Examining the Risk of Cancer and Cancer Mortality in Individuals with Psoriasis

Student thesis: Phd


Aims: This thesis aimed to explore the risk of cancer incidence and mortality in individuals with psoriasis. Methods: Observational studies of the risk of cancer incidence and mortality in individuals with psoriasis were identified through systematic review of six databases. Risk estimates from eligible studies were pooled through Der Simonian-Laird random-effects models. Study heterogeneity was considered through the I2 statistic, and study quality was assessed according to the Newcastle-Ottawa Scale (NOS). Three new cohort studies were undertaken involving participants from the Clinical Practice Research Datalink (CPRD) GOLD and Aurum who were eligible for linkage to hospital episode statistics (HES), Office for National Statistics death registration data (ONS) and the Index of Multiple Deprivation (IMD). Individuals with a record of psoriasis within the study window (01/01/1998 - 30/11/2018) were matched to comparison patients with no previous record of psoriasis based on age, sex, and general practice. Outcomes of interest were any cancer and 26 site-specific cancers. The concordance of cancer recording between linked CPRD, HES and ONS data was assessed for all outcomes. Factors associated with discordance between CPRD and HES were determined through logistic regression. The risk of incident cancer between individuals with psoriasis and matched comparators was determined through the use of Cox proportional hazards models. Cancer mortality risk between the two groups was investigated using flexible parametric models, in order to account for competing risks. Results: 58 unique studies were identified through systematic review - 50 exploring cancer incidence and 15 cancer mortality. The quality of included studies was heterogeneous, with NOS ratings generally greater for electronic health record studies, and lower for studies of cancer incidence compared to those of cancer mortality. Pooled estimates of all cancer incidence were elevated from studies of all psoriasis severities (Relative risk (RR), 1.18; 95% confidence intervals (CI): 1.06, 1.31) and those restricted to moderate-to-severe psoriasis (RR, 1.22; 95% CI: 1.08, 1.39). Elevated incident cancer risk was found at a number of sites, including colon, colorectum, kidney, larynx, liver, lymphoma, non-Hodgkin lymphoma, keratinocyte cancers, oesophagus, oral cavity, and pancreas. Studies of site-specific cancer mortality were infrequent, and limited to those only exploring moderate-to-severe psoriasis. Nonetheless, associations were reported for liver, oesophageal and pancreatic cancers. Notably, the heterogeneity of pooled estimates was often high. 58,904 individuals with psoriasis and 350,592 comparison patients were identified in CPRD GOLD; whereas 213,400 individuals with psoriasis and 1,268,998 comparison patients were included in CPRD Aurum. For any cancer record, concordance between CPRD and HES was greater than 80%. Concordance of cancer recording at the same site was more limited (
Date of Award1 Aug 2022
Original languageEnglish
Awarding Institution
  • The University of Manchester
SupervisorChristopher Griffiths (Supervisor), Darren Ashcroft (Supervisor), Evan Kontopantelis (Supervisor) & Rosa Parisi (Supervisor)


  • Psoriasis
  • Cancer

Cite this