Treatment of Cutaneous T-Cell Lymphoma With Combined Immunomodulatory Therapy: Title and subTitle BreakA 14-Year Experience at a Single Institution FREE

CUTANEOUS T-CELL lymphomas (CTCLs) are skin-invasive, T-cell, non-Hodgkin lymphomas characterized by a clonal proliferation of malignant T lymphocytes.^{1 - 2} Early-stage disease is usually confined to the skin as patches or plaques. As the disease progresses, patients can develop cutaneous tumors or erythroderma and associated peripheral blood, lymph node, and visceral organ involvement.^{3 - 6} The prognosis for patients with CTCL is dependent on stage, as determined by type and extent of skin lesions and degree of extracutaneous involvement.^{6 - 12} Cutaneous tumors, erythroderma, peripheral blood involvement, lymph node or visceral invasion, and eosinophilia denote a poor prognosis.¹⁰ A recent review of 106 cases of CTCL with erythroderma treated with psoralen plus UV-A irradiation (PUVA), radiation therapy, or systemic chemotherapy, alone or in combination, reported a median survival of approximately 43 months, or 3.6 years, from the first treatment date.⁶

A variety of studies have suggested that the Sézary cell is usually derived from the T-helper type 2 subset of CD4⁺ lymphocytes.¹³ This is reflected in the observations that the lymphoid cells from patients with CTCL exhibit depressed cell-mediated immunity, are deficient in interleukin (IL) 2 and interferon gamma production, and show increased production of IL-4, IL-5, and IL-10.^{13 - 14} More recently, the malignant clonal T-cell population in CTCL has been shown to be susceptible to cell-mediated lysis by autologous cytotoxic T lymphocytes.¹⁵ Therefore, therapy for CTCL has been directed at shifting the immune response in favor of the T-helper type 1 or cytotoxic lymphocyte subpopulation in an effort to enhance autologous anticlonotypic activity against the pathogenic Sézary cells.

Extracorporeal photopheresis is a leukapheresis-based immunomodulatory therapy that combines methoxsalen with extracorporeal UV-A irradiation to a fraction of the peripheral blood leukocytes. Photopheresis has been shown to selectively induce apoptosis of malignant or activated lymphocytes with simultaneous enhancement of phagocytic activity by treated macrophages.^{16 - 19} Findings of long-term follow-up of patients with advanced CTCL have suggested that photopheresis produces significant improvement in quality of life and may prolong life in comparison with historical controls with a similar burden of disease.^{20 - 22} More recently, immunostimulatory agents such as interferon alfa, interferon gamma, sargramostim, and systemic retinoids have been added to extracorporeal photopheresis in an effort to further enhance cytotoxic T-lymphocyte function and antigen presentation. However, little is known regarding the efficacy of regimens combining multiple biologic response–enhancing therapeutic modalities.

At our institution since 1985, we have treated 47 patients with advanced CTCL with photopheresis for a minimum of 6 months. In addition, most of these patients received a minimum of 3 months of therapy with interferon alfa, interferon gamma, sargramostim, and/or systemic retinoids in combination with photopheresis. Herein, we review our experience with combination immunomodulatory therapy compared with photopheresis monotherapy and historical controls for the treatment of advanced CTCL.

All patients were treated with the UVAR photopheresis system (Therakos, West Chester, Pa) as previously described.²³ The entire procedure required about 4 hours. Plasma psoralen levels were not routinely measured. However, if a patient failed to respond after 4 or 5 cycles of photopheresis, we measured plasma psoralen levels or empirically increased the psoralen dose.

Forty-seven patients received a minimum of 6 cycles of photopheresis between January 1985 and December 1998. All patients underwent a pretreatment assessment, which included a physical examination, skin biopsy for definitive diagnosis, routine laboratory studies, electrocardiography, and roentgenography of the chest. In addition, all patients had peripheral blood examined for the presence of atypical cells (Sézary cells). This was accomplished by light microscopic examination of 1-µm sections of the buffy coat as previously described.²⁰

Most patients also had fluorescence-activated cell sorter analysis performed on peripheral blood for the cell surface markers CD3, CD4, CD7, and CD8 and/or Southern blot analysis of the β chain gene of the T-cell receptor to determine whether an expanded clonal population of T cells was present. Furthermore, most patients who had palpable lymphadenopathy before treatment had lymph node biopsies performed. On the basis of this evaluation, patients were assigned a disease stage according to a previously defined staging system for CTCL²⁴ (Table 1 and Table 2).

Present criteria for receiving photopheresis include extensive skin involvement and more than 5% circulating Sézary cells detected on buffy coat analysis. However, when photopheresis was in its early development, patients with any stage of cutaneous disease, with or without peripheral blood involvement, were eligible for treatment. Responses in some of these early patients are included in the present study and served to establish the current treatment criteria.

All patients were admitted to the Hospital of the University of Pennsylvania, Philadelphia, for therapy. Typically, patients were treated with 1 cycle of photopheresis at 4-week intervals. Each cycle consisted of a photopheresis treatment on each of 2 consecutive days. Patients were assessed at the time of their hospital admissions by physical examination and routine laboratory studies. Every 3 to 4 months, peripheral blood was collected for fluorescence-activated cell sorter analysis and Sézary counts. In addition, if a clinical remission was suspected and a peripheral blood T-cell clonal population had been detected by Southern blot analysis before initiation of treatment, blood was sent for Southern blot analysis in an effort to verify by molecular analysis the disappearance of the malignant clone from the blood.

Initially, an attempt was made to maintain photopheresis as monotherapy for at least 6 months before the use of adjuvant therapy. If sometime after 6 months of monotherapy a patient's clinical response slowed, was unchanged, or worsened, adjuvant treatment was usually added. As more experience with the common adjuvant agents was obtained and/or when a patient's pretreatment assessment revealed particularly advanced disease defined by a markedly elevated white blood cell count (WBC), bulky lymphadenopathy, high numbers of circulating Sézary cells (>20%), or a markedly increased CD4⁺/CD8⁺ ratio, patients were usually started on a combination of photopheresis with 1 or more forms of adjuvant therapy.

The most common adjuvant agents used were low doses of interferon alfa (1.5-7.5 × 10⁶ U, 3 to 5 times weekly), systemic retinoids (10-50 mg of etretinate daily; 10-25 mg of acitretin daily; 10-30 mg of isotretinoin daily; or 10-20 mg of all-trans retinoic acid daily), sargramostim (75-125 µg subcutaneously 1 hour after photopheresis), and interferon gamma (1-3 × 10⁶ U 3 to 5 times weekly). Immunomodulatory medications and doses were tailored for each patient to minimize adverse effects. Patients were considered members of the combination immunomodulatory therapeutic group if they received both extracorporeal photopheresis and 1 or more of the above-listed agents simultaneously for at least 3 months.

Some of the patients received brief courses of local γ irradiation to cutaneous tumors or lymph nodes and/or focal treatment to affected skin with topical chemotherapeutic agents (0.01% nitrogen mustard in petrolatum or 20-40 mg of carmustine in 100 g of petrolatum) or topical steroids and/or concurrent treatment with PUVA or UV-B irradiation. Since these forms of therapy do not produce systemic immune-enhancing effects, they were not considered forms of adjuvant therapy.

A complete response to therapy was defined as the regression of all evidence of disease for at least 3 months, including skin disease, disappearance of Sézary cells from the peripheral blood if present before therapy, and a change in Southern blot analysis indications from clonal rearrangement to a germline pattern. A partial response was designated as at least a 50% reduction in skin surface area involvement and in numbers of circulating Sézary cells. Patients who experienced less than 50% clearance of skin and peripheral blood disease were classified as nonresponders.

Photopheresis was discontinued if, despite photopheresis and treatment with adjuvant agents, the disease progressed. In those patients who had a complete remission, photopheresis therapy was tapered over several months by increasing the interval between treatment cycles. In patients receiving adjuvant therapy in combination with photopheresis, the medication dose was maintained as the photopheresis was tapered and for a minimum of 6 months after photopheresis therapy had been discontinued. The practice of tapering treatment was based on our experience with 2 patients in whom a prompt relapse occurred after photopheresis was discontinued abruptly.

All 47 patients with biopsy-proven CTCL who received photopheresis for a minimum of 6 months, alone or in combination with adjuvant therapy, between January 1985 and December 1998 at the Hospital of the University of Pennsylvania were included in this series. This study was approved by the Committee on Studies Involving Human Beings of the Hospital of the University of Pennsylvania. Each patient's chart was reviewed by the cutaneous oncology fellow (K.R.S.) to obtain historical data. All patients lost to follow-up were contacted by telephone and interviewed to update our records. Follow-up was unavailable for 3 patients. One patient was a woman with stage IIIB CTCL without circulating Sézary cells who received 6 cycles of photopheresis as monotherapy without a response. The second patient was a man with stage IB CTCL without circulating Sézary cells who received 13 months of photopheresis. Interferon alfa was added to his treatment regimen for the last 4 months of photopheresis. Throughout his treatment, he remained a nonresponder. The last patient had stage IIA disease with 70% circulating Sézary cells and received 21 cycles of photopheresis monotherapy. He had a complete response to therapy and was alive and disease free at 105 months after starting photopheresis but subsequently was lost to follow-up.

Kaplan-Meier actuarial survival curves were calculated for the entire treatment group and separately for patients with stages I and II disease vs stages III and IV disease, and for patients receiving multimodality immunomodulatory therapy or photopheresis monotherapy. Differences between the survival curves were calculated using the Wilcoxon rank sum test.

We used t tests with Satterthwaite correction for unequal variances to determine differences in WBC and CD4⁺/CD8⁺ ratios at the time of treatment initiation between patients who received combination immunomodulatory therapy or photopheresis monotherapy. The t test without correction for unequal variances was used for differences in initial lactic dehydrogenase (LDH) levels and percent Sézary cells on buffy coat analysis in the 2 populations. A χ² contingency table analysis was done to compare differences in disease stage and in the number of patients with circulating Sézary cells at initiation of therapy and to evaluate differences in clinical response rates for patients receiving combination immunostimulatory therapy vs photopheresis monotherapy.

Forty-seven patients with histopathologically confirmed CTCL were treated with at least 6 cycles of photopheresis between January 1985 and December 1998. Table 3 summarizes the baseline characteristics of the patient population. Most patients had advanced disease at the time treatment was initiated: 32 (68%) of the 47 patients had stage III or IV disease; 42 (89%) of the 47 patients had peripheral blood involvement as determined by the presence of more than 5% atypical cells detected by 1-µm section analysis of peripheral blood buffy coats.

Thirty-one (66%) of the 47 patients received combination biologic response therapy whereas 16 (34%) received extracorporeal photopheresis monotherapy. Table 4 summarizes the baseline characteristics of these 2 separate treatment populations. Differences between the 2 populations were statistically significant for the mean CD4⁺/CD8⁺ ratio (P = .006). Differences in the percentage of patients with greater than 5% circulating Sézary cells (P = .07) and for elevated WBC (P = .07) at the time of treatment initiation were nearly statistically significant. Although there seemed to be a trend toward higher Sézary counts seen on buffy coat examination in patients who received combination therapy, there was no statistically significant difference in this parameter between the 2 therapeutic populations (P = .18). There were no significant differences between the 2 treatment groups in mean LDH level (P = .99) or stage of disease at treatment initiation (P = .47).

All 47 patients received a minimum of 6 months of photopheresis therapy. Thirty-one of the 47 patients received 1 or more adjuvant agents in combination with photopheresis for at least 3 months. The most common adjuvant immunostimulatory agent used in combination with photopheresis was interferon alfa (n = 30) followed by systemic retinoids (n = 21) and sargramostim (n = 15) (Table 5). The types of retinoids used included etretinate, acitretin, isotretinoin, and all-trans retinoic acid. Some of the patients received focal treatment with topical nitrogen mustard or carmustine (n = 35), topical steroids (n = 43), or γ irradiation (n = 11). A small number of patients also received concomitant PUVA (n = 5) or UV-B (n = 1) treatment.

Twelve (26%) of the 47 patients experienced a complete remission. Twenty-five (53%) of the patients had a partial remission with at least a 50% reduction of skin and peripheral blood disease. The overall positive response rate, including complete and partial responders, was 37 (79%) of 47. Eight (17%) of the patients were considered nonresponders. Only 2 patients (4%) experienced progression of disease (Table 6). The first patient had stage IVB disease with multiple cutaneous tumors, extensive lymphadenopathy; an enlarged spleen; 7% Sézary cells on buffy coat examination; normal WBC, LDH levels, and CD4⁺/CD8⁺ ratio; and a clonal T-cell gene rearrangement in the peripheral blood by polymerase chain reaction when photopheresis was started. He had been treated with topical nitrogen mustard, topical carmustine, etretinate, PUVA, γ irradiation, interferon alfa, and intralesional kenalog prior to starting photopheresis. He received 7 cycles of photopheresis, 5 of which were combined with sargramostim. The patient's condition progressively declined during that time. All immunomodulatory therapies were discontinued, and the patient began therapy with systemic chemotherapy. He died after receiving 3 months of chemotherapy. The second patient had stage IIIB disease, also with multiple cutaneous nodules, no Sézary cells on buffy coat examination, normal WBC, an LDH level of 626 U/L (normal, ≤620 U/L), and a CD4⁺/CD8⁺ ratio of 2.6 (normal, ≤2) when photopheresis was initiated. She had received γ irradiation, multidrug systemic chemotherapy, PUVA, topical nitrogen mustard, and systemic steroids prior to starting photopheresis. She received 6 cycles of photopheresis monotherapy. Her disease rapidly progressed and she was lost to follow-up soon after her last treatment.

Twenty-six (84%) of the 31 patients in the combination immunostimulatory treatment group experienced either a complete or partial clinical response. Six patients (20%) had a complete response, 20 (65%) had a partial response, 4 (13%) had no response, and 1 (3%) had progression of disease. Of the 16 patients who received photopheresis monotherapy, 12 (75%) had a positive response. In this group, 6 (38%) were complete responders, 6 (38%) were partial responders, 3 (19%) did not respond to therapy, and 1 (6%) had progressive disease (Table 7). The differences in positive response rates between the 2 therapeutic groups were not statistically significant (P = .47). None of the patients experienced greater than grade I toxic effects, according to the common toxicity criteria of the National Cancer Institute.

The median survival of all 47 patients calculated from the first treatment date using the Kaplan-Meier actuarial survival curve was 74 months (6.2 years) (Figure 1). Most (68%) of the patients had stage III or IV CTCL.

Kaplan-Meier survival curves were also calculated to compare survival between patients with early-stage disease (stages I and II) and advanced disease (stages III and IV) (Figure 2). Patients with late-stage disease experienced a median survival of 55 months (4.6 years), whereas patients with early stage disease had a median survival of 92 months (7.7 years). The difference in survival between these 2 groups was statistically significant (P = .03).

To evaluate for a possible difference in survival between patients treated with multimodality biologic response therapy and those who received extracorporeal photopheresis monotherapy, we calculated survival curves separately for the 2 groups (Figure 3). The median survival for the patients undergoing combination therapy was 74 months (6.2 years), whereas the median survival for the monotherapy group was 66 months (5.5 years). Survival differences for the 2 treatment groups were not statistically significant (P = .51).

Extracorporeal photopheresis is an established and effective treatment for advanced CTCL. Several groups have reported positive response rates ranging from 50% to 80% for patients with advanced-stage CTCL.^{20 - 22 ,25 - 27} Photopheresis may also prolong patient survival. Gottleib et al²⁰ reported a median survival of longer than 100 months from the time of diagnosis in patients with advanced-stage disease and peripheral blood involvement treated with a minimum of 6 months of photopheresis compared with a median survival of 30 to 40 months in historical controls.

Although the precise mechanism of action of photopheresis has not been determined, exposure of leukocytes to UV-A following uptake of 8-methoxypsoralen results in cross-linking of DNA and eventual apoptotic cell death.^{16 ,28} Murine skin transplantation models indicate that the treated leukocytes are altered in some fashion such that, when reinfused, a specific anticlonotypic immune response directed toward the pathogenic cells develops.²⁹ Therefore, the overall integrity of the immune response is believed to be a critical factor for patient responsiveness to photopheresis monotherapy.

Interferon alfa is a potent biologic agent that has powerful antiproliferative, immunostimulatory, and differentiation-inducing activities. Clinical trials have shown response rates of 50% to 80% in patients with CTCL receiving systemic interferon alfa.^{30 - 31} It also exerts substantial immune augmentative effects on cytotoxic T-cell function and can reverse cytokine and immune abnormalities in Sézary syndrome.³² Administration of sargramostim may enhance antigen presentation and lead to an augmented antitumor response to the apoptotic T cells, and systemic retinoids can induce T-helper type 1 cytokine production by peripheral blood mononuclear cells in vitro.³³ Because little is known about the use of multiple immunomodulatory agents to treat patients with advanced CTCL, our purpose was to determine the efficacy of this therapeutic approach.

Our previous experience formed the basis for the development of a clinical profile of those CTCL patients most likely to respond to extracorporeal photopheresis monotherapy. These clinical criteria include the presence of modest numbers of peripheral blood Sézary cells (10%-20% of mononuclear cells); short duration of disease (less than 2 years); normal or nearly normal numbers of cytotoxic T lymphocytes with nearly normal CD4⁺/CD8⁺ ratios; normal or modestly elevated WBC; no history of intensive chemotherapy; and absence of bulky lymphadenopathy or overt visceral disease.^{20 ,34} Using these criteria as a guideline, the patients who received combination biologic response therapy in the present study had a worse prognosis at the time of treatment initiation than those who received photopheresis monotherapy, as evidenced by higher CD4⁺/CD8⁺ ratios, higher WBCs, a higher percentage of patients with peripheral blood involvement by Sézary cells, and higher percentages of circulating Sézary cells than those who received photopheresis monotherapy. Although the differences in some of these parameters were not statistically significant, a 4-fold increase in sample size would be needed to achieve an 80% chance of detecting statistical significance. Therefore, we consider the finding of a higher number of poor prognostic factors in the population receiving multimodality immunostimulatory therapy clinically relevant.

For all 47 patients, the overall positive response rate, including complete and partial responders, was 79%. The positive response rate for the 31 patients who received combination therapy was 84%. Although some groups have reported response rates of 76% to 80% with photopheresis monotherapy, these rates were achieved in patient populations with more limited skin disease and/or normal WBCs and CD4⁺/CD8⁺ ratios.^{20 ,25} As most of our patients had multiple poor prognostic factors at the time of treatment initiation, it seems that multimodality biologic response therapy enhances clinical response in patients with an expected worse outcome. In addition, the median survival in our patient population, 68% of whom had stage III or IV CTCL, was 6.2 years. This represents a marked improvement in survival over historical controls, who had median survival times of 4.6 and 1.1 years for stages III and IV, respectively.⁶ Unfortunately, patients with extensive cutaneous tumors experienced poor responses to photopheresis alone or in combination with other immunostimulatory agents.

Survival curves were also calculated to compare survival between patients with limited skin disease and extensive disease. Patients with stages III and IV disease experienced a median survival of 4.6 years, whereas patients with early-stage disease had a median survival of 7.7 years. Nearly all of the patients in our study had peripheral blood involvement, which confers a worse prognosis in any stage of disease. Therefore, our patients with widespread disease experienced a marked improvement in survival over the previously reported median survival of 2.6 years for patients with erythrodermic CTCL and peripheral blood involvement who received PUVA, radiation therapy, or chemotherapy, alone or in combination.⁶ Additionally, Kim et al⁶ have found that age at presentation of 65 years or older, clinical stage of IV, and B stage of 1 are associated with poor prognosis.⁶ In their study, patients receiving traditional nonimmunomodulatory therapy were categorized based on the number of poor prognostic factors present at the time of diagnosis. The median survival was 1.5 years for patients who presented with 2 or 3 poor prognostic factors, and 3.7 years for those who presented with 1 factor.⁶ The average age of presentation for our patients was 62 years, and nearly all had peripheral blood involvement, independent of tumor stage. Thus, since most of our patients had 2 poor prognostic factors independent of tumor stage, treatment with combination biologic response therapy markedly improved survival in all our patients over that of historical controls with similar factors.

We compared survival between patients treated with multimodality immunostimulatory therapy and those who received extracorporeal photopheresis monotherapy to evaluate for a possible improvement in outcome with the addition of immunomodulatory agents. The median survival for the patients undergoing combination therapy was 6.2 years, whereas the median survival for the monotherapy group was 5.5 years. Although neither the differences between the median survival values nor the differences between the 2 survival curves were statistically significant, there may be a trend toward longer survival in patients treated with combination immunomodulatory therapy. Importantly, the patients who received combination immunomodulatory therapy had a worse prognosis at initiation of therapy. Therefore, the findings presented here suggest that multimodality biologic response therapy may be more effective for patients with multiple poor prognostic factors than extracorporeal photopheresis monotherapy.

The review of our experience with combination immunomodulatory therapy suggests that patients with advanced CTCL receiving this therapy experience high clinical response rates and prolonged survival compared with historical controls. In comparison with patients with less severe disease treated with photopheresis monotherapy, patients with multiple poor prognostic factors experience equal or improved clinical outcomes when treated with multimodality immunostimulatory therapy. Although further investigation is needed to better establish the effectiveness of multimodality biologic response therapy for the treatment of cutaneous T-cell lymphoma, the results of our study suggest that it should be considered first-line therapy for patients with advanced-stage disease and multiple poor prognostic variables.

Accepted for publication November 28, 2001.

This study was supported by grants CA80108, CA81022, and MO1RR00040-29 from the National Institutes of Health, Bethesda, Md, and a grant from the Leukemia and Lymphoma Society, White Plains, NY.

This work received the Nelson Paul Anderson Memorial Award of the 2000 Pacific Dermatologic Association, Victoria, British Columbia, and the Johnson-Beerman Award of the 2001 Section on Dermatology, College of Physicians of Philadelphia, Philadelphia, Pa.

Corresponding author and reprints: Alain H. Rook, MD, Department of Dermatology, Hospital of the University of Pennsylvania, 2 Rhoads Pavilion, 3600 Spruce St, Philadelphia, PA 19104-4283 (e-mail: arook@mail.med.upenn.edu).