European dermatology forum – updated guidelines on the use of extracorporeal photopheresis 2020 – part 1

Abstract

Background

Following the first investigational study on the use of extracorporeal photopheresis for the treatment of cutaneous T‐cell lymphoma published in 1983, this technology has received continued use and further recognition for additional earlier as well as refractory forms. After the publication of the first guidelines for this technology in the JEADV in 2014, this technology has maintained additional promise in the treatment of other severe and refractory conditions in a multi‐disciplinary setting. It has confirmed recognition in well‐known documented conditions such as graft‐versus‐host disease after allogeneic bone marrow transplantation, systemic sclerosis, solid organ transplant rejection including lung, heart and liver and to a lesser extent inflammatory bowel disease.

Materials and methods

In order to further provide recognized expert practical guidelines for the use of this technology for all indications, the European Dermatology Forum (EDF) again proceeded to address these questions in the hands of the recognized experts within and outside the field of dermatology. This was done using the recognized and approved guidelines of EDF for this task. All authors had the opportunity to review each contribution as it was added.

Results and conclusion

These updated 2020 guidelines provide at present the most comprehensive available expert recommendations for the use of extracorporeal photopheresis based on the available published literature and expert consensus opinion. The guidelines are divided in two parts: PART I covers cutaneous T‐cell lymphoma, chronic graft‐versus‐host disease and acute graft‐versus‐host disease while PART II will cover scleroderma, solid organ transplantation, Crohn's disease, use of ECP in paediatrics practice, atopic dermatitis, type 1 diabetes, pemphigus, epidermolysis bullosa acquisita and erosive oral lichen planus.

Introduction

Extracorporeal photopheresis (ECP, also known as extracorporeal photochemotherapy, extracorporeal photoimmunotherapy or just photopheresis) is a leukapheresis‐based therapy that is available at more than 200 centres worldwide. ¹ , ² During ECP, the patient's whole blood is processed outside the body: blood is collected via an antecubital vein, or a permanent catheter if vein access is cumbersome; white blood cells are then separated from red blood cells and plasma by centrifugation in a device that is specially constructed for this procedure. White blood cells are exposed to ultraviolet A (UVA) light in a separate plastic chamber and then returned to the patient. ³ In the past, patients treated with ECP were given oral 8‐methoxypsoralen (8‐MOP; methoxsalen) before the blood was leukapheresed. ¹ Thus, during the ECP treatment, patients typically experienced untoward gastrointestinal effects such as nausea and vomiting, or the visual side‐effects of psoralen. Moreover, differences in gastrointestinal absorption due to individual variability resulted in unpredictable blood concentrations of 8‐MOP. ¹ , ⁴ To avoid the problems of oral 8‐MOP administration, a liquid formulation of 8‐MOP (UVADEX^®, Therakos^®) that is added directly to the buffy‐coat/blood fraction was developed. This method of dosing circumvents the potential side‐effects of systemic 8‐MOP administration and eliminates the need to measure for target blood levels. ⁵

The first investigational study of ECP in patients with cutaneous T‐cell lymphoma (CTCL) was completed in 1983. The first ECP apparatus that was approved by the United States Food and Drug Administration in 1988 was a closed system (UVAR^®; Therakos^®). National approvals in Europe and elsewhere followed. Although ECP was initially developed for use in CTCL patients, it has also shown promising efficacy in a number of other severe and difficult‐to‐treat clinical conditions such as graft‐versus‐host disease (GvHD), Crohn's disease, systemic sclerosis and for the prevention and treatment of rejection in solid organ transplantation, particularly in the areas of lung and heart transplantation. ⁶

Several closed and open ECP apparatuses are currently available for clinical use and are compared in Table 1. ⁷ Their clinical efficacy in the treatment of a variety of T‐cell‐mediated diseases is well established. However, the two techniques have not been directly compared in a clinical setting. In a closed ECP apparatus (one‐step method), the blood cell separation, drug photoactivation, and reinfusion stages are fully integrated and automated, and all elements are approved for their combined use, including methoxsalen, a photoactivating agent (Table 2). There is no risk of improper reinfusion when used according to the labelling, and the risk of infection and contamination associated with the medical device itself is very low.

Table 1.

ECP devices in current use in adults and children (adapted from Wong and Jacobsohn) ⁷

Methodology	Automated	Weight limit	Cell separator Extracorporeal volumes	Cell separator technology
One‐step methods
CELLEX® (Therakos®)*	Yes (double needle)	RBC prime needed if >115% ECV	Variable, dependent on Hct, blood volume processed, return bag threshold (lower than UVAR XTS)	IFC (continuous buffy coat collection with intermittent fluid return) (Latham Bowl)
	Yes (single needle)	RBC prime needed if >115% ECV	Variable, dependent on Hct, blood volume processed, return bag threshold (higher than double‐needle method)	CFC (Latham Bowl)
UVAR XTS® (Therakos®) (not available in U.S and Europe)	Yes (single needle)	>40 kg (need to satisfy ECV limits)	Variable, dependent on Hct, number of cycles, and bowl size (225 or 125 mL)	IFC (Latham Bowl)
Two‐step methods †
Spectra OPTIA® (Terumo BCT) and UVA irradiator	Yes (only cell separation)	None	253 mL (Continuous mononuclear cell collection (CMNC), version 1.3); 147 mL (AutoPBSC procedure, Version 3.8)	CFC
Mini‐buffy coat and UVA irradiator	No	Smaller children	None, but limited to 5–8 mL/kg whole blood draw	Standard manual buffy centrifugation technique
Three‐step methods ‡
Spectra OPTIA® (Terumo BCT) & UVAR XTS® (Therakos®)	Yes (only cell separation)	None	See above for MNC and AutoPBSC procedure	CFC

Suitable for low bodyweight patients.

CFC, continuous flow centrifugation; ECV, extracorporeal cell volume; Hct, haematocrit; IFC, intermittent flow centrifugation; MNC, mononuclear cell; PBSC, peripheral blood stem cell; RBC, red blood cell.

^†Only cell separation is automated, while the UVA irradiator is operated manually. Other dedicated continuous or intermittent T‐cell separators may also be used, such as Amicus (Fenwal, MNC kit) and AS104 (Fresenius Kabi) which have extracorporeal volumes of 163 and 175 Ml, respectively.

^‡Three‐step methods involve standard mononuclear cell collection using dedicated continuous cell separators followed by red blood cell priming of the UVAR‐XTS instrument and photoactivation treatment of the 8‐methoxypsoralen treated mononuclear cells within the UVAR‐XTS instrument after programming the instrument that the last ECP cycle has occurred.

Table 2.

European CE mark and FDA approval status of ‘one‐step’ closed photopheresis apparatuses and various cell separation and drug photoactivation devices used in ‘Multistep’ photopheresis procedures

	Company	European CE mark	FDA approval
Closed photopheresis apparatuses
CELLEX®†	Therakos®	✓ For photopheresis	✓ For photopheresis
UVAR XTS®	Therakos®	✓ For photopheresis	✓ For photopheresis
Tubing set (XTS® and CELLEX®)	Therakos®	✓ For photopheresis	✓ For photopheresis
UVADEX	Therakos®	✓ For photopheresis	✓ For photopheresis
Cell separation system (standard apheresis device)
Spectra Optia®	Terumo BCT	✓ For therapeutic plasma exchange, RBC exchange, and WBC collection	✓ For therapeutic plasma exchange, leucocyte collection, and RBC exchange
Com. Tec®	Fresenius Kabi	✓ For therapeutic plasma exchange and WBC collection	✓ For therapeutic plasma exchange and WBC collection
MCS® plus	Haemonetics®	✓ For therapeutic plasma exchange and leucocyte collection	✓ For therapeutic plasma exchange and leucocyte collection
AMICUS	Fenwal	✓ For therapeutic plasma exchange and leucocyte collection	✓ For therapeutic plasma exchange and leucocyte collection
Drug photoactivation system
PUVA light system	Macopharma	CE marked (indicated to treat psoriasis, not dedicated to ECP)	No
MACOGENIC	Macopharma	UVA illumination machine CE 0459	No
MACOGENIC G2	Macopharma	UVA illumination machine CE 0459	No
XUV bag	Macopharma	UVA illumination machine CE 0459	No
8‐MOP	Macopharma	AMM PTA 07.10.109 (indicated for nuclear cell photosensibilization)	No
UVA PIT system	MedTech Solutions	Medical System for photoimmune therapy (body MDC 0483)	No
UVA PIT Kit	MedTech Solutions	Medical System for photoimmune therapy (body IMQ 0051)	No
PUVA Combi‐Light	Cell.Max	CE marked medical device	No
UVA Illuminator	GMBH

CE, Conformité Européenne; RBC, red blood cell; WBC, white blood cell.

^†Suitable for low bodyweight patients.

From a technical aspect, an open apparatus is any disconnected process using a cell separator in combination with a lightbox and a drug. Although the individual components may be Communauté Européenne (CE) marked, they are not explicitly approved for use together in the process of photopheresis. To obtain proper CE marking for photopheresis use, all the components of an apparatus must undergo a validation process prior to being used together in controlled clinical trials and routine therapy. This technology falls under the regulations of cell therapy according to the federal agency L'Agence Nationale de Sécurité du Médicament (ANSM) in France. ⁸ Open apparatuses can only be used by centres that are certified for cell therapy. To obtain the certificate, ANSM requires the filing of a record of authorization describing the entire ECP procedure, including the drug and material to be used, transport, quality controls, traceability, the structure of cell manipulation and much more. Closed apparatuses do not have these restrictions. A closed apparatus is a one‐step method (UVAR‐XTS^® and CELLEX^®; Therakos^®). Critical steps, such as cell separation, drug photoactivation, and reinfusion, are fully integrated and automated processes. All the components are validated for their combined use, including the use with 8‐methoxypsoralen Table 2. Components of closed ECP apparatuses are approved and certified as one functional unit, which may be operated by a single trained person.

One of the critical elements of both open and closed ECP apparatuses is the photoactivation chamber. Closed photopheresis apparatuses are equipped with a microprocessor that allows for a dynamic recirculation of photoactivated cells. All photoactivation elements have a fixed thickness and are tested by UV spectrophotometry to ensure the retention of photodynamic properties (optimal UV transmittance). Adsorption of 8‐methoxypsoralen to the disposable plastic kit is measured and compensated for to ensure proper dosing. Components that are used in open ECP apparatuses are not designed or manufactured for the process of photopheresis, and, therefore, need to be certified prior to their use.

Inconsistent light exposure to targeted cells because of non‐validated plastic films, variation in the fluidity of the solution in the treatment bag, unknown or variable drug adsorption onto plastic components, or stasis of the cells during UVA irradiation could cause partial DNA damage to the cells.

Regardless of the apparatus used, ECP is usually well‐tolerated. There are no reports of grade III‐IV side‐effects (as rated by the World Health Organization (WHO)) following treatment. Transient hypotension or mild anaemia (after multiple treatments) may occur, and thrombocytopenia has also been reported. ECP should not be used as a therapy in patients with a known sensitivity to psoralen compounds such as methoxsalen, or comorbidities, including photosensitivity, a history of heparin‐induced thrombocytopenia, a low haematocrit or cardio‐circulatory failure (note: in some selected patients with cardio‐circulatory failure, the therapy is actually well‐tolerated as response to therapy contributes to stress reduction). It is also contraindicated in pregnancy. Methoxsalen containing ready‐to‐use sterile solutions are contraindicated in patients with aphakia because of the significantly increased risk of retinal damage. In patients with low bodyweight, children, and those with problematic venous access, implantable venous access devices with a proper blood flow per minute should be used. In this regard, peripheral venous catheters appear to be advantageous over central venous devices. In addition, it should be noted that the use of Hickman catheters among erythroderma patients can be quite hazardous due to the high rate of infection with staphylococcus aureus. ⁹

Ideally, ECP should be initiated as soon as clinically indicated, which in most cases is as a second‐line therapy when other first‐line therapies have failed. In general, currently, many centres in Europe perform ECP treatment as inpatient therapy. Monitoring of efficacy before and during ECP treatment should be based on the standards of care for each indication. The use of either heparin or acid citrate dextrose as anticoagulation during ECP depends on the preference of different centres. While the use of UVA protective glassware is recommended during PUVA in combination with oral methoxsalen, it may be unnecessary during ECP therapy due to the very low doses of psoralen used. With the use of UVADEX, there is usually a low to unmeasurable level of 8‐MOP in the blood thus making it unnecessary to completely avoid UVA exposure post‐therapy.

Mode of action

Although ECP has been in clinical use for more than 35 years, its mode of action remains elusive. Current doses and treatment intervals remain almost identical to regimens used in the 1980s. Early studies indicated that ECP induced lymphocyte apoptosis contributed to the therapeutic effect. ¹⁰ , ¹¹ More recent studies have shown that the mechanism of action of ECP is primarily due to an immunomodulatory effect. The principal mechanisms of action comprise of the modulation of dendritic cells, alteration of the cytokine profiles, and induction of particular T‐cell subpopulations. ¹² , ¹³ ECP, like psoralen plus UVA (PUVA), induces psoralen‐mediated DNA crosslinks that cause apoptosis in lymphoid cells, particularly in natural killer (NK) cells and T cells. ¹⁴

However, the therapeutic effect of ECP in Sézary syndrome (SS) cannot be explained by the depletion of malignant cells, as only a relatively low proportion of the entire lymphocyte pool is treated in a photopheresis cycle. Monocytes, which appear to be more resistant to apoptosis, undergo a differentiation process within 2 days, and express surface markers such as CD83, X‐11, ɑ‐V, beta‐ V, or CD1a. ¹⁵ , ¹⁶ , ¹⁷ This differentiation process appears to be independent of the psoralen‐induced photoactivation and is mostly driven by direct contact of the cells with plastic and other synthetic materials during the passage through the ECP apparatus. Apoptotic lymphocytes are phagocytosed and eliminated by immature dendritic cells, which subsequently undergo maturation and present antigenic peptides — a process that has been designated transimmunization. ¹⁸ Thus, it has been suggested that transimmunization may induce an immune response against lymphoma cells, which might explain the beneficial effects of ECP observed in the therapy of SS.

Use of immunosuppressive agents does not have a positive effect on the course of CTCL and may even be hazardous. ECP‐initiated cellular mechanisms of differentiation in contrast are associated with the release of a variety of cytokines including tumour necrosis factor (TNF) and interleukin (IL)‐6, which induce the activation of CD36‐positive macrophages. ¹⁹

Long‐term, beneficial immunologic alterations can be gained through the use of continuous ECP. The severity of CTCL is directly related to the imbalance of the ratio of T‐helper cells 1 to T‐helper cells 2 (Th1/Th2), which leads to the increased release of IL‐4 and IL‐5, the reduced activity of NK cells, and the diminished cytotoxic activity of CD8‐positive T cells. In a study performed in patients with early‐stage CTCL (stage IB) undergoing ECP therapy for 1 year, Di Renzo et al. observed not only an increase in CD36‐positive monocytes in the blood but also a change in the cytokine reaction profile of peripheral blood lymphocytes upon stimulation with phytohaemagglutinin. ²⁰ Both observations imply that ECP reverses the pathologic shift towards a Th2 immune response and restores the Th1/Th2 balance in CTCL patients. Also, anti‐inflammatory cytokines appear to be induced by ECP, whereas blood levels of pro‐inflammatory cytokines are lowered. ²¹

In relation to neutrophils, these also undergo apoptosis resulting in mobilization of neutrophilic myeloid‐derived suppressor cells (MDSC) into the circulation which can dampen Th1 and Th17 responses. ²²

Over the last two decades, ECP has been shown to be beneficial in patients with CTCL, GvHD, transplant rejection, and various autoimmune diseases. The findings mentioned above, however, cannot explain the effects of ECP in these patients, and because these conditions respond to immunosuppressive therapies, it was surmised that ECP might also exert immunosuppressive effects. Furthermore, in patients with GvHD, ECP was shown to induce IL‐10 via the modulation of arginine metabolism. ²³ In contrast to classic immunosuppressive therapy, ECP is not associated with significant side‐effects such as opportunistic infections. It has been postulated that the therapeutic effect of ECP is due to the induction of regulatory T (T‐reg) cells, without causing general immunosuppression. Using a murine contact hypersensitivity model, Maeda et al. ²⁴ demonstrated that T‐reg cells could be induced successfully by an ECP‐like procedure (intravenous injection of leukocytes exposed to 8‐MOP and UVA in vitro). T‐reg cells induced by the combination of 8‐MOP and UVA express CD4, CD25, CTLA‐4, and the transcription factor Foxp3, similar to T‐reg cells induced by UVB. Foxp3 suppresses the activity of other lymphocytes. ²⁵ Furthermore, the release of IL‐10 appears to be involved in this process. ²⁶ The levels of serum B‐cell activating factor (BAFF) were measured in a recent study of 46 patients with chronic GvHD (cGvHD). Serum levels of BAFF determined at 1 month after the start of ECP therapy were predictive of the 3‐month and 6‐month skin responses. Serum levels of BAFF lower than 4 ng/mL were associated with a significant improvement of the skin. ²⁷ In addition, monocytes showed immunoregulatory capacity on CD4+ T cells in a human in vitro model of ECP. Reduced proliferation rates of T cells after co‐culture with ECP‐treated monocytes was dependent on cell‐contact between monocytes and T cells. ²⁸ Also, there is evidence that infusion of lymphocytes treated with 8‐MOP and UVA light induces CD19+ IL‐10+ regulatory B cells and thereby promotes skin allograft survival. ²⁹

The manifestation of acute GvHD (aGvHD) in patients with allogeneic grafts was associated with a low number of T‐reg cells. ³⁰ , ³¹ Hence, several research groups have studied the effects of ECP on counts of T‐reg cells. In a model of murine GvHD, regulatory T cells were shown to be induced by ECP. ³² In the majority of CTCL and GvHD patients, an increase in T‐reg cells was observed after ECP therapy. Also, T‐reg cells showed an enhanced immunosuppressive activity. ³³ , ³⁴ , ³⁵ , ³⁶ , ³⁷ , ³⁸ These findings could explain, at least in part, the beneficial effects of ECP detected in GvHD and autoimmune diseases. In patients with SS, however, reduced counts of T‐reg cells have been reported, and their suppressive activity appears to be impaired. ³⁶ , ³⁹ , ⁴⁰ These observations have led to the notion that T‐reg cells could exert a suppressive impact on CD4‐positive tumour cells in patients with SS.

ECP slightly increases or stabilizes counts of peripheral CD4+ CD25+FoxP3+ T‐reg cells in lung transplant recipients. ⁴¹ Overall, the reinfusion of ECP‐treated leukocytes induced suppression of the humoral and cellular immune responses, and thereby improved and extended the tolerance and survival of transplanted tissues and organs. The mechanism by which ECP counteracts cardiac transplant rejection was studied using a murine model of ECP. ⁴¹ Splenocytes exposed to the combination of 8‐MOP and UVA were injected into syngeneic mice before and after heterotopic cardiac allograft transplantation. None of the mice received immunosuppressive agents. The treatment group showed extended cardiac allograft survival and increased counts of FoxP3‐expressing CD4+ CD25+ T cells when compared to controls. The authors concluded that the murine model of ECP extends graft survival in fully histoincompatible strain combinations with no immunosuppressive agent added. ⁴¹

In Crohn's disease, reinfusion of ECP‐treated apoptotic leukocytes to the patient is hypothesised to induce a tolerogenic response via T‐reg cells. Indeed, recirculation of DNA‐adduct‐positive cells to the intestinal mucosa has been described following ECP. ²⁶ , ⁴² Murine models of inflammatory bowel disease have provided information on the potential therapeutic role of T‐reg cells in overcoming inflammation in the intestine in humans. ⁴³

The effects of ECP on the immune system were studied in a randomized, double‐blind, placebo‐controlled trial in children with type 1 diabetes. ⁴⁴ No significant effects of ECP on lymphocyte populations were observed. However, in the placebo group, the proportions of activated CD4+ (T‐helper cells) and CD8+ cells increased over time, whereas such changes were not seen in the ECP‐treated group. These findings probably reflect the activation of lymphocytes as a part of the natural course of type 1 diabetes and suggest that ECP may exert immunosuppressive effects by preventing lymphocyte activation. ⁴⁵ , ⁴⁶ Patients treated with placebo showed reduced T‐reg cell‐associated activity, which seems to be counteracted by ECP because ECP‐treated patients showed preserved T‐reg cell activity. These data indicate that ECP may help maintain T‐reg cell‐associated activity in recent‐onset type 1 diabetes. ⁴⁷

Although distinct aspects of the mode of ECP action, such as the induction of T‐reg cells, are well understood today, we are still far from a complete understanding of how ECP works. Animal models help us to optimize currently used treatment regimens with respect to the number of cycles, concentrations of 8‐MOP, doses of UVA, and the number of cells treated in one clinical setting. Also, an enhanced understanding of the mechanism of action will finally enable ECP therapy to be directed towards those patients who will most benefit from it.

Methods

The present updated guidelines on the use of ECP were developed based on best medical practices, web review of relevant medical databases and literature, and collected expert opinions on the appropriate use of ECP.

In general, ECP is employed for the therapy of severe refractory disease courses or in situations in which other treatments have failed. However, ECP availability is limited, and evidence for its efficacy is derived from retrospective data, and small cohort or case‐controlled studies. There is a lack of randomized, controlled clinical trials in the literature. Double‐blind trials are challenging to perform and using sham photopheresis may be unethical for patients with severe diseases.

The present guidelines were drawn up to display the indications for which ECP is currently considered useful, as well as other indications where studies have shown promising results. For the main indications of ECP, namely CTCL and GvHD, the recommendations were developed by peers and leaders in the respective diseases. For minor indications, members of expert committees collaborated to examine all available evidence and to make appropriate recommendations. The aim was to answer clinical questions as follows:

• What are the potential indications for the treatment with ECP?

• Are there currently any guidelines/consensus statements on the use of ECP in this indication?

• Which patients should be considered for ECP treatment?

• What is the optimal treatment schedule, and how long should ECP treatment be continued?

• How can therapeutic efficacy be assessed?

For these recommendations, the individual experts in their area of expertise were consulted for their written contribution by email. In addition, individual co‐authors were personally contacted during meetings (St. Gallen, Switzerland, January 26, 2018; Lisbon, Portugal, March 19, 2018; Vienna, Austria, March 22, 2018; Orlando, USA, May 17, 2018; Paris, France, September 15, 2018; St. Gallen, Switzerland, September 28, 2018; Montreux, Switzerland, January 25, 2019) or by email if a meeting in person was not feasible. The document was circulated among all members of the Guidelines Subcommittee before it was submitted to the Guidelines Committee for final approval according to the standard operating procedures of the European Dermatology Forum (EDF).

Cutaneous T‐cell lymphoma

CTCL describes a heterogeneous group of rare lymphoproliferative disorders that are characterized by the accumulation of malignant T‐cell clones that are localized to the skin. ⁴⁸ The most common variants are mycosis fungoides (MF), which accounts for about 60% of CTCL cases, and Sézary syndrome (SS), which accounts for 5% of cases. MF is characterized by the presence of a clonal T‐cell population in the cutaneous environment and, in the early stages of the disease, presents as scaly patches or plaques, which may resemble eczema or psoriasis in appearance and are often associated with pruritus. As the disease progresses, patients may experience the growth of nodular lesions and large tumours, also with severe pruritus, which may ulcerate and result in chronic septicaemia, thrombosis, and pain.

SS is the ‘leukaemic’ form of CTCL, in which the dominant T‐cell population also circulates in the peripheral blood and may affect internal organs such as the lungs and spleen. MF/SS is classified into clinical stages from IA (the earliest stage) to IVB according to the degree of skin, lymph node, peripheral blood, and visceral organ involvement. ⁴⁹ , ⁵⁰

Except for allo‐transplantation which can be curative in some patients, there are no curative therapies for CTCL. Treatment is usually directed towards palliation and the induction of long‐term remissions. The aim is to reduce or clear skin lesions, including tumours, and reduce pruritus, thereby providing symptom relief and improving patient quality of life. ⁴⁸ In the early stages of MF, treatment usually involves skin‐directed therapies such as topical corticosteroids, topical chemotherapy (nitrogen mustard or bis‐chloronitrosourea), or phototherapy (narrow‐band UVB or PUVA). Systemic therapies, including chemotherapy and biological response modifiers (interferon [IFN]‐α, bexarotene), brentuximab vedotin or mogamulizumab, are used if the disease progresses or for those who present with more advanced‐stage disease, often in combination with skin‐directed therapies. ⁵¹

PUVA, in which patients take an oral formulation of 8‐MOP to induce photoactivation followed by exposure of their skin to UVA radiation, is a widely used and effective skin‐directed therapy for early‐stage, skin‐localized CTCL, which can produce relatively long‐lasting remissions. ⁵¹ It is, however, associated with the short‐term side‐effects of oral psoralen intake and possible long‐term complications such as photosensitivity and the potential for the development of skin cancer. ⁴ ECP has enabled the safety profile of PUVA to be improved, avoiding the potential complications associated with long‐term skin exposure to UVA. Thus, the benefits of ECP therapy can be extended to patient populations with more advanced disease stages, including those patients with malignant clones in the peripheral blood. ⁴ Many studies have demonstrated that ECP is of significant value for the treatment of CTCL. ⁵²

However, due to the low prevalence of CTCL and the fact that ECP therapy is only available in specialized centres, there are no prospective, placebo‐controlled, randomized clinical trials that evaluate the impact of ECP treatment on survival available in the literature. Thus, comparisons are usually made with historical controls. The initial ECP study in patients with CTCL resistant to other treatments was reported by Edelson et al. in 1987 and showed it to be a promising therapy. ² Among thirty‐seven patients, twenty‐seven (73%) responded to treatment, with an average decrease of 64% in cutaneous involvement; nine of these patients had a complete response (CR). Data from this study have recently been reanalysed using currently accepted international criteria. The skin overall response rate was 74%; 33% of patients were achieving ≥50% partial skin response, and 41% of patients were achieving ≥90% improvement. ⁵³ An update on the overall survival (OS) of these patients was also provided. Overall survival times were 9.2 and 6.6 years from disease onset and initiation of ECP, respectively.

Since 1987, numerous studies employing ECP have been conducted. A meta‐analysis of nineteen studies covering more than 400 patients at all stages of CTCL reported a combined overall response (OR) rate of 56% for ECP monotherapy and 56% when used in combination with other agents, and a CR rate of 15% and 18%, respectively. ⁵⁴ For erythrodermic disease, the OR rate was 58%, and the CR rate was 15%. Importantly, ECP was effective in SS, showing an OR rate of 43%, with a CR rate of 10%. Table 3, adapted from the UK consensus statement on the use of ECP for the treatment of CTCL and GvHD, provides a summary of the published response rates with ECP in the treatment of CTCL from 1987 to 2011. ⁵⁵ Based on the 30 separate studies in 689 patients published from 1987 to mid‐2007 that were analysed in the UK consensus statement, the mean OR rate in the studies was 63% (range 33–100%), and response rates were generally higher among patients with erythrodermic CTCL. ⁵⁵ The CR rates, where recorded, ranged from 0% to 62% (mean 20%). More recent studies published from late 2007 to 2011 report OR rates ranging from 42% to 80%, with CR rates ranging from 0% to 30%. ⁵⁶ , ⁵⁷ , ⁵⁸ , ⁵⁹ , ⁶⁰ , ⁶¹ , ⁶²

Table 3.

Summary of studies using extracorporeal photopheresis as monotherapy or in combination with other therapies for the treatment of cutaneous T‐cell lymphoma (adapted from Scarisbrick et al. ⁵⁵ )

	Patients (n)	OR (%)	CR (%)	PR (%)	MR (%)
Edelson et al., 1987 2	37 (erythrodermic 29)	73 (27/37) 83 (24/29)	24 (9/37)	35 (13/37)	14 (5/37)
Heald et al., 1989 64	32 (erythrodermic 22)	NK 86 (19/22)	23 (5/22)	45 (10/22)	18 (4/22)
Nagatani et al., 1990 151	7	43 (3/7)	NK	NK
Zic et al., 1992 152	20	55 (11/20)	25 (5/20)	30 (6/20)
Koh et al., 1994 153	34 (erythrodermic 31)	53 (18/34)	15 (5/34)	38 (13/34)
Prinz et al., 1995 154	17 (erythrodermic 3)	71 (12/17)	0 (0/17)	41 (7/17)	29 (5/17)
Duvic et al., 1996 155	34 (erythrodermic 28)	50 (17/34)	18 (6/34)	32 (11/34)
Gottlieb et al., 1996 65	28 (erythrodermic NK)	71 (20/28)	25 (7/28)	46 (13/28)
Stevens et al., 2002 156	17 (erythrodermic)	53 (9/17)	29 (5/17)	24 (4/17)
Zic et al., 1996 66	20 (erythrodermic 3)	50 (10/20)	25 (5/20)	25 (5/20)
Konstantinow et al., 1997 157	12 (erythrodermic 6)	67 (8/12) 50 (3/6)	8 (1/12) 0 (0/6)	42 (5/12) 50 (3/6)	17 (2/12)
Miracco et al., 1997 158	7	86 (6/7)	14 (1/7)	71 (5/7)
Russell‐Jones et al., 1997 159	19 (erythrodermic)	53 (10/19)	16 (3/19)	37 (7/19)†
Vonderheid et al., 1998 160	36 (erythrodermic 29)	33 (12/36) 31 (9/29)	14 (5/36) 10 (3/29)	19 (7/36) 21 (6/29)
Zouboulis et al., 1998 161	20	65 (13/20)	NK	NK
Jiang et al., 1999 162	25 (erythrodermic)	80 (20/25)	20 (5/25)	60 (15/25)
Bisaccia et al., 2000 69	37	54 (20/37)	14 (5/37)	41 (15/37)
Crovetti et al., 2000 163	30 (erythrodermic 9)	73 (22/30) 66 (6/9)	33 (10/30) 33 (3/9)	40 (12/30) 33 (3/9)
Wollina et al., 2000 164	20	65 (13/20)	50 (10/20)	15 (3/20)
Wollina et al., 2001 165	14	50 (7/14)	29 (4/14)	21 (3/14)
Bouwhuis et al., 2002 166	55 SS	80 (44/55)	62 (34/55)	18 (10/55)
Knobler et al., 2002 167	20 (erythrodermic 13)	50 (10/20) 85 (11/13)	15 (3/20) 15 (2/13)	54 (7/13)	15 (2/13)
Suchin et al., 2002 67	47	79 (37/47)	26 (12/47)	53 (25/47)
Quaglino et al., 2004 168	19	63 (12/19)	NK	NK
De Misa et al., 2005 169	10 (advanced SS)	60 (6/10)	10 (1/10)
Rao et al., 2006 170	16	44 (7/16)	NK	NK
Gasova et al., 2007 171	8 (2 with CTCL)	100 (2/2)	NK	NK
Tsirigotis et al., 2007 56	5 (SS 2)	80 (4/5)	20 (1/5)	60 (3/5)
Arulogun et al., 2008 57	13 (all SS; 12 erythrodermic)	62 (8/13)	15 (2/13)	46 (6/13)
Booken et al., 2010 58	12 (all SS)	33 (4/12)	0 (0/12)	33 (4/12)
McGirt et al., 2010 59	21 (18 erythrodermic)	57 (12/21)	14 (3/21)	19 (4/21)	24 (5/21)
Quaglino et al., 2013 62	48 (all erythrodermic;12 MF, 36 SS)	60 (29/48)	13 (6/48)	48 (23/48)
Raphael et al., 2011 61	98 (all erythrodermic)	74 (73/98)	30 (29/98)	45 (44/98)
Talpur et al., 2011 60	19 (all early‐stage MF)	63 (12/19)	11 (2/19)	53 (10/19)

CR, complete response; MF, mycosis fungoides; MR, minor response (>25% improvement in skin scores); NK, not known; OR, overall response (CR + PR); PR, partial response (>50% improvement in skin scores); SS, Sézary syndrome; CTCL, cutaneous T‐cell lymphoma.

^†Combined PR and MR.

ECP is beneficial in the treatment of CTCL. ⁵² However, it is also apparent that there are considerable differences in response rates between centres and the type of device used. Differences in the selection of patients, stage of the disease, prior treatments, treatment schedule of ECP, and the definition of response used might explain the large variability in the study results. ⁵⁵ Similar considerations apply to studies reporting on survival rates of patients with CTCL treated with ECP. Variable median survival data have been reported for SS, ranging from 30 to 60 months. Much longer median survival times for CTCL patients treated with ECP have been reported, but not all patients in these studies had the erythrodermic disease, or they had received other therapies in combination. ⁶³ , ⁶⁴ , ⁶⁵ , ⁶⁶

In most case series, ECP was used as monotherapy or in conjunction with other treatments. Such combination therapies have been investigated to further improve response rates, particularly in patients with a high tumour burden. Raphael et al. ⁶¹ published the most extensive case series of CTCL patients treated with ECP. The group reported on their 25‐year experience from a total of 98 erythrodermic CTCL patients treated with ECP for a minimum of 3 months. A significant clinical improvement was obtained in 75% of patients with a multimodality therapy; 30% achieved CR. Previously, Suchin et al. reported on 47 patients who had received at least six cycles of ECP. In these patients, stage III or IV CTCL was diagnosed in 68%, and malignant T cells were detected in the blood of 89%. ⁶⁷ Thirty‐one patients received treatment with ECP plus other drugs, including IFN‐ɑ, IFN‐ɣ, granulocyte‐macrophage colony‐stimulating factor (GM‐CSF; sargramostim) or systemic retinoids for 3 months at least. Overall, 79% of the patients responded well to therapy; 26% were achieving CR. Among patients receiving combination therapy, 84% responded well to therapy, and 20% were attaining CR; the OR rate with ECP monotherapy was 74% (CR rate was 38%). The median survival times were 74 months for the combination therapy and 66 months for ECP monotherapy; the difference was not statistically significant.

A prospective observational study of 48 patients with erythrodermic CTCL (thirty‐six with SS) reported a response rate of 58% for ECP alone, compared with 64% for combination therapy in patients with more adverse prognostic factors. ⁶² Similarly, Duvic et al. ⁶⁸ reported on a slightly higher response rate among 32 patients treated with ECP in combination with IFN‐ɑ, bexarotene, or GM‐CSF compared with 54% for ECP monotherapy. A number of other studies with ECP plus IFN‐ɑ have been published that report an increased response rate compared with ECP monotherapy. ⁶⁵ , ⁶⁹ However, none of these studies was controlled or randomized, making it difficult to assess how much of the clinical benefit is due to IFN‐ɑ and ECP and what the magnitude of potential synergistic effects is.

In the USCLC review of the 34 patients presenting with SS treated with ECP, IFN, and bexarotene, 30 patients (88.2%) responded to the combined therapy, including eleven patients with CR (32.4%). ⁷⁰ Bexarotene oral dosages ranged from 75 to 450 mg per day. Subcutaneous dosages of IFN‐ɑ and IFN‐γ ranged from 1.5 to 6 MU given three times a week and 40 to 100 µg given five times a week, respectively.

A total of 97 CTCL patients included in five UK sites (2010–2015) were investigated. ⁷¹ Patients tended to be treated early in the course of their disease (median time from diagnosis of CTCL to ECP therapy was 4.6 months). In 45.4% of cases, ECP was used as first‐line systemic therapy. Most patients had advanced disease stage IIIA‐IVA2 at the start of treatment, but three had early‐stage MF (treated for 2, 34 and 148 cycles, respectively). The intention to treat response rate at 6 months was 61.2% (60/97 patients). The median duration of ECP therapy was 9 months (range 1–118 months), and the median number of treatments was 16 cycles (range 1–148). Most patients (72%) were receiving concurrent systemic therapy at the start of treatment. The authors concluded that distinct long‐term responders might have improved survival. However, these results may be confounded by other prognostic factors.

Extracorporeal photopheresis has also been used in combination with total skin electron beam (TSEB) therapy. A retrospective study of 44 patients with erythrodermic MF/SS treated with TSEB with or without ECP reported an overall CR rate of 73%; the 3‐year disease‐free survival rate was 63%. ⁷² Among those patients who were receiving TSEB plus ECP, the 3‐year disease‐free survival rate was 81% compared to 49% for TSEB monotherapy. Based on these data, further studies using the combination of TSEB and ECP are warranted.

Most of the studies with ECP in CTCL have primarily included patients with advanced stages of the disease. Recent guidelines recommend ECP as first‐ or second‐line therapy for erythrodermic MF and SS. ⁵¹ , ⁵⁵ , ⁷³ , ⁷⁴ , ⁷⁵ , ⁷⁶ Its use in early stages of CTCL is controversial but warrants further investigation. A literature review of data from 16 studies with ECP or ECP plus adjuvant therapy performed between 1987 and 2007 included a total of 124 patients with early‐stage CTCL (stage IA, IB, IIA). This study revealed that response rates ranged from 33% to 88% for ECP monotherapy and 50–60% for ECP plus adjuvant therapy. ⁷⁷ Furthermore, many early‐stage patients treated with ECP achieved long‐lasting regression of the disease. In a recent prospective clinical trial, 19 patients with early‐stage MF were treated with one ECP cycle every 4 weeks for 6 months. ⁶⁰ Patients with a partial response (PR) continued with ECP monotherapy for another 6 months, whereas non‐responders were allowed to receive additional therapy with oral bexarotene and/or IFN‐ɑ. The OR rate for ECP monotherapy was 42% (8/19, including 1 CR; 7 PRs), with an overall duration of response of 6.5 months (range 1–48). Seven patients with stable disease at 3 months received additional bexarotene and/or IFN‐ɑ, and four of these patients (57%) responded to therapy. For all 19 patients, the OR rate was 63% (2 CRs, 10 PRs). Most guidelines do not indicate the use of ECP in early‐stage disease, but the National Comprehensive Cancer Network (NCCN) Guidelines recommend ECP in patients with stage IA, IB, and IIA refractory disease. ⁷⁶

In summary, ECP administered as monotherapy or in combination with other immunotherapies can be alternative treatment options that have proven effective and might beneficially impact survival rates in patients with advanced CTCL, i.e., a patient population who is typically resistant to conventional treatments and, therefore, shows poor prognosis. Given the favourable side‐effect profile of ECP compared with other therapies and its demonstrated efficacy in late‐stage CTCL, this treatment modality might also be useful in earlier stages of the disease as recently suggested by Talpur et al. and others. ⁵² , ⁶⁰ However, there is substantial intersubject variability in response to ECP therapy in CTCL disease. Therefore, attempts have been made to characterize and identify those patients who are most likely to respond to therapy.

Baseline predictors of response to photopheresis have recently been summarized (see Table 4). ⁷⁸ Although these criteria are useful in identifying responders to ECP, these criteria consistently need to be adapted and improved. ⁷⁹ A critical factor for the success of ECP therapy is that the patient's immune system must be capable of responding appropriately to malignant cells that have undergone photoactivation. ⁸⁰ , ⁸¹

Table 4.

Baseline predictors of response to photopheresis in the treatment of CTCL

Low tumour load of malignant T cells	Parameter	Reference
Skin	Erythroderma	62, 172
	Plaques <10–15% total skin surface	172, 173
Blood	Lower percentage of elevated circulating Sézary cells	59, 61, 174
	Lower CD4/CD8 ratio <10–15	61, 62, 174, 175
	Lower % CD4+ CD7− <30%	61, 156
	Lower % CD4+ CD26− <30%	61
	Normal LDH levels	62, 175
	B0 or B1 blood‐stage	62
	Lymphocyte count <20 000/µL	173
Lymph nodes	Lack of bulky adenopathy	173
Visceral organs	Lack of visceral organ involvement	173
Peripheral blood involvement
	B1 blood stage > B2 blood stage	62, 81, 173
	Presence of a discrete number of Sézary cells (10–20% mononuclear cells)	172
Relatively intact immune system
	Higher % monocytes >9%	61
	Increased eosinophil count >300/mm3	59
	No previous intense chemotherapy	173, 176
	Short disease duration before ECP (<2 years from diagnosis)	62, 173
	↑ NK cell count at 6 months into ECP therapy	62, 154
	Near‐normal NK cell activity	172
	Normal CD3+ CD8+ cell count >200/mm3	62
	High levels of CD4+ Foxp3+CD25− cells at baseline	177
Other monitored factors
PBMC microRNA levels	↑ miR‐191, ↑ miR‐223, ↑ miR‐342 at 3 months into ECP monotherapy	178
Soluble IL‐2 receptor	↓sIL‐2R at 6 months into ECP	170
Neopterin	↓ neopterin at 6 months into ECP	170
Beta2‐microglobulin	↓ beta2‐microglobulin at 6 months into ECP	170
Response at 5‐6 months of ECP	Predicts durable response and long‐term survival	66, 156

LDH, lactate dehydrogenase, NK natural killer, ECP, extracorporeal photopheresis, PBMC, peripheral blood mononuclear cell.

Adapted from Zic JA. Extracorporeal photopheresis in the treatment of mycosis fungoides and Sézary syndrome. ⁷⁸

Existing clinical guidelines

Several professional organizations have set up guidelines on the management of CTCL and the use of ECP. In 2006, the European Organisation for Research and Treatment of Cancer (EORTC) recommended ECP for the first‐line treatment of SS and MF stage III with a C‐strength of recommendation (on a scale from A to D). ⁵¹ , ⁷³ In MF, the level of evidence was rated 4 (evidence from case series, poor‐quality cohort or case‐control studies), and in SS, 2b (evidence from individual cohort study or poor‐quality, randomized, controlled trial). Although not recommended by EORTC, it was mentioned that ECP treatment is usually performed on two consecutive days at 4‐week intervals, continued for up to 6 months, and followed by maintenance therapy.

The UK Photopheresis Expert Group consensus statement recommends ECP for the treatment of patients with CTCL if they fulfil the criteria of erythroderma and stage III or stage IVA CTCL and at least one of the minor criteria, which are: (i) circulating clonal disease (circulating T‐cell clone proven by polymerase chain reaction or Southern blot analysis), (ii) evidence of circulating Sézary cells (>10% of circulating lymphocytes), and (iii) a CD4/CD8 ratio higher than 10. ⁵⁵ The recommended ECP treatment schedule is one cycle on two consecutive days every 2–4 weeks. It may be administered more frequently in symptomatic patients and those with a high blood tumour burden. At the maximum clinical response, ECP treatment should be tapered to one cycle every 6–12 weeks before it is completely stopped. However, in a very recent update from 2017, the UK Photopheresis Expert Group revised its recommendations and suggested continuing with ECP treatment in patients with complete, partial, or minimal clinical response. ⁸² This revised recommendation is in keeping with other treatments for advanced MF/SS. They should be continued for as long as a clinical benefit is detectable including improvements leading to better quality of life which may encompass improved functionality, reduced symptoms and improved well‐being. Despite not reaching a partial response (better by 50%) in skin diseases, continuous treatment is recommended. However, in some patients, a durable response of more than five years has been observed with ECP, which is markedly better than conventional therapy with a median survival time of about three years in advanced‐stage patients. ⁸²

Guidance on the monitoring of treatment success is also provided. Assessments at 3‐month intervals will allow non‐responders to be offered a combinatory or alternative therapy to ensure that ECP treatment is not unnecessarily prolonged.

In 2006, the British Photodermatology Group and the UK Skin Lymphoma Group reported on the use of ECP in a variety of clinical conditions based on data that were derived between 1987 and 2001. ⁸³ . These groups concluded that there is (i) ‘fair’ evidence of the clinical benefit of ECP in patients with erythrodermic MF/SS (stage III/IVA/B1/0), with a strength of recommendation B (on a scale from A to E) based on a level of evidence of II1 (i.e. derived from well‐designed, non‐randomized, controlled trials); (ii) ‘fair’ evidence that supports the use of TSEB plus ECP for erythrodermic MF/SS patients, with a strength of recommendation B, level of evidence II2 (i.e. well‐designed cohort or case‐control studies); and (iii) ‘poor’ evidence that supports the use of IFN‐α plus ECP for erythrodermic MF/SS, with a strength of recommendation C, level of evidence II2. Per standard protocol, ECP treatments should be performed on two consecutive days per month, continued for up to 6 months, and followed by tapering or maintenance treatment in those patients who have adequately responded. The treatment intervals can be shortened to biweekly cycles in poor responders, or ECP can be combined with other therapeutic agents such as IFN‐α. Recommended time points on patient assessments and appropriate efficacy parameters are also listed. These recommendations have also been updated and adopted in the 2018 British Association of Dermatologists and U.K. Cutaneous Lymphoma Group guidelines for the management of primary cutaneous lymphomas. ⁸⁴

The US National Cancer Institute recommends ECP for the therapy of MF and SS. ⁷⁵ ECP is offered as an option for the treatment of stage III MF/SS and, either alone or in combination with TSEB, for the treatment of stage IV MF/SS. For patients with recurrent MF/SS, it is noted that ECP has produced tumour regression in those patients who were resistant to other therapies. No information was given on the appropriate monitoring of therapy or outcomes.

In 2012, the NCCN clinical guidelines on MF/SS stated that their recommendations are all based on category 2A evidence (lower level evidence). ECP was recommended as first‐line therapy for stage IV SS alone, or in combination with interferon or bexarotene. The guidelines also suggest that ECP may be used in relapsed or refractory stage III disease, and stages IA, IB‐IIA, which are refractory to skin‐directed therapy. ⁷⁶

The United States Cutaneous Lymphoma Consortium (USCLC) reviewed available therapeutic options for SS. ⁷⁰ Based on level II2 evidence, ECP was classified as category A systemic monotherapy. Level II2 evidence means that information was obtained from at least one prospective, well‐designed cohort or a case‐control study, preferably from more than one centre or research group. Similarly, TSEB plus ECP, alone and in combination with IFN‐α, IFN‐γ, or bexarotene and ECP plus bexarotene, IFN‐α, IFN‐γ, or low‐dose methotrexate alone or in combination are alternative therapeutic options.

Finally, the German Association of the Scientific Medical Societies (AWMF) provides guidance on the staging, assessment, diagnosis, and therapy of cutaneous lymphomas. ⁸⁵ , ⁸⁶ ECP was recommended as first‐line therapy for stage III erythrodermic MF and for SS. Their guidelines stated that ECP could be combined with IFN‐α, methotrexate, bexarotene, or PUVA. The AWMF also commented on the excellent safety profile of ECP. No rating of the grade of recommendation or level of evidence was given, and no information was provided on how these guidelines were prepared.

Recommendations

Patient selection

ECP should be considered as first‐line therapy for patients with MF/SS as follows:

• Erythrodermic MF stages IIIA or IIIB (B0 or B1) according to the revised International Society for Cutaneous Lymphomas (ISCL)/ EORTC classification). ⁴⁹

Even though case series have suggested that there is a potential benefit of ECP in patients with early‐stage disease (stage IA, IB, IIA), the consensus decision was that this application should only be considered for clinical trial purposes, as a variety of other effective, safe, and easily accessible treatment options are available for use at these stages. ⁶⁰

• MF/SS Stage IVA1 (T1‐T4, N0‐2, M0, B2)

• MF/SS Stage IVA2 (T1‐T4, N3, M0, B0‐2)

Treatment schedule

The recommended ECP treatment schedule is one cycle (i.e. one ECP procedure per day on two consecutive days) every 2 weeks for the first 3 months followed by an ECP cycle every 3–4 weeks. However, there is no optimal therapy, and other published guidelines have recommended one cycle every 2–4 weeks followed by tapering after the maximum response has been achieved. ⁵⁵

Currently, there are no data in the literature that support the concept of increased clinical efficacy if the frequency of ECP cycles rises. However, based on common clinical experience, it is assumed that an initially higher frequency of ECP treatments may result in a significant improvement of subjective symptoms, particularly in CTCL patients suffering from itchiness and those with B2 staging. Based on the patient's compliance, a standard treatment regimen could also be performed, according to the policies and possibilities at the centre. Treatment of CTCL patients should be continued for 6 months at minimum before the response to ECP is evaluated. At maximum response, treatment should slowly be tapered to one treatment cycle every 4–8 weeks for maintenance therapy. In patients with a favourable response or disease stabilization and good quality of life, ECP treatment should be extended to more than 2 years. Treatment intervals should be progressively prolonged to up to 8 weeks. Patients who do not respond to ECP as first‐line therapy should be considered for combination therapies (i.e. ECP plus other drugs or interventions). IFN and/or bexarotene should be used in combination with ECP. Skincare and topical medications, including topical steroids and emollients, should also be prescribed to help alleviate ongoing symptoms.

In CTCL, patients with leukemic involvement and high white blood cell counts (i.e. >20 000 mm³), a cytoreductive treatment (debulking chemotherapy or alemtuzumab) aimed at decreasing the number of leukemic peripheral cells can be performed prior to the start of ECP therapy (potentially blunting and thus reducing the immune response). Also, local radiotherapy can be performed either before or during ECP to treat localized infiltrated skin lesions. While the combination of ECP with histone deacetylase inhibitors appears potentially useful, there are no published data available which support this approach at present. HDAC inhibitors may also blunt the immune response and might likely inhibit the generation of an optimal response to ECP.

Systemic concurrent therapies can be initiated at any time point. However, the consensus is that ECP monotherapy should be continued for at least 3 months before another drug or therapy is added. If patients are already on other therapies (bexarotene and/or IFN), ECP can be added without the withdrawal of the previous treatment.

Response assessment

Response assessments should be performed every 3 months according to the ISCL/EORTC/USCLC consensus statements. ⁴⁹ , ⁷⁰ , ⁸⁷ Based on clinical experience, responses to ECP therapy are not immediate and may take 3–6 months before a clinical response is observed. Thus, it was agreed that there should be at least 6 months of treatment and evaluation of the response to ECP before conclusions on its efficacy are being drawn. If CR is observed in CTCL patients, ECP treatment should not be stopped. Instead, ECP intervals should be extended to up to 8 weeks. If PR or stable disease is observed, the consensus statement suggests that the efficacy of combining ECP therapy with other treatments or increasing the frequency of ECP treatments should be evaluated. Similar recommendations are made for the case of progressive CTCL disease. Alternatively, ECP may be stopped in favour of other CTCL therapies.

Chronic graft‐versus‐host disease

Chronic GvHD (cGvHD) is a multisystem disorder occurring in the range of 30 to 50% of patients after allogeneic transplant. ⁸⁸

The likelihood of cGvHD rises with the use of unrelated, mismatched, older, or multiparous donors, in older recipients, and with the application of reduced‐intensity conditioning (RIC). RIC transplants are recognized for having haematological malignancies; notably, due to myeloid leukaemia, the number of patients with cGvHD has increased in recent years. ⁸⁹ Non‐myeloablative and RIC treatment regimens enable older patients or comorbid patients presenting with myeloid malignancies to be treated by allogeneic haematopoietic cell transplantation (HCT).

The difficulty of finding the optimal treatment vs. risk balance between cGvHD relapse, significant morbidity, and non‐relapse mortality has been addressed by Kuzmina et al. ⁹⁰ The first report on the successful treatment of cGvHD by use of ECP was published in 1994. ⁹¹ A more recent prospective multicentre study by Arora et al. performed between 2011 and 2014 at thirteen locations in the US reports on a cohort of 911 HCT patients (55% RIC). The authors of this study detected an incidence of 47% (95% confidence interval [CI]: 44%–51%) for cGvHD 2 years after the start of HCT. ⁹² The median time to the onset of cGvHD was 7.4 months or 222 days (range: 0.8–45.1 months). Oral mucosa was the most common site involved (59%), followed by skin (57%) and liver (56%). According to the National Institutes of Health (NIH) Consensus Conference, cGvHD symptoms were classified as mild in 19%, moderate in 53%, and severe in 28% of the patients. Among the 428 cGvHD patients, non‐relapse mortality was 12% (95% CI: 9%–16%). The probability of overall survival was 81% (95% CI: 76%–85%) 2 years after the diagnosis of cGvHD. The 2‐year non‐relapse mortality was 11% (95% CI: 5%–24%) for mild, 8% (95% CI: 5%–13%) for moderate, and 18% (95% CI: 12%–28%) for severe cGvHD. Among all patients with cGvHD, only 11% (95% CI: 8%–16%) were able to discontinue the entire immunosuppression 1 year after cGvHD diagnosis. Patients with severe GvHD were less likely (9%) to discontinue immunosuppression as compared to those with moderate (12%) or mild GvHD (18%).

The pathogenesis of cGvHD remains poorly understood. cGvHD is an immune‐mediated disease resulting from the interactions between the donor graft and the recipient's immune system. The donor T cells are the primary aggressors causing antibody‐mediated damage. There is increasing recognition that B cells may have a role in the initiation and progression of cGvHD pathogenesis by altered B‐cell homeostasis and disruption of tolerance mechanisms. ⁹³ Cytokine dysregulation is implicated with high levels of IL‐6, IFN‐γ, TNF‐ɑ, IL‐12, IL‐17, and low levels of IL‐10. ⁹⁴

The varied manifestations of cGvHD make the diagnosis and monitoring of the multisystem disorder difficult and comparing different clinical studies can be challenging. Criteria for the diagnosis and staging of clinical trials in cGvHD have recently been updated by Jagasia et al. to standardize diagnosis and assessment of response to treatment. Established first‐line treatment of cGvHD is with glucocorticosteroids (~1 mg/kg bodyweight of prednisone equivalent). ⁹⁵ An established first‐line treatment of cGvHD uses the administration of glucocorticosteroids (~1 mg/kg bodyweight of prednisone equivalent). The addition of a calcineurin inhibitor may be considered, if appropriate. ⁹⁶ In some patients, second‐line therapy must be initiated. However, the choice of second‐line agent varies considerably between centres and is often selected on an individual patient basis. Second‐line treatment options include the administration of ECP, mycophenolate mofetil, mTOR inhibitors, methotrexate, imatinib, rituximab and ruxolitinib. ⁹⁷

Extracorporeal photopheresis is an attractive treatment option exerting glucocorticosteroid‐sparing effects and showing response rates of approximately 60% in cGvHD patients. ⁸² In 2008, Scarisbrick et al. ⁵⁵ reviewed 23 studies, including a total of 633 patients presenting with cGvHD who underwent ECP treatment between 1987 and 2001. Response rates were determined based on organ involvement. The mean response rate was 68% (range, 29%–100%) in cutaneous cGvHD as derived from eighteen studies (patients evidencing CR were included in this analysis). In patients with hepatic involvement, the mean response rate was 63% (10 studies). Likewise, the mean response rate was 63% (9 studies) in patients presenting with mucosal involvement. An updated review of the literature reveals that thirteen additional investigations, comprising a total of 492 patients, are available for the analysis of response rates of the skin, liver, and oral manifestations in cGvHD patients. Response rate ranges were 31%–93% for the skin, 29%–100% for the liver, 21%–100% for oral disease, resulting in an overall response rate ranging from 36% to 83% (Table 5).

Table 5.

Extract of studies using extracorporeal photopheresis in adult patients with chronic graft‐versus‐host disease

	Patients (n)	CR/PR skin (%)	CR/PR liver (%)	CR/PR mouth (%)	OR (%)
Greinix et al., 1998 119	15	80	70	100	NK
Apisarnthanarax et al., 2003 179	32	59	0	NK	56
Seaton et al., 2003 180	28	48	32	21	36
Foss et al., 2005 181	25	64	0	46	64
Rubegni et al., 2005 182	32	81	77	92	69
Couriel et al., 2006 183	71	57	71	78	61
Greinix et al., 2006 184	47	93	84	95	83
Flowers et al., 2008 103	48	40	29	53
Dignan et al., 2012 185	82	92	NK	91	74
Greinix et al., 2011 186	29	31	50	70	NK

CR, complete response; NK, not known; OR, overall response; PR, partial response.

These data suggest that ECP is an effective treatment option for patients with cGvHD affecting skin, liver or oral mucosa. However, differences in the selection criteria of patients, and the use of different first‐line therapies, and second‐line treatment combinations may be the reason for the large variability in reported response rates. Alfred et al. investigated the results of 725 adult patients with either steroid‐resistant, steroid‐intolerant or steroid‐dependent cGvHD. ⁸² Response rates for cutaneous cGvHD were available from 23 studies showing a mean response rate of 74%. Response rates for hepatic cGvHD were available from fifteen studies that resulted in a mean response rate of 62%. Also, another twelve studies reported on mucosal cGvHD response rates resulting in a mean value of 62%. Response rates for pulmonary, ocular, and gastrointestinal involvement were 46%, 60%, and 46%, respectively. The overall response rate from a cross‐section of fourteen studies was 68% (Table 6).

Table 6.

Summary of studies using extracorporeal photopheresis in paediatric patients with chronic graft‐versus‐host disease

	Patients (n)	CR/PR skin (%)	CR/PR liver (%)	CR/PR mouth (%)	Comment
Rossetti et al., 1995 187	7	33 (2/6)	100 (1/1)	‐	50% (2/4) lung CR
Dall'Amico et al., 1997 188	4	67 (2/3)	‐	‐	67% (2/3) lung improved
Salvaneschi et al., 2001 114	14	83 (10/12)	67 (6/9)	67 (8/12)	79% OS
Halle et al., 2002 189	8	88 (7/8)	67 (4/6)	‐	100% OS
Perseghin et al., 2002 190	9	88 (7/8)	100 (2/2)	67 (2/3)	‐
Perutelli et al., 2002 191	7	‐	‐	‐	43% (3/7) CR; 57% (4/7) improved
Messina et al., 2003 115	44	56 (20/36)	60 (12/20)	‐	77% OS
Duzovali et al., 2007 192	7	‐	‐	‐	43% (3/7) improved; 43% (3/7) died
Kanold et al., 2007 116	15	75 (9/12)	82 (9/11)	86 (6/7)	67% (10/15) alive
Perseghin et al., 2007 193	25	67 (4/6)	67 (4/6)	78 (7/9)	76% (19/25) alive
Gonzales‐Vicent et al., 2008 117	3	100 (2/2)	100 (2/2)	‐	100% (3/3) alive
Perotti et al., 2010 118	23	96 (22/23)	100 (4/4)	80 (4/5)	83% (19/23) alive at 5 years

CR, complete response; PR, partial response; OS, overall survival.

Jagasia et al. recently reported on a randomized, prospective study investigating ECP use as first‐line therapy in cGvHD, based on the 2015 NIH consensus criteria for diagnosis and response assessment. The addition of ECP to standard of care was compared to standard of care alone. ORR at week 28 was 74.1% (ECP arm) vs. 60.9% (control arm). Patients in the ECP arm tolerated the treatment well while maintaining quality of life (QoL). ⁹⁸ QOL is an important facet of survival post‐HSCT, and scores in cGvHD are comparable to other chronic conditions such as multiple sclerosis and scleroderma. ⁹⁹ Maintaining or improving QoL has also been demonstrated in other ECP studies of cGvHD. ¹⁰⁰ , ¹⁰¹ , ¹⁰² , ¹⁰³ There is also emerging evidence to suggest that ECP helps maintain response to viral infections while also not increasing the risk of relapse, which is of clinical importance in this group of patients. ¹⁰⁴ , ¹⁰⁵

Flowers et al. ¹⁰³ published the first multicentre, randomized, controlled, prospective phase II trial of ECP in 95 patients with steroid‐refractory/‐dependent/‐intolerant cGvHD. The primary efficacy end‐point of the study was a blinded quantitative comparison of percentage change from baseline in Total Skin Score (TSS) of 10 body regions at week 12. The median percentage improvement in TSS at week 12 was 15% for the ECP arm compared with 9% for the control arm – a non‐significant difference. However, significantly more patients in the ECP arm had a complete or partial skin response, as assessed by the clinical investigators (P < 0.001). At week 12, the proportion of patients who had at least a 50% reduction in steroid dose and at least a 25% decrease in TSS was 8% in the ECP arm vs. 0% in the control arm (P = 0.04).

The safety profile of ECP is excellent, with only minimal side‐effects and no long‐term complications. When compared to other immunosuppressive therapies currently available for the treatment of cGvHD, ECP is not associated with organ toxicities, the occurrence of opportunistic infections, treatment‐emergent adverse events or underlying disease relapse. ⁹⁷ , ⁹⁸ , ¹⁰⁴ , ¹⁰⁵

Review of recent guidelines

Extracorporeal photopheresis is recommended as second‐line therapy for steroid‐intolerant, steroid‐refractory or steroid‐dependent cGvHD including but not limited to skin, oral mucosa, and liver involvement. ⁵⁵ , ⁹⁷ , ¹⁰⁶ , ¹⁰⁷ ECP should be performed weekly or every 2 weeks for a minimum of 3 months. The updated NIH criteria for measuring response in cGvHD patients should be used, and treatment should be tapered in responders. ⁸² , ¹⁰⁸

In 2013, an update of the ECP guidelines was provided by the Societa Italiana di Emaferesi e Manipolazione Cellulare (SIdEM) and the Gruppo Italiano Trapianto Midollo Osseo (GITMO) for both adult patients and paediatric patients with steroid‐resistant or steroid‐dependent cGvHD, irrespective of the extent and severity of the disease. ¹⁰¹ Also, it was noted that ECP might exert a potentially steroid‐sparing effect and improve the quality of life in responding patients. SIdEM and GITMO recently published a review article on the assessment of best practices among twenty‐four Italian centres investigated. ¹⁰⁰

In 2017, the UK Photopheresis Society published an update of its 2008 Guidelines. ⁸² For cGvHD of the skin, liver, and oral mucosa, they recommend ECP as second‐line therapy in steroid‐refractory, steroid‐intolerant or steroid‐dependent patients. Two treatments per week (one cycle) performed at 2‐week intervals and a monitoring schedule according to the updated NIH criteria are stipulated. ¹⁰⁸

The American Society for Apheresis recommends ECP for second‐line therapy of cutaneous and non‐cutaneous cGvHD (level of evidence cII), either as monotherapy or in conjunction with other therapies. ¹⁰⁹

Recommendations

Patient selection

Extracorporeal photopheresis should be considered as second‐line therapy in patients with steroid‐dependent, steroid‐intolerant or steroid‐resistant cGvHD and those with recurrent infections or a high risk of relapse of their underlying disease.

Patients ineligible for ECP include those with leucocyte counts below 1.0 G/L, intolerance to methoxsalen, heparin, or citrate products, and haemodynamic instability due to life‐threatening infections.

Treatment schedule

Extracorporeal photopheresis cycles are recommended weekly (two treatments; one cycle) for the first 3 months (or until GVHD stabilizes) followed by one cycle twice per month and then tapered depending on clinical response. The time schedule is largely dependent on the severity of cGvHD and the documented response. If cGvHD progresses, a change in the treatment strategy should be considered.

Response assessment

Serial response assessments should be carried out using NIH assessment criteria and performed by appropriately trained staff. ¹⁰⁸

Serial quality of life assessments, in addition to clinical response criteria, are recommended. Concurrent steroid and other immunosuppressive drug doses should be recorded at each assessment.

Acute graft‐versus‐host disease

Acute GvHD is a serious complication of allogeneic Haematopoietic Stem Cell Transplantation (HSCT) and a fundamental cause of transplant‐related morbidity and mortality, mainly due to severe infections and organ toxicities. ¹¹⁰ Furthermore, aGvHD is an important risk factor and determinant for the development of cGvHD. Currently, the standard first‐line therapy comprises the application of corticosteroids. However, <50% of patients respond to corticosteroid therapy, and thus a substantial proportion of patients presenting with aGvHD require salvage treatment. ¹¹⁰ , ¹¹¹ , ¹¹² , ¹¹³ So far, not a single immunosuppressive agent has been approved by regulatory agencies for the treatment of steroid‐refractory aGvHD; as a consequence, there is large variation in its management and treatment. Martin et al. published recommendations by the American Society of Blood and Marrow Transplantation (ASBMT) for the treatment of aGvHD based on a comprehensive and critical review of published reports. ¹¹⁰ Across the 67 studies selected, a total of nineteen different agents or medical devices were investigated. Besides horse anti‐thymocyte globulin (ATG), ECP was the most frequently studied therapeutic intervention. ECP was applied in approximately 300 patients with steroid‐refractory aGvHD, and these numbers have been increasing over time. ¹¹⁴ , ¹¹⁵ , ¹¹⁶ , ¹¹⁷ , ¹¹⁸ , ¹¹⁹ , ¹²⁰ , ¹²¹ , ¹²² , ¹²³ , ¹²⁴ , ¹²⁵ , ¹²⁶ , ¹²⁷ , ¹²⁸ , ¹²⁹ , ¹³⁰ , ¹³¹ , ¹³² , ¹³³ , ¹³⁴ , ¹³⁵

Overall, the median rates for CR and PR of cutaneous manifestations of patients are 75% each (range 50–100%). Accordingly, the median rates for CR and PR of patients with hepatic involvement are 47% each (range 0–100%), while the median rates for CR and PR of gastrointestinal manifestations are 58% each (range 0–100%). ECP was tolerated excellently as the side‐effects observed were only mild in severity, consisting primarily of reversible, temporal drops in peripheral blood cell counts after the first courses of ECP.

The results of studies that employed ECP as second‐line treatment of aGvHD are summarized in Table 7. ¹¹⁴ , ¹¹⁵ , ¹¹⁶ , ¹¹⁷ , ¹¹⁸ , ¹¹⁹ , ¹²² , ¹²⁶ , ¹²⁷ , ¹²⁸ , ¹³⁰ Promising first results from a preliminary study were confirmed by a pilot study performed on twenty‐one aGvHD patients. ¹¹⁹ , ¹³⁶ . Subsequently, Greinix et al. conducted a phase II study on ECP in fifty‐nine steroid‐refractory or steroid‐dependent adult patients with severe aGvHD. ¹²⁶

Table 7.

Summary of studies using extracorporeal photopheresis in the second‐line treatment of acute graft‐versus‐host disease

	Patients (n)	CR skin (%)	CR liver (%)	CR gut (%)	OS (%)
Salvaneschi et al., 2001 114	9	67 (6/9)	33 (1/3)	60 (3/5)	67
Dall'Amico et al., 2002 122	14	71 (10/14)	57 (4/7)	60 (6/10)	57
Messina et al., 2003 115	33	76 (25/33)	60 (9/15)	75 (15/20)	69 at 5 years
Garban et al., 2005 127	12	67 (8/12)	0 (0/2)	40 (2/5)	42
Greinix et al., 2006 126	59	82 (47/57)	61 (14/23)	60 (9/15)	47 at 5 years
Kanold et al., 2007 116	12	90 (9/10)	56 (5/9)	83 (5/6)	75 at 8.5 months
Calore et al., 2008 130	15	92 (12/13)	−	100 (14/14)	85 at 5 years
Gonzales‐Vicent et al., 2008 117	8	100 (8/8)	100 (2/2)	57 (4/7)	38
Perfetti et al., 2008 128	23	65 (15/23)	27 (3/11)	40 (8/20)	48 at 37 months
Perotti et al., 2010 118	50	83 (39/47)†	67 (16/24)†	73 (8/11)†	64 at 1 year
Jagasia et al., 2013 139	57	67 (38/57)†	67 (38/57)†	67 (38/57)†	59 at 2 years
Calore et al., 2015 133	72	78 (50/64)	84 (10/12)	76 (42/55)	71 at 5 years

CR, complete response; OS, overall survival; PR, partial response.

^†Combined CR and PR.

In contrast to the pilot study, an intensified schedule of ECP was applied in the respective phase II study, consisting of 2–3 treatments per week until a maximum response was observed. ¹²⁵ , ¹²⁶ By using this ‘intensified’ ECP schedule, CR rates improved in patients with grade IV aGvHD (60% vs. 12%) and gastrointestinal involvement (73% vs. 25%) compared to results from the pilot study. ¹²⁵ , ¹²⁶ , ¹³⁶ . The best response rates to ECP were observed after a median treatment duration of 1.3 months (range 0.5–6), and no flare‐ups were detected after tapering and discontinuation of corticosteroid therapy. In ECP‐responders, corticosteroid therapy was discontinued after a median of 55 days (range 17–284 days) after the start of ECP. In subsequent univariate analyses, the following parameters were identified as significantly affecting the outcome of aGvHD patients treated with ECP: (i) the grade of aGvHD, (ii) the number of organs involved at the start of ECP and first‐line therapy with corticosteroids, and (iii) the cumulative corticosteroid dose given prior to ECP. However, in logistic regression analyses, a low grade of aGvHD at the start of ECP therapy and the late onset of corticosteroid drugs after HSCT were the only variables that affected CR outcomes positively. Three months after the start of ECP, the cumulative incidence of transplant‐related mortality at 4 years was 14% in patients achieving CR of steroid‐refractory aGvHD, compared to 73% in patients without CR (P < 0.0001). Patients with CR of steroid‐refractory aGvHD with ECP had a significantly improved OS rate of 59%, compared to 11% in patients without CR (P < 0.0001). The cumulative incidence of relapse at 4 years was 28%, which was thus not increased when compared with HSCT patients not receiving ECP. In general, treatment with ECP was well‐tolerated.

Perotti et al. ¹¹⁸ recently reported on excellent response rates in 50 patients with steroid‐refractory aGvHD and confirmed the corticosteroid‐sparing effect of ECP. There was a policy of early intervention in place in patients with aGvHD, so the median time from onset of symptoms to the start of ECP therapy was only 9 days. The OR rate was 68% (32% CR and 36% PR), with almost similar response rates for skin (83%), liver (67%), and gastrointestinal tract (73%). Furthermore, the survival of ECP‐responders was significantly improved (62% vs. 6%) in aGvHD patients compared to ECP non‐responders (P < 0.001). The ability to decrease the corticosteroid dose 30 days after the start of ECP therapy was associated with a significantly reduced mortality rate, confirming the importance of sparing corticosteroid doses in aGvHD patients. Other authors have noted that the decrease of dose of corticosteroids after 30 days of therapy reflects a major advantage of ECP in the prevention of long‐term complications in children. ¹¹⁵ , ¹¹⁶

Several ECP studies conducted in paediatric patients with aGvHD have shown similar results to those obtained in adults. For instance, in a large, multicentre, retrospective study of thirty‐three paediatric patients with steroid‐refractory aGvHD, the overall rates were 54% for CR and 21% for PR. ¹¹⁵ The CR rates were 76% for skin symptoms, 75% for gastrointestinal manifestations, and 60% for liver involvement. The 5‐year OS rate was significantly better for responders than for non‐responders (69% vs. 12%; P = 0.001). Due to ECP therapy, immunosuppressive treatment was discontinued in eight out of nineteen survivors (42%) and reduced in another seven patients (36%). The median Karnofsky performance score improved significantly from 60% before ECP therapy to 100% (range 80–100%) after the completion of ECP therapy.

Supportive data were derived from subsequent small studies using a twice‐weekly ECP treatment regimen. ¹¹⁷ , ¹³⁷ In fifteen paediatric patients with steroid‐refractory aGvHD, the strongest predictor of response to ECP treatment was the stage of the disease itself: there was a 100% response rate for stage II, 75% for stage III, and 0% for stage IV. The stage of aGvHD and the response to ECP therapy both turned out to be significant predictors of transplant‐related mortality. A direct comparison of ECP and steroid therapy also showed somewhat better results for ECP in paediatric patients. ¹³⁰ Following ECP treatment, 73% of fifteen patients showed CR; the remaining 27% showed PR. CR was recorded in 92% of patients with skin manifestations, 71% with gastrointestinal tract manifestations, and 100% with liver disease. In comparison, 56% of the 16 patients receiving steroid therapy showed CR and 31% had PR; two patients had persistent cGvHD after 1 year. CR rates were 46% for skin, 57% for gastrointestinal tract, and 67% for liver. Transplant‐related mortality at day 100 of treatment was 6% for steroid therapy, but no patient died in the ECP group, and the 2‐year OS rates were numerically, but not significantly, higher in the ECP groups (85%) as compared to the steroid therapy group (57%). ¹³⁰

Several authors have pointed out that the use of ECP in children might be challenging because of low bodyweight, difficult vascular access, high extracorporeal volume, metabolic and haematological problems, and psychological intolerance. ¹¹⁵ , ¹¹⁶ , ¹³¹ Nevertheless, Messina et al. ¹¹⁵ were able to treat patients with bodyweights as low as 10 kg without detecting significant side‐effects. Kanold et al. reported on the follow‐up with paediatric patients diagnosed with GvHD. The authors put particular emphasis on the technical aspects of the ECP therapy. ¹¹⁶ The efficacy results were similar to those from other studies (7/12 patients [58%] with aGvHD showed CR, and 3/12 [25%] showed PR). They observed good tolerability of the treatment in patients with low bodyweight and emphasized the importance of a dedicated paediatric environment and care team to manage challenges such as difficult vascular access and psychological intolerance. ¹¹⁶

Calore et al. consecutively treated 72 paediatric patients (21 steroid‐refractory, 42 steroid‐dependent, 9 steroid‐naïve) between 1997 and 2013, achieving CR in 72% and PR in 11% of the patients. ¹³³ Transplantation‐related mortality was 3% and 20% among responders and non‐responders to ECP (P < 0.0001), respectively. The 5‐year overall survival showed a significant difference between responders and non‐responders (78% vs. 30%, P = 0.0004). ¹³³

The challenge of treating paediatric patients of low‐body‐weight (as low as 15 kg) was addressed in a study of patients presenting with aGvHD or cGvHD. ¹³¹ In contrast to many groups that have used an ‘offline’ two‐stage technique for mononuclear cell collection and irradiation, this group reported on the use of a sterile, closed‐loop procedure in which patients received fluid boluses of normal saline or 5% albumin to boost blood volume before and, if needed, during the ECP treatment. ¹¹⁶ , ¹¹⁷ , ¹¹⁸ This procedure allows for the use of continuous flow ECP even in patients with low bodyweight.

In a retrospective analysis of 128 patients with steroid‐refractory or steroid‐dependent aGvHD treated with ECP as second‐line therapy on a weekly basis, Das‐Gupta et al. reported 6‐month freedom from treatment failure of 77.3% and a 2‐year survival rate of 56%. ¹³⁸ Higher grades of aGvHD (grade 2 vs. grades 3‐4) at the start of ECP were predictive of poor clinical outcome as determined by survival analysis (hazard ratio [HR] 2.78, P < 0.001); non‐relapse mortality (HR 2.78, P = 0.001); and 6‐month freedom from treatment failure (HR 3.05, P < 0.002). Jagasia et al. ¹³⁹ compared ECP vs. anti‐cytokine therapy as second‐line treatment for steroid‐refractory aGvHD in a retrospective analysis. Overall response rates were 66% and 32% in the ECP and the anti‐cytokine cohort, respectively. The respective rates for CR were 54% and 20%. ECP was an independent predictor of response (HR 3.42, P = 0.007) and survival (HR 2.12, P = 0.018). In patients with steroid‐refractory aGvHD grade II, the use of ECP was associated with superior survival rates (HR 4.6, P = 0.016). Furthermore, the administration of ECP was associated with lower non‐relapse mortality (HR 0.45, P = 0.018). These promising results warrant confirmation in a prospective clinical study.

In a systematic review of six prospective studies including a total of 103 patients with steroid‐refractory aGvHD given ECP as salvage therapy, Abu‐Dalle et al. ¹⁴⁰ reported an overall response rate of 69%, including 84% for cutaneous, 65% for gastrointestinal, and 55% for hepatic manifestations. In a meta‐analysis of seven prospective studies on ECP treatment in patients with steroid‐refractory aGvHD, Zhang et al. included a total of 121 patients. The reported overall response rate was 71%, and the CR rate was 71%. ¹⁴¹ The efficacy rates of ECP on the skin, liver, and gut manifestations of aGvHD were 86%, 60%, and 68%, respectively.

To reduce the incidence of aGvHD, several studies investigated the use of ECP as part of the myeloablative conditioning regimen prior to HSCT. For instance, Miller et al. showed an unexpectedly low incidence of severe aGvHD if ECP was used as part of a novel ‘reduced‐intensity’ conditioning regimen on –day 6 and –day 7 (prior HSCT). No disease relapse or negative effects on the engraftment were observed. ¹⁴² However, the results from a phase II study revealed that after the addition of ECP to cyclosporine and methotrexate (all given as aGvHD prophylaxis as part of a standard myeloablative regimen), the incidence of aGvHD was similar to that found by other studies. ¹⁴³ The comparison of the ECP‐treated group to historical controls indicated a somewhat lower incidence of aGvHD grades II‐IV and improved OS of patients when ECP was introduced during the conditioning phase. ¹⁴³

Recently, Michallet et al. performed a prospective multicentre phase II study to evaluate the safety and efficacy of prophylactic ECP soon after the start of reduced‐intensity conditioning (RIC)‐HCT in 20 adult patients with haematological malignancies. ¹⁴⁴ ECP was started on day twenty‐one and was given twice per week for the first 2 weeks and then once per week for the following 4 weeks for a total of eight ECP courses. The cumulative incidence of aGvHD grades II‐IV on day 100 was 15%. The 2‐year OS and progression‐free survival (PFS) were 84% and 74%, respectively. ECP was tolerated well, and no adverse effects related to ECP were reported.

Kitko et al. investigated the combination of etanercept and ECP for GvHD prophylaxis in a prospective phase II study in 48 patients undergoing RIC‐HCT. ¹⁴⁵ The cumulative incidence of aGvHD grades II‐IV was 46% on day 100. The 1‐year OS was 73% because of low rates of non‐relapse mortality (21%) and relapse (19%). However, this strategy was ineffective in preventing chronic GvHD and late deaths. Therefore, the 2‐year survival probability declined to 56%. The preventive use of ECP may have some benefits, but data from more patients with a longer duration of follow‐up are needed for confirmation.

In conclusion, ECP is well‐tolerated, shows an excellent safety profile in children and adults, and is highly effective in aGvHD. The early start and use of an intensified ECP schedule consisting of 2–3 treatments per week and rapid tapering of corticosteroids in steroid‐refractory patients are necessary actions that might exert a significant impact on the patients' survival rate. However, more prospective clinical studies are warranted, including those studies investigating the use of ECP for prophylactic purposes.

Existing clinical guidelines

The American Society for Apheresis (ASFA) reviewed available data on ECP in aGvHD patients and concluded that OR rates for steroid‐refractory aGvHD in paediatric and adult patients range from 52% to100%. Corresponding response rates for the skin, gastrointestinal tract and liver were ranging from 66% to 100%, 40% to 83%, and 27% to 71%, respectively. ¹⁴⁶ ASFA recommends that ECP be used weekly on two consecutive days (one series) until disease response is maximized and then be tapered to every other week before discontinuation.

A joint working group established by the haemato‐oncology subgroup of the British Committee for Standards in Haematology (BCSH) and the British Society for Blood and Marrow Transplantation (BSBMT) provided guidelines on the diagnosis and management of aGvHD. ¹¹³ Based on the findings of this group, ECP is recommended as second‐line therapy for the treatment of steroid‐refractory aGvHD (level of evidence 2C). The BCSH/BSBMT commented on the excellent tolerability of ECP but concluded that the optimal treatment schedule and duration have not yet been established. However, Das‐Gupta et al. reported on a regimen of weekly ECP cycles for a minimum of 8 weeks continued until maximum response or CR is observed. ¹⁴⁷ Of note, no other immunosuppressive agent recommended by the BCSH/BSBMT obtained a higher level of evidence.

In a recent update of a consensus statement from the UK Photopheresis Society, the promising role of ECP in the treatment of steroid‐refractory aGvHD was confirmed. ⁸² Based on expert opinions, analyses of current practices and published results, in 2007, Kanold et al. ¹¹⁶ released clinical practice guidelines for physicians caring for children with aGvHD. In this guideline article, ECP is recommended for paediatric aGvHD patients who are unresponsive to corticosteroids as defined by the absence of clinical and biologic improvement after 1 week of corticosteroid therapy (prednisolone or methylprednisolone up to 2–5 mg/kg/day). However, the authors commented that they were considering ECP as early as 48 h after the initiation of corticosteroid therapy in severe cases of aGvHD. Thus, ECP was suggested as second‐line therapy of aGvHD in paediatric patients presenting either with steroid‐intolerant, steroid‐refractory or steroid‐dependent severe aGvHD. In detail, given the excellent safety profile of ECP, Kanold et al. considered ECP as first‐line therapy for paediatric patients with grade IV aGvHD (in combination with conventional immunosuppressive therapies) and as second‐line therapy in steroid‐refractory aGvHD grades II‐III. The authors recommended that ECP therapy be started at three times a week until a maximum response has been achieved, followed by individual progressive tapering of immunosuppressive treatment. Recommendations on vascular access and ECP technique in children were also provided.

Martin et al. ¹¹⁰ published the recommendations of the American Society of Blood and Marrow Transplantation (ASBMT) for the treatment of aGvHD based on a comprehensive and critical review of published reports. In total, data from 67 reports on 6‐month survival, CR and PR of aGvHD have been reviewed. Among the five studies with outliers in the 6‐month survival rate, the clinical trial by Messina et al. was particularly prominent. Since only children were treated (median age of 9.6 years), Martin et al. ¹⁴⁸ concluded that these outliers could be the result of age differences between patient cohorts, as the benchmark study had used horse ATG and included patients with a median age of 27 years. In conclusion, no specific agent was recommended or suggested to be avoided in the second‐line therapy of steroid‐refractory aGvHD.

The ASBMT reported on blood loss, hypocalcaemia, mild cytopenia and catheter‐associated bacteraemia due to ECP therapy but did not identify an increased risk of infections compared to other treatments. In particular, the ASBMT stated that there are no concerns about viral reactivations during ECP treatment. A typical ECP treatment schedule consists of three administrations during the first week followed by two administrations per week thereafter. According to ASBMT, the appropriate choice of second‐line treatment regimens should be guided by factors such as the potential toxicity of drugs, drug interactions, the experience of the physician with the agents, tolerability, and drug costs. Due to the excellent safety profile of ECP and the lack of interactions with other agents, ECP compares favourably to alternative immunosuppressive strategies, supporting the concept of its frequent use as second‐line therapy of steroid‐refractory aGvHD.

The Italian Society of Hemapheresis and Cell Manipulation (SIdEM) and the Italian Group for Bone Marrow Transplantation (GITMO) stated that ECP is a valuable option for patients with aGvHD who are unresponsive to steroids and calcineurin inhibitors. ¹⁰¹ GITMO and SIdEM recommend the use of ECP in both adults and children. The early start of ECP therapy, particularly in children and recipients of haploidentical or unrelated donor HCTs is suggested. In a recent survey of twenty‐four Italian HCT centres, more than 85% of these GITMO accredited centres agreed with best practice recommendations including the use of ECP. ¹⁰⁰

Recommendations

Patient selection

Patients presenting with aGvHD but not responding to first‐line therapy with corticosteroids at 2 mg/kg/day (progression of aGvHD after ≥3 days or lack of response after ≥7 days of corticosteroid treatment) should receive adjunct ECP as second‐line therapy.

Treatment schedule

Patients should undergo ECP cycles every week, comprising 2 to 3 treatments per week. At present, there is no evidence that maintenance ECP therapy is necessary. Thus, as soon as patients achieve CR, ECP should be discontinued.

Response assessment

The activity of aGvHD should be assessed at seven‐day intervals with staging according to published criteria. ¹⁴⁹ , ¹⁵⁰ Clinical assessments should relate to organ involvement, and data on the quality of life should also be collected.

Summary/conclusions

The first results from an international, prospective, multicentre clinical study on the use of ECP for the treatment of CTCL were published by Edelson et al. ² almost 32 years ago. Based on these data, the US FDA approved ECP as the first cellular immunotherapy for cancer. This approval triggered many investigators to test ECP in the prevention and treatment of a variety of T‐cell mediated diseases as outlined in the present guideline document. Over the last two decades, a large body of data has been derived from retrospective or prospective single and multicentre clinical trials with ECP that allow for the provision of recommendations on treatment schedules for different patient populations.

Extracorporeal photopheresis is a well‐tolerated therapy with an excellent safety profile. No significant side‐effects have been reported in any of the conditions reviewed here except for the short‐term effects of oral 8‐MOP observed in the earlier studies. Unlike other immunosuppressive therapies, ECP has not been associated with an increased incidence of infections. New technical developments and advances have substantially shortened the cycle duration and qualified ECP for the use in children. Initially, ECP had only been used empirically in clinical settings. However, recent preclinical and clinical research activities are throwing more light on the complexities of its mechanisms of action. Also, promising data on the identification of potential surrogate markers that are considered predictive of clinical response to ECP therapy are emerging.

Recent technical advances and a large body of data on the usefulness, safety and efficacy of ECP have established this method as a well‐recognized and accepted immunomodulatory second‐line therapy in a variety of dermal and non‐dermal diseases.