Darunavir and Trimethoprim-Sulfamethoxazole Cross-Reactivity
Darunavir and Trimethoprim-Sulfamethoxazole Cross-Reactivity
This retrospective cohort study showed that patients with a known TMP-SMX allergy were at greater risk for a DRV-related rash upon initiation of DRV/r-containing cART. The occurrence of a DRV-related rash led to discontinuation of DRV in only 1% of the patients, whereas none of the patients developed a potentially life-threatening allergic reaction on administration of DRV. Finally, no predicting factors for the occurrence of a DRV allergy among patients with a TMP-SMX allergy were identified.
These findings are worth noting because they contradict the results of previously published small studies in Asian patients. Nishijima et al.. and Lin et al.. found no association between TMP-SMX allergy and DRV allergy in 145 and 74 HIV-infected patients with a history of TMP-SMX use, respectively. However, higher incidences of DRV-induced rash of 5.3 and 6.8% were found in these studies compared with 2% in our study. This difference in incidence may be partly explained by the prospective study design and by the different ethnicities included in these studies compared with ours. Furthermore, our study is distinctive because it included a very large number of patients allergic to TMP-SMX from two large HIV-specialized hospitals in the Netherlands. Furthermore, combining data from the Dutch HIV registry with electronic hospital databases, data regarding patient or HIV-specific characteristics were rarely missing. Finally, in our study the Naranjo probability scale for drug allergy was used for a more objective classification of DRV allergies, whereas both other studies relied on the diagnosis reported in the patient chart.
Despite the large number of included patients, the number of patients allergic to DRV was small occurring in only 2% of the patients with DRV discontinuation in 1%. Therefore, DRV allergy may not be of great clinical concern among HIV-infected patients and as a result, cross-reactivity with TMP-SMX is even less of a problem. Additionally, multivariate analysis could not be performed because of the small number of DRV-allergic patients.
A limitation of our study is its retrospective design that might have caused inconsistencies in reporting information about allergic reactions to DRV. In order to assess and support the causality of DRV in an objective manner, the Naranjo probability scale was used. Using this probability scale meant that none of the allergic reactions could be classified as 'definite' because some questions of the scoring system could not be answered because of the retrospective design of this study. For example, administration of placebo or detection of DRV levels in blood lacked in all cases. This resulted in a classification of probable and possible DRV-allergic patients. When analyzing the whole cohort of DRV/r-treated HIV-infected patients irrespective of TMP-SMX exposure, a similarly low incidence of DRV-allergy of 2.5% was found, demonstrating that the reporting of allergies is not influenced by TMP-SMX exposure. Allergies were also dose independent, considering almost half of the TMP-SMX allergies occurred in patients receiving prophylactic doses (49.4%) and only two of the eight (25%) DRV-allergic patients were administered DRV 800 mg twice daily.
In conclusion, although DRV allergy was documented infrequently in our cohort, cross-reactivity with TMP-SMX was significantly observed. Importantly, DRV allergy was rarely of clinical relevance. As no potentially lethal allergic reactions were observed and most allergic reactions were treatable, it may be safe to say that DRV can be administered in patients allergic to TMP-SMX.
Discussion
This retrospective cohort study showed that patients with a known TMP-SMX allergy were at greater risk for a DRV-related rash upon initiation of DRV/r-containing cART. The occurrence of a DRV-related rash led to discontinuation of DRV in only 1% of the patients, whereas none of the patients developed a potentially life-threatening allergic reaction on administration of DRV. Finally, no predicting factors for the occurrence of a DRV allergy among patients with a TMP-SMX allergy were identified.
These findings are worth noting because they contradict the results of previously published small studies in Asian patients. Nishijima et al.. and Lin et al.. found no association between TMP-SMX allergy and DRV allergy in 145 and 74 HIV-infected patients with a history of TMP-SMX use, respectively. However, higher incidences of DRV-induced rash of 5.3 and 6.8% were found in these studies compared with 2% in our study. This difference in incidence may be partly explained by the prospective study design and by the different ethnicities included in these studies compared with ours. Furthermore, our study is distinctive because it included a very large number of patients allergic to TMP-SMX from two large HIV-specialized hospitals in the Netherlands. Furthermore, combining data from the Dutch HIV registry with electronic hospital databases, data regarding patient or HIV-specific characteristics were rarely missing. Finally, in our study the Naranjo probability scale for drug allergy was used for a more objective classification of DRV allergies, whereas both other studies relied on the diagnosis reported in the patient chart.
Despite the large number of included patients, the number of patients allergic to DRV was small occurring in only 2% of the patients with DRV discontinuation in 1%. Therefore, DRV allergy may not be of great clinical concern among HIV-infected patients and as a result, cross-reactivity with TMP-SMX is even less of a problem. Additionally, multivariate analysis could not be performed because of the small number of DRV-allergic patients.
A limitation of our study is its retrospective design that might have caused inconsistencies in reporting information about allergic reactions to DRV. In order to assess and support the causality of DRV in an objective manner, the Naranjo probability scale was used. Using this probability scale meant that none of the allergic reactions could be classified as 'definite' because some questions of the scoring system could not be answered because of the retrospective design of this study. For example, administration of placebo or detection of DRV levels in blood lacked in all cases. This resulted in a classification of probable and possible DRV-allergic patients. When analyzing the whole cohort of DRV/r-treated HIV-infected patients irrespective of TMP-SMX exposure, a similarly low incidence of DRV-allergy of 2.5% was found, demonstrating that the reporting of allergies is not influenced by TMP-SMX exposure. Allergies were also dose independent, considering almost half of the TMP-SMX allergies occurred in patients receiving prophylactic doses (49.4%) and only two of the eight (25%) DRV-allergic patients were administered DRV 800 mg twice daily.
In conclusion, although DRV allergy was documented infrequently in our cohort, cross-reactivity with TMP-SMX was significantly observed. Importantly, DRV allergy was rarely of clinical relevance. As no potentially lethal allergic reactions were observed and most allergic reactions were treatable, it may be safe to say that DRV can be administered in patients allergic to TMP-SMX.