Introduction

Actinic keratosis (AK; solar keratosis or squamous cell carcinoma in situ), is a localized area of dysplasia with malignant potential and regarded as a strong predictor of a subsequent squamous cell carcinoma (Memon et al., 2000). Estimates of progression of AK to invasive squamous cell carcinoma range from <1 to 16% (Persaud et al., 2002). It is impossible to predict the point at which an individual AK lesion will evolve into invasive squamous cell carcinoma, so most clinicians advocate the treatment of all AK lesions (Lebwohl et al., 2004). Squamous cell carcinoma of the skin has potential to metastasize and may account for up to 20% of deaths from skin cancer (Szeimies et al., 2004).

Most AKs are diagnosed clinically (red, scaly papules or plaques, 2–5 mm in diameter, occurring in sun-exposed sites such as face, ears, neck, forearms, hands, and balding scalp) and rarely confirmed histologically. A positive predictive value of 74% for diagnosis of AK suggests that the threshold for biopsy of suspect lesions in patients with a history of skin cancer should be low and warrant further evaluation (Venna et al., 2005).

AK development is associated with individual susceptibility (i.e. fairness of skin), cumulative ultraviolet radiation exposure and advancing age (Perras, 2004). Although the highest incidence is reported in Australia (40–60% in white-skinned adults; Lebwohl et al., 2004), a recent study in England demonstrated that sun exposure received in "normal" life is sufficient for potentially malignant skin damage in a significant proportion of the population (Memon et al., 2000).

Current treatment of AK involves surgical or non-surgical interventions or a combination. Cryosurgery and curettage are used for small areas with few lesions. Other treatment options include chemical peels, dermabrasion, laser therapy, excision, and photodynamic therapy (Perras, 2004), as well as retinoid therapy and intralesional interferon. Currently, prescribed therapies can be painful and may result in scarring and changes in skin pigmentation (Stockfleth et al., 2004). A recent meta-analysis of three randomized trials of topical diclofenac with sodium hyaluronate (Pirad et al., 2005) reported the outcome of complete resolution of all lesions 30 days after the end of treatment in 39% of patients, with a number needed to treat (NNT) of 3.7 (95% confidence interval 2.8–5.5). Another review reported the same outcome with 5-fluorouracil in 53% of patients, with an indicated NNT of about 2 (Jorizzo, 2005). Direct comparisons between them are not available, and indirect comparison is limited by few trials and patients.

Imiquimod 5% cream has been available since 1997 for the treatment of genital and perianal warts (Moore et al., 2001). Imiquimod, an imidazoquinoline amine, is an immune-response modifier that stimulates both innate and adaptive arms of the immune system (Sauder, 2003). This novel therapy therefore targets a critical component that current therapies do not: the suppression of the immune response (Lober and Fenske, 2004; Korman et al., 2005). Differences in gene expression between AK and uninvolved skin have been demonstrated, and recent evidence suggests that imiquimod causes pharmacological modulation of expression in this context (Lysa et al., 2004). Treatment with 75 mg three times per week for 16 weeks demonstrated minimal systemic absorption and good safety margins for topical administration (Harrison et al., 2004), although adverse effects include headaches, influenza-like symptoms, and myalgia.

This is the first meta-analysis of short-term randomized-controlled trials evaluating the efficacy and safety of imiquimod 5% cream for the treatment of AK.

Results

Literature searches identified 19 publications, 14 of which were excluded (Table S1) and five included (details in Table S2). Of 1,293 included patients, almost 90% were men. The mean age ranged from 64 to 71 years, with individual patients aged from 37 to 94 years. More patients had Fitzpatrick skin types I and II, but individuals with types I–IV were included. All studies diagnosed AK by clinical examination, supplemented by biopsy and histology in two trials (Stockfleth et al., 2004; Szeimies et al., 2004).

Two studies (Stockfleth et al., 2004; Chen et al., 2003) were small, with about 40 patients, while the other three (Lebwohl et al., 2004; Szeimies et al., 2004; Korman et al., 2005) each had 280–500 patients. Trial duration was 12–16 weeks. Quality scores were high; all trials reported being both randomized and double blind, scoring 3 or more of the maximum 5 points. All five trials used one sachet of 5% imiquimod cream or vehicle cream (placebo) twice or three times a week; none used an active control. Cream was applied to specified areas of sun-exposed skin, usually 20–25 cm² on the face and balding scalp, but including neck, forearms, and hand in one trial (Stockfleth et al., 2004). All trials were conducted in the setting of home administration after initial professional examination and advice.

Complete clearance occurred in 50% of patients with imiquimod, compared with 5% with vehicle (Figure 1; Table 1). The NNT for one patient to have keratoses completely cleared after 12–16 weeks was 2.2 (95% confidence interval 2.0–2.5). For partial clearance the NNT was 1.8 (1.7–2.0) (Figure 1; Table 1). For both complete and partial clearance, results were consistent between individual trials. NNTs for the three trials reporting formal lesion-counting methods were lower (better) than for all trials combined (Table 1).

Figure 1.

Complete and partial clearance with imiquimod and vehicle in individual trials at 12–16 weeks. Each symbol represents a single trial, and the size of trial is shown in the inset scale. The diagonal line shows the line of equivalence between imiquimod and vehicle.

Full figure and legend (18K)

Table 1 - Results for clearance of keratosis, withdrawals, and adverse events.

Full table

Adverse event information was available on over 1,200 patients. The proportion of patients with any adverse event, local adverse event, or treatment-related adverse event was substantially higher with imiquimod than with vehicle (Table 1), and number needed to harm to produce one additional adverse event with imiquimod over 12–16 weeks ranged from 3.2 to 5.9. Serious adverse events were not significantly different between imiquimod or vehicle control, with a relative risk of 1.2 (0.7–2.0).

Most adverse events were local, most frequently erythema, scabbing, and flaking. Figure 2 shows the percentage of patients reporting each of a number of local adverse events with imiquimod. All of these, apart from vesicles, occurred significantly more frequently with imiquimod than with vehicle.

Figure 2.

Percentage of imiquimod-treated patients reporting particular severe local adverse events.

Full figure (30K)

All-cause withdrawals occurred about twice as frequently with imiquimod than with vehicle (Table 1), with an NNT to cause one additional withdrawal of 20 (12–55). Withdrawals due to adverse events were not significantly different between imiquimod and vehicle, with a relative risk of 1.5 (0.8–2.7). There were no data concerning withdrawals due to lack of efficacy.

Discussion

This systematic review is the first comparison of randomized-controlled trials for the use of imiquimod 5% cream in the treatment of AK. The five trials, with 1,293 patients were of high reporting quality minimizing the risk of bias, and were valid, involving patients with multiple lesions, using home application to mimic clinical practice, and of sufficient duration to measure both benefits and harm, though not recurrence.

There were limitations, however. Two of the five studies used histological rather than clinical diagnoses of AK. Although histology has a positive predictive value of 74% for the diagnosis of AK (Venna et al., 2005), in practice clinical rather than histological diagnosis is likely. Studies were clinically homogeneous, with 97% of patients having ketatoses on the face and balding scalp only; one small study (Stockfleth et al., 2004) recruited some patients with lesions on thicker skin of neck or forearms. The reliability of using AK lesion counts as an outcome has been questioned (Weinstock et al., 2001). Studies dealt with this potential problem by using photographs or diagrams and the same investigator counting lesions (Stockfleth et al., 2004; Chen et al., 2003) and one (Szeimies et al., 2004) favorably compared clinical counting with histology, and had lower, better, NNTs than those that did not report using these methods (Table 1).

The NNT of 2.2 (2.0–2.5) for complete clearance of AK lesions, and 1.8 (1.7–2.0) for partial clearance, shows that imiquimod 5% cream is an effective therapy, at least in the short term. This compares with NNTs of 3.7 and 2.0 for topical diclofenac and 5-fluorouracil for complete clearance about 4 weeks after the end of treatment.

Although treatment with imiquimod cream is associated with adverse events, particularly local skin reactions, this could be deemed important in arriving at the optimal efficacy. Better clearance was seen in subjects with the most severe adverse events (Stockfleth et al., 2004). Chen et al. (2003), found that better clearance occurred in subjects with larger increases in AK lesions during initial therapy, and Korman et al. (2005), found that both an increase in the number of lesions and intensity of local skin reaction was associated with better clearance rates.

Optimal frequency and duration of treatment remain to be determined. A small open-label trial in 25 subjects evaluated cycle therapy using 5% imiquimod cream, suggesting that alternative dosing regimens may reduce adverse effects (Salasche et al., 2002).

Long-term effectiveness was addressed by only one study in this review (Stockfleth et al., 2004). Twenty-four of 27 patients randomized to imiquimod (three applications per week) were reviewed 1 year after final application, and clinical recurrence was found in two, giving a recurrence rate of about 10%. One further study (Lee et al., 2005) followed up patients with complete clearance from four Phase III studies, after a median of 16 months, in which 82% of eligible patients underwent clinical examination. Recurrence occurred in 19/77 (25%) of those treated with imiquimod three times per week, 23/54 (43%) of those treated twice a week, and 7/15 (47%) treated with vehicle cream. No clinically meaningful long-term adverse events were reported, and skin quality in the treated area showed no long-term adverse changes.

This review demonstrates the effective use of imiquimod 5% cream in the short-term treatment of AK. Future investigations should be aimed at elucidating optimal dosing regimens to minimize adverse events without detriment to efficacy, and outcomes beyond 16 weeks are also needed.

Materials and Methods

QUORUM guidelines were followed (Moher et al., 1999). Full journal publications in any language of randomized trials of imiquimod for AK were sought. Eligible reports were identified from MEDLINE (to August 2005), Cochrane Library (Issue 2, 2005), and PubMed (to August 2005), using the terms: (imiquimod or aldara) and ((actinic or solar) and keratosis) and (random OR randomized). Review articles (Lober and Fenske, 2004; Perras, 2004) and reference lists were also used. The manufacturers of Aldara™ (imiquimod 5% cream; 3 M, Loughborough, Leicestershire, UK) were asked about any published or unpublished trials. Authors were not contacted regarding unpublished reports or additional information from published reports.

Included were full publications of randomized double-blind trials investigating imiquimod for AK with efficacy or safety data. Excluded were reviews with clinical information published elsewhere, biochemical or immunological studies, abstracts, and studies in conditions other than AK.

Each report was scored for quality using a three-item quality scale (Jadad et al., 1996). A maximum of five points were awarded to studies according to whether they were randomized, double blind, and accounted for withdrawals or drop-outs, and whether the methods of randomization and double blinding were described and appropriate.

Information was extracted independently by two reviewers (G.H., S.D.), and checked by the third (R.A.M.). From each trial we extracted the number of patients treated per group, dosing regimes, study design (including the rigor of lesion-counting methods, such as use of photography, and a single assessor), and the number of patients with efficacy and/or safety outcomes. Three main efficacy outcomes and five harm outcomes were sought from the trials, using as denominator the number of patients randomized, so that results were on an intention-to-treat basis. They were:

• Complete clearance of lesions in the treatment area.
• Partial clearance; at least 75% reduction in the number of lesions in the treatment area.
• Complete clearance of AK and no recurrence thereafter.
• Patients withdrawing for any cause.
• Patients withdrawing from the study because of treatment-related adverse effects.
• Patients withdrawing from the study because of lack of efficacy.
• Patients with at least one adverse event.
• Patients with serious, treatment-related, and particular local or systemic adverse effects.

Relative benefit and risk estimates were calculated with 95% confidence intervals using a fixed effects model (Morris and Gardner, 1995). Heterogeneity tests were not used as they have previously been shown to be unhelpful (Gavaghan et al., 2000). Publication bias was not assessed using funnel plots as these tests have been shown to be unhelpful (Sterne et al., 2000; Terrin et al., 2005); homogeneity was assessed visually (L'Abbe et al., 1987). The NNT and number needed to harm with confidence intervals were calculated by the method of Cook and Sackett (1995). Relative benefit or risk was considered to be statistically significant when the 95% confidence interval did not include 1. NNT or number needed to harm values were only calculated when the relative risk or benefit was statistically significant, and are reported with the 95% confidence interval.

Conflict of Interest

The authors state no conflict of interest.

References

1. Chen K, Yap LM, Marks R, Shumack S (2003) Short-course therapy with imiquimod 5% cream for solar keratoses: A randomized control trial. Aus J Dermatol 44:250–255 | Article |
2. Cook RJ, Sackett DL (1995) The number needed to treat: a clinically useful measure of treatment effect. BMJ 310:452–454 | PubMed | ChemPort |
3. Gavaghan DJ, Moore RA, McQuay HJ (2000) An evaluation of homogeneity tests in meta-analyses in pain using simulations of individual patient data. Pain 85:415–424 | Article | PubMed | ISI | ChemPort |
4. Harrison LI, Skinner SL, Marbury TC, Owens ML, Kurup S, McKane S et al. (2004) Pharmacokinetics and safety of imiquimod 5% cream in the treatment of actinic keratoses of the face, scalp, or hands and arms. Arch Dermatol Res 296:6–11 | PubMed | ISI | ChemPort |
5. Jadad AR, Moore RA, Carroll D, Jenkinson C, Reynolds DJ, Gavaghan DJ et al. (1996) Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials 17:1–12 | Article | PubMed | ISI | ChemPort |
6. Jorizzo JL (2005) Current and novel treatment options for actinic keratosis. J Cutan Med Surg 8(Suppl 3):13–21 | Article |
7. Korman N, Moy R, Ling M, Matheson R, Smith S, McKane S et al. (2005) Dosing with 5% imiquimod cream 3 times per week for the treatment of actinic keratosis: results of two phase 3, randomized, double-blind, parallel-group, vehicle-controlled trials. Arch Dermatol 141:467–473 | Article | PubMed | ISI | ChemPort |
8. L'Abbe KA, Detsky AS, O'Rourke K (1987) Meta-analysis in clinical research. Ann Intern Med 107:224–233 | PubMed | ChemPort |
9. Lebwohl M, Dinehart S, Whiting D, Lee PK, Tawfik N, Jorizzo J et al. (2004) Imiquimod 5% cream for the treatment of actinic keratosis: results from two phase III, randomized, double-blind, parallel group, vehicle-controlled trials. J Am Acad Dermatol 50:714–721 | Article | PubMed | ISI |
10. Lee PK, Harwell WB, Loven KH, Phillips TJ, Whiting DA, Andres KL et al. (2005) Long-term clinical outcomes following treatment of actinic keratosis with imiquimod 5% cream. Dermatol Surg 31:659–664 | PubMed | ISI |
11. Lober BA, Fenske NA (2004) Optimum treatment strategies for actinic keratosis (intraepidermal squamous cell carcinoma). Am J Clin Dermatol 5:395–401 | Article | PubMed | ISI |
12. Lysa B, Tartler U, Wolf R, Arenberger P, Benninghoff B, Ruzicka T et al. (2004) Gene expression in actinic keratoses: pharmacological modulation by imiquimod. Br J Dermatol 151:1150–1159 | Article | PubMed | ISI | ChemPort |
13. Memon AA, Tomenson JA, Bothwell J, Friedmann PS (2000) Prevalence of solar damage and actinic keratosis in a Merseyside population. Br J Dermatol 142:1154–1159 | Article | PubMed | ISI | ChemPort |
14. Moher D, Cook DJ, Eastwood S, Olkin I, Rennie D, Stroup DF (1999) Improving the quality of reports of meta-analyses of randomized controlled trials: the QUOROM statement. Quality of Reporting of Meta-analyses. Lancet 354:1896–1900 | Article | PubMed | ISI | ChemPort |
15. Moore RA, Edwards JE, Hopwood J, Hicks D (2001) Imiquimod for the treatment of genital warts: a quantitative systematic review. BMC Infect Dis 1:3 | Article | PubMed | ChemPort |
16. Morris JA, Gardner MJ (1995) Calculating confidence intervals for relative risk, odds ratios and standardised ratios and rates. In: Statistics with Confidence – Confidence Intervals and Statistical Guidelines (Gardner MJ, Altman DG, eds), London: British Medical Journal, 50–63
17. Perras C (2004) Imiquimod 5% cream for actinic keratosis. Issues Emerg Health Technol 61:1–4 | PubMed |
18. Persaud AN, Shamuelova E, Sherer D, Lou W, Singer G, Cervera C et al. (2002) Clinical effect of imiquimod 5% cream in the treatment of actinic keratosis. J Am Acad Dermatol 47:553–556 | Article | PubMed | ISI |
19. Pirad D, Vereecken P, Mélot C, Heenen M (2005) Three percent diclofenac in 2.5% hyaluron gel in the treatment of actinic keratosis: a meta-analysis of the recent studies. Arch Dermatol Res 297:185–189 | PubMed |
20. Salasche SJ, Levine N, Morrison L (2002) Cycle therapy of actinic keratoses of the face and scalp with 5% topical imiquimod cream: An open-label trial. J Am Acad Dermatol 47:571–577 | Article | PubMed | ISI |
21. Sauder DN (2003) Imiquimod: modes of action. Br J Dermatol 149(Suppl 66):5–8 | Article | PubMed | ISI | ChemPort |
22. Sterne JA, Gavaghan D, Egger M (2000) Publication and related bias in meta-analysis: power of statistical tests and prevalence in the literature. J Clin Epidemiol 53:1119–1129 | Article | PubMed | ISI | ChemPort |
23. Stockfleth E, Christophers F, Benninghoff B, Sterry W. (2004) Low incidence of new actinic keratoses after topical 5% imiquimod cream treatment: a long-term follow-up study. Arch Dermatol 140:1542 | Article | PubMed | ISI |
24. Szeimies RM, Gerritsen MJ, Gupta G, Ortonne JP, Serresi S, Bichel J et al. (2004) Imiquimod 5% cream for the treatment of actinic keratosis: results from a phase III, randomized, double-blind, vehicle controlled, clinical trial with histology. J Am Acad Dermatol 51:547–555 | Article | PubMed | ISI |
25. Terrin N, Schmid CH, Lau J (2005) In an empirical evaluation of the funnel plot, researchers could not visually identify publication bias. J Clin Epidemiol 58:894–901 | Article | PubMed | ISI |
26. Venna SS, Lee D, Stadecker MJ, Rogers GS (2005) Clinical recognition of actinic keratoses in a high-risk population. How good are we? Arch Dermatol 141:507–509 | Article | PubMed | ISI |
27. Weinstock MA, Bingham SF, Cole GW, Eilers D, Naylor MF, Kalivas J et al. (2001) Reliability of counting actinic keratoses before and after brief consensus discussion: the VA topical tretinoin chemoprevention (VATTC) trial. Arch Dermatol 137:1055–1058 | PubMed | ISI | ChemPort |

Acknowledgments

The study was supported with Pain Research funds and by the Oxford Pain Relief Trust. The manufacturers of imiquimod were not involved in any way with this review, and provided no direct or indirect funding.