D3.86 - ALEX versus ISAC. The same story or an improvement for allergologic molecular diagnostic?
Background
ALEX2 and ISAC are widely used multiplex IgE platforms, but head-to-head agreement across shared targets needs clarification.
Method
Retrospective paired study including 97 patients tested with both ALEX2 and ISAC. Positivity was defined as IgE ≥0.3. Analyses were restricted to allergens shared by both platforms and predefined source categories (mites, pollens, moulds, epithelia/latex, milk, egg, fish, shellfish, Anisakis, cereals, legumes, nuts/seeds, fruits/vegetables). For each platform and target, 2x2 tables versus clinician-established diagnosis were constructed to derive sensitivity, specificity, PPV and NPV with exact 95% confidence intervals (Clopper–Pearson). ALEX2 versus ISAC results were compared using the exact McNemar test; paired odds ratios were reported. Agreement was quantified with Cohen’s kappa. Discordant cases were reviewed to identify drivers of platform disagreement. All analysis were performed in IBM SPSS statistics 25.
Results
Performance varied by allergen and source. Paired McNemar comparisons identified asymmetric positivity for 7 shared targets. ISAC yielded more positives for Act d 2 (OR 0.043, 95%CI 0.003–0.738), Jug r 3 [LTP] (OR 0.077, 95%CI 0.010–0.588), Bet v 2 [profilin] (OR 0.059, 95%CI 0.003–1.019), Der f 1 (OR 0.059, 95%CI 0.003–1.019), Cup a 1 (OR 0.067, 95%CI 0.004–1.167) and Cyn d 1 (OR 0.067, 95%CI 0.004–1.167). Conversely, ALEX2 yielded more positives for Cor a 8 [LTP] (OR 13.000, 95%CI 0.732–230.775). At source level, epithelia/latex favoured ISAC (OR 0.077, 95%CI 0.004–1.366). Overall discordance was mainly driven by aeroallergen components and LTP-related markers.
| Target | ALEX Se | ALEX Sp | ALEX PPV | ALEX NPV | ISAC Se | ISAC Sp | ISAC PPV | ISAC PPN | Direction | OR | 95% IC | p (McNemar) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Act d 2 | 0.0% | 97.6% | 0.0% | 86.3% | 23.1% | 88.1% | 23.1% | 88.1% | ISAC>ALEX | 0.043 | 0.003–0.738 | <0.001 |
| Jug r 3 | 26.7% | 98.5% | 88.9% |
75.0% |
60.0% | 95.5% | 85.7% | 84.2% | ISAC>ALEX | 0.077 | 0.010–0.588 | 0.0018 |
| Bet v 2 | 37.5% | 88.8% | 23.1% | 94.0% | 66.7% | 83.0% | 28.6% | 96.1% | ISAC>ALEX | 0.059 | 0.003–1.019 | 0.0078 |
| Der f 1 | 38.5% | 81.7% | 43.5% | 78.4% | 57.7% | 77.5% | 48.4% | 83.3% | ISAC>ALEX | 0.059 | 0.003–1.019 | 0.0156 |
| Cup a 1 | 62.5% | 77.5% | 20.0% | 95.8% | 66.7% | 70.5% | 18.8% | 95.4% | ISAC>ALEX | 0.067 | 0.004–1.167 | 0.0313 |
| Cyn d 1 | 62.5% | 69.7% | 15.6% | 95.4% | 77.8% | 63.6% | 17.9% | 96.6% | ISAC>ALEX | 0.067 | 0.004–1.167 | 0.0313 |
| Cor a 8 | 56.7% | 97.0% | 89.5% | 83.3% | 36.7% | 97.0% | 84.6% | 77.4% | ALEX>ISAC | 13.000 | 0.732–230.775 | 0.0313 |
| Epithelia/latex | 90.0% | 67.5% | 41.9% | 96.3% | 95.0% | 61.0% | 38.8% | 97.9% | ISAC>ALEX | 0.077 | 0.004–1.366 | 0.0313 |
Conclusion
In this cohort of 97 paired patients, ALEX2 and ISAC produced broadly similar overall profiles, but some allergens showed clear platform differences. Findings support platform-aware interpretation since specific components, such as Act d 2 and Jug r 3 for ISAC and Cor a 8 for ALEX2, exhibit critical variations in their sensitivity and positivity rates that can impact clinical diagnosis.
