Beta-lactam and sulfonamide allergy testing should be a standard of care in immunocompromised host

The increasing threat of antimicrobial resistance continues to drive development of novel programs to improve the appropriate use of antibiotics and aid antimicrobial stewardship (AMS)^{1, 2}. The high prevalence of patient-reported antibiotic allergies (so-called antibiotic allergy labels [AALs])^3–5, in particular penicillin allergies, in combination with the known associations between AALs and antimicrobial resistance^{6, 7}, has driven calls for antibiotic allergy “de-labelling” to be a standard fixture of AMS programs.^{1, 2} This is a particularly relevant for high-antibiotic usage populations⁸, such as immunocompromised hosts and those with pending immunosuppression where the prevalence of AALs is highest and so too is the rate of broad-spectrum antibiotic utilization and antimicrobial resistance.⁹ Although antibiotic allergy testing has demonstrated that over 85% of patients can be de-labelled with skin testing followed by oral provocation^{10, 11}, the utility of antibiotic allergy testing (AAT) in different populations of immunologically altered hosts has until recently been ill defined and supportive data absent. Frequently, this cohort of patients had been excluded from antibiotic allergy testing studies despite being the group most likely to benefit from appropriate antibiotic options.

The prevalence of AALs has frequently been demonstrated to be highest in immunocompromised hosts, up to 35.1% in two tertiary referral hospitals in the US (2010–15)¹². Additionally, we reported an AAL prevalence of 24% in a cohort of Australian cancer patients admitted with an infective episode between 2013–14 to a stand-alone cancer hospital¹³. Both the Australian and US studies demonstrated significant healthcare AAL impacts - including higher readmission rates, increased utilization of restricted antibiotics and longer antibiotic course duration^{12, 14}. The combination of high prevalence and associated inferior hospital outcomes should focus our attention to this group for targeted testing and de-labelling programs, in particular prior to planned immunosuppression.¹⁵

Antibiotic prophylaxis strategies are severely hampered by the reported presence of both penicillin and “sulfa” allergies that also happen to be the two most common antibiotic allergy labels globally. Penicillins are the preferred prophylaxis strategy to prevent infection with encapsulated organisms in asplenia¹⁶ and allogenic bone marrow transplantation¹⁷. Identification of penicillin allergic individuals should trigger referrals to antibiotic allergy testing programs to enable prophylaxis with penicillin family drugs. In the setting of a confirmed penicillin hypersensitivity, desensitization is an appropriate tool to utilize particularly when long-term uninterrupted dosing of penicillins is indicated. For patients reported an non-sulfonamide allergy, there is no cross-reactivity with antibiotic sulfonamides, hence trimethoprim-sulfamethoxazole (TMP-SMX) should be employed in these patients reporting “sulfur” allergy.¹⁸ In those with a historical antibiotic-sulfonamide allergy, oral rechallenge should be performed via a two-step challenge (TMP-SMX; 8mg/40mg followed by 72mg/360mg) or single-step challenge (TMP-SMX, 80mg/400mg), for mild immediate hypersensitivities or mild delayed hypersensitivities, respectively. This protocol was successfully employed in a total of 46 immunocompromised patients without any adverse event in the North America¹⁹ and Australia²⁰. In those with confirmed antibiotic sulfonamide cross-reactivity, dapsone despite being an antibiotic sulfonamide can be employed in most instances due to the low rate of cross-reactivity (9–13%).²¹ A step-wise approach to alternatives in the setting of moderate-severe antibiotic sulfonamide allergy are available to clinicians^{20, 22}.

We have demonstrated the benefits of antibiotic allergy testing, primarily for beta-lactam and sulfonamide AALs, to the immunocompromised host. In a multicentre study of 118 AAL patients undergoing protoliced antibiotic allergy testing, of which 48% (n = 57) were immunocompromised, 84% were “de-labelled” and a resultant 12-fold increase in antibiotic appropriateness was observed²³. In this study immunocompromised was defined as hematological malignancy, oncological malignancy, solid organ or stem cell transplant recipient, autoimmune disease, condition requiring >15 mg steroid daily for 1 month. This study examined patients reporting any antibiotic allergy, predominantly penicillins (54.4%), cephalosporins (18.1%) and sulfonamide antibiotics (7.5%). This finding was further validated in a cohort of 59 cancer patients with a similar AAL distribution, were no adverse events from testing where reported and a 4-fold increase in appropriate prescribing and 21-fold increase in preferred beta-lactam/beta-lactamase prescribing was noted¹³. A recent study by Taremi et al. adds to the published literature already supporting penicillin allergy testing in the immunocompromised host, and demonstrates a 95% de-labelling rate and post-testing utilization of penicillins in 51%²⁴. This collective work demonstrates that antibiotic allergy testing, in particular for beta-lactams, is both safe and effective in immunocompromised hosts. The question for clinicians caring for immunocompromised hosts should no long be whether they offer testing to their patients but how such testing can be incorporated into the care plans of such patients.

If we are to implement antibiotic allergy testing into the routine care of the immunocompromised host, risk-stratification is required to determine who can move to direct oral provocation and who requires initial skin testing. In hematological malignancies and stem cell transplantation there is an increased rate of maculopapular exanthema,^{12, 25} potentially related to a drug allergy “mimicker” (e.g. engraftment syndrome, viral infection, skin graft versus host disease), and these should be actively pursed to potentially allow direct de-labelling in patients where an alternative diagnosis is established. There is growing literature to support the safe use of direct oral rechallenge in those that are identified as low-risk – childhood exanthema, unknown reactions, family history, pruritus without rash)^{15, 26–28}. This may be particularly useful when identifying patients before the initiation of immunosuppressive therapy. Although data is primarily from the outpatient and pediatric setting, there is no current evidence to indicate that an immunocompromised hospitalized patients should be managed differently. This is supported by our pilot work, where the implementation of a low-risk criteria allowed the safe oral challenge in 48 patients, including cancer patients²⁹. In patients of moderate or high-risk (e.g. anaphylaxis or immediate symptoms to suggest IgE-mediated hypersensitivity), antibiotic allergy testing such as prick and intradermal testing can be performed safely in this cohort²³. There are emerging data to support the role of patch and delayed intradermal skin testing in severe cutaneous adverse drug reactions in any host, however reassuringly there is emerging data that even performing this “higher risk” testing in immunocompromised host is safe and improves beta-lactam utilization³⁰. Although immunosuppression has the theoretical potential to cause a “false negative” skin prick or intradermal test primarily in delayed testing, in our tested cohorts we did not see adverse events to previously de-labelled antibiotics in the 90-day post period.^{13, 23, 29} It is also relevant that these populations are complex and in many their immune system may not normalize either due to their associated therapies or underlying immune dysregulation, so it becomes a practical issue of being able to provide them clear guidance for future safe antibiotic options. Indeed the vast majority of them who are labelled as allergic to beta lactam or sulfa antibiotics at the time of their initial diagnosis would have tested negative prior to the onset of immunocompromising therapy. We recommend that prednisone therapy be at the lowest manageable dose at the time of testing where feasible. To avoid this issue ideally wherever possible patients prior to immunosuppression should be referred and appropriately evaluated. While larger cohort studies are required to demonstrate the safety of risk-stratified antibiotic allergy testing in immunocompromised host, the current body of evidence suggests that such testing is feasible and efficacious.

The immunocompromised host is an emerging cohort with heightened antibiotic needs compounded by a large AAL burden. The AAL, in particular toward penicillin, can no longer be ignored and the key targets need addressing (Figure 1). The result of attending to AALs in immunocompromised hosts is improved antibiotic appropriateness. Future work is required to determine the direct benefit to patient efficacy and safety and health economic and antimicrobial resistance benefits of antibiotic allergy testing in this cohort and ways in which risk-stratification can afford access to testing for more patients. However, for now, clinicians caring for this cohort should be asking if their patients are” truly” allergic to antibiotics and how they can proactively be “de-labelled”.

Figure 1 –

Key points to approaching antibiotic allergy in the immunocompromised host

Adapted from references:^{15, 19, 20, 23, 29–32}