenotypic, or targeted based on the composition of matter or method of drug discovery as described[56]. (C) The relationship between publications on cancer (log scale) and annual approvals. Trend lines are shown for 1960–1985 (not significant) and 1985 to the present (p<0.005). Data points 1970 and 1986 are indicated for reference only.

Fig 3. The relationship between drug approval dates and growth of associated technologies.

Timelines for the growth of technologies are shown as blue bars from the initiation point (Ti), through the exponential growing stage, to the point at which the technology is established (Te). The number of approvals associated with each technology is shown for phenotypic drugs (open, orange circles) and targeted or biologic products (closed, green circles). Approval dates for drugs associated with multiple technologies are shown more than once.

Fig 4. Timelines of drug approval and growth of associated technologies.

(A) Years from initiation of associated technology (Ti) and approval of phenotypic drugs (orange bars) and biologic and targeted products (green bars). (B) Years from technologies becoming established (Te) and approval of phenotypic drugs (orange bars) and biologic and targeted products (green bars). Each drug is shown only once in association with the Ti and Te of the lagging technology.