Review article
Characterizing the phenomenon of radiation recall dermatitis

https://doi.org/10.1016/S0167-8140(01)00328-0Get rights and content

Abstract

Radiation recall represents the ‘recalling’ of an effect similar in appearance to that of an acute radiation reaction in a previously irradiated field. The recall is triggered by the administration of certain drugs days to years after the exposure to ionizing radiation. This review focuses almost exclusively on the skin manifestations of radiation recall to assemble the largest data base upon which to discuss this rare phenomenon. No absolute radiation dose threshold is apparent, but rather an interplay between dose and time before drug exposure seems to affect both the risk and speed of onset of recall. Recall usually occurs on first exposure to a particular recall-triggering drug. The skin reaction develops within minutes to days. The time to develop the reaction may be slightly longer for oral than intravenously administered drugs reflecting their bioavailability. Most drugs associated with recall are cytotoxics, but several other drugs may elicit the phenomenon. Individuals exposed to a number of potentially recall-triggering drugs reveal the marked drug specificity characteristic of the phenomenon. Skin reactions usually settle within a few days of stopping the triggering drug. The role of steroids or anti-histamines in affecting resolution is unclear. Drug rechallenge tends to produce either only a mild recurrence or no recurrence of recall. Steroids or dose reduction may favour uneventful rechallenge. A number of aetiological hypotheses on radiation recall exist. Using the available evidence these hypotheses are critically reviewed and a novel hypothesis based on radiation affecting local cutaneous immunological responses proposed.

Introduction

Radiation recall is well recognized as a concept in the field of radiation oncology but rarely clinically seen. It represents the development of an inflammatory reaction throughout a previously irradiated area, precipitated by the administration of certain drugs. The drugs are often, but not exclusively, intravenous cytotoxics (cf. Section 6). There may be a considerable time period between the radiation exposure and the administration of a recall-triggering drug (cf. Section 3).

The phenomenon was first described at the Childrens Medical Center in Boston in association with actinomycin D, when D'Angio and co-workers [11] reported that:

Actinomycin D therapy alone can reactivate ‘latent’ radiation effects in normal tissues...In skin previously irradiated but normal in appearance an erythema may develop during actinomycin D therapy, identical in type to that produced by roentgen irradiation. This response is sharply restricted to the areas previously treated by X-rays. The severity of the reaction varies from a mild reddening to pronounced desquamation.

In the intervening 40 years, many other drugs have been associated with radiation recall reactions (Table 1), but the recall itself remains a loosely characterized and poorly understood phenomenon. When cases occur physicians often have to decide between running the risk of further recall reactions or discontinuing therapeutically useful drugs with little published data upon which to base their decisions. This review attempts to draw together what is known about the phenomenon, based largely on what has only been described in case reports, and what can be hypothesized given the available evidence. The vast majority of cases described in the literature involve dermatitis as the overt manifestation of radiation recall. To limit the discussion to an unequivocal phenomenon as much as possible, the clinical data reviewed will therefore concentrate almost exclusively on radiation recall dermatitis (RRD). It is important to note however, that radiation recall pneumonitis, oesophagitis and mucositis have also been described [5], [18], [35], with appropriately serious clinical consequences.

Section snippets

Towards a tighter timing definition

No formal grading system for RRD exists. Where reported cases of RRD have been described in enough detail to analyse (Table 2), we have graded each reaction using a scale based on the RTOG Acute Radiation Morbidity Criteria for skin (taking into account the terms ‘pruritus,’ ‘vesiculation’ and ‘urticaria’ common in the descriptions of RRD but not contained in the original RTOG system):

    Grade 1 (mild):

    Erythema±pruritus±dry desquamation

    Grade 2 (mild-moderate):

    Grade 1 with pain, oedema, urticaria

Range of timing intervals

The longest recorded interval for any recall phenomenon was that reported by Burdon et al. [5] in a man with adriamycin-induced stomatitis 15 years after radiotherapy treatment of a palatal sarcoma. The longest interval between radiotherapy and drug triggering a recall dermatitis is 7 years [26]. In this patient, however, the previously irradiated skin showed significant late radiation change prior to drug administration. In addition, the recall reaction only manifested itself after six cycles

Incidence

There is little published directly on the incidence of RRD, as cases tend to be described with no information on the numbers treated similarly who have not developed RRD. Even if the background information is provided, most reports include only a single case of RRD, therefore overestimates of the incidence are likely. Certain drugs do, however, seem to be associated with RRD more frequently than others. Given these caveats, estimates of incidence may be made for individual drugs.

At one extreme

Radiation characteristics

The majority of case reports on RRD specify the radiation dose to the tumour, rather than to the skin where the RRD is later manifest. The tumour dose in cases of RRD ranges from 10 to 61.2 Gy (Table 2). Even allowing for the uncertainty regarding skin dose, a lower estimate of the radiation dose required for RRD based on this information remains inappropriate to calculate. This is because the effect of time elapsed from radiation exposure also has to be considered. A radiation dose necessary

Drug characteristics

The majority of drugs (Table 1) reported in association with RRD have been given as monotherapy. When drugs, notably chemotherapy drugs, have been given in combination, no particular interactions either increasing or decreasing the risk of RRD have been documented. The features that the drugs in Table 1 share that explain their association with RRD are unknown. Interestingly, RRD has never been reported in association with a number of commonly used chemotherapy drugs, such as cyclophosphamide

The nature of the skin reaction

Pronounced acute radiation reactions during radiotherapy are not required for subsequent RRD, indeed their existence is unusual [34], [43]. Moreover, patients having experienced RRD may have other areas of skin irradiated subsequently with no abnormal radiation reaction [36]. There is no known association between the development of RRD and any subsequent late skin reactions.

The speed of onset of RRD reactions is particularly interesting (Table 2). The range with intravenous drugs is from a few

Rechallenge with recall-triggering drugs

The majority of patients reported with RRD are not rechallenged with the RRD implicated drug, presumably due to fears about recurrence. In the literature of RRD evidence to justify or refute this is limited, but 15 patients rechallenged with their RRD-inducing drug can be confidently identified [7], [14], [15], [17], [25], [29], [30], [31], [34], [35], [38], [47]. Of these, ten showed recurrence of RRD [14], [15], [17], [25], [30], [34], but the RRD was milder than the initial presentation in

Potential aetiologies

Any viable aetiological hypothesis must be consistent with the key characteristics of RRD suggested in this review. A number of different hypotheses have been proposed for RRD although with little evidence base to support any of them. Essentially the hypotheses focus on either vascular, epithelial stem cell inadequacy, epithelial stem cell sensitivity or drug hypersensitivity reactions as the mechanism for RRD. Each will be covered in turn.

Conclusions

The ability of certain drugs to elicit inflammatory reactions in previously irradiated skin long after the initial effects of the radiation have apparently disappeared has been recognized for over 40 years.

Potential areas of confusion exist between radiation recall, radiosensitization and interference with ongoing acute radiation reactions. Guidelines to distinguish these have been proposed in this review.

The aetiology of RRD remains unknown. A number of hypotheses have been described. We have

References (47)

  • V.T. Raghavan et al.

    Taxol and radiation recall dermatitis

    Lancet

    (1993)
  • G.G. Steel

    The search for therapeutic gain in the combination of radiotherapy and chemotherapy

    Radiother Oncol

    (1988)
  • B.U. Wintroub et al.

    Cutaneous drug reactions: pathogenesis and clinical classification

    J Am Acad Dermatol

    (1985)
  • W. Yeo et al.

    Radiation-recall dermatitis with docetaxel: establishment of a requisite radiation threshold

    Eur J Cancer

    (1997)
  • S.A. Aristizabal et al.

    Adriamycin-irradiation cutaneous complications

    Int J Radiat Oncol Biol Phys

    (1977)
  • A. Bostrom et al.

    Radiation recall: another call with tamoxifen

    Acta Oncol

    (1999)
  • J. Burdon et al.

    Adriamycin-induced recall phenomenon 15 years after radiotherapy

    J Am Med Assoc

    (1978)
  • H.J. Burstein

    Side effects of chemotherapy. Case 1. Radiation recall dermatitis from gemcitabine

    J Clin Oncol

    (2000)
  • D.R. Camidge et al.

    Docetaxel-induced radiation recall dermatitis and successful rechallenge without recurrence

    Clin Oncol

    (2000)
  • J.R. Cassady et al.

    Radiation-adriamycin interactions: preliminary clinical observations

    Cancer

    (1975)
  • D. Castellano et al.

    Side effects of chemotherapy. Case 2. Radiation recall reaction induced by gemcitabine

    J Clin Oncol

    (2000)
  • G.J. D'Angio et al.

    Potentiation of X-ray effects of actinomycin D

    Radiology

    (1959)
  • G.J. D'Angio

    Clinical and biological studies of actinomycin D and roentgen irradiation

    Am J Roentgenol

    (1962)
  • Cited by (255)

    • Radiation recall dermatitis induced by ibuprofen

      2023, Anais Brasileiros de Dermatologia
    • Radiation Recall Dermatitis after Donor Lymphocyte Infusion for Adult T-Cell Leukemia Lymphoma Post–Allogeneic Stem Cell Transplant

      2023, Practical Radiation Oncology
      Citation Excerpt :

      Its mechanism remains unknown. Camidge and Price proposed that the rarity, speed of onset, unpredictable effect, and extreme drug specificity of RRD was most consistent with an idiosyncratic drug hypersensitivity reaction.7 Irradiation may lower the inflammatory response threshold through continued low-level secretion of the inflammation-mediating cytokines.

    • Management of skin toxicity

      2023, Palliative Radiation Oncology
    • Late toxicities management

      2023, Palliative Radiation Oncology
    • Stochastic Stomatitis: Radiation Recall Reactions in the Era of Immune Checkpoint Blockade

      2023, International Journal of Radiation Oncology Biology Physics
    • Radiation recall dermatitis induced by COVID-19 vaccination in breast cancer patients treated with postoperative radiation therapy

      2022, Breast
      Citation Excerpt :

      In our dataset we found the highest risk of having RRD when the first dose of COVID-19 vaccine was administered within the first 30 days after treatment with postoperative EBRT. Even though there is no data regarding the specific timing for RRD development, one paper estimated that most cases occurred within 40 days Camidge et al. [18], in agreement with our results. As reported in the literature [19], we also found that the risk of having RRD did not depend on the postoperative radiation dose (dose schedule or boost addiction), age, breast reconstruction or previous chemotherapy or by a previous COVID-19 infection.

    View all citing articles on Scopus
    View full text