Between 1950 and 1990 cancer research underwent a radical transformation. Drug screening programs in the 1950s tested single compounds on multiple cancers in heterogeneous patient samples. Yet in the decades that followed the first chemotherapy trials, researchers turned to studies of narrow subcategories of cancer (e.g. metastatic breast cancer in postmenopausal women) to test complex therapies involving surgery, radiotherapy and drug combinations. The shift between these research designs, that are practically mirror images of one another, poses puzzling questions: How has cancer research shifted its structure and orientation? How and when did 'cancer' stop being a meaningful clinical unit? How did multiple cancer diagnoses emerge and come to replace it?
This project examines the structuring of cancer research to explain the multiplication of cancer diagnoses. The researchers hypothesize that the effort to improve anticancer drugs through clinical trials was a mechanism for creating new diagnoses. The introduction of statistical expertise pressured cancer scientists to homogenize patient samples so that they could compare treatment and control groups that were as similar as possible. This compelled researchers to narrow down trials' recruitment criteria and to transform diagnostic schemes. The study examines whether prognostic variables used in clinical trials to predict patients' responses to treatment turned into refined recruitment criteria in subsequent trials and were eventually translated into new diagnoses. The increasing specification of diagnosis limited patient pools and slowed down the recruitment of patients that met inclusion criteria. The researchers thus also hypothesize that diagnosis specification increased the collaboration between institutions that could recruit patients, reinforced the need for standardization and sophisticated statistical tools, allowing researchers to conduct multicenter trials and to work with small patient samples, and reinforced specialization in oncology. The researchers conduct comprehensive archival research and analyze materials from all major institutions involved in cancer research: The National Cancer Institute, The Food and Drug Administration, Medicare, Cooperative Research Groups from across the U.S, The U.S. National Library of Medicine and The Mary Lasker Paper Collection.
This project offers an important contribution both to sociology of diagnosis and to sociology of standardization and evidence-based-medicine. Sociology of diagnosis asks how are new diagnoses created. The researchers suggest a novel mechanism that drives the creation of diagnoses. That is, medical drugs create opportunities and incentives to develop new diagnoses not only through pharmaceutical marketing and through medicalization of previously non-medical conditions, but also that drugs function to extract data that serve to split up patient populations and existing diagnoses. Sociology of evidence-based-medicine asks how clinical trials became the 'gold standard' of biomedicine. This project suggests that the endogenous effects of clinical trials acted to reinforce trials' golden status by fragmenting diagnosis clinical trials restructured cancer research and reinforced the need for standardization and statistical expertise. This study also offers a critical contribution to the understanding of how scientific procedures (homogenization) interact with exclusion and with disparities in the production of medical knowledge about different population groups.
