Comparison of the Unsaturated Iron-Binding Capacity with Transferrin Saturation as a Screening Test to Detect C282Y Homozygotes for Hemochromatosis in 101 168 Participants in the Hemochromatosis and Iron Overload Screening (HEIRS) Study

The diagnosis of hemochromatosis was previously based on a combination of clinical and laboratory assessments that included history and physical examination, increased transferrin saturation (TS) and serum ferritin, liver biopsy, the amount of iron removed by phlebotomy, and pedigree studies identifying other family members with iron overload (1). Since the discovery of the hemochromatosis gene (HFE) in 1996 (2), most studies from referral centers have shown that >90% of typical hemochromatosis patients are homozygous for the C282Y mutation of the HFE gene (3). Before the availability of DNA-based testing, it was assumed that most hemochromatosis patients have increased TS. However, recent population screening studies incorporating HFE genotyping have now shown that many C282Y homozygotes will have a normal TS and may never develop clinical signs and symptoms related to iron overload (4)(5)(6)(7)(8). TS has been recommended in many studies as the most clinically useful screening test for hemochromatosis because it is widely available and may be increased even in young adults with a genetic predisposition to hemochromatosis. Another potential advantage over DNA-based testing as an initial screening test is that TS may detect many types of iron overload other than those associated with HFE mutations. In addition, screening for iron overload instead of performing DNA-based testing may reduce the risks of potential genetic discrimination that some authors suggest is associated with identification of a C282Y homozygote with normal serum iron tests (9)(10)(11). The TS is a 2-step assay in which serum iron is the numerator and the denominator is either total iron-binding capacity (TIBC), [serum iron + unsaturated iron-bonding capacity (UIBC)] or an adjusted serum transferrin. The UIBC is a 1-step automated colorimetric assay that has been reported to have similar or better operating characteristics than TS for the detection of C282Y homozygotes (12)(13)(14)(15). In this study, UIBC is compared directly with the TS (as measured by serum iron/serum iron + UIBC) for the detection of C282Y homozygotes in a large primary care population. The study design and overall results of the Hemochromatosis and Iron Overload Screening (HEIRS) Study have been reported previously (16)(17). Participants were recruited from 5 field centers that serve ethnically and socioeconomically diverse populations. The study recruited all participants ≥25 years of age who gave informed consent, and it was approved by all local Institutional Review Boards. All participants had nonfasting testing for serum UIBC, serum iron, and serum ferritin and were genotyped for the C282Y and H63D mutations of the HFE gene. All C282Y homozygotes were notified and offered genetic counseling and advice on treatment options. In this analysis, participants who reported a previous diagnosis of hemochromatosis or iron overload, whether reported to be treated or untreated, were excluded because phlebotomy therapy or other interventions potentially could affect the serum TS and UIBC. ROC curves were generated for males and females to compare TS and UIBC as tests for the detection of C282Y homozygotes. Baseline cut points were determined from the intersection of the sensitivity and specificity plots from the ROC curves. Men and women were analyzed separately. TS was calculated from the ratio serum iron/(serum iron + UIBC) and expressed as a percentage. Samples from field centers located in the United States were tested at the Fairview-University Medical Center at the University of Minnesota in Minneapolis, MN, and those from Canada were tested at MDS Laboratory Services, Toronto, Canada. Serum iron and UIBC were measured by a ferrozine-based colorimetric assay on a Hitachi 911 (Fairview-University) or 917 (MDS) with reagents supplied by Roche (Iron Prod #1970743 and UIBC Prod #1030600; Roche Diagnostics Corp.). Internal quality-control pool results from both laboratories are shown in Table 1 of the Data Supplement that accompanies the online version of this Technical Brief at http://www.clinchem.org/content/vol51/issue6/. Method biases were assessed 3 times yearly by use of external proficiency testing samples provided by the College of American Pathologists Surveys (Northfield, IL) and by use of blind replicate samples that were collected from 2% of all participants and analyzed in both laboratories. In addition, comparisons between MDS Laboratory Services and the Central Laboratory were done before the start of testing, and 2% of the MDS samples were repeated at the Central Laboratory throughout the study. Testing for HFE C282Y and H63D alleles was performed with DNA obtained from EDTA–whole-blood samples by a modification of the Invader assay (Third Wave Technologies) that increases the allele-specific fluorescent signal by including 12 cycles of locus-specific PCR before the cleavase reaction (16). ROC curves and 95% confidence intervals (CIs) were generated with S-PLUS (Insightful Inc.). The nonparametric method of Delong for correlated ROC curves was used to compare the curves for UIBC and TS (18). The HEIRS study recruited 101 168 participants from February 2001 through February 2003. A total of 1216 participants were excluded from this analysis because they reported a previous diagnosis of hemochromatosis or iron overload (including 97 C282Y homozygotes). In addition, 47 participants had a missing UIBC. Among the remaining 99 905 participants included in the analysis, 236 undiagnosed C282Y homozygotes were detected (91 men and 145 women). Non-C282Y homozygotes included 37 002 men and 62 667 women. The median age of all participants in this study was 50 years (range, 25–100 years). By self-identified race/ethnicity, the participants included 44% Caucasian, 27% African-American, 13% Asian, 13% Hispanic, 0.7% Pacific Islander, 0.6% Native American, and 2% mixed or unknown race; 97% of the C282Y homozygotes were Caucasian. There were 4 Hispanic and 2 African-American C282Y homozygotes. An increased serum ferritin (>300 μg/L in men, >200 μg/L in women) was found in 88% of the male C282Y homozygotes and in 57% of the female homozygotes. In men, the area under the ROC curve (AUC) was significantly greater for UIBC (0.96; 95% CI, 0.94–0.98) than TS (0.94; 95% CI, 0.90–0.97; P <0.001; Fig. 1A1 ). In women, the AUC for the ROC was also significantly greater for UIBC (0.93; 95% CI, 0.91–0.96) than for TS (0.90; 95% CI, 0.86–0.93; P <0.001; Fig. 1B1 ). In this study we demonstrated that UIBC has a greater AUC than TS for the detection of C282Y homozygotes. The cut points shown in Table 11 were selected as a balance between sensitivity and specificity (Figs. 1 and 2 in the online Data Supplement). These cut points can be raised or lowered to reduce false-positive and false-negative results as appropriate (19)(20)(21). Both TS and UIBC are better for the detection of C282Y homozygotes with an increased serum ferritin than the detection of all unselected C282Y homozygotes. In screening for hemochromatosis, adjusting the cut points to increase the sensitivity (lower TS, higher UIBC) increases the number of C282Y homozygotes detected but also increases the number of non-C282Y homozygotes requiring further evaluation. This follow-up evaluation may include DNA-based testing for HFE mutations. Adjusting the cut points in the other direction (higher TS, lower UIBC) will detect fewer C282Y homozygotes. This may be less important because many such undetected C282Y homozygotes with a normal serum ferritin never develop iron overload (Table 11 ) (8). Both TS and UIBC likely are efficacious in the detection of non-HFE iron overload. However, because confirmation of iron overload in such cases requires liver biopsy or quantitative phlebotomy, it was beyond the scope of this study to determine the operating characteristics of TS and UIBC in the detection of non-HFE iron overload. Preliminary results of other studies of pedigrees with ferroportin mutations have suggested that serum TS is less commonly increased than serum ferritin (22). Possible explanations for the improved performance of UIBC over TS include the reduced analytical error in 1 assay vs 2 assays or the possibility that a circadian rhythm in serum iron affects TS more than UIBC (23). Information collected from both laboratories used in this study has estimated the cost of UIBC testing to be slightly less than TS testing, probably less than US $1.00, depending on how one cost accounts the preanalytical (e.g., specimen collection, processing, and loading of the serum specimen on the automated analyzer) and postanalytical (e.g., results-reporting) steps. The reagent costs are also slightly higher for the 2-step TS analysis compared with the 1-step UIBC analysis. These analytical cost estimates are similar or slightly lower than those reported for previous studies (12)(13)(14)(15). The ROC curves for UIBC and TS have only small differences and could be considered to be almost equivalent, but the advantages of a single test at a lower cost would make UIBC the preferred test. In summary, this study has demonstrated in a large primary care population that the UIBC is a useful test for the detection of C282Y homozygotes. The UIBC has a greater AUC for the ROC curve compared with TS. Because it is a 1-step automated test, which is somewhat less expensive to perform than TS testing, UIBC may be the preferred biochemical screening test for C282Y-linked hemochromatosis. ROC curves comparing UIBC (dashed line) with TS (solid line) in undiagnosed male (A) and female (B) C282Y homozygotes. (A), the AUC was significantly greater for UIBC (0.96) than TS (0.94; P <0.001). (B), the AUC was significantly greater for UIBC (0.93) than TS (0.90; P <0.001). Cut points for UIBC and TS. The cut points represent specificities greater than or equal to 90%, 92.5%, 95%, and 97.5%, respectively, for each group and test. Number of C282Y homozygotes not detected by the TS or UIBC tests who also had increased ferritin. These data are provided to illustrate how many C282Y homozygotes with possible iron overload may have been missed at the various cut points. PPV, positive predictive value; LR+, positive likelihood ratio. Cut points for UIBC and TS. The cut points represent specificities greater than or equal to 90%, 92.5%, 95%, and 97.5%, respectively, for each group and test. Number of C282Y homozygotes not detected by the TS or UIBC tests who also had increased ferritin. These data are provided to illustrate how many C282Y homozygotes with possible iron overload may have been missed at the various cut points. PPV, positive predictive value; LR+, positive likelihood ratio.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.