Randomized, Vehicle-controlled Trials of Topical 5-fluorouracil Therapy for Actinic Keratosis Treatment

An Overview

Maral Rahvar; Sonia A Lamel; Howard I Maibach


Immunotherapy. 2012;4(9):939-945. 

In This Article


The inciting factor in the development of AKs is UV radiation exposure eliciting DNA mutation. A thymine for cytosine substitution in certain areas of DNA in the P53 gene is an identified mutation common to both AKs and SCCs that leads to an alteration in DNA repair and subsequent apoptosis.[4,26] One study revealed that a P53 mutation is present in 53% of AKs and 69–90% of SCCs, thus implicating the potential involvement of P53 mutations in the malignant conversion of AK to SCC.[27] Some risk factors for the development of AKs include light skin types, excessive UV exposure, older age, a high-risk occupation, location of residence, a personal history of AKs and/or skin cancers and gender (male). Among these factors, advanced age, a history or presence of other AKs, and a history of non-melanoma skin cancer are the most significant.[7,28]

The 5-FU molecule is an analog of thymine, which works as a false base. After entering a cell like all other natural nucleotides, it gets phosphorylated and ribosylated. Next, by bonding to thymidylate synthetase enzyme, it inhibits synthesis of thymidine nucleotides from deoxyuridine nucleotides. Thymidine depletion eventually results in reduction of DNA and cell apoptosis.[18,29] An additional benefit of 5-FU is its ability to make clinically imperceptible lesions visible and therefore more amenable to treatment.[30] In patients with extensive sun damage and presence of lesions, treatment with 5-FU may be beneficial to more effectively diagnose additional lesions and more accurately monitor treatment response.

In the USA, topical 5-FU is available in 0.5, 1, 2 and 5% concentrations, formulated in either a solution or cream. Some commonly reported adverse effects include irritation, rash, xerosis and allergy.[31,101] As with other topical medications, formulations with higher 5-FU concentrations result in greater absorption and thus systemic exposure. In a comparison of patients applying 0.5% topical 5-FU once daily with patients applying 5% topical 5-FU twice daily, the urine concentrations varied by a factor of 40.[32] However, another study showed that the percentage of 5-FU in the skin after a 24-h exposure is 86–92% for 0.5% formulations versus 54% for 5% formulations (p < 0.001).[33]

In this review of randomized, vehicle-controlled trials evaluating the efficacy of 5-FU in the treatment of AKs, the rate of vehicle responders achieving total clearance of AKs was close to zero. The number of vehicle responders reaching partial clearance, however, was much higher, with treatment responses ranging from 21.6 to 34.4%. The vehicle responses observed in topical 5-FU trials appear to corroborate those in efficacy evaluations of other topical AK treatments. One efficacy evaluation of 5% imiquimod cream as compared with its vehicle showed a median reduction of lesions of 86.6% in the imiquimod-treated group and 14.3% for the vehicle-treated group.[34] Another efficacy investigation of 5% imiquimod cream compared with its vehicle, both applied 3 days a week for either 1 or 2 weeks, reported a 61% partial reduction of lesions in the active drug group versus a 25.2% partial reduction in the vehicle group.[35]

In an efficacy study of 3% diclofenac gel for AKs, percentage reduction in lesion size of those in the diclofenac-treated group was 64.7 versus 34.3% for vehicle-treated lesions after a 3-month treatment course.[36] In a study examining topical tretinoin for the treatment of AKs, 66% of patients who applied tretinoin cream had a more than 30% reduction in lesion size compared with 45% of patients who applied vehicle.[37] In another study evaluating 0.05% tretinoin, 1% tretinion and vehicle cream for 1265 patients with histologically confirmed AKs, 40% of patients in the vehicle group showed reduction of AKs.[38] The significant responses in patients treated with a presumed biologically inert substance imply that influences other than improvement in the disease state of AKs are responsible for the vehicle treatment responses.

Potential reasons for the observed improvement in outcomes in vehicle groups include the natural regression of AK lesions, variation in investigator-determined lesion counts and differences in the measurements of AK clearance. The assessment of AK lesions is an imperative and arduous aspect of clinical study design when evaluating AK treatments. Furthermore, interpretation of the data in making efficacy statements is highly dependent on the reliability and accuracy of investigators' assessments of lesion sizes and counts. For instance, different physicians might employ different approaches in assessing the number and size of individual lesions.

In a clinical examination of the reliability of direct counting methods of visible AKs, multiple experienced dermatologists counted the lesions of multiple patients individually.[39] Comparison of their lesion assessments showed poor reliability. The dermatologists then discussed the discrepancies in their assessments before examining another group of patients with AKs again separately. While there were fewer inconsistencies in lesion counts, poor reliability was again documented.

In a prospective, single-blinded study examining the reliability of investigators' body surface area assessments and direct counting of AK lesions, two investigators assessed 37 subjects at baseline and 2 weeks later. Sufficient reliability of the assessments made by both the body surface area and counting methods was noted; however, only two investigators provided measurements, with one investigator having trained the other.[40] Furthermore, the four trials analyzed in our study employed the total (absolute) clearance and percentage reduction to measure treatment changes and two trials additionally utilized Physician Global Assessment of Improvement ratings as a parallel method. Data evaluating this as a ten-point scale are not currently present in the literature. The paucity and implications of the existing data on the accuracy and reliability of AK lesion assessment highlight the need for further investigation as to how these measures affect efficacy data and how disease assessments can be improved.

An additional hurdle in the assessment of AKs is the presence of subclinical lesions. It is known that the application of 5-FU often renders lesions not previously visually identifiable more prominent.[41] This implies that, during trial periods, patients in 5-FU treatment groups may appear to develop new lesions that were not present in baseline disease assessments. The natural regression of AK lesions may also occur over time. One study showed that up to 65% of AKs may clinically regress over 5 years.[17] Without accurate means of documenting the natural disease course of AKs as well as their response to treatment, efficacy measures are obscured by the responses documented in clinical trials. The addition of 'no treatment' groups may aid in determining a natural disease course reference.

The diagnosis of AKs in the majority of trials is done clinically. However, a limited number of studies have investigated the validity of diagnosing these lesions clinically. In general, the predictive value of the clinical diagnosis of AKs is good. One study evaluated the accuracy of clinically diagnosing AKs by confirmation with histopathology. In 34 out of 36 patients (94%), the diagnoses were correct. In another validation study, 22 patients were randomly selected and clinical diagnosis of AK was histologically confirmed in 20 of the participants by a single dermatopathologist.[42] While these studies imply that investigators can rely on clinical diagnosis, there is still a need to investigate further the accuracy of clinical diagnosis.

Importantly, the findings from this study imply that there is a need for improvements in study design, greater scrutiny of the measures used to capture efficacy and determine outcomes, and an enhanced knowledge of the overall 'placebo effect', as well as the natural disease course of AKs. The results of randomized controlled trials evaluating the efficacy of their treatment are subject to much variation. For example, the exclusion of patients with comorbidities such as immunosuppression, current non-melanoma skin cancers and recent treatment for AKs may result in elimination of a patient population more indolent to standard treatments. In clinical practice, the condition of these patients and the presence of AKs may warrant topical means for treatment more than patients without these comorbidities. This may lead to the employment of treatment lacking clinical evidence proving its safety.

The results of this study must be interpreted in the context of the study design. Because of the inclusion and exclusion criteria of the studies evaluated, the observed efficacies and adverse effects are only applicable to the population examined. Additionally, the studies only evaluated the treatment of AKs present on the face, and therefore the results may not be relevant to the treatment of AKs on other anatomical regions. Furthermore, as only studies evaluating 5-FU in the treatment of AKs were included, the limited data did not allow for conclusive remarks about the impact of study design factors on the observed efficacies in the active treatment and vehicle groups.