Extracorporeal photochemotherapy for graft versus host disease in pediatric patients

1. Introduction 

Graft versus host disease (GvHD) is still a major complication of allogeneic stem cell transplantation responsible for a large part of the transplant related mortality [1], [2]. Standard first-line approach for GvHD include corticosteroids and cyclosporine A [3]. Refractory diseases are treated with other therapeutics [2], [4]: thalidomide [5], azathioprine [3], psoralen and ultraviolet A light [6], [7], antithymocyte globulin and monoclonal antibodies [8], [9], [10], mycophenolate mofetil [11], tacrolimus [11], [12]. Unfortunately, despite these aggressive treatments GvHD remains uncontrolled in a more than 50% of affected patients. Moreover these therapies favor the development of life-threatening infections, secondary malignancies and multiorgan injury.

In view of incomplete response rates and the toxicity of conventional treatments, alternative approaches are being investigated as second line therapy in patients with refractory GvHD.

Extracorporeal photochemotherapy (ECP) based on the immunomodulating action of UV-A irradiation on autologous blood mononuclear cells collected by apheresis and photosensitized by 8-methoxypsoralen (8-MOP), is one of these relatively new nonpharmacologic therapeutic options.

In adult patients, in the last decade, ECP has shown considerable efficacy in the treatment of steroid-resistant, acute GvHD or chronic GvHD. Complete responses have been reported in patients with skin, mucosal, liver, lung and gut involvement [13], [14], [15], [16], [17], [18], [19].

In children, where the growing organism is particularly vulnerable to the consequences of GvHD itself and of prolonged steroids or other immunosuppressive treatment, the use of ECP may be of particular interest and has also been investigated [20], [21], [22], [23], [24], [25], [26], [27], [28], [29], [30], [31], [32]. Moreover, recently published data show that 64% of pediatric teams agree to this therapeutic approach as a component of the second line GvHD therapy in children [2]. Although worldwide experience with ECP has grown enormously over the past decade, only a few pediatric centers have experience with ECP and there have been only a few published study specifically evaluated the response to ECP of children with chronic GvHD (54 children reported so far) [20], [21], [22], [23], [24], [25], [26], [28], [29], [30], [31], [32] or acute GvHD (19 children reported so far) [21], [23], [26], [27], [29], [32] (Table 1).

Table 1. Publications on pediatric experience with ECP for GvHD

	Author (reference)	Year	Institution	Total children (<18 yrs)	GvHD type	Age (year)	ECP technique	Duration of ECP
	Rossetti [20]	1995	Department of Pediatric, Padua, Italy	1	Chronic	14	UVAR, Therakos	24 ECP/12 months
	Balda [28]	1996	Department of Dermatology, Augsburg, Germany	1	Chronic	13	UVAR, Therakos	15 ECP/12 months
	Schooneman [38]	1996	Blood Transfusion Centre, Nancy, France	1	Chronic	12	UVAR, Therakos	56 ECP/12 months
	Rossetti [21]	1996	Department of Pediatric, Padua, Italy	8	Acute (1 pts)	3–17	UVAR, Therakos	aGvHD: 2 months
					Chronic (7 pts)			cGvHD: 3–30 months
	Dall Amico [22]	1997	Department of Pediatric, Padua, Italy	4	Chronic	6–16	UVAR, Therakos	6–24 months
	Looks [27]	1997	Department of Dermatology, Jena, Germany	1	Acute	16	UVAR, Therakos	8 ECP/6 months
	Besnier [29]	1997	Department of Nephrology, Nantes, France	4	Acute (2 pts)	10–18	UVAR, Therakos	aGvHD: 8 & 9 ECP/3 weeks
					Chronic (2 pts)			cGvHD: 13 & 15 ECP/2–3 months
	Perotti [23]	1999	Policlinico San Matteo, Pavia, Italy	5	Acute (2 pts)	7–16	Cobe Spectra and UV-MATIC	aGvHD: 1–6 ECP
					Chronic (3 pts)			cGvHD: 2–18 ECP
	Biagi [30]	2000	Clinica Pediatrica, Milan, Italy	2	Chronic	5 & 14	Cobe Spectra and UV-MATIC	21 & 12 ECP/5 & 2 months
	D’Incan [25]	2000	Pédiatrie CHU, Clermont Ferrand, France	3	Chronic	5–8	Cobe Spectra and UV-MATIC	6–34 months
	Salvaneschi [26]	2001	Policlinico San Matteo, Pavia, Italy	23	Acute (9 pts)	5–18	Cobe Spectra and UV-MATIC	aGvHD: 0.5–13.4 months
					Chronic (14 pts)			cGvHD: 1–32 months
	Perseghin [32]	2002	Servizio di ImmunoematologiaMonza, Italy	13	Acute (4 pts)	4–18	Cobe Spectra and UV-MATIC	aGvHD: 6–12 ECP
					Chronic (9 pts)			cGvHD: 2–28 ECP
	Perutelli [31]	2002	Hematology LaboratoryGenoa, Italy	7	Chronic	6–16	UV-MATIC	24 ECP
	Halle and Kanold [24]a	2002	Pédiatrie CHU, Clermont Ferrand, France	13	Acute (2 pts)	5–15	Cobe Spectra and UV-MATIC	aGvHD: 16–18 ECP/2 & 3 months
					Chronic (11 pts)			cGvHD: 11–66 ECP/2–20 months

a And personal communication.

The available information on the safety and toxicity of using this method in small children are even more anecdotal. The principal reasons for the restriction in the use of ECP in children are: technical difficulties of leukapheresis procedures (venous access, hemodynamic, metabolic and hematological tolerance) and the necessity of a specially adapted pediatric patient approach to improve the psychological tolerance of this long term repetitive and time consuming treatment.

This review article will remain focused on the efficacy and the safety aspect of ECP in pediatric patients to provide information about the steps that should be taken to overcome the difficulties with ECP use in children with acute or chronic GvHD.

2. Inclusion and evaluation criteria 

As in adults patients, children included to the ECP trials are those with extensive chronic or grade II–IV acute GvHD, not responding to conventional treatment or responding to conventional treatment but with unacceptable treatment-related side effects. However, definitions of “not responding GvHD” and of “unacceptable treatment-related side effects” varied. For example: chronic GvHD not responding to at least one line of treatment, or absence of clinical and biological improvement of chronic GvHD after at least 2 months of conventional treatment, or steroid-resistant (a lack of stable clinical improvement after at least 7 days treatment with MPDN at a dose of 2–5 mg/kg/day) acute GvHD etc.

Complete hematologic remission, stable donor cell engraftment, no concomitant treatment with either ALG or MoAbs causing lymphocyte lysis are also required as inclusion criteria.

In all reported pediatric series GvHD was classified and the response to the ECP procedure was established according to published criteria on the treatment of patients with either acute or chronic GvHD [13], [14], [33], [34]. Initial and on-treatment clinical evaluation was generally done by a single observer.

The median interval between the occurrence of GvHD and the start of ECP was 22 days (range 9–47) for acute [21], [26] and 12 months (range 1–110) for chronic GvHD [20], [21], [22], [24], [26] with patients enrolled very late after GvHD onset––up to 8 years after transplant [30]. As patients with chronic GvHD are concerned, this interval being longer for patients enrolled at the beginning of both Italian [26] and French studies [24], [25].

In almost all reported patients, immunosuppressive therapy before the initiation of ECP consisted of steroids up to 5 mg/kg body weight and cyclosporine A or tacrolimus. The majority of patients with chronic GvHD have also received at least one agent from the second-line therapeutic group.

3. Venous access 

The difficulty of peripheral venous access is a crucial problem, particularly in children with severe skin involvement. On the other hand, the risk of central venous catheter-related complications has to be considered in these immunosuppressed patients. Therefore, the advantages and the disadvantages of each approach must be weighed. Patient acceptance of the chosen venous access being also of importance.

In opposition to adult trial reports which suggest the use of peripheral veins as the access of choice [13], [14], in pediatric patients, a central or a combined central/peripheral access are generally used. Rossetti and Dall’Amico use venous portacatheters or jugular vein catheters [20], [21], [22]. In Salvaneschi series a central venous catheter (when available, the same that had been used for transplant) was used for collection when peripheral venous access was inadequate. In detail, 20 of 23 patients were treated with ECP using a central venous catheter, and an additional central venous lane was required in nine [26]. In our center, a simple lumen central venous catheter together with peripheral vein was used for 70% of ECP-apheresis (previously inserted for other treatment or specially inserted for the ECP-aphereses) and a dual-lumen permanent central catheter access was used for 25% of ECP-apheresis [24], [25].

Contrary to our experience with PBSC collection in small children [35], for ECP-aphereses we have preferentially used a combined access: peripheral venous with a single-lumen permanent central catheter. In view of the fact that (1) an adequate vascular accesses for ECP must be ensure (defined in our centre as ⩾1 ml/kg/min) and (2) only 2 catheter-related infectious episodes are occurred in our patients (254 ECP, 2310 catheter-days), we recommend this approach for pediatric patients [24].

We believe that venous access choice should be made on a case by case basis, the child’s venous capital, cutaneous status and psychological acceptance being of importance.

4. Cell collection and irradiation 

In the most important pediatric series cell are collected according to a technique developed by Andreu et al. [36] using a Cobe Spectra separator (Cobe, Lakewood, CO, USA), on mononuclear cell program version 4 or 6 and a Vilber Lourmat irradiator [24], [26], [32].

The ECP-apheresis duration range from 1 to 4 h, generally about 2–3 h with 2–3 patient’s blood volumes processed. However the median total time of ECP procedure (apheresis, irradiation, reinfusion) was 4 h.

No extracorporeal line priming was performed, even in the smallest of patients. Hypothermia and hypocalcemia were systematically prevented.

Dall’Amico and Rossetti used the UVAR, Therakos instrument with reduced buffy coat and plasma volume [21], [22].

In the pediatric context the ECP technique developed by French groups [36] has several advantages over the classical technique: minimization of the extracorporeal volume, shortening of the time of procedure and the reduction of the final collection volume. However, future study may be of interest to compare the two systems in the pediatric context.

In our series, we have details the information about pre-collection blood cell count and about collection efficiency.

At the start of each ECP-apheresis the median number of white blood cells, lymphocytes and platelets were: 2.3×109/l (range 0.2–63.6), 0.9×109/l (range 0.03–5.9) and 395×109/l (range 22–784), respectively. Hemoglobin ranged from 71 to 140 g/l (median 103). A negative correlation was found between lymphocyte collection efficacy of ECP-apheresis (median=38%, range 6–148) and pre ECP-apheresis lymphocyte count (r=0.4, p=0.00001). There was no significant correlation between the lymphocyte collection efficacy and pre ECP-apheresis WBC count. The median number of total nucleated cells and lymphocytes irradiated and reinfused to patients in each procedure was 0.8×108/kg (range 0.05–5.8) and 5×107/kg (range 0.1–50) respectively. At least 1×107 lymphocytes/kg is derived from 2 patients blood volumes processed in 95% of ECP if the pre ECP-apheresis lymphocyte count is ⩾0.5×109/l. A target value of 5×107 lymphocytes/kg is obtained from a single ECP-apheresis in all patients if the pre ECP-apheresis lymphocyte count is ⩾2.0×109/l [24].

To date no evidence of a correlation between the number of lymphocytes reinfused and clinical efficacy has been reported. However, in Perseghin series patients who received a higher dose of mononuclear cells showed a better and more rapid clinical response [32].

In agreement with Perotti [23], we confirm that a low white blood cell count and thrombocytopenia are not absolute contraindications to ECP even in small children.

5. Toxicity and adverse events 

Treatment with ECP was generally associated with minimal side effects, even in the smallest of patients. In particular, extra-corporeal apheresis caused some episodes of hypotension, which, however, were only rarely symptomatic. No signs of citrate toxicity or metabolic complications were noted [20], [21], [22], [24], [26].

The changes of peripheral blood counts following each ECP-apheresis were significant in our study, with median platelet decrease of 17% (range 0–71) (p=0.0001) and median hemoglobin level decrease of 15 g/l (range 0–31) (p=0.0001) [24]. However, platelet depletion of more than 50% was noted only in 5% of procedures and contrary to other studies [13], [14] none of the patients had profound thrombocytopenia or anemia. Eight percent of ECP-apheresis were preceded or followed by red blood cell transfusion and for 2% of ECP-apheresis platelet procedure-related transfusions were needed [24]. In Italian study, a decrease in Hb level greater than 2 g per dl from the pre-ECP value was observed during treatment in all patients with acute GvHD and in 13 of the 14 children with chronic GvHD. Seven of the nine children with acute GvHD required RBC transfusion during ECP, whereas only one child with chronic GvHD did so [26]. This increase of transfusional demand may be avoided by processing fewer blood volumes. The management of transfusional demand is also, in part, operator dependant and the progressive mastering of the ECP technique may allow its reduction.

No child experienced life-threatening bloodstream infections related to the procedure. In Salvaneschi study the percentage of children with acute GvHD treated with ECP who experienced reactivation of human CMV infection was comparable to that of pediatric patients with acute GvHD [37]. In our series of children with chronic GvHD no reactivation of latent CMV infection occurred [24]. No patient died of human CMV or developed EBV-related lymphoproliferative disorder.

When oral 8-MOP was used nausea and vomiting were noted in 13% of patients [24].

In our study, the principal complaint by patients was the need to remain relatively sedentary for 3–4 h required for the ECP procedure and the repetitiveness of long term treatment. In one patient, a 14-year-old girl, ECP treatment had to be discontinued after 14 ECP because she refused to continue [24].

No adverse events were associated with the irradiated cell product reinfusion.

Even though, side effects related to ECP treatment were more frequently observed in patient with acute GvHD or lower body weight, the overall tolerance of ECP-aphereses was excellent. We believe that having a dedicated pediatric environment in which to carry out this technique together with an experienced, motivated and specifically pediatric team is of crucial importance for improving patient’s acceptance of this long-term therapeutic program. Support and explanation from the nursing and medical staff involved in the procedure is required over several months of treatment.

6. ECP timetable and treatment duration 

The timing-schedule of ECP in the GvHD treatment is still under discussion.

In patients with acute GvHD, ECP was initially performed three times a week (alternate days) until clinical improvement [26], [29]. Thereafter, responding patients were treated on 2 consecutive days at 2-week intervals for 3 months [26] or according to the course of disease [29]. Different schedule was proposed by Perseghin: two ECP/week for 4 weeks, followed by two ECP/2 weeks for a month for a total of 20 ECP [32].

In patients with chronic GvHD ECP timetables were less consistent. ECP sessions were performed: (1) three procedures a week for 3 weeks with tapering according to the course of disease [29]; (2) on 2 consecutive days at 2-week intervals for 3 months and than (when improved) for 2 consecutive days at 3-week intervals for 3 additional months [26]; (3) twice a week (two days intervals between two ECP) for 2 weeks and than once a week or every two weeks during the next 3 months [24]; (4) two ECP/week for 3 weeks, than two ECP/2 weeks for a month, and finally two ECP/month, up to 20 ECP [32]; (5) at 3 weeks intervals for the first 6 months and than intervals between treatment have been lengthen of a week every 3 months [22].

It has been suggested that ECP administrated on 2 consecutive days is more effective than single-day treatment courses [17]. However, in the pediatric setting this approach would seem to be problematical as the associated constraints would be unacceptable to these young patients and the risk of hematological and metabolic intolerance may be probably increased.

In all series ECP was progressively tapered for those patients who were stabilized or showed improvement and discontinued on a case by case basis, depending on treatment efficacy and tolerance.

It has also been suggested that successful ECP treatment depends on its early initiation [21], [22], specially in patients with acute GvHD [26]. However, as far as chronic GvHD is concerned, in children with severe long-term refractory progressive disease (up to 8 years post transplant) excellent responses to ECP treatment were observed [24], [25], [30]. These observations suggested that although it may be preferable to begin ECP early as second line therapy, ECP treatment may be beneficial in patients with late-stage disease.

The optimal duration of ECP treatment for GvHD is also unclear in children as it is in adults. The median treatment duration for acute GvHD was 4.6 months (range 0.5–13.4) in Salvaneschi study [26] and 6–12 procedures in patients reported by Perseghin [32].

Several reports have indicated that more that 6 months of ECP-treatment were necessary to improve severe chronic GvHD: 17 months (range 1–32) in Salvaneschi study [26], 6 months (range 3–30) in children reported by Rossetti [21]. In our study, all but one patient were treated for more than 6 months and in 3 patients treatment duration exceeded one year [24]. This confirms that multiple treatments several months apart may be required.

7. Response to ECP 

Studies reporting clinical outcome in pediatric patients with GvHD treated by ECP comprises a very limited number of children. Apart 3 Italian [20], [21], [22], [26], [32] and one French [24] series, there are mainly “a case” reports with only few information described. Also the response to ECP is difficult to compare because the criteria for defining a response are often not comparable.

In children with acute GVHD, data concerning 19 patients treated with ECP have been reported so far (Table 2). Overall response was as followed: complete (n=9) or partial (n=3) response was observed in 12/19 (63%) patients, in 7/19 patients, acute GVHD was stable (n=2) or worsened (n=5). Like in adults, a maximal response was observed after a median of 9 ECP (1 month of treatment) [24], [26], [33]. The best responders to ECP were children with cutaneous Grade I–III manifestations whereas the rate of response in grade IV cutaneous acute GVHD or liver and gut manifestations was lower. Response rates of different organs are shown in Table 2. Skin lesions improved in 13 of 17 (76%) patients, liver involvement improved in 3 of 8 patients. In contrast to previous published adult study [13], in pediatric patients we are unable to confirm that gut involvement predicts a poor response to ECP, as 4 of 9 children with gut involvement had either partial or complete response.

Table 2. Response to ECP in children with GvHD

	Author (reference)	No. of patients <18 yrs		Response	Characteristics of responses
	Rossetti [20], [21], Dall Amico [22]	Acute	1 pt	1 stable	Skin 0/1
		Chronic	7 pts	3 improved, 4 stable	Skin 2/6, lung 2/4, liver 1/1
	Balda [28]	Chronic	1 pt	1 resolved	Skin 1/1, liver 1/1
	Schooneman [38]	Chronic	1 pt	1 improved	Skin 1/1
	Looks [27]	Acute	1 pt	1 resolved	Skin 1/1
	Besnier [29]	Acute	2 pts	2 worsened	Skin 0/2, gut 0/2, liver 0/1
		Chronic	2 pts	2 improved	Skin 1/1, lung 1/1
	Biagi [30]	Chronic	2 pts	1 resolved, 1 improved	Skin 2/2
	Perseghin [32]	Acute	4 pts	1 resolved, 1 improved, 1 stable, 1 worsened	Skin 2/3, gut 0/1, liver 0/2
		Chronic	9 pts	5 resolved, 2 improved, 1 stable, 1 worsened	Skin 7/8, mucosal 2/3, liver 2/2
	Perutelli [31]	Chronic	7 pts	3 resolved, 4 improved	–
	Salvaneschi [26], Perotti [32]	Acute	9 pts	5 resolved, 2 improved, 2 worsened	Skin 8/9, gut 3/5, liver 1/3
		Chronic	14 pts	4 resolved, 5 improved, 2 stable, 3 worsened	Skin 10/12, joints 2/4, mouth 8/12, liver 6/9, lung 1/1
	Halle and Kanold [24]a, D’Incan [25]	Acute	2 pts	2 resolved	Skin 2/2, gut 1/1, liver 2/2
		Chronic	11 pts	3 resolved, 6 improved, 1 stable, 1 worsened	Skin 8/10, joints 2/2, mucosal 6/7, liver 8/9, gut 5/5

a And personal communication.

In children with chronic GVHD, 10 series have been considered for a total of 54 patients treated with ECP (Table 2). Overall response was as followed: in 17 of 54 (31%) children disease resolved completely, in 24 of 54 (44%) improvement was observed, 8 of 54 experiencing stabilization of the disease and in 5 of 54 GvHD worsened. In all responded patients 2–9 ECP procedures were needed before patient felt better and before a clinical improvement was noted [21], [24], [30], [32]. A maximal response was observed after a median of 6 months of treatment [26], [32]. Unfortunately, details of response rates of different organs are not available for all patients (Table 2).

As far as skin condition evolution under ECP therapy is concerned, 33 of 42 (75%) patients experienced improvement in their cutaneous status. In some patients with scleroderma-like lesions complete regression of skin retraction was noted, in others regression was only partial but important enough to improve patient’s quality of life, normal mobility, running and cycling became possible and Karnofsky performance scores improved after ECP [21], [24].

There is a controversy as to whether ECP treatment is beneficial on mucosal involvement. In published series, mucosal affection due to chronic GvHD resolved or improved in 8 of 10 patients. However, our long-term data suggest that ECP does not prevent an evolution of oral and genital lesions to atrophic lichen-like plaques and local steroids are needed to treat residual inflammation. Therefore, these patients should be carefully followed because a possible transformation into carcinoma cannot be excluded [24].

Efficacy of ECP on liver and gut disease was observed in pediatric patients. Complete or partial resolution of liver was seen in 18 of 22 (81%) patients. Improvement of lung involvement was observed in 4/6 patients.

Patients with ocular manifestation had only partial improvement with persistence of keratoconjunctivitis and photophobia.

Like in adults [13], [14], [15], [17], [18], in all responded pediatric patients, both with acute and chronic GvHD, ECP allowed reduction or discontinuation of the concomitant pharmacological immunosuppressive therapy without an increase in GvHD activity.

It should also be noted that like adults, responded children described a significant subjective improvement in their conditions and their general feeling of well being.

8. Long term follow up 

Only little information is available on the long term follow up of pediatric patients after ECP treatment.

As far as patients with acute GvHD are concerned the follow up range between 1 and 17 months [21], [24], [26], [29]. Six of 14 patients died of GvHD, relapse or infection.

For children with chronic GvHD follow up range between 10 and 153 months with 11 of 14 patients alive in Salvaneschie series [26]. In our series, 10 of 11 children are alive and well, with follow-up time between 3 and 65 months after termination of ECP. Besides cutaneous and mucosal sequels they have normal physical examination, blood and lymphocyte subsets counts, liver tests and serum chemistry. Nine of ten have normal growth rate (“catch up” growth was observed) and development, normal school and sport activity and a good quality of live. Two of them have thyroid hormone substitution. One patient with chronic wasting syndrome has persistent growth and weight deficit. She has growth hormone and thyroid hormone substitution and hyper-caloric supplements. No late infection episodes have been noted in these children [24].

9. Conclusion 

The data of this retrospective review demonstrate that ECP is beneficial and well tolerated in children with acute or chronic GvHD. It can be safely used even in young children with low body weight and a poor performance status when it was performed by a qualified pediatric team.

The use of ECP is of particular interest in children as its offers the possibility of either reducing or discontinuing steroids or other immunosuppressive treatment responsible of infectious complications, severe catabolic damage, hyperglycemia, a vascular necrosis of bone, a detrimental effect on growth and favoring the development of secondary neoplasia.

The observations concerning the response rate and onset suggest that in children with acute GVHD, ECP should be started early in the course of disease and employed over a relatively short period of time. As far as chronic GvHD is concerned, despite the fact that it is preferable to begin ECP early as second line therapy, it may also be beneficial in patients with late-stage disease.

Since ECP safety and feasibility have been demonstrated, larger trials (randomized prospective and multicenter) will be required to determine the proper place of ECP in the GvHD prophylaxis and treatment in children.