Prospective Randomized Multicenter Clinical Trial on the Use of Interferon -2a Plus Acitretin Versus Interferon -2a Plus PUVA in Patients With Cutaneous T-Cell Lymphoma Stages I and II

CUTANEOUS T-CELL lymphomas (CTCL) represent a heterogeneous group of T-cell malignancies of the skin including classic mycosis fungoides and Sézary’s syndrome as the most common forms.1 In 1984, interferon α (IFNα) was first reported to be effective in the treatment of advanced and heavily pretreated CTCL.2 Because of high toxicity of high IFN doses reported in this early investigation, several studies have since been conducted with lower doses, showing IFNα to be similarly effective.3 To increase the response rates of 52% and an average duration of response of 4 to 28 months with IFN monotherapy, combinations of IFN with oral photochemotherapy (PUVA) or with retinoids were introduced into the treatment of CTCL. These combinations demonstrated higher response rates of about 96% with IFNα plus PUVA and 60% with IFNα plus retinoid.3However, these data were derived exclusively from uncontrolled trials.4-10 To prove combination therapies including IFN as a potentially high effective substance in the treatment of CTCL, a prospective randomized multicenter clinical trial, in cooperation with 21 departments of dermatology in Germany, Austria, and Switzerland, was conducted comparing the efficacy and tolerability of the combination of IFNα plus PUVA versus IFNα plus retinoid (acitretin).

Ninety-eight patients with histologically confirmed small to medium sized pleomorphic T-cell lymphoma or mycosis fungoides stage I or II were enrolled in this study. In addition, diagnoses were confirmed by T-cell receptor rearrangement.

Staging procedures included history and physical examination, chest x-ray, abdominal and lymph node ultrasound, complete blood count, electrolytes, creatinine, uric acid, total bilirubin, alkaline phosphatase, lactate dehydrogenase (LDH), serum glutamate oxalacetate transaminase (SGOT), serum glutamate pyruvate transaminase (SGPT), γ-glutamyl transferase (γ-GT), total serum proteins and electrophoresis, triglycerides, and total serum cholesterol. In case of elevation, differentiation of lipoproteins was performed.

After obtaining informed consent, the patients were randomized in a central institution (Estimate GmbH, Ausgsburg, Germany) either to treatment with the combination IFN α-2a plus oral photochemotherapy (49 patients) or IFN α-2a plus acitretin (49 patients). Treatment groups were stratified by pretreatment.

The primary study objective was the rate of complete remissions (CRs) in both groups. Secondary study objectives were rate of partial remissions (PRs), time to CR, tolerability of the treatments, and side effects. Initially, 65 evaluable patients were planned in both treatment groups, with an interim analysis after half of the patients had been followed for at least 48 weeks, showing significantly higher complete response rates in the IFN+PUVA group.

From the initially randomized 98 patients in this study, 6 patients did not receive any treatment within the study, 6 patients’ data were insufficiently documented by their study centers, and 4 entries resulted from staging errors.

Finally, 40 patients were evaluable in the IFN+PUVA group and 42 in the IFN+acitretin group. All patients had received a minimum of 4 weeks of treatment and photographs were taken before and after treatment.

Patients’ age, duration of disease, time between last treatment and start of the study, mode of pretreatment, and performance status were equal in both groups. According to Bunn and Lamberg,11stage of disease as the most important prognostic variable within this group of mycosis fungoides or small to medium sized pleomorphic T-cell lymphoma shows a homogeneous distribution between both treatment groups (Table 1). There were more males (P= .007) in the IFN+acitretin group; however, no influence was found regarding treatment result. Despite stratification, there was a higher number of pretreated patients (P = .037) in the IFN+acitretin group. This imbalance resulted from excluding nonevaluable patients.

In both groups, treatment was initiated with increasing doses of IFN α-2a (Roferon; Hoffmann-La Roche AG, Grenzach-Whylen, Germany) in the first week and continued with 9 MU 3 times a week subcutaneously. Photochemotherapy was performed with oral 8-methoxypsoralen (0.6 mg/kg) 2 hours before UVA radiation 5× weekly during weeks 1 through 4, then 3 days a week until week 23, and 2× weekly up to 48 weeks. UVA doses were started with 0.25 J/cm², increasing up to the minimal erythema dose. Acitretin (Neotigason; Hoffmann-La Roche AG) was administered daily in a dosage of 25 mg during the first week and 50 mg from weeks 2 through 48. Photochemotherapy and acitretin therapy were performed as long as IFN treatment. Depending on tolerability, treatment interruptions for a maximum of 2 weeks or dose adjustments were allowed.

Duration of therapy in both treatment groups was proposed until CR defined as no signs and symptoms of CTCL and no pathological findings in histological examination. Maximum treatment duration was 48 weeks in both groups. Because it was not the aim of this study to analyze time to treatment failure or time to progression, no recommendations concerning instructions after CR were made. Therefore, no data answering these questions are available.

For statistical analysis of the responses, the χ² test and multivariate logistic regression were performed. Time to events was calculated by Kaplan-Meier methods and tested using the log-rank test.

With the combination IFN+PUVA, 28 of 40 (70.0%) patients achieved a CR, with 26 of these patients in stage I and 2 in stage II disease. This result was significantly better than the 16 of 42 (38.1%) CRs observed in the IFN+acitretin group (Table2). Intent to treat analysis, including all 92 patients receiving any treatment within this study, still shows a significantly higher rate of complete responses in the IFN+PUVA group (P < .05). The probability of success was found to be significantly increased (P = .005) by the combination of IFN+PUVA after adjustment for pretreatment in a multivariate logistic regression.

PRs were seen in 4 patients with IFN+PUVA therapy and in 9 patients with IFN+acitretin therapy. Overall response (CRs+PRs) was 80.0% in the IFN+PUVA group and 59.5% in the IFN+acitretin group. Time to CR was significantly shorter in the IFN+PUVA group with 18.6 weeks versus 21.8 weeks in the IFN+acitretin group (P = .0015). In addition, time to response (CRs+PRs) was significantly shorter (10.9 weeks) in the IFN+PUVA group compared with 12.6 weeks in the IFN+acitretin group (P = .026; Table 2). However, these data may be of less importance for patients’ treatment.

Total IFN dose was similar in both treatment groups, with 569 and 609 MU, respectively, showing no difference in patients responding with CR (569 MU) or not (611 MU). Mean cumulative acitretin dose was 6,556 mg and was lower in the responding patients with 5,962 versus 6,927 mg in the nonresponding patients. Total PUVA dose was 139 J/cm²in a total of 58.5 treatment days, with responding patients receiving 158.6 J/cm² in 61.2 days versus 94.8 J/cm² in 51.5 days in patients failing to show a complete response (Table 3).

Life-threatening adverse events were not observed in either treatment group. Mild and moderate (World Health Organization [WHO] grade 1 and 2) adverse events were seen in 55.0% of the IFN+PUVA group and in 52.4% of the IFN+acitretin patients. Most common were flu-like symptoms, and less frequent were reductions of red and/or white blood counts and gastrointestinal disorders and elevations of triglycerides (retinoid group) and liver enzymes. In some patients, dryness of the skin, hair loss, and neurologic or psychiatric symptoms were observed (Table 4). Four (10.0%) patients of the IFN+PUVA group and 13 (31.0%) of the IFN+acitretin group developed severe side effects (WHO grade 3). Most adverse events could be managed satisfactorily by adjusting doses, although 2 patients of the IFN+PUVA and 9 of the IFN+acitretin group had to be withdrawn from the study.

Determination of neutralizing IFN antibodies, performed in all patients before, during, and after treatment, showed no correlation between the development of antibodies, the treatment group, and the response. In fact, 3 patients with high neutralizing antibodies in the IFN+PUVA group were responders.

CTCLs represent a heterogeneous group of malignant diseases arising mostly from CD4⁺ T cells in the skin, including mycosis fungoides, Sézary’s syndrome, pleomorphic T-cell lymphoma, and other, more rare entities.1 A number of treatment options, such as oral photochemotherapy, total skin electron beam, topically applied cytotoxic substances, or systemic chemotherapy, are available, although none of them has been proven to be curative.12Therefore, new agents, such as IFNs and retinoids, have been introduced into the treatment of CTCL. IFNα has been shown to be effective even in advanced and heavily pretreated patients,2 with an overall response rate in the most important studies not exceeding 52%.3 To improve these results, combinations of IFNα with PUVA or retinoids were investigated, but only in small nonrandomized trials.4-10 To prove combination therapies containing IFNα as an effective substance in the treatment of CTCL, we conducted this prospective randomized multicenter trial to compare the combination of IFN α-2a plus PUVA with that of IFN α-2a plus acitretin.

Our results within the IFN+PUVA group show similar complete response rates (70%) to those published by Kuzel et al4 (62%) and the 73% observed by us in an earlier study,5 despite differences in patient selection and maximum dose of IFNα. The number of partial responses is lower in our study, resulting in an overall response rate of only 80%, compared with 90% by Kuzel et al.4 

In comparison, complete response results with photochemotherapy alone are varying between 40% and 100% in several retrospective studies concerning stage I and II disease.13 

The results with the IFN+acitretin treatment must be compared with publications on the combination of IFNα and etretinate, because acitretin is the main active metabolite of etretinate and has been substituted for etretinate on the market for safety reasons. With 38.1%, we reached a better complete response rate than Dreno et al,6 who treated 26 patients with 9 MU IFNα plus 0.5 mg/kg body weight etretinate, and in comparison to results of Thestrup-Pedersen et al,7 who used 18 MU IFNα and 0.7 mg/kg body weight etretinate in 7 patients. In contrast to our study, these investigators included patients with advanced CTCL stages III and IV, which might explain the differences in response rates. Braathen and McFadden10 reached 50% CR, although in only a small number of patients. Our response rate of 59.5% is similar to data published by others.3 

For retinoids such as etretinate, 13 cis-retinoic acid, and arotinoid-ethylesther as monotherapy, complete response rates between 12% and 30% are published.3 

Because the mean total IFN dose of 569 MU for IFN+PUVA versus 609 MU for IFN+acitretin was similar in both treatment groups, the difference in response between these two groups cannot depend on the IFN dose. Within the IFN+acitretin group, total IFN and acitretin doses were similar in patients with complete or no CR, giving evidence that CR seems not to be dependent on these variables. In the IFN+PUVA group, there was a difference in total PUVA dose comparing patients with complete and no CR, but an influence of this difference on the outcome of the present study could not be determined.

In contrast to the report of Kuzel et al,4 we observed high titers of neutralizing anti-IFN antibodies in 3 patients, contradicting their statement that immunosuppressive effects of PUVA therapy might prevent the development of anti-IFN antibodies. Furthermore, the present data speak against an impact of neutralizing anti-IFN antibodies on the treatment results in CTCL patients, as suggested by Rajan et al.14 

In conclusion, IFNα plus PUVA is superior compared with IFNα plus acitretin in the induction of CR in patients with CTCL stages I and II. In addition, response was reached earlier and side effects seemed to be comparable, with slight advantages in favor of IFN+PUVA treatment.

The role of IFN in combination with PUVA is currently evaluated in a randomized controlled study with IFN+PUVA versus PUVA alone.