Key points

Introduction

The term “drug reactions” is relevant to dermatology in three categories of reactions: (1) cutaneous drug reactions without systemic features, (2) cutaneous drug reactions with systemic features, and (3) systemic drugs prescribed by a dermatologist with systematic adverse effects. This article is not intended to be comprehensive, but instead uses three examples from each of these categories to illustrate several important principles central to drug reaction diagnosis and management. There are several important areas of overall drug safety that are not included in this article, including drug interactions and medical-legal risk management.

Introduction

The term “drug reactions” is relevant to dermatology in three categories of reactions: (1) cutaneous drug reactions without systemic features, (2) cutaneous drug reactions with systemic features, and (3) systemic drugs prescribed by a dermatologist with systematic adverse effects. This article is not intended to be comprehensive, but instead uses three examples from each of these categories to illustrate several important principles central to drug reaction diagnosis and management. There are several important areas of overall drug safety that are not included in this article, including drug interactions and medical-legal risk management.

Categories of drug reactions discussed

Purely cutaneous drug reactions ( Table 1 ) include morbilliform reactions (synonyms include exanthematous and maculopapular reactions), fixed drug eruption, and linear IgA bullous dermatosis.

Cutaneous drug reactions with systemic features include Stevens-Johnson syndrome (SJS)/toxic epidermal necrolysis (TEN) spectrum, drug-induced hypersensitivity syndrome also known as drug reaction with eosinophils and systemic symptoms (DRESS), and acute generalized exanthematous pustulosis.

Dermatologic systemic drugs with systemic adverse effects include methotrexate (MTX)-induced chronic liver disease, cyclosporine-induced chronic kidney injury, and biologic therapy–induced hepatitis B reactivation.

General principles

In general, clinicians work with less than 100% certainty in making most medical decisions. I have found the following decision triad to have significant clinical value in reaching the highest level of certainty possible ( Table 2 ).

• •
  

  Literature experience: with all the pitfalls discussed next, realizing there is no subject for which all articles agree

  
  
• •
  

  Personal experience: direct (clinician’s own experience) and indirect (shared experience of mentors and colleagues)

  
  
• •
  

  Biologic plausibility: mechanistically what is logical or makes sense

There are several general limits of this decision triad. In most clinical scenarios, diagnostic level of certainty falls well short of 100% ( Table 3 ). For example, in clinical trials, a significant P = .05 still leaves 5% possibility that the findings were caused by chance alone. Smaller P values (≤0.01) substantially improve the level of certainty. That example is substantially more precise than diagnosing a drug reaction in an individual patient. In general, clinicians make drug reaction diagnostic decisions using the preponderance of evidence that is available at the time the decision must be made.

A second general concept when facing a clinical decision involving a possible drug reaction is the risk/benefit ratio, where the risk of possible drug reaction is compared with the potential benefit of continuing the drug/drugs in question. Given that most of the time this clinical scenario occurs in an outpatient office setting, at least briefly discontinuing the drug/drugs in question is wise until further diagnostic evidence becomes available. A noteworthy exception to this general rule is with prednisone (or other forms of systemic corticosteroid) therapy in which abrupt cessation of therapy has significant potential risks including an addisonian crisis. However, such corticosteroid therapy is seldom central to the diagnostic decisions of any drug reactions.

Pitfalls and limitations in general

Before looking closely at the proposed causation algorithm, a discussion of some general pitfalls of medical decision making concerning drug reactions is presented next. The following are no doubt influenced by the clinician’s temperament and his or her inherent level of awareness for information inherent to medical decision making involving drug reaction diagnosis.

There are two types of general errors: errors of underconcern and errors of overconcern. The errors of underconcern include (1) failure to monitor for liver toxicity with dapsone and azathioprine, (2) failure to screen for possible hepatitis B reactivation with biologic therapeutics for psoriasis, and (3) minimal to no MTX surveillance in an individual with multiple predictors of nonalcoholic steatohepatitis. Errors of overconcern are excessive concern for a morbilliform reaction in a patient not receiving drugs known to induce DRESS syndrome; or managing morbilliform drug reactions without systemic findings of DRESS syndrome, such as the absence of eosinophilia, liver transaminase changes, or findings of acute kidney injury (urinalysis changes or rise in creatinine). The ideal level of concern is difficult to define, but includes appropriately cautious measures to monitor for drug reactions discussed in this review.

In addition, there are two other distinct general errors clinicians commonly make. The first is failure to adapt to compelling new literature information (complete skepticism); this is in essence saying “don’t confuse me with the facts, my mind is made up.” The second is no careful analysis or scrutiny for any new literature information (complete gullibility); an example of this approach includes “the last article is always right,” believing the last article published automatically negates the validity all prior articles on the same topic. A happy medium to sort out literature trends is finding a healthy skepticism. Tools for sorting out such trends are detailed in later sections.

Causation determination through an algorithm

Even with careful use of the following algorithm, diagnostic certainty typically falls well short of 100% (See Table 3 ). This is the reality all clinicians face, yet once again the preponderance of evidence needs to be used in the necessary clinical decisions; making no decision (while awaiting possible future definitive studies) generally is not a viable option. The basic algorithm with some adaptation of the original terminology to simplify phraseology, is as follows

• •
  

  Challenge : this is a retrospective component assessing the composite of (1) literature reputation of the drug/drugs involved in causing the specific drug reaction, (2) the clinician’s personal experience both direct and indirect (as previously defined), (3) the individual patient’s prior experience with the drug/drugs involved, and (4) duration the patient has taken the drug/drugs in question.

  
  
• •
  

  Dechallenge : this is a prospective step concerning what happens to the clinical presentation of the drug reaction when the drug/drugs in question are stopped.

  
  
• •
  

  Rechallenge : this prospective step is the most definitive step of the algorithm concerning clinical response after the drug/drugs in question are restarted (see limitations below in particular with this step)

  
  
• •
  

  Exclusion : this prospective step includes excluding other etiologies for the same clinical presentation (ruling out viral hepatitis in patients with possible drug-induced liver injury) and testing for systemic aspects of a given clinical presentation (eg, testing for hematologic, liver, renal, and thyroid elements of DRESS syndrome)

Table 3

Causation algorithm components

Components	Subcomponents	Pitfalls
Kramer Criteria Algorithm
Challenge	1. Literature reputation of suspected drug 2. Clinician’s personal experience 3. Patient experience with drug/drugs of concern 4. Duration patient has been taking the drug in question	1. Literature reputation of drug may be unclear 2. Delayed timing from initiation of drug therapy
Dechallenge	1. Clinical response of drug reaction after drug cessation 2. Keep drug metabolism half-life in mind	1. May be partial or no resolution 2. Inherent disease activity may complicate decision making 3. Prolonged half-life in serum or fat
Rechallenge	1. Clinical response to restarting drug/drugs in question 2. Infrequently use intentionally (at times done unintentionally)	1. Never rechallenge potentially life-threatening drug reactions 2. Seldom rechallenge reaction with significant morbidity 3. Perform if no suitable alternative exits for potentially serious disease
Exclusion	1. Exclude nondrug causes of reaction pattern 2. Exclude systemic aspects of drug reaction	1. Some nondrug causes not easily diagnosed 2. Some drug reaction systemic components may be delayed
Additional FDA criteria
Biologic plausibility	Mechanism of drug correlates with drug reaction mechanism	Most drug reactions do not have clear-cut mechanism
Class effect	Members of a drug class typically cross-react	Tetracyclines in particular differ in types of reactions
Dose relationship	Subtoxic doses often correlate with likelihood of reaction	Most reactions without clear dose correlation

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

Psoriasis Trends and Practice Gaps Practice Gaps and Training Gaps: Delineating What We Need to Fix Clinical and Educational Gaps in Diagnosis of Nail Disorders Practice and Educational Gaps in Dermatology Practice and Educational Gaps in Cosmetic Dermatologic Surgery Practice and Educational Gaps in Light, Laser, and Energy Treatments

Stay updated, free articles. Join our Telegram channel

Join