- Email: [email protected]
Mold testing by RAST and skin test methods in patients with allergic fungal sinusitis
Mold testing by RAST and skin test methods in patients with allergic fungal sinusitis RICHARD L. MABRY, MD, FAAOA, BRADLEY F. MARPLE, MD, and CYNTHIA S. MABRY, RN, BSN, CORLN, Dallas, Texas
RAST tests have traditionally been considered less sensitive than skin tests during investigation of atopy involving molds. This has been attributed to technical problems such as difficulty in binding the mold antigen to the carrier substrate. Ten patients with proven allergic fungal sinusitis were evaluated for sensitivity to 11 important molds by both RAST and dilutional intradermal testing. A predictable correlation between RAST and skin test scores was observed in many, but not all, cases. Most often this disparity was in the form of greater sensitivity indicated by skin testing than by RAST, sometimes differing by as many as 3 classes. The lack of concordance was not confined to testing for the fungi cultured from the sinuses, nor was it more or less pronounced in the case of dematiaceous fungi. The most likely causes for the disparity noted in this series are subtle differences in antigens used in skin test material and for RAST standards. Skin tests allow for evaluation of delayed and late-phase reactions, a measurement not possible by specific IgE testing with RAST. Delayed skin test reactions were not noted in this series of patients. An additional important finding was the sensitivity of patients with allergic fungal sinusitis to virtually every fungal antigen to which they were tested. (Otolaryngol Head Neck Surg 1999;121:252-4.)
It is now generally accepted that allergic fungal sinusitis (AFS) represents a problem that is atopic rather than infectious.1 One treatment modality that continues to show great promise in the management of this recalcitrant disorder is immunotherapy with fungal antigens
From the Department of Otorhinolaryngology, University of Texas Southwestern Medical Center. Winner of the 1998 Sam Sanders Basic Science Award, given by the American Academy of Otolaryngic Allergy. Presented at the Annual Meeting of the American Academy of Otolaryngic Allergy, San Antonio, TX, September 10-12, 1998. Reprint requests: Richard L. Mabry, MD, FAAOA, Department of Otorhinolaryngology, UTSMC, 5323 Harry Hines Blvd, Dallas, TX 75235-9035. Copyright © 1999 by the American Academy of Otolaryngology– Head and Neck Surgery Foundation, Inc. 0194-5998/99/$8.00 + 0 23/1/93873 252
(ie, molds).2 Such immunotherapy is obviously dependent on accurate testing to determine the sensitivity of each patient to relevant fungal organisms. It has been postulated that AFS represents both Gell and Coombs type I and III hypersensitivity. However, the exact nature of the immunologic process involved, in both the production of disease and the benefits provided by immunotherapy, remains unproven and a matter of conjecture.3 In an attempt to shed further light on the process, we compared sensitivity levels obtained by dilutional intradermal testing (skin end-point titration [SET]) and RAST for 11 fungal antigens in 10 patients with AFS. METHODS Within the past year, 10 consecutive patients meeting the criteria for AFS4 were tested by both RAST and SET for 11 relevant fungal antigens. All molds of significance in allergy are classed as fungi imperfecta. Although AFS is generally caused by members of the darkly pigmented, or dematiaceous, fungi, we chose to test for the molds that, in our experience and that of others,5 were most often positive in patients with mold allergy of any type. RAST was performed by the Fadal-Nalebuff modified RAST method. Because Bipolaris antigen is not commercially available, we continued our practice of using Helminthosporium antigen for this purpose because mycologic taxonomy indicates that these 2 organisms may be the same. Antigens were obtained from the Meridian Division of ALK Corp and were prepared in sequential 5-fold dilutions according to the usual procedure for SET. Skin testing was performed initially at a strength one 5-fold dilution weaker than the corresponding RAST score (RAST – 1 level), along with the application of appropriate positive and negative controls. Additional tests were applied as necessary to exactly define an end point.6 Skin test reactions were read at 10 minutes. Patients were subsequently contacted 24 hours after testing to evaluate for delayed skin reactions. RESULTS
Ten patients were tested for 11 fungal antigens. The antigens tested and the concordance between RAST and SET scores are reflected in Table 1. No positive latephase skin reactions were identified in this series of patients. A striking finding was that all patients tested demon-
Otolaryngology– Head and Neck Surgery Volume 121 Number 3
MABRY et al
Table 1. SET end point compared with RAST class
Antigen
Under
Same
Dematiaceous fungi Alternaria Cladosporium* Helminthosporium Stemphylium Pullularia Epicoccum Curvularia Penicillium Aspergillus Mucor Fusarium
1 1 — 1 — 1 — — — — —
3 2 3 3 1 1 — 1 2 3 3
*In
1 2 ≥3 class classes classes over over over
4 5 7 3 4 4 — 8 4 4 1
2 — — 2 2 1 4 — 4 3 2
— — — 1 3 — 6 1 — — 1
2 patients negative for Cladosporium by RAST, SET was not done.
Table 2. Number of positive RAST responses per class Class
No. of responses
0 0/I I II III IV V
2 8 28 37 21 13 1
strated some degree of sensitivity to almost every fungal antigen tested. Ten patients were subjected to RAST for 11 fungal antigens. Of these 110 tests, only 2 were class 0, and 8 others were class 0/I (borderline) (Table 2). The distribution of positive responses was heaviest in the midrange (class I through III), with only 1 class V response. Fungal cultures were done in all patients. Seven of the 10 were positive: 3 for Curvularia, 3 for Bipolaris, and 1 for Aspergillus. No consistent alteration in either RAST or skin test results was noted for the particular fungus cultured. DISCUSSION
When we initially embarked on this protocol of treating patients with AFS with both relevant fungal and relevant nonfungal antigens, we used skin testing for evaluation of the degree of sensitivity to molds. We chose to do this because conventional wisdom had indicated that RAST testing yielded lower scores for molds than did skin tests. Furthermore, because RAST measures only allergen-specific IgE, whereas there have been sugges-
253
Table 3. SET levels within 1 class of RAST score Antigen
%
Cladosporium Helminthosporium Penicillium Alternaria Mucor Stemphylium Aspergillus Epicoccum Fusarium Pullularia Curvularia
100 100 90 80 70 70 60 60 50 50 0
Table 4. SET levels 3 or more classes higher than RAST score Antigen
%
Curvularia Pullularia Fusarium Penicillium Stemphylium Epicoccum
60 30 20 10 10 10
tions that skin tests may also measure other factors, we believed it prudent to use this method. Finally, delayed hypersensitivity to molds is frequently described and can be evaluated only by skin testing, not by RAST. Although in this very small series we found no patients with delayed reactions, we believe it is still important to look for these reactions because, when present, they may necessitate significant alterations in a dosage advancement schedule. It is generally accepted that SET results will parallel Fadal-Nalebuff modified RAST results, with SET being slightly more sensitive than RAST. However, this relationship has been found to be highly variable, more so for molds than other antigens.7 When immunotherapy is based on skin testing, the antigens used in both testing and treatment are identical, coming from the same stock vial. Unfortunately, this relationship does not exist for immunotherapy based on in vitro testing. For these reasons, in vitro–based treatment must always be preceded by some form of bioassay or vial test.8 This is necessary to establish the exact relationship of the antigens being injected to those used on the disk or other carrier used in the in vitro test. Although prolonged usage of a given in vitro system may provide the clinician sufficient familiarity to construct a correlation table between in vitro results and anticipated SET levels for various anti-
254
Otolaryngology– Head and Neck Surgery September 1999
MABRY et al
gens, a vial test of some sort is always necessary when immunotherapy commences. In most patients in this series, the results of RAST and SET are parallel. In a few instances, however, a rather significant difference between the 2 tests was noted (Tables 3 and 4), in the form of underclassification (or less often overclassification) by RAST. This lack of concordance was not limited to (or even apparently associated with) the particular fungus cultured from the sinuses, nor was it attributable to altered sensitivity to only dematiaceous fungi, which are the causative organisms in AFS. Although it is tempting to postulate that patients with AFS somehow behave differently than patients with uncomplicated allergy, the more likely conclusion is that the disparity is simply related to a difference in antigens between those used in RAST and those used in skin tests. This difference appears to be more marked for some antigens than for others. In a previous paper dealing with testing patients with allergy (but not AFS), a tendency for RAST to overclassify (ie, indicate a greater sensitivity than reflected by SET) was noted in a very small proportion of patients.9 In that series of 53 patients, sensitivity for Helminthosporium was overclassified by RAST in almost 10% of patients, overclassification for Cladosporium was 4%, and sensitivity to Alternaria was underclassified and overclassified in equal proportions (2.7% each). In our study overclassification was rare. However, underclassification by RAST of sensitivity to a number of fungal antigens occurred. The most striking example is Curvularia, in which the RAST showed a degree of sensitivity that was 2 or more classes lower than the SET end point in all 10 patients. This series presents an exceedingly small sample, and obviously no statistical conclusions can be drawn from it. However, on the basis of this series, one logical inference is that different commercially available fungal antigens have a rather variable relationship to the antigens bound to a carrier substrate for RAST testing. Furthermore, it is certainly possible that, as postulated by others in the past, skin tests (especially for molds) depend on mechanisms that are not simply IgE-mediated atopy. Our findings in this small series certainly indicate the need for further investigation of the relationship between skin test and in vitro results in testing for sensitivity to molds, not only in patients with AFS but in all situations. Another finding that should not be overlooked is that all of these patients with AFS were sensitive to virtually every fungal antigen with which they were tested, both by RAST and skin tests. Cross-reactivity between fungal antigens is not well understood, but it
is generally considered to be poor. Believing that decreasing the total allergic load should be a goal of therapy in the management of AFS, we find that sensitivity to multiple fungal antigens supports our decision to test for and treat all relevant fungal antigens, not just the particular organism that might have been cultured from the sinuses. Chrzanowski et al10 have shown recognition by Western blot of an 18-kD protein in 11 patients with AFS, which they suggest may represent a fungal panallergen. It is conceivable that patients with AFS may have a unique hypersensitivity to this protein, explaining their sensitivity to virtually all fungal antigens for which they are tested. We eagerly anticipate seeing the results of further work in this area. CONCLUSION
Based on a small sample, it appears that there is not always a consistent relationship between levels of sensitivity to molds reflected in skin testing (with dilutional intradermal tests) and RAST in patients with AFS. This lack of concordance seems to vary among differing antigens. At present, although RAST testing remains an excellent screening modality, some type of quantitative skin testing is necessary for these antigens before immunotherapy is begun. As is often the case, the results in this small preliminary series indicate the need for further study in a larger series of patients. We hope that this article will stimulate others to add their experience to ours. REFERENCES 1. DeShazo RD, Chapin K, Swain RE. Fungal sinusitis. N Engl J Med 1997;337:254-9. 2. Mabry RL, Marple BF, Folker RJ, et al. Immunotherapy for allergic fungal sinusitis: three years’ experience. Otolaryngol Head Neck Surg 1998;119:648-51. 3. Mabry RL, Marple BF. Allergic fungal sinusitis. Curr Opin Otolaryngol Head Neck Surg 1997;5:49-52. 4. Bent JP III, Kuhn FA. Diagnosis of allergic fungal sinusitis. Otolaryngol Head Neck Surg 1994;111:580-8. 5. King WP. Clinical significance of molds newly available for radioallergosorbent testing. Otolaryngol Head Neck Surg 1989; 101:1-4. 6. King HC, Mabry RL, Mabry CS. Testing methods for inhalant allergy. In: Allergy in ENT practice: basic guide. New York: Thieme Medical Publishers; 1998. p. 103-54. 7. Corey JP, Liudahl JJ, Yound SA, et al. Diagnostic efficacy of in vitro methods vs skin testing in patients with inhalant allergies. Otolaryngol Head Neck Surg 1991;104:299-302. 8. Mabry RL. Blending skin endpoint titration and in vitro methods in clinical practice. Otolaryngol Clin North Am 1992;25:61-70. 9. Tandy JR, Mabry RL, Mabry CS. Correlation of modified radioallergosorbent test scores and skin test results. Otolaryngol Head Neck Surg 1996;115:42-5. 10. Chrzanowski RR, Rupp NT, Kuhn FA, et al. Allergenic fungi in allergic fungal sinusitis. Ann Allergy Asthma Immunol 1997;79: 431-5.
RAST tests have traditionally been considered less sensitive than skin tests during investigation of atopy involving molds. This has been attributed to technical problems such as difficulty in binding the mold antigen to the carrier substrate. Ten patients with proven allergic fungal sinusitis were evaluated for sensitivity to 11 important molds by both RAST and dilutional intradermal testing. A predictable correlation between RAST and skin test scores was observed in many, but not all, cases. Most often this disparity was in the form of greater sensitivity indicated by skin testing than by RAST, sometimes differing by as many as 3 classes. The lack of concordance was not confined to testing for the fungi cultured from the sinuses, nor was it more or less pronounced in the case of dematiaceous fungi. The most likely causes for the disparity noted in this series are subtle differences in antigens used in skin test material and for RAST standards. Skin tests allow for evaluation of delayed and late-phase reactions, a measurement not possible by specific IgE testing with RAST. Delayed skin test reactions were not noted in this series of patients. An additional important finding was the sensitivity of patients with allergic fungal sinusitis to virtually every fungal antigen to which they were tested. (Otolaryngol Head Neck Surg 1999;121:252-4.)
It is now generally accepted that allergic fungal sinusitis (AFS) represents a problem that is atopic rather than infectious.1 One treatment modality that continues to show great promise in the management of this recalcitrant disorder is immunotherapy with fungal antigens
From the Department of Otorhinolaryngology, University of Texas Southwestern Medical Center. Winner of the 1998 Sam Sanders Basic Science Award, given by the American Academy of Otolaryngic Allergy. Presented at the Annual Meeting of the American Academy of Otolaryngic Allergy, San Antonio, TX, September 10-12, 1998. Reprint requests: Richard L. Mabry, MD, FAAOA, Department of Otorhinolaryngology, UTSMC, 5323 Harry Hines Blvd, Dallas, TX 75235-9035. Copyright © 1999 by the American Academy of Otolaryngology– Head and Neck Surgery Foundation, Inc. 0194-5998/99/$8.00 + 0 23/1/93873 252
(ie, molds).2 Such immunotherapy is obviously dependent on accurate testing to determine the sensitivity of each patient to relevant fungal organisms. It has been postulated that AFS represents both Gell and Coombs type I and III hypersensitivity. However, the exact nature of the immunologic process involved, in both the production of disease and the benefits provided by immunotherapy, remains unproven and a matter of conjecture.3 In an attempt to shed further light on the process, we compared sensitivity levels obtained by dilutional intradermal testing (skin end-point titration [SET]) and RAST for 11 fungal antigens in 10 patients with AFS. METHODS Within the past year, 10 consecutive patients meeting the criteria for AFS4 were tested by both RAST and SET for 11 relevant fungal antigens. All molds of significance in allergy are classed as fungi imperfecta. Although AFS is generally caused by members of the darkly pigmented, or dematiaceous, fungi, we chose to test for the molds that, in our experience and that of others,5 were most often positive in patients with mold allergy of any type. RAST was performed by the Fadal-Nalebuff modified RAST method. Because Bipolaris antigen is not commercially available, we continued our practice of using Helminthosporium antigen for this purpose because mycologic taxonomy indicates that these 2 organisms may be the same. Antigens were obtained from the Meridian Division of ALK Corp and were prepared in sequential 5-fold dilutions according to the usual procedure for SET. Skin testing was performed initially at a strength one 5-fold dilution weaker than the corresponding RAST score (RAST – 1 level), along with the application of appropriate positive and negative controls. Additional tests were applied as necessary to exactly define an end point.6 Skin test reactions were read at 10 minutes. Patients were subsequently contacted 24 hours after testing to evaluate for delayed skin reactions. RESULTS
Ten patients were tested for 11 fungal antigens. The antigens tested and the concordance between RAST and SET scores are reflected in Table 1. No positive latephase skin reactions were identified in this series of patients. A striking finding was that all patients tested demon-
Otolaryngology– Head and Neck Surgery Volume 121 Number 3
MABRY et al
Table 1. SET end point compared with RAST class
Antigen
Under
Same
Dematiaceous fungi Alternaria Cladosporium* Helminthosporium Stemphylium Pullularia Epicoccum Curvularia Penicillium Aspergillus Mucor Fusarium
1 1 — 1 — 1 — — — — —
3 2 3 3 1 1 — 1 2 3 3
*In
1 2 ≥3 class classes classes over over over
4 5 7 3 4 4 — 8 4 4 1
2 — — 2 2 1 4 — 4 3 2
— — — 1 3 — 6 1 — — 1
2 patients negative for Cladosporium by RAST, SET was not done.
Table 2. Number of positive RAST responses per class Class
No. of responses
0 0/I I II III IV V
2 8 28 37 21 13 1
strated some degree of sensitivity to almost every fungal antigen tested. Ten patients were subjected to RAST for 11 fungal antigens. Of these 110 tests, only 2 were class 0, and 8 others were class 0/I (borderline) (Table 2). The distribution of positive responses was heaviest in the midrange (class I through III), with only 1 class V response. Fungal cultures were done in all patients. Seven of the 10 were positive: 3 for Curvularia, 3 for Bipolaris, and 1 for Aspergillus. No consistent alteration in either RAST or skin test results was noted for the particular fungus cultured. DISCUSSION
When we initially embarked on this protocol of treating patients with AFS with both relevant fungal and relevant nonfungal antigens, we used skin testing for evaluation of the degree of sensitivity to molds. We chose to do this because conventional wisdom had indicated that RAST testing yielded lower scores for molds than did skin tests. Furthermore, because RAST measures only allergen-specific IgE, whereas there have been sugges-
253
Table 3. SET levels within 1 class of RAST score Antigen
%
Cladosporium Helminthosporium Penicillium Alternaria Mucor Stemphylium Aspergillus Epicoccum Fusarium Pullularia Curvularia
100 100 90 80 70 70 60 60 50 50 0
Table 4. SET levels 3 or more classes higher than RAST score Antigen
%
Curvularia Pullularia Fusarium Penicillium Stemphylium Epicoccum
60 30 20 10 10 10
tions that skin tests may also measure other factors, we believed it prudent to use this method. Finally, delayed hypersensitivity to molds is frequently described and can be evaluated only by skin testing, not by RAST. Although in this very small series we found no patients with delayed reactions, we believe it is still important to look for these reactions because, when present, they may necessitate significant alterations in a dosage advancement schedule. It is generally accepted that SET results will parallel Fadal-Nalebuff modified RAST results, with SET being slightly more sensitive than RAST. However, this relationship has been found to be highly variable, more so for molds than other antigens.7 When immunotherapy is based on skin testing, the antigens used in both testing and treatment are identical, coming from the same stock vial. Unfortunately, this relationship does not exist for immunotherapy based on in vitro testing. For these reasons, in vitro–based treatment must always be preceded by some form of bioassay or vial test.8 This is necessary to establish the exact relationship of the antigens being injected to those used on the disk or other carrier used in the in vitro test. Although prolonged usage of a given in vitro system may provide the clinician sufficient familiarity to construct a correlation table between in vitro results and anticipated SET levels for various anti-
254
Otolaryngology– Head and Neck Surgery September 1999
MABRY et al
gens, a vial test of some sort is always necessary when immunotherapy commences. In most patients in this series, the results of RAST and SET are parallel. In a few instances, however, a rather significant difference between the 2 tests was noted (Tables 3 and 4), in the form of underclassification (or less often overclassification) by RAST. This lack of concordance was not limited to (or even apparently associated with) the particular fungus cultured from the sinuses, nor was it attributable to altered sensitivity to only dematiaceous fungi, which are the causative organisms in AFS. Although it is tempting to postulate that patients with AFS somehow behave differently than patients with uncomplicated allergy, the more likely conclusion is that the disparity is simply related to a difference in antigens between those used in RAST and those used in skin tests. This difference appears to be more marked for some antigens than for others. In a previous paper dealing with testing patients with allergy (but not AFS), a tendency for RAST to overclassify (ie, indicate a greater sensitivity than reflected by SET) was noted in a very small proportion of patients.9 In that series of 53 patients, sensitivity for Helminthosporium was overclassified by RAST in almost 10% of patients, overclassification for Cladosporium was 4%, and sensitivity to Alternaria was underclassified and overclassified in equal proportions (2.7% each). In our study overclassification was rare. However, underclassification by RAST of sensitivity to a number of fungal antigens occurred. The most striking example is Curvularia, in which the RAST showed a degree of sensitivity that was 2 or more classes lower than the SET end point in all 10 patients. This series presents an exceedingly small sample, and obviously no statistical conclusions can be drawn from it. However, on the basis of this series, one logical inference is that different commercially available fungal antigens have a rather variable relationship to the antigens bound to a carrier substrate for RAST testing. Furthermore, it is certainly possible that, as postulated by others in the past, skin tests (especially for molds) depend on mechanisms that are not simply IgE-mediated atopy. Our findings in this small series certainly indicate the need for further investigation of the relationship between skin test and in vitro results in testing for sensitivity to molds, not only in patients with AFS but in all situations. Another finding that should not be overlooked is that all of these patients with AFS were sensitive to virtually every fungal antigen with which they were tested, both by RAST and skin tests. Cross-reactivity between fungal antigens is not well understood, but it
is generally considered to be poor. Believing that decreasing the total allergic load should be a goal of therapy in the management of AFS, we find that sensitivity to multiple fungal antigens supports our decision to test for and treat all relevant fungal antigens, not just the particular organism that might have been cultured from the sinuses. Chrzanowski et al10 have shown recognition by Western blot of an 18-kD protein in 11 patients with AFS, which they suggest may represent a fungal panallergen. It is conceivable that patients with AFS may have a unique hypersensitivity to this protein, explaining their sensitivity to virtually all fungal antigens for which they are tested. We eagerly anticipate seeing the results of further work in this area. CONCLUSION
Based on a small sample, it appears that there is not always a consistent relationship between levels of sensitivity to molds reflected in skin testing (with dilutional intradermal tests) and RAST in patients with AFS. This lack of concordance seems to vary among differing antigens. At present, although RAST testing remains an excellent screening modality, some type of quantitative skin testing is necessary for these antigens before immunotherapy is begun. As is often the case, the results in this small preliminary series indicate the need for further study in a larger series of patients. We hope that this article will stimulate others to add their experience to ours. REFERENCES 1. DeShazo RD, Chapin K, Swain RE. Fungal sinusitis. N Engl J Med 1997;337:254-9. 2. Mabry RL, Marple BF, Folker RJ, et al. Immunotherapy for allergic fungal sinusitis: three years’ experience. Otolaryngol Head Neck Surg 1998;119:648-51. 3. Mabry RL, Marple BF. Allergic fungal sinusitis. Curr Opin Otolaryngol Head Neck Surg 1997;5:49-52. 4. Bent JP III, Kuhn FA. Diagnosis of allergic fungal sinusitis. Otolaryngol Head Neck Surg 1994;111:580-8. 5. King WP. Clinical significance of molds newly available for radioallergosorbent testing. Otolaryngol Head Neck Surg 1989; 101:1-4. 6. King HC, Mabry RL, Mabry CS. Testing methods for inhalant allergy. In: Allergy in ENT practice: basic guide. New York: Thieme Medical Publishers; 1998. p. 103-54. 7. Corey JP, Liudahl JJ, Yound SA, et al. Diagnostic efficacy of in vitro methods vs skin testing in patients with inhalant allergies. Otolaryngol Head Neck Surg 1991;104:299-302. 8. Mabry RL. Blending skin endpoint titration and in vitro methods in clinical practice. Otolaryngol Clin North Am 1992;25:61-70. 9. Tandy JR, Mabry RL, Mabry CS. Correlation of modified radioallergosorbent test scores and skin test results. Otolaryngol Head Neck Surg 1996;115:42-5. 10. Chrzanowski RR, Rupp NT, Kuhn FA, et al. Allergenic fungi in allergic fungal sinusitis. Ann Allergy Asthma Immunol 1997;79: 431-5.