SEARCH

SEARCH BY CITATION

Keywords:

  • nonimmediate reactions;
  • penicillins;
  • skin tests

Abstract

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Funding of the study
  7. References

To cite this article: Romano A, Gaeta F, Valluzzi RL, Caruso C, Rumi G, Bousquet PJ. The very limited usefulness of skin testing with penicilloyl-polylysine and the minor determinant mixture in evaluating nonimmediate reactions to penicillins. Allergy 2010; 65: 1104–1107.

Abstract

Background:  The contribution of skin testing with penicilloyl-polylysine (PPL) and the minor determinant mixture (MDM) to the diagnosis of hypersensitivity reactions to penicillins differs greatly according to the type of reaction: immediate (occurring within 1 h after the last drug administration) or nonimmediate (occurring more than 1 h after the last drug administration).

Objective:  To assess the contribution of skin testing with PPL and MDM to the diagnosis of nonimmediate reactions to penicillins.

Methods:  We evaluated 162 adults who had had 232 nonimmediate reactions to penicillins, mostly aminopenicillins, and presented positive skin and/or patch tests to one or more penicillin reagents: PPL, MDM, benzylpenicillin, ampicillin, and amoxicillin, as well as any responsible penicillins.

Results:  A total of 157 subjects (96.9%) displayed patch-test and/or delayed-reading intradermal-test positivity to penicillin reagents, which indicates a cell-mediated hypersensitivity; six of them also presented immediate-reading skin-test positivities. All 157 patients with a cell-mediated hypersensitivity were positive to the responsible penicillins (parent drugs); 16 of them also displayed delayed-reading intradermal-test positivity to MDM. Five (3.1%) of the 162 patients displayed only immediate-reading skin-test positivity (four to PPL and one to amoxicillin). Overall, 158 subjects (97.5%) presented positive responses to the responsible penicillins, while only 9 (5.5%) and 17 (10.5%) were positive to PPL and MDM, respectively.

Conclusions:  The contribution of skin testing with PPL and MDM in diagnosing nonimmediate hypersensitivity reactions to penicillins, especially cell-mediated ones, is very limited. This finding could be useful at a time when PPL and MDM are not available in all countries.

Abbreviations
ENDA

European Network for Drug Allergy

g

gram

IgE

immunoglobulin E

IU

international unit

MDM

minor determinant mixture

mg

milligram

ml

milliliter

mm/l

millimol/liter

PPL

penicilloyl-polylysine

In both the American practice parameters (1) and the European Network for Drug Allergy (ENDA) guidelines (2–4), skin testing with penicilloyl-polylysine (PPL) and the minor determinant mixture (MDM) represents the first-line method for evaluating hypersensitivity reactions to β-lactams. Nevertheless, both Hollister-Stier and Allergopharma ceased the production of penicillin reagents in 2004, severely hampering the diagnosis of β-lactam hypersensitivity (5). However, literature data regarding studies distinguishing between subjects with immediate (i.e. occurring within 1 h after the last drug administration) and nonimmediate (i.e. occurring more than 1 h after the last drug administration) reactions to penicillins indicate that the contribution of skin testing with PPL and MDM to the diagnosis differs greatly according to the type of reaction.

In a Spanish study concerning 290 subjects with immediate reactions to penicillins (6), 111 (54.7%) of the 203 subjects with immediate-reading skin-test positivity were positive to PPL and/or MDM. In a recent study of ours (7), of the 300 French and Italian subjects with histories of immediate reactions to penicillins and with positive skin tests, 82 (27.3%) were positive to PPL and 100 (33.3%) to MDM, while 64 (21.3%) were positive only to either PPL or MDM, or to both of them, and not to any other penicillin reagents.

On the other hand, in studies regarding subjects with nonimmediate reactions to penicillins and a diagnosis of cell-mediated hypersensitivity (8–10), the rate of positive responses to patch tests and/or delayed-reading intradermal tests with PPL and MDM ranged from 0% (9) to 7.1% (8).

In the present study, using information from our database, we assessed the contribution of skin testing with PPL and MDM to the diagnosis of nonimmediate reactions to penicillins.

Methods

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Funding of the study
  7. References

Patient selection

We evaluated 162 adults (age range, 14–79 years) with a well-demonstrated hypersensitivity to penicillins. These 162 subjects constituted 37.4% of an outpatient population of 433 adults recruited prospectively between January 1994 and June 2009 in the Allergy Units of the Complesso Integrato Columbus and Oasi Maria Santissima because of histories of nonimmediate reactions to penicillins. We carefully questioned patients to identify the suspect antibiotic. Symptoms were classified as previously described (8). We applied a specific diagnostic protocol, which has been described in previous studies of ours (8, 11), in which 259 of the 433 subjects were first evaluated. The inclusion criterion required positive skin and/or patch tests to one or more of the penicillin reagents tested. The protocol was approved by the respective institutional review boards.

Skin and patch tests

On the first day, prick and intradermal tests were carried out using PPL (Allergopharma, Reinbeck, Germany), MDM (Allergopharma), and benzylpenicillin. The final concentrations were, respectively, 5 × 10−5 mm/l, 2 × 10−2 mm/l, and 10 000 IU/ml. Because Allergopharma ceased production of penicillin reagents, from July 2005 we used Diater S.A. (Madrid, Spain) ones: PPL (final concentration: 1.07 × 10−2 mm/l) and MDM (benzylpenicillin, sodium benzylpenicilloate, and benzylpenicilloic acid; final concentration: 1.5 mm/l). Patch tests were also administered with benzylpenicillin, ampicillin, and amoxicillin (5% in petrolatum); piperacillin (at a concentration of 200 mg/ml in normal saline) was also used in subjects with adverse reactions to it. Occlusion time was 48 h.

Two days later, ampicillin and amoxicillin, at concentrations of 1 and 20 mg/ml, after dilution in normal saline, were used for prick and intradermal tests. Piperacillin, at concentrations of 1 and 20 mg/ml, after dilution in normal saline, was also used in subjects with adverse reactions to it.

Readings of patch tests were made 15 min after removal of the strips and 48 h later, while those of skin tests were carried out after 20 min, as well as after 48 and 96 h, as previously described (3, 12).

Results

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Funding of the study
  7. References

Our 162 patients had experienced a total of 232 reactive episodes. The responsible β-lactams and the clinical manifestations are shown in Table 1. One hundred and fifty-three subjects (94.4%) reported adverse reactions to aminopenicillins (amoxicillin, ampicillin, and bacampicillin), and nine of them also to other β-lactams, such as cephalosporins and unspecified penicillins. The most frequent clinical manifestations were maculopapular exanthemas.

Table 1.   Clinical data of patients
 All patients (= 162)
  1. *Time (months) elapsed between last adverse penicillin reaction and current allergologic examination.

Age (y), mean (SD)43.9 (16.5)
Women, n (%)112 (69.1)
Time since last penicillin reaction*, median (range [25th–75th percentile])18 (1–540 [4, 108])
Family history of allergic diseases, n (%) 61 (37.6)
Personal history of allergic diseases, n (%)39 (24)
Responsible β-lactams, n (%)All reactions (= 232)
 Amoxicillin112 (48.2) [44 plus clavulanic acid]
 Ampicillin75 (32.3) [4 plus sulbactam]
 Bacampicillin28 (12.1)
 Piperacillin7 (3)
 Benzylpenicillin2 (0.9)
 Benzathine penicillin2 (0.9)
 Cephalexin1 (0.4)
 Ceftriaxone1 (0.4)
 Unspecified4 (1.8)
Manifestations, n (%)All reactions (= 232)
 Maculopapular exanthema106 (45.7)
 Maculopapular exanthema plus edema91 (39.2)
 Erythema11 (4.7)
 Bullous exanthema6 (2.6)
 Erythema plus edema5 (2.1)
 Local reaction4 (1.8)
 Urticaria3 (1.2)
 Toxic epidermal necrolysis2 (0.9)
 Others4 (1.8)

Of the 162 subjects – who, according to the inclusion criterion, had positive patch and/or skin tests to at least one of the penicillin reagents tested (Table 2) – 157 (96.9%) displayed patch-test and/or delayed-reading intradermal-test positivity to penicillin reagents, which indicates a cell-mediated hypersensitivity; six of them also presented immediate-reading skin-test positivities (four to PPL, one to both PPL and MDM, and one to piperacillin). Four of these six subjects had experienced maculopapular exanthemas during aminopenicillin therapy and the remaining two local reactions a few hours after intramuscular benzathine penicillin and piperacillin, respectively.

Table 2.   Patterns of skin test and patch test reactivity in 162 patients allergic to penicillins
Patch test results, n (%)Skin test results, n (%)Pattern, n (%)
BPAMAXResponsible§PPLMDMBPAMAXResponsible§
  1. +Positivity of patch tests or positivity of delayed-reading skin tests. +*Positivity of immediate-reading skin tests; PPL, penicilloyl-polylysine; MDM, minor determinant mixture; BP, benzylpenicillin; AM, ampicillin; AX, amoxicillin. §Responsible penicillin (parent drug). 150 aminopenicillins, 5 piperacillin, and 3 benzylpenicillin. In these 4 subjects, the responsible penicillins were amoxicillin (two patients), benzathine penicillin (1), and bacampicillin (1).

++++++74 (45.7)
++++++++24 (14.8)
+++++++23 (14.3)
+++++++++13 (8)
+++6 (3.7)
 +*  4 (2.5)
++++2 (1.3)
+++ +*+++2 (1.3)
++++ +*++++1 (0.6)
+++++++1 (0.6)
+++++++1 (0.6)
+++++1 (0.6)
++++++++1 (0.6)
+++++++1 (0.6)
+++++1 (0.6)
++++++1 (0.6)
+++++1 (0.6)
++1 (0.6)
+ +*1 (0.6)
 +*+++1 (0.6)
 +* +*+++1 (0.6)
 +* +*1 (0.6)
40 (24.7)146 (89.5)141 (87)144 (88.9)9 (5.5)17 (10.5)66 (40.7)154 (95.1)156 (96.3)158 (97.5)162

Five (3.1%) of the 162 patients presented only immediate-reading skin-test positivity (four to PPL and one to amoxicillin): two had had an urticarial reaction 2 h after a benzathine penicillin intramuscular injection and 5 h after the first amoxicillin tablet, respectively, another had experienced an erythema 4 h after the second tablet of bacampicillin, and the remaining two had had a maculopapular rash during amoxicillin therapy.

Table 2 shows the patterns of skin- and patch-test reactivity in the 162 patients allergic to penicillins. Patch-test and delayed intradermal-test positivity to ampicillin and amoxicillin constituted the main pattern of reactivity, followed by patch-test and delayed intradermal-test positivity to benzylpenicillin, ampicillin and amoxicillin. Fifty-one subjects presented positive results in delayed-reading skin tests with benzylpenicillin and negative ones with MDM. All these subjects were positive to intradermal tests with the highest concentration of benzylpenicillin.

Overall, 158 (97.5%) of the 162 subjects displayed positive responses to the responsible penicillins (parent drugs): 150 to aminopenicillins (of whom, 149 delayed-reading and one immediate-reading), 5 to piperacillin (four delayed-reading and one immediate-reading), and 3 to benzylpenicillin (all delayed-reading). Only 9 (5.5%) and 17 (10.5%) subjects were positive to PPL (all immediate-reading) and MDM (16 delayed-reading and one immediate-reading), respectively.

Discussion

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Funding of the study
  7. References

The diagnostic workup allowed us to diagnose a cell-mediated hypersensitivity in 157 (96.9%) of the 162 subjects on the basis of positive responses to patch tests and/or delayed-reading intradermal tests. Six of these 157 subjects had a mixed pattern of skin reactivity, because they also presented immediate-reading skin-test positivities. The aforesaid 157 patients represent 36.2% of the 433 subjects evaluated between January 1994 and June 2009 because of histories of nonimmediate reactions to penicillins. Therefore, the percentage of patients with a cell-mediated hypersensitivity, diagnosed on the basis of positive responses to patch tests and/or delayed-reading intradermal tests, found in the present study (36.2%, 157 out of 433) is very similar to that (37.8%, 98 out of 259) found in a previous study of ours (8), which evaluated 259 patients, who were included in the present study.

An IgE-mediated hypersensitivity was diagnosed in the remaining 5 (3.1%) of the 162 positive patients of the present study on the basis of positive responses to immediate-reading skin tests. Four of them were positive only to PPL. According to the aforesaid guidelines (1–4), these four patients would have been missed in the absence of this classic penicillin reagent. Thus, the present study confirms that PPL is a useful reagent in diagnosing an IgE-mediated hypersensitivity to penicillins. Considering that maculopapular rashes generally are not IgE-mediated, the two aforementioned subjects, who had experienced a maculopapular rash during amoxicillin therapy and were positive only to the immediate-reading intradermal test with PPL, may have had a false positive response; however, they were not challenged.

On the other hand, the contribution of skin testing with PPL and MDM in diagnosing cell-mediated hypersensitivity reactions to penicillins actually was null. Indeed, although we observed 16 patients with delayed-reading intradermal-test positivity to MDM, all the 157 patients with a cell-mediated hypersensitivity (including the aforesaid 16 and the 6 who also displayed immediate skin-test positivity) presented positive responses to the responsible penicillins (Table 2): 149 to aminopenicillins, 5 to piperacillin, and 3 to benzylpenicillin. This finding could be useful at a time when PPL and MDM are not available in all countries. However, it should be noted that skin testing with PPL and MDM is in any case essential for assessing cross-reactivity.

With regard to the 51 patients with a cell-mediated hypersensitivity to penicillins, displaying negative results in delayed-reading skin testing with MDM – which contains benzylpenicillin – and positive ones in delayed-reading skin testing with benzylpenicillin, as previously noted, they actually reacted to intradermal tests with the highest concentration of benzylpenicillin (10 000 IU/ml = 16.84 mm/l). Therefore, this phenomenon could be explained by the fact that the aforesaid concentration of benzylpenicillin is higher than those of the Allergopharma and Diater MDM, 1.68 and 1.40 mm/l, respectively.

Finally, the low rate of positive responses to PPL (9 of 162 subjects, 5.5%) and MDM (17 of 162, 10.5%) found in the present study could be because of the fact that 94.4% of the 162 subjects had reacted to aminopenicillins, and it is known that in cell-mediated allergy to aminopenicillins – which was diagnosed in 149 of our 162 patients – the amino-benzyl group of the side chain plays a predominant role in sensitization (3). However, additional comprehensive studies in large samples are required to provide a definitive assessment of the usefulness of skin testing with PPL and MDM in evaluating subjects with nonimmediate reactions to penicillins.

Funding of the study

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Funding of the study
  7. References

Funding was obtained from MURST, (Italian) Ministry for University, Scientific and Technological Research.

References

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Funding of the study
  7. References