Friday 26 September 2014
Peanut allergy diagnosis and management: the role of allergen components
Promotor: Prof. dr C.A.F. Bruijnzeel - Koomen
Defence: 26 September 2014
Peanut and soy are both legumes and therefore phylogenetically related. Both peanut and soy should be addressed in the diagnostic work-up of a suspected allergy to either of these foods because of possible cross-reactivity. Allergic symptoms to both foods can range from mild oral allergy symptoms to severe respiratory or even cardiovascular symptoms. Therefore, it is important that peanut or soy allergy is not missed in the diagnostic procedure and an adequate elimination diet is prescribed. An incorrect diagnosis of food allergy on the other hand significantly impairs the quality of life and increases the risk of impaired growth and inadequate nutrient intake in case of children. Skin prick test (SPT) and sIgE reactivity to peanut extract are two methods that are generally used to measure sensitization. Since sensitization is often not clinically relevant, test results are often false-positive when used as a diagnostic test for peanut or soy allergy. The reference standard ! on the other hand, a double-blind, placebo-controlled food challenge, has many disadvantages: it is time-consuming, has high costs, is stressful for the patient, might result in severe clinical reactions and it requires high standard hospital facilities. Therefore, there is strong need for an accurate diagnostic test that is cheap, non-invasive and ideally can discriminate between mild and severe allergy as well. One way of improving diagnostics is by combining several predictors into one prediction model. Another way is by measuring sIgE to specific allergenic peanut proteins, also called components, such as Ara h 1, Ara h 2, Ara h 3, Ara h 6, Ara h 8, and Ara h 9.
In Chapter 2 we systematically reviewed current literature for studies addressing the diagnostic accuracy of the current diagnostic tests (SPT and sIgE to peanut extract) or measuring sIgE to one or more peanut components in diagnosing peanut allergy. All eligible studies were scored for risk of bias and concerns regarding applicability. Our results showed that sIgE to Ara h 2 had the best diagnostic accuracy of all diagnostic test methods that were analyzed. Specific IgE to Ara h 2 was superior in both diagnosing as well as excluding peanut allergy, while SPT and sIgE to peanut extract were primarily useful in excluding peanut allergy.
The diagnostic accuracy of peanut components can be analyzed by different techniques. In Chapter 3 we compared four different techniques with each other; i.e. multi-plexed microarray immunoassay, single-plexed IgE-assay, SPT and immunoblot. All techniques were analyzed in both peanut allergic adults and children. Comparable sensitivity values were found between all four techniques in both adults and children, with the highest sensitivity for Ara h 2. One exception was found however; the multi-plexed assay to Ara h 1 demonstrated a higher sensitivity compared with the other 3 techniques in children, but this might be explained by a technical issue, since two different preparations of Ara h 1 were used in these tests.
In Chapter 4 we validated a previously published, promising diagnostic prediction model in a pediatric population suspected of peanut allergy. This prediction model consisted of 6 predictors: sex, history, SPT and sIgE to peanut extract, total IgE minus sIgE to peanut extract, and age. Validation of the model in 100 subjects showed good discrimination between peanut allergy and peanut tolerance but poor calibration (agreement between predicted and observed frequencies of peanut allergy). We updated the model because of the poor calibration and to analyze whether sIgE to peanut components could increase the diagnostic accuracy of the prediction model. In the updating process, age, history and sIgE to peanut components did not significantly increase discrimination, leaving only 4 predictors of the original model: sex, SPT and sIgE to peanut extract, and total IgE minus sIgE. However, when building a model with only sIgE to peanut components as candidate predictors, Ara h 2 wa! s the only predictor that was left in the model, with a discriminative ability almost equal to the updated prediction model. Cutoff values with 100% positive and negative predictive values could be calculated for both the updated model and sIgE to Ara h 2. In this way, the need for peanut challenges in a pediatric population could be reduced by at least 50%.
Consequently, the diagnostic accuracy of the updated pediatric prediction model was validated in an adult population suspected of peanut allergy in Chapter 5. Validation showed poor discrimination between peanut allergy and peanut tolerance but good calibration. We also showed in this adult population that sIgE to Ara h 2 had the best diagnostic accuracy of all peanut components, and even better than the diagnostic accuracy of sIgE to peanut extract and SPT. It could accurately diagnose peanut allergy in 28% of patients, but could not be used to exclude a peanut allergy in an adult population because of the low negative predictive value.
The diagnostic value of sIgE to peanut components in Chapters 4 and 5 was analyzed by the ImmunoCAP. Specific IgE to Ara h 6 however, can currently only be tested by the ImmunoCAP ISAC. Ara h 6 is a 2S albumin, like Ara h 2. Therefore, we compared the diagnostic value of sIgE to Ara h 2 and sIgE to Ara h 6, as measured by the ImmunoCAP ISAC, in an adult population suspected of peanut allergy. The diagnostic value of sIgE to Ara h 6 on population level was as good as sIgE to Ara h 2. Combining both tests did not result in a better diagnostic accuracy. On individual level, however, 5% of the subjects showed contradicting results between both tests using a cutoff of 0.3 ISU/l, leading to a risk of misdiagnosis if only one of both tests is used.
The diagnostic accuracy of sIgE to soy components in diagnosing soy allergy was addressed in Chapter 7. Adult patients with a suspected soy allergy were included. Specific IgE to Gly m 2S albumin (Gly m 8) had the best diagnostic accuracy, which was comparable to SPT and sIgE to soy extract. All patients were sensitized to either soy extract or Gly m 4. Specific IgE to soy extract, Gly m 5, and Gly m 6 was significantly higher in patients with mild symptoms. Patients only reacting to soy milk products had significantly higher levels of sIgE to Gly m 4.
For risk assessment in food allergy, and to improve labeling guidelines, information on eliciting doses (ED) for peanut on population level is needed. In Chapter 8, we established EDs for objective and subjective symptoms and analyzed the effect of sensitization levels and other patient characteristics on threshold distribution curves. Threshold data from 100 adults and 262 children with a positive food challenge were analyzed and fitted to a threshold distribution curve from which the ED05 value, the value to which 5% of patients react, could be extracted. Subjective and objective threshold distribution curves were different between adults and children, but objective ED05 values were comparable, meaning that threshold data from children and adults can be combined for risk assessment. Higher sIgE levels to Ara h 2 and peanut extract were associated with a larger proportion of both patient groups to react to a certain ED.
In conclusion, this thesis shows that sIgE to Ara h 2 has the best diagnostic accuracy of all diagnostic tests in both adults and children to diagnose peanut allergy. In clinical practice this means that sIgE to Ara h 2 should be measured in all subjects with a suspected history of peanut allergy. In case of unknown prior ingestion, SPT or sIgE to peanut extract can be used as a screening tool to assess whether there is a suspicion of peanut allergy, since both tests have high sensitivity values. Consequently, sIgE to Ara h 2 should be measured in case of a positive SPT or sIgE to peanut extract to assess the risk of peanut allergy. In children, this sIgE to Ara h 2 measurement can be used to accurately diagnose or exclude a peanut allergy, while in adults sIgE to Ara h 2 should primarily be used to accurately diagnose peanut allergy. A negative sIgE result for Ara h 2 in adults does not accurately exclude a possible peanut allergy. Therefore, in case of a negative Ara h 2 o! utcome in adults but a highly suspected peanut allergy, it is advisable to test Ara h 6 and Ara h 8. Although sIgE to Ara h 8 was found to be a poor predictor for peanut allergy, some patients with mild symptoms were monosensitized to Ara h 8. In suspected soy allergy, sIgE to Gly m 2S albumin (Gly m 8) has the best diagnostic accuracy in diagnosing soy allergy, which is comparable to the current SPT and sIgE to soy extract.