REPORT
`
`ASBM1
`
`American Society for Blood
`and Marrow Transplantation
`
`Consensus Conference on Clinical Practice in Chronic
`Graft-versus-Host Disease (GVHD): First-Line and
`Topical Treatment of Chronic GVHD
`
`Daniel Wolff,1 Armin Gerbitz,2 Francis Ayuk,3 Alexander Kiani,4
`Gerhard C. Hildebrandt,1 Georgia B. Vogelsang,5 Sharon Elad,6 Anita Lawitschka,7
`Gerard Socie,8 Steven Z. Pavletic,9 Ernst Holler,1 Hildegard Greinix10
`
`Chronic graft-versus-host disease (cGVHD) after allogeneic hematopoietic stem cell transplantation is still
`associated with significant morbidity and mortality. First-line treatment of cGVHD is based on steroids of
`1 mg/kg/day of prednisone. The role of calcineurin inhibitors remains controversial, especially in patients
`with low risk for mortality (normal platelets counts), whereas patients with low platelets at diagnosis and/
`or high risk for steroid toxicity may be treated upfront with the combination of prednisone and a calcineurin
`inhibitor. Additional systemic immunosuppressive agents, like thalidomide, mycophenolic acid, and azathio-
`prine, failed to improve treatment results in the primary treatment of cGVHD and are in part associated with
`higher morbidity, and in the case of azathioprine, with higher mortality. Despite advances in diagnosis of
`cGVHD as well as supportive care, half of the patients fail to achieve a long-lasting response to first-line treat-
`ment, and infectious morbidity continues to be significant. Therefore, immunomodulatory interventions with
`low infectious morbidity and mortality such as photopheresis need urgent evaluation in clinical trials. Beside
`systemic immunosuppression, the use of topical immunosuppressive interventions may improve local re-
`sponse rates and may be used as the only treatment in mild localized organ manifestations of cGVHD.
`
`Biol Blood Marrow Transplant 16: 1611-1628 (2010) Ó 2010 American Society for Blood and Marrow Transplantation
`
`KEY WORDS: Allogeneic hematopoietic stem cell transplantation, Graft-versus-host disease, Immunosup-
`pressive therapy
`
`INTRODUCTION
`
`Chronic graft-versus-host disease (cGVHD) con-
`tinues to be associated with significant morbidity and
`
`From the 1Department of Hematology and Clinical Oncology, Uni-
`versity of Regensburg, Germany; 2Department of Immunology,
`Campus Benjamin Franklin, Charite´dUniversity Hospital Ber-
`lin, Germany; 3Interdisciplinary Clinic for Stem Cell Trans-
`plantation, University Cancer Center Hamburg (UCCH),
`Germany; 4Department of Internal Medicine I, University of
`Dresden, Germany; 5Johns Hopkins Oncology Center, Balti-
`more, Maryland; 6Department of Oral Medicine, The Hebrew
`UniversityeHadassah School of Dental Medicine, Jerusalem,
`Israel; 7St. Anna Children’s Hospital, Vienna, Austria; 8Service
`d’Hematologie Greffe & Inserm Hospital Saint-Louis, Paris,
`9Experimental Transplantation and Immunology
`France;
`Branch, Center of Cancer Research, National Cancer Institute,
`Bethesda, Maryland; and 10Department of Internal Medicine I,
`Medical University of Vienna, Austria.
`Financial disclosure: See Acknowledgments on page 1625.
`Correspondence and reprint requests: Daniel Wolff, MD, Depart-
`ment of Hematology and Oncology, University of Regensburg,
`F.J. Strauss Allee 11, 93053 Regensburg, Germany (e-mail:
`daniel.wolff@klinik.uni-regensburg.de).
`Received February 10, 2010; accepted June 21, 2010
`Ó 2010 American Society for Blood and Marrow Transplantation
`1083-8791/$36.00
`doi:10.1016/j.bbmt.2010.06.015
`
`is the leading cause for late mortality after allogeneic he-
`matopoietic stem cell transplantation (allo-HSCT)
`[1,2]. Moreover, because of rising recipient age, and
`the use of unrelated donors as well as peripheral blood
`stem cells (PBSCs) as a graft source, the incidence of
`cGVHD has been increasing [3]. Although major prog-
`ress has been achieved in understanding the pathophys-
`iology of acute GVHD (aGVHD), cGVHD is far less
`defined. Current concepts include the persistence of
`alloreactive T cells, a Th1-Th2 shift of the cellular im-
`mune response, defective peripheral, and central toler-
`ance mechanisms (ie, failure of control by regulatory
`T cells and/or impaired negative selection of T cells in
`the thymus), replacement of antigen presenting cells
`(APCs) of the host by APCs of the donor leading to in-
`direct antigen presentation of allo-antigens, an increas-
`ing role of B cells producing auto- and allo-antibodies
`against the host, and unspecific mechanisms of chronic
`inflammation leading to fibrosis of involved organs [4].
`First-line treatment of cGVHD consists mainly of pred-
`nisone with a starting dose of 1 mg/kg/day, often com-
`bined with a calcineurin inhibitor (CNI). Evidence for
`first-line treatment options is based on controlled trials
`with the exception of severe cGVHD, which continues
`to be associated with interior survival [1]. Until recently,
`
`1611
`
`

`

`1612
`
`D. Wolff et al.
`
`Biol Blood Marrow Transplant 16:1611-1628, 2010
`
`no valid criteria for the diagnosis and staging of cGVHD
`severity were available, which limits the value of most re-
`ported trials on the treatment of cGVHD. The National
`Institutes of Health (NIH) consensus criteria on diag-
`nosis and staging of cGVHD as well as on treatment re-
`sponse criteria, reported in 2005, provide defined
`criteria that should improve the validity of reported re-
`sults on treatment of cGVHD in the future [5-9].
`Despite available evidence from controlled studies, no
`consensus has been achieved on first-line treatment
`of cGVHD. The Consensus Conference on Clinical
`Practice in Chronic GVHD held in fall of 2009 in Re-
`gensburg, Germany (complete program provided at
`www.gvhd.de), aimed to summarize the current avail-
`able evidence for first-line and topical treatment and
`to provide practical guidelines for the use of treatment
`modalities. The presented consensus was based on a re-
`view of published evidence and a survey on the current
`clinical practice including transplant centers from
`Germany, Austria, and Switzerland. Moreover, the
`consensus was circulated among all transplant centers
`performing allo-HSCT in Germany, Austria, and
`Switzerland, and was discussed during the Consensus
`Conference meetings. The Consensus Conference was
`organized under the auspices of the German working
`group on bone marrow and blood stem cell transplanta-
`tion (DAG-KBT) and the German Society of Hematol-
`ogy and Oncology (DGHO), the Austrian Stem Cell
`Transplant Working Group of the Austrian Society of
`Hematology and Oncology, the Swiss Blood Stem Cell
`Transplantation Group (SBST), and the German-
`Austrian Paediatric Working Group on SCT.
`The evaluation of evidence and the subsequent rec-
`ommendation was graded according to the system used
`in grading of supportive care published by Couriel [10].
`The evidence of the majority of treatment options in
`cGVHD is sparse, and, therefore, for most of the thera-
`peutic options the strength of recommendation falls
`into category C. In addition, category C and evidence
`III level were further specified as shown in Tables 1
`and 2. All recommendation and evidence levels were
`first rated by an expert panel and subsequently rated
`by all participants of the consensus process. Only
`evidence from the use in cGVHD was included in the
`evaluation.
`According to the number and severity of organs
`involved with cGVHD, the NIH consensus defined
`mild cGVHD as mild involvement of 2 organs only,
`excluding lung involvement, moderate cGVHD as
`mild involvement of more than 2 organs, or moderate
`organ involvement excluding moderate lung involve-
`ment, and severe cGVHD as any severe organ manifes-
`tation or moderate lung manifestations [8].
`Here, we discuss first-line and topical treatment op-
`tions for cGVHD. We mainly focus on reported clinical
`trials and retrospective analyses. The literature search
`was performed by the working group on first-line
`
`Table 1. Strength of Recommendation
`
`Strength of
`Recommendation
`Level
`
`A
`B
`C
`
`C-1*
`C-2*
`
`D
`
`Definition of Recommendation Level
`
`Should always be offered
`Should generally be offered
`Evidence for efficacy is insufficient to support
`for or against, or evidence might not outweigh
`adverse consequences, or cost of the
`approach. Optional
`Use in first-line treatment justified
`Use in equal to or greater than second-line
`treatment justified
`Moderate evidence for lack of efficacy or
`for adverse outcome supports
`a recommendation against use.
`Should generally not be offered.
`
`*Only applied for topical treatment of chronic graft-versus-host disease
`(cGVHD).
`
`treatment within the Consensus Conference using the
`Pubmed database. Only English literature was consid-
`ered. Abstracts from the Bone Marrow Transplantation
`Tandem meetings, the European Bone Marrow Trans-
`plantation meetings, and the American Society of He-
`matology meetings were cited, but were not included
`in the evidence rating.
`
`Principles of First-Line Treatment of cGVHD
`
`As the diagnosis of cGVHD has major conse-
`quences on the further clinical course of the patient,
`the diagnosis needs to be based on either diagnostic
`clinical signs of cGVHD or requires confirmation by
`histology as described by the NIH consensus as well
`as the consensus within the German/Austrian/Swiss
`Bone Marrow Transplantation Group [8]. Once diag-
`nosis of GVHD is established, the first step is to distin-
`guish classic cGVHD from overlap syndrome or late
`aGVHD. Especially in the latter situation as for overlap
`syndrome with dominating acute features, treatment
`should be applied according to standard practice in
`treatment of aGVHD (ie, treatment with steroids in
`
`Table 2. Quality of Evidence Supporting the Recommendation
`
`Strength of
`Evidence Level
`
`Definition of Evidence Level
`
`I
`II
`
`III
`
`III-1*
`
`III-2*
`
`III-3*
`
`Evidence from $1 properly randomized, controlled trial
`Evidence from >1 well-designed clinical trial without
`randomization, from cohort or case-controlled analytic
`studies (preferable from >1 center) or from multiple
`time series or dramatic results from uncontrolled
`experiments
`Evidence from opinions of respected authorities based on
`clinical experience, descriptive studies, or reports from
`expert committees
`Several reports from retrospective evaluations or small
`uncontrolled clinical trials
`Only 1 report from small uncontrolled clinical trial
`or retrospective evaluations
`Only case reports available
`
`*Only applied for topical treatment of chronic graft-versus-host disease
`(cGVHD).
`
`

`

`Biol Blood Marrow Transplant 16:1611-1628, 2010
`
`First-Line and Topical Treatment of Chronic GVHD
`
`1613
`
`combination with a CNI). Standards for treatment of
`overlap syndrome with evenly balanced symptoms still
`need to be defined. While aGVHD is defined by the
`presence of exclusive features of aGVHD, the diagnosis
`of overlap syndrome is based on the simultaneous pres-
`ence of symptoms of aGVHD and distinctive or diag-
`nostic features of cGVHD [8]. The diagnosis of
`classic cGVHD requires the presence of either diagnos-
`tic or distinctive symptoms of cGVHD in the absence
`of features of aGVHD [8].
`The Consensus Conference on Clinical Practice of
`Chronic GVHD focused on treatment of classic
`cGVHD. Prognostic features at diagnosis of cGVHD
`have been described. The presence of thrombocytope-
`nia or direct progression from aGVHD have been
`associated with adverse outcome [1]. The value of
`“progressive onset” as a risk factor is limited by the
`fact that traditionally any GVHD being present at
`day 100 was documented as cGVHD. However, 2 stud-
`ies reclassifying GVHD according to the NIH criteria
`revealed a significant proportion of patients being tra-
`ditionally classified as cGVHD instead of late aGVHD
`[11,12]. The risk factor “thrombocytopenia” has been
`identified in cohorts receiving a myeloablative (MA)
`conditioning regimen and mainly bone marrow (BM)
`as a graft source [1]. Therefore, it remains to be shown
`whether low platelets remain as a risk factor in patients
`receiving nonmyeloablative regimens and PBSCs as
`a graft source. Additional risk factors are extensive
`skin disease (.50% body surface) as well as severe
`cGVHD (NIH grading) [1]. A detailed classification
`of cGVHD severity according to the NIH consensus
`is delineated in Table 3 [8].
`As for treatment, prognosis of overlap syndrome is
`a matter of debate as well; Jagasia and colleagues [13]
`reported a significantly worse survival of patients
`with any features of aGVHD after day 100 of HSCT
`compared with cGVHD. This is in contrast to reports
`by Arora et al. [11] and Cho et al. [12], stating no sig-
`nificant survival difference in patients with overlap
`syndrome compared to classic cGVHD and to a retro-
`spective analysis published by Vigorito et al. [14], dem-
`onstrating no significant survival differences between
`patients with late aGVHD and cGVHD.
`Currently, no uniformly accepted definition of ste-
`roid refractory cGVHD is available. Generally, ac-
`cepted criteria for steroid refractory cGVHD are (1)
`progression despite immunosuppressive treatment us-
`ing 1 mg/kg/day of prednisone for 2 weeks, (2) stable
`disease if 4 to 8 weeks on $0.5 mg/kg/day of predni-
`sone, and (3) inability to taper below 0.5 mg/kg/day
`of prednisone. Treatment duration may vary depend-
`ing on clinical manifestation (sclerosis requires longer
`to respond) or toxicity (shorter duration in the presence
`of significant toxicity) [9,15]. In the presence of primary
`treatment failure, alternative treatment options need to
`be started.
`
`Table 3. Severity Grading of cGVHD
`
`Severity
`
`Mild
`
`Moderate
`
`Severe
`
`Number of involved
`organs
`Severity of organ
`manifestations
`
`1-2
`
`$3
`
`$3
`
`1 (excluding lung)
`
`2 (or lung 1)
`
`3 (or lung 2)
`
`Treatment of Mild cGVHD
`
`During the Consensus Conference on Clinical
`Practice of Chronic GVHD, an agreement was
`achieved that mild cGVHD may be treated either
`with topical immunosuppressive agents or with sys-
`temic steroids alone. In the scenario of solely topical
`immunosuppression, a close follow-up and screening
`for any potential manifestation of cGVHD is crucial
`to detect systemic progression of cGVHD during top-
`ical treatment. An additional factor influencing the de-
`cision of treatment of choice is the risk for relapse of
`the underlying malignancy, supporting topical treat-
`ment in the presence of a high relapse risk.
`In pediatric patients, 2 additional considerations
`have to be taken into account. Side effects of systemic
`steroid therapy can be deleterious on a growing child.
`On the other hand, patients with nonmalignant under-
`lying diseases have no benefit from the cGVHD-
`associated graft-versus-leukemia (GVL) effect even
`in mild disease courses. Therefore, topical therapy
`should be offered as often and as early as possible.
`Mild manifestations of cGVHD that cannot be suf-
`ficiently treated by topical treatment such as hepatic
`manifestations or fasciitis may be treated with systemic
`corticosteroids alone. Again, lower initial doses than 1
`mg/kg/day of prednisone may be used, but evidence for
`or against a reduced dose of steroids is virtually absent.
`In the presence of a high risk of relapse, an approach
`using supportive treatment with either nonsteroidal
`anti-inflammatory drugs (involvement of fascia or
`joints) or ursodeoxycholic acid (hepatic disease) may
`be suitable treatment options as long as a close
`follow-up to detect progression is guaranteed.
`Because treatment is rather symptomatic and does
`not aim to control a systemic process, topical treat-
`ment should be continued as long as symptoms are
`present and may be tapered and withdrawn in the pres-
`ence of remission of symptoms. The same applies for
`systemic treatment, although treatment for at least 4
`to 8 weeks should be given to avoid frequent relapses
`of symptoms of cGVHD.
`
`Treatment of Moderate cGVHD: Role of
`Prednisone (A I)
`
`Treatment of moderate cGVHD requires systemic
`immunosuppression. Additional topical treatment may
`be applied to speed up the response or to improve local
`response rates, but it does not replace the requirement
`
`

`

`1614
`
`D. Wolff et al.
`
`Biol Blood Marrow Transplant 16:1611-1628, 2010
`
`Table 4. First-Line Treatment Options in cGVHD
`
`Recommendation
`(citation number
`of references)
`
`Agent
`
`Steroids
`
`A [16-18]
`
`CNI
`
`C [16,17]
`
`MMF in triple agent
`combinations
`Azathioprine
`
`D [19]
`
`D [18]
`
`Thalidomide
`
`D [20,21]
`
`Evidence
`
`Side Effects
`
`Comments
`
`I
`
`II
`
`II
`
`II
`
`II
`
`Osteoporosis, avascular necrosis
`of the bone, diabetes
`
`Renal toxicity, hypertension
`
`GI complaints, infectious and relapse risk
`
`Hematologic toxicity, infectious risk
`
`Neurotoxicity, sedation, constipation,
`thrombosis
`
`Important but need to spare steroids because
`of side effect profile, generally sufficient in primary
`treatment of mild cGVHD as single agent, may be
`used in combination with CNI in moderate
`or severe cGVHD
`Only be used in combination with steroids, spares steroids,
`lower rate of avascular necrosis of the bone, may be
`considered in combination with steroids in primary
`treatment of severe cGVHD as well as in CNI
`dependent moderate cGVHD
`Failure to improve efficacy in a randomized trial
`documented
`Adverse outcome in a randomized trial in combination
`with steroids
`May be used in concomitant relapse of multiple myeloma
`
`CNI indicates calcineurin inhibitors; MMF, mycophenolate mofetil; GI, gastrointestinal; cGVHD, chronic graft-versus-host disease.
`
`for systemic immunosuppression. Standard treatment
`is 1 mg/kg/day of prednisone or an equivalent dose of
`methylprednisolone. So far, no other treatment option
`replacing steroids in first-line treatment has been eval-
`uated, resulting in a grade A recommendation with an
`evidence grade of I, although a steroid-free approach
`has never been applied [16-18]. Steroid dependence
`of the majority of patients failing first-line treatment
`
`indicates the central role of steroids in treatment of
`cGVHD (Tables 4).
`A first report in the early 1980s indicated that
`prednisone alone or in combination with other immu-
`nosuppressive agents (particularly azathioprine and
`cyclophosphamide) could improve the outcome of pa-
`tients who required treatment for extensive cGVHD
`[66]. A randomized double-blinded study comparing
`
`Table 5. Topical Immunosuppressive Treatment Options in cGVHD
`
`Organ
`
`Agent (citation number
`of references)
`
`Skin
`
`Topical steroids [7]
`
`Tacrolimus/Pimecrolimus [22-25]
`
`PUVA [26-31]
`
`UVA [32-35]
`
`UVB [36]
`
`GI
`Lung
`
`Topical steroids [37-39]
`Topical steroids
`
`Oral
`
`Topical steroids [40-44]
`
`Topical tacrolimus /cyclosporine [45-51]
`
`Topical PUVA/UVB [44,52-54]
`
`Eye
`
`Topical steroids [55,56]
`
`Topical cyclosporine [57-60]
`
`Vaginal Topical steroids [61-64]
`
`Topical tacrolimus/
`cyclosporine/pimecrolimus [63-65]
`
`cGVHD indicates chronic graft-versus-host disease.
`
`Recommendation
`
`Evidence
`
`Side Effects
`
`Comments
`
`C-1
`
`C-1
`
`C-1
`
`C-1
`
`C-1
`
`C-1
`B
`
`C-1
`
`C-2
`
`C-2
`
`C-1
`
`C-1
`
`B
`
`B
`
`III-1
`
`Skin atrophy
`
`III-1
`
`III-1
`
`III-1
`
`III-2
`
`III-1
`III-2
`
`III-1-III-3
`
`III-1
`
`III-1
`
`III-1
`
`III-1
`
`III-3
`
`III-3
`
`Long-term risk for
`cutaneous malignancies
`Phototoxicity, risk for
`cutaneous malignancies
`Phototoxicity, risk for
`cutaneous malignancies
`
`Phototoxicity, risk for
`cutaneous malignancies
`
`Best results with topical
`budesonide
`Burning
`
`Optional treatment option for
`refractory manifestations
`Risk for corneal thinning and
`infectious keratitis
`Burning, stinging
`
`Increased risk for infectious
`complications and atrophy
`Burning
`
`Neck down: mid-strength steroids if no
`response upper strength steroids,
`face: hydrocortisone 1%
`Should be given twice daily
`
`Should not be used with phototoxic
`medication
`Requires no UV protection after
`treatment, should not be used with
`phototoxic medication
`Only effective in lichenoid cGVHD
`
`Either budesonide or beclomethasone
`May be combined with betamimetic
`agents
`Requires oral hygiene and possibly
`topical antifungals
`Potentially increases risk for oral
`malignancies, may be combined
`with topical steroids
`Psoralene may be given topically
`or systemically
`Duration of exposure should be limited
`
`Fewer long-term side effects
`compared to steroids, high
`long-term efficacy
`Topical estrogen application and
`antifungal prophylaxis suggested
`Less well tolerated but better
`long-term efficacy
`
`

`

`Biol Blood Marrow Transplant 16:1611-1628, 2010
`
`First-Line and Topical Treatment of Chronic GVHD
`
`1615
`
`prednisone and placebo versus prednisone and azathi-
`oprine in patients with platelet counts .100,000/mL
`showed better outcome with prednisone alone, and
`thus established prednisone as the treatment of choice
`for patients with standard-risk extensive cGVHD [17].
`The central role of prednisone was further confirmed
`by a randomized trial comparing prednisone alone ver-
`sus prednisone and cyclosporine (CsA) in patients with
`extensive cGVHD and platelet counts .100,000/mL
`showing no difference in overall survival (OS) in the
`2 arms and no better control of cGVHD [16].
`Starting in the 1980s, the standard initial steroid
`dose for the treatment of cGVHD has been 1 mg/kg/
`day, regardless of whether prednisone was used alone
`or in combination with other drugs [17,18,66]. There
`are no randomized studies comparing this dose with
`higher or lower initial doses. Recent retrospective analy-
`ses of patients with aGVHD indicate that 1 mg/kg/day
`could be at least as effective as 2 mg/kg/day for
`patients with grades I-1I aGVHD [67]. Considering
`the need for protracted treatment of cGVHD, it may
`be worthwhile exploring lower doses of steroids. Pend-
`ing such studies 1 mg/kg daily is considered the standard
`initial dose.
`The Seattle group suggests to maintain this dose for
`2 weeks and then to taper to 1 mg/kg every other day
`over a period of 6 to 8 weeks if symptoms are stable or
`improving, and then either maintain this dose for 2 to
`3 months or continue straight on to taper by 10% to
`20% per month [68]. The survey sent to all centers
`participating in the Consensus Conference revealed
`that 26 of 31 centers (84%) start to reduce the steroid
`dose after 2 weeks of treatment if symptoms are inactive.
`G. Vogelsang [15], from the Johns Hopkins group,
`reported that 90% of responding patients would have
`done so within 3 months after achieving the alternate-
`day dose; thus, a reevaluation of patients at this stage
`should guide further tapering. Patients with complete
`responses (CRs) should be further tapered 10% to
`20% monthly, whereas those still responding should
`stay on 1 mg/kg for about another 3 months after
`achieving maximum response and then slowly be ta-
`pered as described. If symptoms flare during tapering,
`increasing the steroid dose may again induce response.
`Patients who by the 3-month reevaluation have not re-
`sponded should be considered for alternative treatment
`strategies [15,68].
`Since the early 1960s, alternate dosing of steroids
`has been considered an effective regimen for the treat-
`ment of many immune-mediated disorders. The dose-
`spacing is thought to maintain efficacy while reducing
`toxicity of the applied steroids [69]. However, there are
`no randomized studies comparing daily and alternate-
`day strategies
`in cGVHD.
`In kidney transplant
`patients, the alternate-day dosing reduces the level of
`plasma lipids [70]. Likewise, administration of steroids
`as a single dose in the morning instead of a split dose
`
`is meant to match the circadian cycle and reduce side
`effects. Randomized studies in children treated with
`prednisolone for nephrotic syndrome and adults
`treated for proctocolitis showed similar efficacy of sin-
`gle compared to split-dose strategies [71,72]. Whether
`this holds true for patients being treated for cGVHD
`still has to be demonstrated in prospective studies.
`There are no studies comparing the effects and side
`effects of prednisone to other
`systemic steroid
`preparations such as methylprednisolone in patients
`with cGvHD.
`
`Role of CNIs (C II)
`
`Although the role of steroids in first-line treatment
`is well established, the role of CNIs is less clear. The
`potential benefit of the CNIs CsA and tacrolimus
`(FK506)
`in the treatment of cGVHD has been
`addressed in a small number of studies [16,17]. In
`a nonrandomized trial conducted more than 20 years
`ago, Sullivan et al. [17] added an alternating-day sched-
`ule of CsA (6 mg/kg twice a day) to a previously estab-
`lished alternating-day regimen of 1 mg/kg prednisone
`to treat 40 high-risk patients with newly diagnosed mul-
`tiorgan cGVHD and thrombocytopenia\100,000/mL.
`After 9 months a CR rate of 33% and 4-year survival of
`51% were reported. These results compared favorably
`to a 16% CR and 26% survival rate of a cohort of 38 pa-
`tients with cGVHD and thrombocytopenia treated by
`the same center in a similar period of time. This study
`constituted the basis for the inclusion of CNIs to the
`therapeutic regimens for cGVHD in clinical practice
`for many years [15,68,73].
`However, this practice was challenged by a random-
`ized trial, in which Koc and colleagues [16] compared
`CsA alternating with prednisone every other day to
`alternate-day prednisone alone in 287 patients with
`newly diagnosed cGVHD and a platelet count
`.100,000/mL [16]. The primary endpoint of this study
`was
`the incidence of
`treatment-related mortality
`(TRM) at 3 years, which was not different between
`both groups. There was also no significant difference
`with respect to the incidence of secondary therapy,
`the discontinuation of immunosuppression, the inci-
`dence of recurrent malignancy, or OS in this study. In
`contrast, patients treated with CsA and prednisone
`showed a significantly inferior survival without recur-
`rent malignancy (progression-free survival [PFS]) com-
`pared to patients treated with prednisone alone. In
`addition, a small subset of high-risk patients with
`progressive onset cGVHD displayed a tendency for
`increased TRM and inferior survival at 5 years in the
`CsA plus prednisone (16 patients) versus the prednisone
`alone (29 patients) arm. A significantly decreased rate of
`avascular necroses in patients treated in the combina-
`tion arm was observed, suggesting that the addition of
`CsA to the therapeutic regimen resulted in lower
`
`

`

`1616
`
`D. Wolff et al.
`
`Biol Blood Marrow Transplant 16:1611-1628, 2010
`
`cumulative doses of prednisone and a decreased inci-
`dence of complications related to steroid treatment.
`Although the study published by Koc et al. provided im-
`portant information, it needs to be mentioned that all
`patients in the study received BM grafts following MA
`conditioning. Because it is known that the use of non-
`myeloablative conditioning regimens influences the
`presentation of cGVHD and that PBSC grafts result
`in prolonged need for treatment of cGVHD when com-
`pared to BM, at this point it remains unclear whether
`the combination of steroids with a CNI is of advantage
`in the latter population [74,75]. Moreover, within the
`published trials, CNIs have been dosed with an every
`other day schedule without dose correction according
`to plasma levels, which differs from current practice
`with daily application and dose adjustments according
`to plasma levels.
`Limited published experience is available with
`respect to the role of FK506 for first-line treatment
`of cGVHD.
`In summary, evidence for the use of CNIs for treat-
`ing newly diagnosed cGVHD patients is scarce and,
`based on the randomized trial by Koc et al. [16], cannot
`generally be recommended. In standard-risk patients
`(de novo or quiescent cGVHD, platelet counts
`.100.000/mL), the use of CsA in addition to steroids
`may be considered for those patients who are at high
`risk for glucocorticoid-usage-related complications
`(eg, based on age, sex, and/or the presence of comor-
`bidities). Even though a direct comparison of CsA
`and FK506 in the treatment of cGVHD is lacking, clin-
`ical and indirect evidence suggests that both CNIs
`may be equally effective in this setting [19,76]. Serum
`concentrations of CsA or FK506, as well as creatinine
`and other clinical or laboratory signs of adverse
`events, should be monitored regularly because of
`associated drug toxicity. As there are no data showing
`a superiority of the alternating-day schedule, daily
`administration of CsA and FK506 is generally em-
`ployed in clinical practice [16,17]. Caution should be
`formulations of
`advised, however, that some oral
`CsA (eg, Sandimmun optoralÒ) provide an improved
`bioavailability of the drug, and therefore, may result
`in increased serum concentrations. FK506 clearance
`is age dependent in pediatric patients, and especially
`children younger than 6 years of age have a higher
`clearance [77].
`The role of CNIs in high-risk cGVHD (ie, pro-
`gressive onset and/or with thrombocytopenia [platelet
`count \100,000/mL]) is unclear. The initially promis-
`ing historic data of Sullivan et al. [17] on patients
`including
`with cGVHD and thrombocytopenia,
`a more recent update [78], cannot be readily transferred
`to current patient cohorts, considering the change in
`transplantation practice in the past 20 years. Therefore,
`more randomized studies are clearly warranted to
`clarify these issues.
`
`Treatment of Severe cGVHD
`
`In general, treatment of severe cGVHD follows the
`same rules as treatment of moderate cGVHD. Severe
`cGVHD has been associated with an increased mortal-
`ity and may require prolonged immunosuppression
`[79]. Therefore, the combination of steroids with
`a CNI may be of potential advantage in severe cGVHD
`to spare steroids [16]. Furthermore, patients with
`severe de novo or quiescent onset of cGVHD after
`withdrawal of the CNI may be CNI dependent and
`may benefit from treatment with a combination of ste-
`roids with a CNI, but data evaluating this approach are
`virtually absent. Although the randomized trials on
`triple-agent first-line treatment of cGVHD did not
`differentiate the response rate according to the extent
`of organ manifestations, there is no indication that
`a triple-agent approach leads to an improved response
`rate and outcome in severe cGVHD [19-21].
`Whether a 2-agent combination consisting of
`prednisone and an additional non-CNI agent such as
`mycophenolate mofetil (MMF), mTOR inhibitors, or
`extracorporeal photopheresis (ECP) improves the out-
`come has not been evaluated yet. Because the combina-
`tion of prednisone and azathioprine resulted in an
`improved response rate but significantly worse survival,
`any alternative combination urgently requires its eval-
`uation in a clinical trial [18]. As ECP has not been asso-
`ciated with an increased risk for infectious morbidity
`and mortality, it may be a promising candidate to be
`evaluated as a new combination partner with cortico-
`steroids and may be a treatment option in first-line
`treatment, when contraindication for CNIs exist and
`high steroid-related toxicity must be expected [80,81].
`
`Treatment of Progressive Onset of cGVHD
`
`Progressive onset of cGVHD is characterized by
`direct progression of active symptoms of aGVHD into
`symptoms of cGVHD during treatment of aGVHD
`and has been historically associated with a dismal out-
`come [1]. The majority of patients with progressive onset
`of cGVHD are on a CNI and steroids during onset of
`symptoms of cGVHD. Therefore, other treatment
`strategies need to be considered. Most of the 31 trans-
`plant centers responding to the survey on treatment of
`cGVHD reported, that when progressive onset cGVHD
`evolved during the taper of steroids and despite being on
`a CNI, treatment involved a temporary increase of the
`steroid dose and the addition of a new agent such as
`MMF or ECP to the CNI. The reported increased
`morbidity and mortality because of infectious complica-
`tions when using the triple-agent combination of pred-
`nisone, a CNI, and MMF indicate the limitation of
`this approach [19]. Because of the low risk for infectious
`complications, ECP therefore may be a promising can-
`didate for clinical trial evaluation treating cGVHD de-
`veloping during treatment of aGVHD [80,81]. Other
`
`

`

`Biol Blood Marrow Transplant 16:1611-1628, 2010
`
`First-Line and Topical Treatment of Chronic GVHD
`
`1617
`
`potential alternatives may be the replacement of the CNI
`by a mTOR inhibitor.
`
`Azathioprine (DII)
`
`Although responses of cGVHD to azathioprine
`have been documented, a double-blinded randomized
`trial comparing prednisone 1 placebo versus predni-
`sone 1 azathioprine in standard risk cGVHD (platelets
`.100 000/mL) revealed a significantly increased nonre-
`lapse mortality (40% versus 21%) and a significantly
`decreased OS (47% versus 61%) [18]. Therefore, aza-
`thioprine should not be used as first-line therapy in
`combination with steroids in patients with cGVHD.
`
`Thalidomide (DII)
`
`Although thalidomide displays therapeutic activity
`in second-line treatment of cGVHD, 2 randomized tri-
`als evaluating the efficacy of thalidomide in first-line
`treatment of cGVHD failed to show any advantage for
`the addition of thalidomide to standard immunosup-
`pression with CsA and prednisone [20,21]. Arora et al.
`[20] published the results of a randomized trial compar-
`ing standard treatment with CsA and predniso