Histologic findings are not pathognomonic for exposure to ionizing radiation [36]. Diagnosis essentially relies on clinical findings and an appropriate history of radiation exposure to the area of concern [13]. A skin biopsy may not be necessary if the presentation is very suggestive. Characteristic signs include erythema matching the position of the radiation port, and skin reaction evolving in typical temporal patterns. The histologic appearance of radiation dermatitis depends on the clinical presentation and on the phase of injury, and is too extensive to be dealt with in detail [37, 39, 41, 45, 53]. The main changes in the early phase are marked edema, degeneration of the basal cell layer, and inflammatory reaction. These changes are associated with progressive pyknosis of nuclei and cell death. Dilated blood vessels with endothelial proliferation, arteriolar thrombosis, and extravasation of RBCs can be seen. Hyperkeratosis is observed in dry desquamation. Intraepidermal blisters and loss of corneum are features of moist desquamation. In the late phase, the epidermis is irregular, with areas of atrophy and relative hyperplasia [39]. Dermal thinning with atrophy of the adnexal structures and dilatation of blood vessels is observed. Thickened, hyalinized collagen bundles and relative increase of elastic fibers are signs of dermal fibrosis that can lead to palpable induration of the skin. Atypical large stellate fibroblasts with enlarged hyperchromatic nuclei may be found at random in the dermis. These cells are a sign of radiation damage and usually are not observed in other skin diseases that also lead to end-stage dermal fibrosis [36, 39, 41, 47, 53].

The degree of skin response to radiation varies for different body sites. Kalz [54] found that skin sensitivity for acute reactions, from most to least sensitive, is as follows: anterior aspect of the neck, antecubital, and popliteal spaces; flexor surfaces of the extremities, chest, and abdomen; the face; the back and extensor surfaces of the extremities; the nape of the neck; the scalp; and the palms and soles. Hair follicles of the scalp appear to be more radiosensitive than those in other parts of the body [41].

A variety of reports have been published that suggest a correlation between exaggerated reactions after radiotherapy and connective tissue diseases, especially scleroderma, systemic and discoid lupus erythematosus, and mixed connective tissue disease [24]. Although a causative relationship for these rare observations is assumed, definite evidence is lacking. Diabetes mellitus and hyperthyroidism have also been associated with increased skin response to irradiation [36, 41, 55]. Wagner et al. [24] describe a patient with multiple health problems, including mixed connective tissue disease and diabetes mellitus, who underwent a transjugular intrahepatic portosystemic shunt procedure and later developed a severe necrotic ulceration. Patients carrying the homozygous form of the ataxia telangiectasia gene are also known to exhibit significant hypersensitivity to radiation [36, 42].

Skin sensitivity to radiation can be increased by various chemotherapeutic agents, such as actinomycin D, adriamycin, bleomycin, 5-fluorouracil, and methotrexate [36, 41, 56]. An early reaction that has healed over time can even manifest itself again in the same skin area when actinomycin D is given some weeks or months after the irradiation. This effect is known as a recall reaction. A similar reaction has been described in one patient after simvastatin administration [57]. Ciprofibrate, a fibric acid derivative used for the treatment of hyperlipidemia (not available in the United States), has been recently suggested to play a role in radiation-induced skin injury [9]. The patient in this case, however, also had lupus erythematosus, which may have contributed to the event, although the authors of the report noted that the disease could not be stimulated by ultraviolet radiation at that time.

Some patients discussed in this review did not seek the attention of their physician for radiation-induced changes. In our series [30] (Table 2, patients 1-7), we found that four of seven patients with chronic lesions were not aware of them at the time of discovery by physical examination. The skin lesions are frequently on the back and not directly visible to the patient. Moreover, many patients have limited mobility and may be less likely to notice skin changes. When lesions are minor, they may go unnoticed. A number of patients have had skin lesions that were visible but did not cause any pain. These patients disregarded the finding until the lesion became painful.

Patients usually do not return to the physician who initially performed the procedure when they seek medical advice for their skin problem [30]. Usually they present at some stage to a dermatologist but do not give a history of prior fluoroscopy because they assume it is irrelevant for the complaint or have forgotten about it. This fact is especially true when there is a significant latent period. The physician will not be given the correct history indicating the cause of the lesion unless the physician specifically asks. In some cases, even if the patient mentions prior fluoroscopy, the dermatologist has disregarded fluoroscopy as a possible cause because of lack of experience with the high doses from these procedures. In our review of 73 patients [30], an initial diagnosis of a fixed drug eruption, morphea (circumscribed cutaneous scleroderma), contact dermatitis, viral or bacterial infection, or a spider bite was made, [7, 11, 12, 24], including four from this report. Consequently, the correct diagnosis was delayed. In some cases the correct diagnosis was made after a delay of 2 years [13] (Table 2, patient 14) and 5-6 years [4] after the first appearance of the lesion.

A skin biopsy was taken in 27 patients to confirm the diagnosis [30]. However, as previously stated, the diagnosis can be made from a careful medical history and the appearance of the lesion. A biopsy may be helpful if other skin conditions must be considered; however, biopsy is not always necessary. Pezzano et al. [15] discourage skin biopsies because they leave a defect that heals poorly and that can result in a chronic ulcer. We concur with this advice.

In general, physicians have difficulty recognizing the cause of fluoroscopy-induced skin injuries because such injuries are rare, and modern fluoroscopy typically has not been associated with such injuries. Inability of physicians to recognize radiation-induced skin changes has led to misdiagnoses and prolonged and uncertain courses of treatment. Dermatologists and interventionalists must be aware of the potential for skin injuries and recognize the characteristics of such injuries to avoid misdiagnosis.

Biopsy is frequently performed. However, the results are not pathognomonic for radiation changes. Fluoroscopy-induced injuries can be recognized by the location of the injury as being congruent to the entrance of the X-ray beam and by the temporal pattern of the injury in relation to the fluoroscopy. Additionally, the injury often shows well-defined borders, which occur when the beam is not moved or resized during prolonged fluoroscopy over one site. A biopsy is usually not necessary and is not recommended because it may result in a nonhealing ulcer.

Some patients may be at greater risk for injury because of preexisting health conditions such as collagen vascular disease, diabetes mellitus, or ataxia telangiectasia, or because of a high radiation dose from a previous procedure. Good communication with the patient is essential. Interviewing the patient for potential high-risk conditions before a procedure is recommended, as is appropriate counseling. A short skin examination should be considered for patients who have had previous procedures. Further irradiation of any previous injury should be kept to a practical minimum and the patient counseled appropriately.

If a procedure is prolonged or the dose to the skin is known to be high, the patient should be advised to examine him- or herself about 2-3 weeks after the procedure to look for skin changes and to contact the interventionalist if any are observed. This information is not only good for patient care, it also assists in quality control because it indicates when dose levels have reached certain thresholds.

Address correspondence to L. K. Wagner.