SA Holme, AV Anstey, NM Badminton*, GH Elder*
Departments of Dermatology and *Medical Biochemistry and Immunology, University of Wales, School of Medicine, Heath Park, Cardiff. CF14 4XN
There are no large-scale, prospective studies on erythropoietic protoporphyria (EPP). Previous research examining liver dysfunction in EPP is hard to interpret due to retrospective design of studies, small sample size and/or selection bias. The aim of this UK-wide study was to collate clinical and biochemical features (molecular features, reported elsewhere) in a large cohort of EPP patients in order to better define the condition and to identify possible risk factors for liver disease. Ethical approval was obtained prior to the onset of the study. Patients with biochemically proven EPP were identified from records held at units in Cardiff, Dundee, Leeds, London and Manchester. All study subjects were seen by a single study investigator. 223 EPP patients (114 female; 109 male) were recruited to the study over a six-month period, 41 of whom were siblings and 6 other relations. Median age was 34 years (range 5-87 years). 203 complete blood sample sets were available for analysis. After initial descriptive statistics, six patients’ results were withdrawn from further analysis due to: metastatic malignancy (one colonic carcinoma, one testicular seminoma), liver failure, orthotic liver transplant recipient, and two pronounced alcoholic liver disease. All independent factors were analysed using Spearman’s correlation coefficient for continuous quantitative data, or the Mann-Whitney test for nominal data, to identify significant associations (where significance was taken to be p<0.05, or where the association approached significance, p<0.1). Linear regression analysis was subsequently used to investigate associations between these factors. 34% of individuals were anaemic, affecting similar proportions of males and females (34 females, 36 males), 37% had a low MCV, and 47% a low MCH. A related-samples t-test showed significance between low haemoglobin and MCV (p<0.005) and MCH (p<0.005). Low ferritin was seen in 42% individuals, low serum iron and transferrin saturation in 35%, and a raised TIBC in 17%. Low ferritin (p<0.0005), low serum iron (p=0.027) and low transferrin saturation (p=0.019) were all significantly associated with anaemia. Abnormal liver function, taken to be an elevated ALT, was seen in 15% individuals. It was significantly associated with male gender (p>0.0005), total porphyrin (p=0.016), haemoglobin (p>0.0005), ferritin (p=0.006), albumin (p=0.009) and gGT (p<0.0005).
In summary, this study has identified abnormal liver function in 15% of subjects. Male gender, raised gGT and total porphyrin were all significantly associated with abnormal liver function. More surprisingly, haematological features of iron deficiency anaemia were identified in one third of cases, with no sex difference. The occurrence of iron deficiency anaemia in males is highly unusual, and suggests a fundamental defect in iron metabolism in EPP. However, it is unclear from this study what role, if any, iron is playing in the abnormal liver function seen in EPP.