Extracorporeal photopheresis (ECP) is an effective treatment for cutaneous T-cell lymphoma. Controversy has arisen regarding its ability to improve survival rates in Sézary syndrome (SS). We describe our experience with ECP in the treatment of SS and erythrodermic mycosis fungoides, with particular emphasis on early predictors of long-term outcome.
We included 17 patients (15 with SS and 2 with erythrodermic mycosis fungoides) who received ECP as initial treatment. Four of these patients were moribund on presentation (Eastern Cooperative Oncology Group Performance Status score, 4) and underwent only 1 to 2 cycles of ECP. The median survival was 56 months for the 11 patients with SS and an Eastern Cooperative Oncology Group Performance Status score of less than 4. If all 15 patients with SS are considered, median survival was 34 months. Response after 5 months of ECP correlated with long-term survival. A low number (<6.0 ×103/µL) of circulating CD4+CD7− lymphocytes correlated with response after 5 months of ECP.
Extracorporeal photopheresis is a safe, effective, and well-tolerated treatment for erythrodermic mycosis fungoides and SS. Low numbers of CD4+CD7− cells in the circulation and a positive response after 5 months of therapy predicted long-term survival. Moribund patients are much less likely to benefit from ECP.
SÉZARY SYNDROME (SS) is the leukemic variant of cutaneous T-cell lymphoma (CTCL) characterized by erythroderma, lymphadenopathy, and the presence of atypical lymphocytes with cerebriform nuclei (Sézary cells) in the circulation. This syndrome carries a poor prognosis, with a median survival time of 31 months and a 5-year survival rate of 33.5%.1 A multivariate analysis of 51 SS cases identified the following to be poor prognostic factors: (1) the presence of cytoplasmic inclusions on results of periodic acid–Schiff staining, (2) a CD7− phenotype, and (3) the presence of large (>15 µm) Sézary cells.1 Another study that analyzed data from 106 patients with SS or erythrodermic mycosis fungoides showed that patient age at presentation, overall stage, and peripheral blood involvement were significant prognostic factors.2
Extracorporeal photopheresis (ECP) has been a standard treatment for CTCL and has been specifically recommended as first-line treatment for SS on the basis of multicenter reports indicating a prolonged median survival among ECP-treated patients with SS.3 The ratio of peripheral blood CD4/CD8 cell count, time from CTCL diagnosis to initiation of ECP, number of photopheresis sessions, and response at 6 to 8 months of treatment have been shown to be of predictive value with regard to long-term outcome.4- 7 A higher baseline lymphocyte count and a higher Sézary cell count as a percentage of total white blood cell count have also been associated with a positive response after 6 months of ECP.8 However, the absence of randomized controlled trials, the small numbers of patients described in case series, and the lack of rigorous entry criteria such as T-cell receptor gene rearrangement (TCGR) data have resulted in controversy regarding the benefits of ECP.9,10
The purpose of this study is to describe our experience with ECP as first-line treatment for SS and erythrodermic mycosis fungoides, with a particular focus on early predictors of outcome.
The study was approved by the Institutional Review Board of the University of Michigan Medical Center, Ann Arbor. From August 1, 1987, through June 20, 1994, a total of 18 patients with the diagnosis of SS (n = 16) or erythrodermic mycosis fungoides (n = 2) entered the ECP program of the Cutaneous Lymphoma Clinic at the University of Michigan. The diagnosis and stage (as determined by the criteria of the CTCL Cooperative Group11) were established by results of physical examination; histology; immunophenotypic analysis of blood and skin biopsy specimens; TCGR analyses (by means of Southern blotting of skin and blood specimens); peripheral blood smear review; chest x-ray computed tomographic scans of the chest, abdomen, and pelvis; cytologic or histologic review of suspicious lymph nodes; routine complete blood cell counts; and chemistry panels as previously reported.12 Patients likewise underwent screening via serologic tests for human T-lymphotropic virus 1 and human immunodeficiency virus. One patient with SS transferred to another institution after only 2 months of treatment and was therefore excluded from further analysis. Data are reported through December 31, 2000.
A UV-A radiation system (UVAR System; Therakos, Inc, Exton, Pa) was used. Patients received 0.6 mg/kg of methoxsalen 90 to 120 minutes before phlebotomy. Approximately 6 cycles of discontinuous centrifugation were required to obtain the requisite 240 mL of leukocyte-enriched blood, which was then combined with 300 mL of plasma and 200 mL of sterile isotonic sodium chloride solution. This mixture was then exposed to the UV-A (320-400 nm) light source for 1.5 to 2.0 hours and reinfused into the patient. This procedure was performed on 2 consecutive days twice monthly for the first 2 months, and monthly thereafter. Patients who maintained a complete remission for 1 year continued treatment every other month.
We used Kaplan-Meier analysis and defined outcome as survival. Intervening outcomes at 5 months of ECP were defined as follows: complete remission, no evidence of disease on results of physical examination, computed tomographic scans, blood examination for circulating Sézary cells, or abnormal immunophenotype by means of flow cytometry; partial remission, greater than 50% reduction in the skin severity score.13 Responders are those who achieved a partial or a complete remission, whereas stable disease indicated less than a 50% reduction in the skin severity score. Data were analyzed by χ2 and 2-tailed Fisher exact test.
Patient characteristics are summarized in Table 1. Although 5 patients in the SS group did not have 100% total body surface area involvement as reflected by their T stage, they all had severe erythema over at least 90% of their total body surface area. Significant blood involvement was documented in the 15 patients with SS; 14 showed positive findings for clonality in the peripheral blood as shown by results of TCGR analysis, and 1 had a CD4+CD7− population consisting of 80% circulating T cells. Four patients were moribund at the time of enrollment, with an Eastern Cooperative Oncology Group Performance Status (ECOG PS) score of 4.14 These patients received only 1 or 2 cycles of ECP and died within the first 2 months of therapy. The median survival for patients with an ECOG PS score of less than 4 was 56 months (Figure 1). When the moribund patients were included in the analysis, median survival of all 15 shifted to 34 months (Figure 1).
Survival of patients with Sézary syndrome (SS) receiving extracorporeal photopheresis (ECP) by length of treatment. The solid line indicates the survival plot of patients with SS and an Eastern Cooperative Oncology Group Performance Status (ECOG PS) score of less than 4 (n = 11); dashed line, patients with SS regardless of ECOG PS score (n = 15); crosses on each plot, the number of months of ECP for each of the 4 subjects who were still being followed up as of December 2000.
Three subpopulations emerged among the study patients. Patients in the first group, as mentioned above, were in the end stages of aggressive disease and died early; the second group had recalcitrant disease with an indolent course and died 2 to 3 years after initiation of ECP; and the third group showed a rapid response to ECP (within 5 months) and survived longer than 3 years. Most of the third group are current survivors (>5 years). To confirm whether early response correlated with long-term outcome, the patient status after 5 months of ECP was analyzed and compared with survival. Results of this analysis showed that 5 of 9 five-month responders are long-term survivors (ie, >5 years). In contrast, none of the 8 subjects whose response was less than 50% after 5 months of ECP survived beyond 37 months from initiation of ECP. The number of circulating CD4+CD7− lymphocytes was then compared with status at 5 months (Table 2). Patients who had fewer (<6.0 ×103/µL) CD4+CD7− lymphocytes were more likely to have achieved a response after 5 months of ECP (P = .046). The presence or absence of nodal disease was shown to be of significance when χ2 analysis was used (P = .03). However, using the Fisher exact test, which more accurately accommodates small cell sizes, this variable did not demonstrate statistically significant predictive value (P = .10). Similarly, the following variables were not found to be predictive of outcome in our patients: age, sex, time from onset of skin symptoms to initiation of ECP, stage of disease, skin severity score, and complications during therapy (data not shown).
The 2 patients with erythrodermic mycosis fungoides were not included in the survival analysis. One of them is in complete remission, whereas the other died after 39 months of ECP, secondary to chronic obstructive pulmonary disease.
No major adverse effects were noted during the course of ECP. Two patients with preexisting cardiac problems experienced transient pulmonary edema, which responded to diuresis. Minor problems related to the intravenous access site, such as bleeding and poor wound healing, were managed appropriately.
Controversy exists regarding the efficacy of ECP in SS. We believe the appropriate patients to assess ECP efficacy in SS are those with involvement of the skin and blood. We accept morphologic, phenotypic, or molecular biological evidence of blood involvement. We do not view ECP as salvage therapy; therefore, we do not believe that including results from moribund patients (ECOG PS score, 4) is appropriate. In the present study, 11 patients had significant skin and blood involvement and an ECOG PS score of less than 4, and their median survival was 56 months (Figure 1).
Although many reports support the notion of benefit of ECP in patients with SS, a more recent report questions these findings.9 The authors suggested that the presence of a demonstrable T-cell clone, determined by Southern blotting, should define SS. Using this definition, they found a 39-month survival. A more recent study by these authors and others suggests benefit of ECP for patients with SS.8 The ECOG PS scores of the patients were not indicated. Applying this criterion to our patients (ie, including only patients with positive findings for TCGR and disregarding performance status) similarly yields a median survival of 31 months. However, we believe that the benefit of ECP accrues over months of treatment. Other studies on ECP have also emphasized the need for a reasonable number of treatments before its efficacy can be evaluated.15,16 Thus, exclusion of moribund patients is appropriate. Our median survival for patients with positive findings for TCGR in the blood, with ECOG PS of less than 4, is 47 months. However, because false-negative results of gene rearrangement tests have been found,17 we do not require this test result to be positive to diagnose SS. Our patient with negative findings for TCGR was more likely to have had a false-negative finding, particularly given the large number of circulating CD4+CD7− cells and the Sézary cell count of greater than 15%.
Our study population could be divided between those who demonstrated a response within 5 months of ECP (n = 9) and those who died before or did not show any response after 5 months of treatment (n = 8). Results showed that status at 5 months is a surrogate marker for long-term outcome, which is similar to the findings of Zic et al.6 Adding an extra treatment course for each of the first 2 months, as we have done, might reduce the response time, or might be an expected variation given small sample sizes.
The dominant T-cell clone in CTCL is believed to be of the CD4+CD7− phenotype.18 To our knowledge, the CD4+CD7− level is a variable that has never been evaluated in previous investigations involving the use of ECP for SS. Nevertheless, the CD7− phenotype has been reported to be correlated with poor prognosis.1 The exact reason for this is unknown. However, because CD7+ cells have been found to enhance humoral and cell-mediated immunity, it has been speculated that a high proportion of CD7− cells could result in reduced or impaired antitumor immune responses.1 Our finding that patients with a higher burden of CD4+CD7− cells have a worse prognosis supports this concept. The increase in CD4+CD7− cells in patients with CTCL has also been associated with altered expression of certain activation-, differentiation-, and homing-related antigens and with the presence of a specific subset of circulating malignant cells (CD3dim [ie, having diminished intensity of CD3 expression] or large-sized lymphoblasts) among patients with SS, all of which are associated with more aggressive disease.18
The presence or absence of nodal disease surprisingly was not found to be a significant predictor of outcome based on results of the Fisher exact test. Results of χ2 analysis, however, suggested that this variable might have a predictive value. A larger sample size is necessary to confirm the value of these tests.
In our cohort of patients, we confirmed that early response is a predictor of survival in ECP-treated patients with SS. We also found that high levels of circulating CD4+CD7− cells are a poor prognostic indicator and that moribund patients (ECOG PS score, 4) are much less likely to respond to ECP. These findings are consistent with the notion that to respond, patients must have enough immunologic reserve to be augmented by ECP. Similarly, the patients must survive long enough (≥3 months) to generate ECP-induced immune modulation. We believe that appropriate patient selection for ECP is critical for success and suspect that differences in reported outcomes derive in part from such patient selection.
Accepted for publication March 26, 2002.
Corresponding author: Seth R. Stevens, MD, Department of Dermatology, Case Western Reserve University, 11100 Euclid Ave, Cleveland, OH 44106 (e-mail: firstname.lastname@example.org).
Stevens SR, Baron ED, Masten S, Cooper KD. Circulating CD4+CD7− Lymphocyte Burden and Rapidity of ResponsePredictors of Outcome in the Treatment of Sézary Syndrome and Erythrodermic Mycosis Fungoides With Extracorporeal Photopheresis. Arch Dermatol. 2002;138(10):1347-1350. doi:10.1001/archderm.138.10.1347