Drug Criteria & Outcomes-Understanding antimicrobial desensitization protocols
By Franklin Messer, PharmD
Patricia Hanks, PharmD
Veterans Affairs Medical Center
Sepulveda, CA
Introduction
An understanding of the different types of hypersensitivity reactions is essential to proper usage of desensitization protocols. Hypersensi tivity reactions are generally classified into four major categories: Types I, II, III, and IV.
The life-threatening reactions of the Type I hypersensitivity are often referred to as anaphylactic reactions. These reactions involve a typical immune response whereby an antigen enters the body and is identified as foreign. If the patient is antigen naïve, symptoms may not be expressed for one to two weeks, if at all. Otherwise, the reaction is immediate. In response to this antigen, humans will isolate each antigenic particle and present this antigen via antigen-presenting cells (usually macrophages or T lymphocytes) to T lymphocytes.
For Type I reactions, T lymphocytes initiate the clonal expansion of B lymphocytes into exaggerated immunoglobulin E (IgE)-producing plasma cells on primary exposure. On secondary exposure, IgE bound to mast cells and eosinophils attach to antigen, and mast cell degranulation ensues. Degranulation releases cytokines (cellular communicators) like chemotactic factors (such as NCF, LTB4, PAF, ECF-A) and spasmogens (histamine, SRS, PG, PAF), which result in the clinical picture of anaphylaxis.
Types II and III hypersensitivity reactions are mediated by IgF. Both of these reactions are more delayed in their presentation and are usually less severe than Type I reactions in that they do not cause immediate death. The mechanism of these reactions is related to antibody destruction of drug-bound to cell-bound or solubilized protein. This combination of drug and protein is called a hapten. For example, when IgG is formed against penicillin bound to a protein on a red blood cell's surface, the antibody destroys the cell causing hemolytic anemia.
Type II, III, and IV reactions occur on secondary exposure to a drug. Type IV reactions are usually less severe and often can be treated while the offending agent is continued.
Tolerance to an antigen is desirable when exposure is frequent or necessary. When a patient requires treatment with a pharmacological agent to which the patient is hypersensitive, tolerance may actually be critical to patient survival or cure of the infection. Immunologic tolerance to a drug (antigen or hapten) can be defined as one's ability to endure sustained exposure to the drug without adverse events. In a patient who has an "allergy" or "hypersensitivity" to a drug, tolerance can be achieved by decreasing the sensitivity of B or T lymphocytes.
Induction of tolerance in T or B lymphocytes can be achieved by four common methodologies. Clonal abortion is the loss of the specific B-lymphocyte clone that recognizes the antigen in question. This is easiest to achieve in a neonate for B lymphocytes when the B cells are less developed and less sensitive to antigen exposure. T lymphocytes respond equally well at all ages in life to clonal abortion. Clonal abortion is achieved by long-term, low-dose exposure of the antigen.
A second methodology is clonal exhaustion, which involves depletion of the clone's metabolic capabilities leading to cytokine and antibody depletion or exhaustion. Exhaustion can also be acquired again by long-term, low-dose exposure of the antigen.
Functional cell deletion is a recent phenomenon available due to molecular biology. Development of antibodies to specific cell lines (e.g. OKT3, OKT8) allows for the deletion of one or multiple cell lines within animals.
Finally, T-suppressor cell induction is also effective, but it is much slower than other methods. Induction of T-suppressor cells to a specific T lymphocyte's activity can suppress T-helper cells or B lymphocyte responses, yet it requires continued presence of the antigen and is also short-lived after antigen is removed.
A method not commonly used is blockage of antibody secretion by plasma cells. Suppression of both B and T lymphocytes is also obtained by the use of immunosuppressive agents. Cyclophosph amide would effectively induce tolerance by suppressing lymphocyte cellular function. However, in a patient with an infectious disease, the use of an immunosuppressive to induce tolerance to an anti-infective is undesirable.
An understanding of the pharmacodynamics of tolerance is critical for pharmacotherapeutic induction of tolerance or "desensitization." Administration of a slow upward titration of antigen controls mast cell degranulation. Degranulation should occur slowly, leading to milder or undetectable response to cytokine expression. This methodology induces T-suppressor cell induction and is probably the safest technique to use for most desensitizations. Methods that induce induction of B-lymphocyte tolerance can take four to 15 days and require two- to three-log higher doses of antigen. The duration of desensitization is at least as long as antigen presence with T-suppressor cell induction. In addition, T-lymphocyte tolerance persists longer than B-lymphocyte tolerance.
The goal of therapy is to block T-lymphocyte-dependent, B-lymphocyte response and exaggerated IgE expression. T-suppressor lymphocytes will also become class specific and prevent B-lymphocyte clonal expansion, antibody production, and thereby mast cell degranulation.
The desensitization protocols presented in this article deal primarily with IgE-mediated hypersensitivity reactions to common antimicrobial agents. It is not acceptable to extrapolate these published protocols to Type II and III reactions. Type IV reactions do not need desensitization.
Materials/methods
A MEDLINE search was performed using "antibiotics," "desensitization," and "protocol" as key terms. The identified articles were obtained, and a second search through their bibliographies was performed. All articles were then classified according to the class of antimicrobial agent and the route of administration for desensitization. For each reported class of antimicrobial, a desensitization protocol was identified or derived from these reports. When more than one desensitization protocol was identified, all authors reviewed the protocols, and a consensus was reached about the most effective, safest, and most convenient protocol. In some instances, slight variations in the original protocol were made, primarily to simplify the preparation of desensitization solution(s).
Beta-lactam antibiotics
(penicillins and cephalosporins)
Many desensitization protocols have been published for use in various patient populations with a positive penicillin skin test to the major (penicilloyl) or minor (penicilloic acid, penillic acid) determinants. Basic formulas for desensitization begin with very small oral dosages (i.e. 100 units of penicillin G, 60 mcg of carbenicilllin, 60 mcg of nafcillin; note that 1 mg penicillin G = 900-1050 penicillin G units [USP] [see table 1]) and double the dose every 15 minutes until 800,000 units is reached, then change to parenteral administration and advance until a full therapeutic dosage is achieved. Successful desensitization has been reported in 33 of 33 patients when using this approach.
| Table 1-Oral Penicillin Desensitization Protocol | ||
| Preparation of Oral Penicillin Desensitization Solutions | ||
| Solution 1 400,000 units/f mL | ||
| Solution 2 [2 mL solution 1 + 14 mL sterile water] 10,000 units/mL | ||
| Solution 3 [2 mL solution 2 + 18 mL sterile water] 1,000 units/mL | ||
| Dosing and Administration Schedule | ||
Time (in hours) |
Penicillin (oral units) |
Dose (mL) |
0 |
100 |
0.1 |
0.25 |
200 |
0.2 |
0.5 |
400 |
0.4 |
0.75 |
800 |
0.8 |
1 |
1,600 |
1.6 |
1.25 |
3,200 |
3.2 |
1.5 |
6,400 |
6.4 |
1.75 |
12,000 |
12 |
2 |
25,000 |
25 |
2.25 |
50,000 |
50 |
2.5 |
100,000 |
100 |
2.75 |
200,000 |
200 |
3 |
400,000 |
400 |
3.25 |
800,000 |
800 |
A rapid IV protocol for desensitization to a variety of penicillins and cephalosporins used an extremely small initial dosage (i.e. 0.5 mcg) infused over 20 minutes and increased the dose tenfold every 30 minutes until a full therapeutic dosage was reached.
Successful desensitization was achieved in 12 patients: four receiving ticarcillin, four receiving penicillin G, and one each receiving oxacillin, piperacillin, cefamandole and cephalothin. Two patients did experience delayed reactions requiring treatment with corticosteroids or antihista - mines. However, no patient required cessation of the desired therapy (table 2).
| Table 2 Intravenous Penicillin Desensitization Protocol | ||||
| Preparation of Intravenous Penicillin Densensitization Solutions | ||||
| Solution 1 500 mg per 50 mL (a)(10 mg/mL) | ||||
| Solution 2 [5 mL solution 1 + 50 mL diluent] 0.9 mg/mL | ||||
| Solution 3 [5 mL solution 2 + 50 mL diluent] 0.08 mg/mL | ||||
| Solution 4 [5 mL solution 3 + 50 mL diluent] 0.007 mg/mL | ||||
| Solution 5 [5 mL solution 4 + 50 mL diluent] 0.0007 mg/mL | ||||
| Solution 6 [5 mL solution 5 + 50 mL diluent] 0.00006 mg/mL | ||||
| Solution 7 [5 mL solution 6 + 50 mL diluent] 0.000006 mg/mL | ||||
| (a) stock solution 1 prepared by solubilizing > 1 g of antibiotic with diluent to a final concentration of 10 mg/mL | ||||
| Dosing and Administration Schedule | ||||
Solution (#) |
Time (hr:min) |
Concentration of infused solution (mg/mL) |
Infusion rate (mL/hr) |
Total dose delivered (mg) |
7 |
0:0 |
0.000006 |
167 |
0.00033 |
6 |
0:20 |
0.00006 |
167 |
0.0033 |
5 |
0:40 |
0.0007 |
167 |
0.039 |
4 |
1:00 |
0.007 |
167 |
0.39 |
3 |
1:20 |
0.08 |
167 |
4.4 |
2 |
1:40 |
0.9 |
167 |
49.5 |
1(b) |
2:00 |
10 |
150 |
500 |
| (b) after solution 1 infused, patient can receive remainder of dose | ||||
A protocol effective for IV cefotaxime desensitization was used in one patient with severe lumbar osteomyelitis. The protocol began with a 1 mg/day dose that was doubled daily until a 2 g q12h schedule was achieved by day 14. This titration is unnecessarily slow, compared with penicillin protocols, because it does not correlate with the pharmacokinetic properties of the drug in question, and therefore a much faster desensitization should be possible.
Additionally, most patients requiring treatment with this or similar type agents could not wait for such a prolonged period of time to achieve a therapeutic dosing regimen.
Sulfonamides (trimethoprim-sulfamethoxazole)
Multiple reports address the issue of trimethoprim-sulfame thoxazole desensitization. As one would expect, the majority of concern for hypersensitivity and the need for such desensitization occurs most often in patients with HIV infection or AIDS, in which the frequency of allergic reactions to drugs is said to be increased.
Although differences do exist, most studies/case reports use serial dilutions, administered orally at various intervals. The largest report described desensitizing 22 patients using tenfold serial dilutions.
Nineteen patients were successfully desensitized. Three patients experienced nausea and/or chills, requiring cessation of the desensitization. These adverse events resolved upon treatment. Eighteen patients were evaluated long term, and of those, 15 tolerated for a mean of 14 months, until therapy was discontinued for reasons other than allergic reactions (table 3).
| Table 3 - Trimethoprim/Sulfamethoxazole Oral Desensitization Protocol | |||
| Preparation of Oral Desensitization Solutions | |||
| Solution 1 40/200 mg/5mL | |||
| Solution 2 [0.5 mL solution 1 + 4.5 mL sterile water] 4/20 mg/5mL | |||
| Solution 3 [0.5 mL solution 2 + 4.5 mL sterile water] 0.4/2 mg/5mL | |||
| Solution 4 [0.5 mL solution 3 + 4.5 mL sterile water] 0.04/0.2 mg/5mL | |||
| Solution 5 [0.5 mL solution 4 + 4.5 mL sterile water] 0.004/0.02 mg/5 mL | |||
| Dosing and Administration Schedule | |||
Hours |
Solution |
Oral Dose |
Dose (mL) |
0 |
5 |
0.004/0.02 mg |
5 |
1 |
4 |
0.04/0.2 mg |
5 |
2 |
3 |
0.4/2 mg |
5 |
3 |
2 |
4/20 mg |
5 |
4 |
1 |
40/200 mg |
5 |
5 |
* |
160/800 mg |
5 |
| * commercially available tablet | |||
Only one case report was identified addressing IV desensitization. The patient tolerated the treatment doses but developed chills, rigor, and rash within hours of the full dose infusion of 240/1,200 mg.
No explanation was given regarding this delayed reaction. Table 4 presents a slight modification, for solution preparation and administration purposes, of the protocol used in that case.
| Table 4 - Trimethoprim/Sulamethoxazole Intravenous Desensitization Protocol | ||
Minutes |
Volume of Solution to be administered (mL) |
Resulting IV Dose (mg) |
0 |
0.2 |
0.16/0.8 |
20 |
2 |
1.6/8 |
40 |
10 |
8/40 |
60 |
20 |
16/80 |
80 |
100 |
80/400 |
100 |
170 |
136/680 |
120 |
300 |
240/1200 |
Sulfadiazine
Only one case report on this desensitization was identified. The protocol used 1 mg, 10 mg, 100 mg, 500 mg, 1,000 mg, and 1,500 mg given every four hours orally. The patient experienced a recurrence in the maculoerythematous rash after 10 days of therapy. The rash was controlled with prednisone and the patient was able to continue therapy.
Vancomycin
For vancomycin, the first step should be to accurately classify the reaction as an allergic reaction in contrast to the more common "red-man syndrome." This syndrome is thought to be due to an infusion rate-dependent increase in histamine release, not immune mediated. Patients should not be labeled "vancomycin allergic" unless the possibility of red-man syndrome has been eliminated because vancomycin may be the only effective agent against some pathogens.
Several protocols have been used to date for this purpose. One described a rapid desensitization protocol applied to seven patients, five of whom required a modification (decrease fusion rate) in the protocol. Desensitization required two hours to several days and was successful for all patients. Slower protocols also have been used, with successful desensitization within four hours to nine days, depending upon the protocol used. A modification of these protocols, to facilitate solution preparation and administration, has been used in two patients, both of whom tolerated desensitization without adverse effects (table 5).
| Table 5 - Vancomycin Desensitization Protocol | |||
| Preparation of Desensitization Solutions | |||
| Solution 1 2 mg/mL (a) (2:1) | |||
| Solution 2 [25 mL solution 1 + 25 mL D5W] 1mg/mL (1:1)(b) | |||
| Solution 3 [ 5 mL solution 2 + 50 mL D5W] 0.09 mg/mL (1:10)(b) | |||
| Solution 4 [5 mL solution 3 +50 mL D5W] 0.008 mg/mL (1:100)(b) | |||
| Solution 5 [5 mL solution + 50 mL D5W] 0.0008 mg/mL (1:1000)(b) | |||
| Solution 6 [5 mL solution 5 +50 mL D5W] 0.00007 mg/mL (1:10,000)(b) | |||
| (a) Vancomycin 500 mg diluted in 250 mL of 5% dextrose in water (2 mg/mL). | |||
| (b) Approximate dilutions - 50 mL, rather than 45 mL, of diluent used to facilitate preparation (i.e. solution can be added to 50 mL minibag). | |||
| Dosing and Administration Schedule | |||
Solution (#) |
Time (hr:min) |
Infusion Rate ML/hr (mL/min) |
Volume Infused (mL) |
6 |
0:0 |
60 (1) |
15 |
5 |
0:15 |
15 (0.25) |
3.75 |
5 |
0.30 |
30 (0.5) |
7.5 |
5 |
0.45 |
60 (1) |
15 |
4 |
1:0 |
15 (0.25) |
3.75 |
4 |
1:15 |
30 (0.5) |
7.5 |
4 |
1:30 |
60 (1) |
15 |
3 |
2:0 |
15 (0.25) |
3.75 |
3 |
2:15 |
30 (0.5) |
7.5 |
3 |
2:30 |
60 (1) |
15 |
2 |
3:0 |
15 (0.25) |
3.75 |
2 |
3:15 |
30 (0.5) |
7.5 |
2 |
3:30 |
60 (1) |
15 |
1 |
3:45 |
225 (3.75) |
225 |
Aminoglycosides
An IV protocol was reported for tobramycin utilized in one cystic fibrosis patient. The first dose was 0.001 mg and was doubled every 30 minutes until a therapeutic dosage of 80 mg was obtained.
Pentamidine
Two reports of pentamidine desensitization were reviewed. Both used a serial dilution protocol. The starting dilution was 1:10,000 in the first, and 1:1,000 in the second report. Both patients were successfully treated, but they died shortly thereafter due to an underlying disorder (table 6).
| Table 6 - Pentamidine Desensitization Protocol | ||||
| Preparation of Desensitization Solutions | ||||
| Solution 1 0.8 mg/mL(a) | ||||
| Solution 2 [1 mL solution 1 + 9 mL D5W] 0.08 mg/mL | ||||
| Solution 3 [1 mL solution 2 + 9 mL D5W] 0.008 mg/mL | ||||
| Solution 4 [1 mL solution 3 + 9 mL D5W] 0.0008 mg/mL | ||||
| Solution 5 [1 mL solution 4 + 9 mL D5W] 0.00008 mg/mL | ||||
| (a) Pentamidine iseothionate 200 mg diluted in 250 mL of 5% dextrose in water (0.8 mg/mL) | ||||
| Dosing and Administration Schedule | ||||
Time (Min) |
Solution (#) |
Dilution of solution |
Volume of Infused (mL) |
Pentamidine delivered (mg) |
0 |
5 |
1:10,000 |
2 |
0.00016 |
15 |
4 |
1:1,000 |
2 |
0.0016 |
30 |
3 |
1:100 |
2 |
0.016 |
45 |
2 |
1:10 |
2 |
0.16 |
60 |
1 |
full strength |
250 during 2 hrs. |
200 |
| (b) Infuse over 2 minutes. The line should be flushed following each administration to ensure complete delivery of medication. | ||||
Clindamycin
One report of clindamycin desensitization was identified. A female patient with HIV, diagnosed with toxoplasmosis encephalitis, experienced a cutaneous reaction to clindamycin. The maculopapular reaction was confirmed by rechallenge. Treatment with sulfadiazine had to be discontinued due to neutropenia.
Subsequently, clindamycin desensitization was attempted. The patient received 20 mg of oral clindamycin every eight hours, with the dose doubling every day, until a full therapeutic dose of 600 mg was achieved. The dose was then increased to 600 mg every six hours, which she had continued to tolerate at the time the report was published.
Anti-tubercular agents
Reactions to anti-tubercular agents have been a significant problem since the 1960s when these agents were first marketed. Rates of reaction as high as 17% have been observed with para-amin osalicylic acid (PAS). Streptomycin (1.2%-10.3%) and isoniazid (INH) (0.4%-1.5%) also have been frequently implicated. Overall, reaction rates of 4% to 5% are reported for all anti-tuberculars, which have led to drug discontinuation.
The onset of reactions, usually rash, is usually four to 24 hours after starting anti-tuberculars. Reactions as late as three to nine weeks also have been reported. Data on the patients reacting after two days suggest that they still could be having a Type I hypersensitivity reaction and could be candidates for desensitization. Measurement of IgE can be performed in these cases to verify the presence of a true Type I hypersensitivity.
However, most of these reactions are probably Type II or Type III reactions, which are not responsive to desensitization.
Rifampin also has been implicated as causing hypersensitivity reactions. However, caution is warranted in making the diagnosis of a hypersensitivity reaction because many noncompliant patients will have a flu-like syndrome associated with stopping and restarting rifampin. Fever and rash can mislead a diagnostician into thinking that the patient is allergic to the medication. It is also difficult to distinguish which of many anti-tuberculars is the offending agent.
Discussion
The common theme in all desensitization protocols presented in this review is a three-step process. The first step it so confirm the lack of alternatives to treat the infectious disease of the patient. The second step is to determine the nature of the allergy. The desensitization process is the third and final step in these protocols.
Determining the nature of the allergy is performed by careful history of past exposures to the suspected agent, followed by a skin test, if necessary. Skin testing for IgE reactions is well-defined only for penicillin. The major determinant of penicillin (penicilloyl) is available commercially, and the minor determinants used are penicillin G and penicilloic acid, which is obtained by alkaline hydrolysis of penicillin G.
The first procedure in skin testing is the skin prick test. If this is negative, intradermal injections are placed on the anterior aspect of the forearm. The test is considered positive if the induration is greater than 5 mm after 15 to 20 minutes. A clinical caveat is to ensure that the patient is not taking a drug that would blunt the skin test reaction, such as corticosteroids or antihistamines. If the test is positive, desensitization should be initiated in a controlled environment, with cardiopulmonary resuscitation equipment and personnel available (preferably an intensive care unit) to treat any severe reaction that may occur.
Most protocols reviewed involved a serial dilution procedure to desensitization. Therefore, we propose a template desensitization protocol using tenfold serial dilutions. The recommended starting dilution is 1:10,000 and progressing toward a normal dose at intervals of 30 minutes (table 7). Desensitization should be initiated with oral doses whenever possible.
| Table 7 - Antimicrobial Desensitization Protocol | ||
| Preparation of Desensitization Solutions | ||
| Solution 1 10 mg/mL (10:1) | ||
| Solution 2 [1 mL solution 1 + 9 mL diluent] 1 mg/mL (1:1) | ||
| Solution 3 [1 mL solution 2 + 9 mL diluent] 0.1 mg/mL (1:10) | ||
| Solution 4 [1 mL solution 1 + 9 mL diluent] 0.01 mg/mL (1:100) | ||
| Solution 5 [1 mL solution 1 + 9 mL diluent] 0.001 mg/mL (1:1000) | ||
| Solution 6 [1 mL solution 1 + 9 mL diluent] 0.0001 mg/mL (1:10,000) | ||
| Dosing and Administration Schedule | ||
Solution (#) |
Time (minutes) |
Dilution of solution |
6 |
0:0 |
1:10,000 |
5 |
0:30 |
1:1,000 |
4 |
1:0 |
1:100 |
3 |
1:30 |
1:10 |
2 |
2:0 |
1:1 |
1 |
2:30 |
10:1 |
- |
3:00 |
Full therapeutic dose* |
| Administer slowly; recommend over 2 hours | ||
A missed dose during the procedure creates an ambiguous situation. It is unknown how many doses can be missed and what the consequences or risk for anaphylaxis are for a missed dose. Fortu nately, Type I hypersensitivity reactions occur infrequently enough that an assessment of risk is almost impossible. Unfortu nately, because of the low incidence of reactions, we do not have data to make recommendations. Therefore, a conservative approach should be taken. Begin the entire desensitization protocol at the beginning if there is a single missed dose.
It is noteworthy that a successful desensitization does not mean that the patient is free from future reactions to the same agent — only that the present course of therapy is unlikely to result in an immediate, life-threatening reaction.
[The American College of Allergy, Asthma & Immunology released a supplement to its journal that may be useful. The title is Disease Management of Drug Hypersensitivity: A Practice Parameter. It is available for free at www.annallergy.org (December 1999 issue). The authors of the preceding article can be reached at the VA Medical Center in Sepulveda, CA, at (818) 891-7711.]
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content