High-dose chemotherapy followed by autologous stem cell transplantation for patients with relapsed/refractory Hodgkin lymphoma

  • Review
  • Intervention

Authors

  • Michaela Rancea,

    Corresponding author
    1. University Hospital of Cologne, Cochrane Haematological Malignancies Group, Department I of Internal Medicine, Cologne, Germany
    • Michaela Rancea, Cochrane Haematological Malignancies Group, Department I of Internal Medicine, University Hospital of Cologne, Kerpener Str. 62, Cologne, 50924, Germany. michaela.rancea@uk-koeln.de.

    Search for more papers by this author
  • Ina Monsef,

    1. University Hospital of Cologne, Cochrane Haematological Malignancies Group, Department I of Internal Medicine, Cologne, Germany
    Search for more papers by this author
  • Bastian von Tresckow,

    1. University Hospital of Cologne, Department I of Internal Medicine, Cologne, Germany
    Search for more papers by this author
  • Andreas Engert,

    1. University Hospital of Cologne, Cochrane Haematological Malignancies Group, Department I of Internal Medicine, Cologne, Germany
    Search for more papers by this author
  • Nicole Skoetz

    1. University Hospital of Cologne, Cochrane Haematological Malignancies Group, Department I of Internal Medicine, Cologne, Germany
    Search for more papers by this author

Abstract

Background

Hodgkin lymphoma (HL) is one of the most common malignancies in young adults and has become curable for the majority of patients, even in advanced stage. After first-line therapy, 15% to 20% do not respond to treatment and relapse. For those patients, high-dose chemotherapy (HDCT) followed by autologous stem cell transplantation (ASCT) is a frequently used therapy option.

Objectives

To find the best available treatment with HDCT followed by ASCT for patients with relapsed or refractory HL after first-line treatment.

Search methods

We searched the Central Register of Controlled Trials (CENTRAL), MEDLINE, and relevant conference proceedings up to January 2013 for randomised controlled trials (RCTs). We also contacted experts for unpublished data.

Selection criteria

We included RCTs comparing HDCT followed by ASCT versus conventional chemotherapy without ASCT, or versus additional sequential HDCT (SHDCT) followed by ASCT. We also included RCTs with different HDCT regimens before ASCT in patients with relapsed or primary refractory HL after any first-line therapy.

Data collection and analysis

Two review authors (MR, NS) independently selected relevant studies, extracted data and assessed trial quality. We used hazard ratios (HR) for overall survival (OS) and progression-free survival (PFS), and we calculated risk ratios (RR) for the other outcomes. We presented all measures with 95% confidence intervals (CI).

We assessed the quality of evidence using GRADE methods.

Main results

Our search resulted in 1663 potentially relevant references, of which we included three trials with 14 publications, assessing 398 patients. Overall, we judged the quality of the trials as moderate. The trials were all reported as randomised controlled and open-label. We included two RCTs assessing the effect of HDCT followed by ASCT compared with conventional chemotherapy in a meta-analysis. The number of studies was very low, therefore, the quantification of heterogeneity was not reliable. We included one further RCT (one assessing additional SHDCT followed by ASCT versus HDCT followed by ASCT), which was not compatible with our meta-analysis. For this trial, we performed further analyses.

Two trials showed a non-statistically significant trend that HDCT followed by ASCT compared to conventional chemotherapy increases OS (HR 0.67; 95% CI 0.41 to 1.07; P value = 0.10, 157 patients, moderate quality of evidence). However, the increase in PFS was statistically significant for people treated with HDCT followed by ASCT (HR 0.55; 95% CI 0.35 to 0.86; P value = 0.009, 157 patients, moderate quality of evidence). Adverse events were reported in one trial only and did not differ statistically significant between the treatment arms. We were not able to draw conclusions regarding treatment-related mortality (TRM) because of insufficient evidence (RR 0.61; 95% CI 0.16 to 2.22; P value = 0.45, 157 patients, moderate quality of evidence).

For the second comparison, SHDCT plus HDCT followed by ASCT versus HDCT followed by ASCT there was no difference between the treatment arms regarding OS (HR 0.93; 95% CI 0.5 to 1.74; P value = 0.816, three-year OS: 80% SHDCT versus 87% HDCT, 241 patients), or PFS (HR 0.87; 95% CI 0.58 to 1.30; P value = 0.505, 241 patients). Seven patients died in the SHDCT arm and one in the HDCT arm due to increased toxicity of the treatment. Adverse events were increased with SHDCT plus HDCT followed by ASCT after two cycles of dexamethasone plus high-dose cytarabine plus cisplatin (DHAP) (88% SHDCT versus 45% HDCT, 223 patients, P value < 0.00001). Overall, more statistically significant World Health Organization (WHO) grade 3/4 infections occurred with SHDCT (48% SHDCT versus 33% HDCT; P value = 0.002, 223 patients).

Authors' conclusions

The currently available evidence suggests a PFS benefit for patients with relapsed or refractory HL after first-line therapy, who are treated with HDCT followed by ASCT compared to patients treated with conventional chemotherapy. In addition, data showes a positive trend regarding OS, but more trials are needed to detect a significant effect.

Intensifying the HDCT regime before HDCT followed by ASCT did not show a difference as compared to HDCT followed by ASCT, but was associated with increased adverse events.

Résumé scientifique

Chimiothérapie à haute dose suivie d'une greffe de cellules souches autologues pour les patients atteints d'un lymphome de Hodgkin récidivant/réfractaire

Contexte

Le lymphome de Hodgkin (LH) est l'une des tumeurs malignes les plus courantes chez les jeunes adultes et peut désormais se soigner pour la majorité des patients, même à un stade avancé. Après un traitement de première intention, 15 % à 20 % des patients ne réagissent pas au traitement et rechutent. Pour ces patients, une chimiothérapie à haute dose (CTHD), suivie d'une greffe de cellules souches autologues (GCSA), est une option de traitement fréquemment utilisée.

Objectifs

Trouver le meilleur traitement disponible avec une CTHD suivie d'une GCSA pour les patients atteints d'un LH récidivant ou réfractaire après un traitement de première intention.

Stratégie de recherche documentaire

Nous avons effectué des recherches dans le registre Cochrane des essais contrôlés (CENTRAL), MEDLINE, ainsi que dans les actes de conférences pertinents jusqu'à janvier 2013 afin de trouver des essais contrôlés randomisés (ECR). Nous avons également contacté des experts pour obtenir des données non publiées.

Critères de sélection

Nous avons inclus les ECR comparant la CTHD suivie d'une GCSA versus chimiothérapie conventionnelle sans GCSA ou versus CTHD séquentielle (CTHDS) supplémentaire suivie d'une GCSA. Nous avons également inclus les ECR avec différents schémas de CTHD avant GCSA chez des patients atteints d'un LH récidivant ou d'un LH primaire réfractaire après un traitement de première intention.

Recueil et analyse des données

Deux auteurs de la revue (MR, NS) ont sélectionné les études pertinentes, extrait les données et évalué la qualité des essais de manière indépendante. Nous avons utilisé les hazard ratios (HR) pour la survie globale (SG) et la survie sans progression (SSP) et avons calculé les risques relatifs pour les autres critères de jugement. Nous avons présenté toutes les mesures avec des intervalles de confiance (IC) à 95 %.

Nous avons évalué la qualité des preuves en utilisant les méthodes GRADE.

Résultats principaux

Nos recherches ont fourni 1 663 références potentiellement pertinentes, parmi lesquelles nous avons inclus trois essais, avec 14 publications, évaluant 398 patients. Dans l'ensemble, nous avons estimé que la qualité des essais était modérée. Les essais ont tous été présentés comme des essais contrôlés randomisés et ouverts. Nous avons inclus deux ECR évaluant l'effet de la CTHD suivie d'une GCSA comparé à une chimiothérapie conventionnelle dans une méta-analyse. Le nombre d'études était très faible, par conséquent la quantification de l'hétérogénéité n'était pas fiable. Nous avons inclus un ECR supplémentaire (un essai évaluant une CTHDS suivie d'une GCSA versus CTHD suivie d'une GCSA) qui n'était pas compatible avec notre méta-analyse. Pour cet essai, nous avons procédé à d'autres analyses.

Deux essais ont montré une tendance statistiquement non-significative selon laquelle la CTHD suivie d'une GCSA comparé à une chimiothérapie conventionnelle augmentait la SG (HR 0,67 ; IC à 95 % 0,41 à 1,07 ; valeur P = 0,10, 157 patients, qualité des preuves modérée). Cependant, l'augmentation de la SSP a été statistiquement significative pour les personnes traitées par une CTHD suivie d'une GCSA (HR 0,55 ; IC à 95 % 0,35 à 0,86 ; valeur P = 0,009, 157 patients, qualité des preuves modérée). Les événements indésirables n'ont été rapportés que dans un seul essai et n'ont pas montré de différence statistiquement significative entre les bras de traitement. Nous n'avons pas pu établir de conclusions concernant la mortalité liée au traitement (MLT) en raison de l'insuffisance de preuves (RR 0,61 ; IC à 95 % 0,16 à 2,22 ; valeur P = 0,45, 157 patients, qualité des preuves modérée).

Pour la deuxième comparaison, la CTHDS plus CTHD suivie d'une GCSA versus CTHD suivie d'une GCSA, il n'y a eu aucune différence entre les bras de traitement concernant la SG (HR 0,93 ; IC à 95 % 0,5 à 1,74 ; valeur P = 0,816, SG à trois ans : 80 % avec CTHDS versus 87 % avec HDCT, 241 patients) ou la SSP (HR 0,87 ; IC à 95 % 0,58 à 1,30 ; valeur P = 0,505, 241 patients). Sept patients sont décédés dans le bras de CTHDS et un patient dans le bras de CTHD en raison d'une toxicité accrue du traitement. Les événements indésirables ont été accrus avec la CTHDS plus CTHD suivie d'une GCSA après deux cycle de dexaméthasone plus cytarabine à haute dose plus cisplatine (DHAP) (88 % avec CTHDS versus 45 % avec CTHD, 223 patients, valeur P < 0,00001). Globalement, un plus grand nombre d'infections statistiquement significatives de classe 3/4 selon l'Organisation mondiale de la Santé (OMS) sont survenues avec la CTHDS (48 % avec SHDCT versus 33 % avec HDCT ; valeur P = 0,002, 223 patients).

Conclusions des auteurs

Les preuves disponibles actuellement suggèrent un bénéfice concernant la SSP pour les patients atteints d'un LH récidivant ou réfractaire après un traitement de première intention qui sont traités avec une CTHD suivie d'une GCSA comparé aux patients traités avec une chimiothérapie conventionnelle. De plus, les données montrent une tendance positive concernant la SG, mais davantage d'essais doivent être réalisés pour détecter un effet significatif.

L'intensification du schéma de traitement par CTHD avant CTHD suivie d'une GCSA n'a pas montré de différence comparé à la CTHD suivie d'une GCSA, mais a été associée à une augmentation des événements indésirables.

Plain language summary

High-dose chemotherapy followed by autologous stem cell transplantation for patients with relapsed/refractory Hodgkin lymphoma

Hodgkin lymphoma (HL) is a malignancy of single lymph nodes, the lymphatic system, and might affect other additional organs. It is a relatively rare disease, accounting for two or three people per 100,000 every year in Western countries, but it is one of the most common cancers in young adults between 20 and 30 years of age. The second peak of the disease is after the age of 60 years. Treatment options for HL have improved since the 1980s, so that even patients in advanced stages may be cured with adequate therapy. Treatment approaches include chemotherapy, radiotherapy or chemotherapy combined with radiotherapy (combined-modality treatment), of which the combined-modality treatment is standard for most patients nowadays. Nevertheless, 15% to 20% of patients do not reach complete remission and have refractory disease or relapse. For these patients high-dose chemotherapy (HDCT) followed by autologous stem cell transplantation (ASCT) has become the optimal treatment option. However, the impact of this regimen on overall survival is still unclear. Therefore, we conducted a Cochrane Review on efficacy and safety of HDCT followed by ASCT in patients with primary refractory or relapsed HL. We searched several important medical databases (the Cochrane Central Register of Controlled Trials and MEDLINE) and summarised and analysed evidence from randomised controlled trials (RCTs). We identified three RCTs corresponding to our pre-defined inclusion criteria treating 398 patients. We included two trials that compared HDCT followed by ASCT versus conventional chemotherapy alone, and one trial evaluating additional sequential HDCT (SHDCT) followed by ASCT against HDCT followed by ASCT.

Both trials assessing HDCT followed by ASCT versus conventional chemotherapy showed no significant improvement in overall survival, however progression-free survival was significantly improved with HDCT followed by ASCT. Only one trial reported adverse events and showed no difference between the treatment arms. The other trial was prematurely closed as patients refused randomisation and requested ASCT.

Only one trial evaluated the effect of SHDCT before HDCT plus ASCT, compared to HDCT plus ASCT. Overall survival and progression-free survival were similar in both arms. However, after three years, there was a negative trend for the SHDCT arm regarding mortality as well as significantly increased adverse events.

In summary, the currently available evidence suggests a benefit for patients with relapsed HL treated with HDCT followed by ASCT compared to conventional chemotherapy.

Résumé simplifié

Chimiothérapie à haute dose suivie d'une greffe de cellules souches autologues pour les patients atteints d'un lymphome de Hodgkin récidivant/réfractaire

Le lymphome de Hodgkin (LH) est une tumeur maligne des ganglions lymphatiques (du système lymphatique) et peut affecter d'autres organes supplémentaires. Il s'agit d'une maladie relativement rare, touchant deux ou trois personnes sur 100 000 chaque année dans les pays occidentaux, mais c'est l'un des cancers les plus courants chez les jeunes adultes entre 20 et 30 ans. Le second pic de la maladie se situe après 60 ans. Les options de traitement du LH se sont améliorées depuis les années 1980, de sorte que même les patients aux stades avancés peuvent être soignés par une thérapie adéquate. Les approches de traitement comprennent la chimiothérapie, la radiothérapie ou la chimiothérapie associée à la radiothérapie (traitement à modalités combinées), parmi lesquelles le traitement à modalités combinées est aujourd'hui la référence pour la plupart des patients. Néanmoins, 15 % à 20 % des patients ne parviennent pas à une rémission complète et sont atteints d'une maladie réfractaire ou victimes d'une rechute. Pour ces patients, une chimiothérapie à haute dose (CTHD), suivie d'une greffe de cellules souches autologues (GCSA), est devenue l'option de traitement optimale. Cependant, l'impact de ce schéma de traitement sur la survie globale reste incertain. Par conséquent, nous avons réalisé une revue Cochrane sur l'efficacité et la sécurité de la CTHD suivie d'une GCSA chez les patients atteints d'un LH primaire réfractaire ou récidivant. Nous avons effectué des recherches dans plusieurs bases de données médicales importantes (le registre Cochrane des essais contrôlés et MEDLINE) et avons résumé et analysé les preuves issues d'essais contrôlés randomisés (ECR). Nous avons identifié trois ECR portant sur 398 patients et correspondant à nos critères d'inclusion prédéfinis. Nous avons inclus deux essais qui comparaient la CTHD suivie d'une GCSA versus chimiothérapie conventionnelle seule et un essai évaluant la CTHD séquentielle (CTHDS) supplémentaire suivie d'une GCSA par rapport à une CTHD suivie d'une GCSA.

Les deux essais évaluant la CTHD suivie d'une GCSA versus chimiothérapie conventionnelle n'ont montré aucune amélioration significative en termes de survie globale, cependant la survie sans progression s'est améliorée significativement avec la CTHD suivie d'une GCSA. Un seul essai a rapporté des événements indésirables et il n'a montré aucune différence entre les bras de traitement. L'autre essai a été arrêté prématurément, car les patients ont refusé la randomisation et ont demandé une GCSA.

Un seul essai a évalué l'effet de la CTHDS avant CTHD plus GCSA comparé à la CTHD plus GCSA. La survie globale et la survie sans progression ont été semblables dans les deux bras. Cependant, après trois ans, on a observé une tendance négative pour le bras de la CTHDS concernant la mortalité, ainsi qu'une augmentation significative des événements indésirables.

En résumé, les preuves actuellement disponibles suggèrent un bénéfice pour les patients atteints d'un LH récidivant traités par une CTHD suivie d'une GCSA comparé à une chimiothérapie conventionnelle.

Notes de traduction

Traduit par: French Cochrane Centre 7th August, 2013
Traduction financée par: Pour la France : Minist�re de la Sant�. Pour le Canada : Instituts de recherche en sant� du Canada, minist�re de la Sant� du Qu�bec, Fonds de recherche de Qu�bec-Sant� et Institut national d'excellence en sant� et en services sociaux.

Summary of findings(Explanation)

Summary of findings for the main comparison. High-dose chemotherapy followed by autologous stem cell transplantation for patients with relapsed/refractory Hodgkin lymphoma
  1. 1 Both trials were closed prematurely.

High-dose chemotherapy followed by autologous stem cell transplantation for patients with relapsed/relapsed Hodgkin lymphoma
Patient or population: Patients with refractory/relapsed Hodgkin lymphoma
Settings:
Intervention: High-dose chemotherapy followed by autologous stem cell transplantation
OutcomesIllustrative comparative risks* (95% CI)Relative effect
(95% CI)
No of Participants
(studies)
Quality of the evidence
(GRADE)
Comments
Assumed riskCorresponding risk
Control High-dose chemotherapy followed by ASCT
Mortality
(after 3 years)
Study population HR 0.67
(0.41 to 1.07)
157
(2 studies)
⊕⊕⊕⊝
moderate 1
-
18 per 1000 13 per 1000
(8 to 20)
Moderate
  

Relapse or death

(at 3 years)

Study population HR 0.55
(0.35 to 0.86)
157
(2 studies)
⊕⊕⊕⊝
moderate 1
-
14 per 1000 8 per 1000
(5 to 13)
Moderate
  

Treatment-related mortality - overall analysis

(at 3 years)

Study population RR 0.61
(0.16 to 2.22)
157
(2 studies)
⊕⊕⊕⊝
moderate 1
-
66 per 1000 40 per 1000
(11 to 146)
Moderate
45 per 1000 27 per 1000
(7 to 100)
Adverse events
not estimable
Study populationNot estimable100
(1 study)
See commentOnly 1 trial reported this outcome, therefore not assessable
See commentSee comment
Moderate
  
Quality of life
not reported
Study populationNot estimable0
(0 studies)
See commentNeither study provided data with regard to this outcome
See commentSee comment
Moderate
  
*The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: confidence interval; HR: hazard ratio; RR: risk ratio.
GRADE Working Group grades of evidence
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

Background

Description of the condition

Hodgkin lymphoma (HL) is a malignancy of the lymph nodes and lymphatic system with possible involvement of other organs. The annual incidence is approximately two or three cases per 100,000 inhabitants in Western countries, so HL is a comparatively rare disease; however, it is one of the most common malignancies in young adults. Incidence increases with age with a first peak in the third decade and a second peak after the age of 60 years (Swerdlow 2003; Thomas 2002). In 2010, the American Cancer Society reported 8490 new cases of HL and 1320 deaths due to HL; however, since the early 1980s, death rates have continually decreased (ACS 2010). Men are affected slightly more often than women with a decreasing rate of 0.6% per year, while prevalence rates in women have increased by 1.1% per year since the early 1980s. In addition, HL affects more white people than black people (ACS 2010; Thomas 2002). The aetiology of HL is still unknown although epidemiological studies assume a viral influence. Patients with a history of Epstein-Barr virus infection have a two- to three-fold higher risk of developing HL and approximately half of the HL patients carry this virus (Weiss 2000). In 2013, 94% of HL in childhood and 80% to 90% of HL in adults can be cured with combined modality treatment. The five-year overall survival (OS) for HL is over 90% (ACS 2010; Eich 2010; Smith 2011; Von Tresckow 2011).

After a histologically confirmed diagnosis, HL is graded using the Ann-Arbor staging system (Carbone 1971):

  • stage I (involvement of a single lymph node region or lymphoid structure);

  • stage II (involvement of two or more lymph node regions on the same side of the diaphragm or localised involvement of extralymphatic organs);

  • stage III (involvement of two or more lymph node regions on both sides of the diaphragm, possibly accompanied by involvement of spleen);

  • stage IV (diffuse or disseminated involvement of one or more extralymphatic organs or tissues).

In addition, B-symptoms (night sweats, unintended weight loss of more than 10% of total body weight during the previous six months, fever of over 38 °C without known cause) and risk factors (e.g. mediastinal mass > 10 cm bulky nodal disease, extranodal involvement) can occur. According to Ann-Arbor staging, B-symptoms and risk factors, patients are assigned to early favourable stage (Ann-Arbor stages I and II without risk factors), early unfavourable stage (Ann-Arbor stages I and II with risk factors) and advanced stage (Ann-Arbor stages III and IV and selected stages II) (Engert 2007). Treatment of choice for first-line therapy is combined-modality treatment for most patients. Reducing the toxicities of the treatment, minimising the number of secondary malignancies and improving quality of life (QoL) are major aspects of treatment (Franklin 2010). After first-line treatment, about 15% to 20% of people cannot be cured and will relapse, and a smaller proportion will be primary refractory (defined as progression of disease during induction treatment or within 90 days after end of treatment) (Connors 2005; Von Tresckow 2011). In advanced-stage HL, the relapse rate can reach 30% after primary treatment (Kurvilla 2009). In the following, we will use the term 'relapsed' for both.

For relapsed patients, high-dose chemotherapy (HDCT) followed by autologous stem cell transplantation (ASCT) has become the optimal treatment option (Engert 2009; Linch 1993; NCCN 2012; Schmitz 2002). Josting et al. have yielded an important aspect for the treatment of patients with relapsed HL by identifying risk factors (early first relapse (between three and 12 months after treatment) or primary progressive disease, Ann-Arbor stage III or IV and anaemia at the time of relapse represent negative prognostic factors) that affect the prognosis of these patients. Patients experiencing all three factors showed a freedom from second treatment failure (FF2F) of 17% and an OS of 27%, while patients without risk factors had an FF2F of 48% and OS of 85% (Josting 2002). Moskowitz identified additional prognostic markers (extranodal disease, complete response of less than one year, B-symptoms) for event-free survival (EFS) after ASCT. In patients with none or one factor five-year EFS was 83% and OS was 90%, which decreased to five-year EFS of 10% and OS of 25%, if all factors were present (Moskowitz 2001).

Description of the intervention

There is no consensus about the gold standard for people with relapsed HL, but different treatment options are possible: conventional chemotherapy, salvage chemotherapy followed by HDCT and ASCT, HDCT followed by ASCT, salvage radiotherapy alone, allogeneic stem cell transplantation (allo-SCT), reduced-intensity allo-SCT, and different non-curative treatment options (Kuruvilla 2011). However, treatment with HDCT followed by ASCT achieved a durable remission of approximately 50% (Kuruvilla 2011; Smith 2011).

It remains unclear how to proceed with non responding patients after salvage therapy with dexamethasone plus high-dose cytarabine plus cisplatin (DHAP); however, intensifying salvage therapy or a second-line salvage regimen with Dexa-carmustine plus etoposide plus cytarabine plus melphalan (Dexa-BEAM) with or without mini-BEAM or IGEV (ifosfamide plus gemcitabine plus vinorelbine plus prednisolone) are some of the possibilities (Santoro 2007). A risk-adapted strategy with tandem autologous transplantation was analysed in the prospective multicenter H96 trial, resulting in single ASCT for intermediate-risk patients, while high-risk patients benefited from tandem ASCT (Morschhauser 2008). Allo-SCT or reduced-intensity allogeneic stem cell transplantation (RIC-allo) might be another treatment option for patients with multiple relapses, but regarding the high treatment-related mortality (TRM), a second ASCT should be rather considered (Sureda 2012). For patients who relapse after ASCT, newer drugs, such as panobinostat, are currently being assessed in clinical trials. The antibody-drug conjugate brentuximab vedotin has already been approved in the US by the Food and Drug Administration (FDA) and also in Europe in July 2012 by the European Medicines Agency (EMA) (Sureda 2010; Younes 2010).

How the intervention might work

Treatment of refractory and relapsed HL is challenging for clinicians as optimal treatment options still remain controversial and so far no difference in OS has been shown. To improve tumour control in patients with relapsed or refractory HL, HDCT was applied efficiently. But HDCT is associated with severe toxicity and complications for the patient, the major side effect being myeloablation, which is the near-complete deletion of bone marrow. To solve that problem, ASCT was introduced in the 1990s and was soon implemented. While the use of different doses and different chemotherapeutical drugs is still unclear, the role of ASCT is to avoid relapses and cure patients with relapses or refractory disease (Von Tresckow 2011).

Salvage therapy should enable patients to proceed to ASCT leading to a high-response rate combined with acceptable haematological and non-haematological toxicity. One overview of the literature (Kuruvilla 2011) reported 10 published phase II studies (Baetz 2003; Bertlett 2007; Bonfante 2001; Fermé 2002; Josting 2002a; Martin 2001; Moskowitz 2001; Proctor 2003; Schmitz 2002; Weaver 1998), analysing the impact of salvage-therapy before ASCT. Overall response rates of salvage therapy in different patient populations ranged between 60% and 87% with overlapping 95% confidence intervals (CI), without any study being superior. Even though discussion continues with regards to which salvage treatment is preferable, IGEV and ICE (ifosfamide plus carboplatin plus etoposide) (or DHAP) are the ones most frequently used before ASCT and, therefore, they are considered as standard regimens (Kuruvilla 2011; Moskowitz 2001; Santoro 2007). Salvage therapy with two cycles of DHAP chemotherapy is an effective and safe treatment option for patients with relapsed HL (Josting 2010).

Why it is important to do this review

Based on published randomised controlled trials (RCTs), the use of HDCT followed by ASCT is considered as standard therapy for patients with relapsed HL. The main challenge in treating relapsed patients is to find the best treatment that combines optimal efficacy with acceptable toxicity and decreases late effects such as secondary neoplasms, infertility, or cardiovascular and pulmonary effects. Implementation of the best treatment option in clinical pathways requires clear information about therapy regimens and possible adverse events (AEs) (Bhatia 2005; Connors 2005; Horning 2007). Although HDCT and autografting is recommended for patients with relapsed HL in clinical guidelines, it remains investigational (Engert 2009; Rancea 2013). To date, no systematic review evaluating the efficacy of the different chemotherapeutic options for patients with relapsed HL is known to the review authors. By systematically identifying all randomised studies and meta-analysing similar trials, we will overcome the statistical limitations of individual trials. In addition, a systematic identification of all RCTs conducted to date and a critical review of the reliability and validity of the studies is required.

Objectives

To assess and summarise the currently available evidence in terms of efficacy and safety of HDCT followed by ASCT for patients with primary refractory or first relapsed HL.

Methods

Criteria for considering studies for this review

Types of studies

We considered RCTs and included full-text publications only, as we found no publications in abstract form.

Types of participants

We included trials of patients with primary refractory or first relapse histologically confirmed HL without restrictions on age, gender or ethnicity. If we had identified trials with different haematological malignancies (mixed populations), we would have only used data from the refractory/relapsed HL subgroups. If subgroup data for these patients had not been provided (after contacting the authors of the trial), we would have excluded the trial if less than 80% of patients were relapsed or had refractory HL.

Types of interventions

The main intervention was HDCT followed by ASCT for patients with relapsed or primary refractory HL. We considered different treatment approaches for control comparisons: HDCT without ASCT, conventional chemotherapy without ASCT, additional SHDCT followed by HDCT before ASCT, and different HDCT regimens followed by ASCT.

We examined the following intervention:

  • any HDCT followed by ASCT.

We examined the following types of comparisons:

  • different chemotherapy regimens without ASCT;

  • different HDCT regimens (sequential or conventional) before ASCT.

We would have analysed other treatment approaches such as different HDCT regimens without ASCT, radiotherapy, combined-modality treatment, and different dosages or time schedules of HDCT, but we found no trials assessing these treatments.

Types of outcome measures

Primary outcomes
  • OS was defined as the time interval from random treatment assignment or entry into the study to death from any cause or to last follow-up.

Secondary outcomes
  • PFS was defined as the time interval from random treatment assignment/entry to the study to first progression or relapse, death from any cause or the last follow-up.

  • Response rates were measured as overall response rate (ORR), complete response rate (CRR) and partial response rate (PRR).

  • TRM was defined as death related to the intervention under investigation.

  • AEs were defined as any AE occurring under study medication or during follow-up, or both; serious adverse events (SAEs) as defined by the authors of the primary studies or the World Health Organization (WHO).

We would have analysed QoL with reliable and valid instruments, but we did not find any trials assessing this outcome.

Search methods for identification of studies

Electronic searches

We adapted the search strategies suggested in the Cochrane Handbook of Systematic Reviews of Interventions (Lefebvre 2011). To reduce language bias, we did not apply language restrictions.

We searched the following databases/sources of medical literature:

  • the Central Register of Controlled Trials (CENTRAL), 2013, Issue 1 (see Appendix 1 for search strategy);

  • MEDLINE (Ovid) (1946 to January 2013) (see Appendix 2 for search strategy).

Searching other resources

We searched conference proceedings of annual meetings of the following societies not included in CENTRAL (1990 to January 2013):

  • American Society of Hematology (ASH);

  • American Society of Clinical Oncology (ASCO);

  • European Hematology Association (EHA).

We searched electronic databases of ongoing trials:

We also handsearched references of all identified trials, relevant review articles and current treatment guidelines (www.g-i-n.net).

In order to receive additional data for our main analysis, we contacted Dr. Horst Müller from the German Hodgkin Study Group (GHSG). With his help, we were able to include additional follow-up information and data regarding PFS from the HDR1 in our analysis.

Data collection and analysis

Selection of studies

After the first screen of all titles and abstracts of the identified studies, two review authors (MR, NS) independently excluded all studies that were clearly ineligible. We assessed selected studies by using an eligibility form regarding study design and compliance with inclusion criteria (Higgins 2011b). If there was any doubt, we obtained the full-text publication of the study and discussed eligibility. If disagreement had persisted, we would have asked a third review author to finalise a decision, but this was not necessary. We preferred to include more studies rather than to lose relevant data. According to PRISMA, we used a flow chart (see Figure 1) to document the study selection process, showing the total numbers of retrieved references, and the numbers of included and excluded studies (Moher 2009).

Figure 1.

Study flow diagram.

Data extraction and management

Two review authors (MR, NS) independently extracted data according to the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011b) by using the standardised data extraction form with following items:

  • general information: author, title, source, publication date, country, language, duplicate publications;

  • quality assessment: allocation concealment, blinding (patients, personnel, outcome assessors), incomplete outcome data, selective outcome reporting, other sources of bias;

  • study characteristics: trial design, setting, dates, source of patients, inclusion/exclusion criteria, comparability of groups, subgroup analysis, statistical methods, power calculations, treatment cross-overs, compliance with assigned treatment, length of follow-up, time point of randomisation;

  • participant characteristics: age, gender, ethnicity, number of patients recruited/allocated/evaluated, patients lost to follow-up, additional diagnoses, risk factors (i.e. time to relapse, response to salvage chemotherapy, B-symptoms, stage of disease, presence of anaemia at relapse);

  • interventions: setting, type of (multiple drug) chemotherapy (intensity of regimen, number of cycles, with or without radiotherapy), transplantation with or without growth factor support, transplant details, infection prophylaxis, duration of follow-up;

  • outcomes: OS, PFS, response rate, TRM, AEs, QoL.

Assessment of risk of bias in included studies

To assess the methodological quality and the risk of bias, we used a questionnaire using the recommendations of the Cochrane Handbook for Systematic Reviews of Interventions for the following criteria (Higgins 2011a):

  • sequence generation;

  • allocation concealment;

  • blinding (patients, personnel, outcome assessors);

  • incomplete outcome data;

  • selective outcome reporting;

  • other sources of bias.

For every criterion, we made a judgement using one of three categories:

  • 'low risk': if the criterion was adequately fulfilled in the study, that is, the study was at a low risk of bias for the given criterion;

  • 'high risk': if the criterion was not fulfilled in the study, that is, the study was at high risk of bias for the given criterion;

  • 'unclear': if the study report did not provide sufficient information to allow for a judgement of 'yes' or 'no' or if the risk of bias is unknown for one of the criteria listed above.

Measures of treatment effect

For binary outcomes, we calculated risk ratios (RR) with 95% CI for each trial. For time-to-event data, we estimated treatment effects of individual studies as hazard ratios (HR) using the methods described by Parmar 1998 and Tierney 2007.

Dealing with missing data

As suggested in Chapter 16 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011), there are many potential sources of missing data that had to be taken into account: at study level, at outcome level, at summary data level, at individual level and study level characteristics (e.g. for subgroups analysis). First, it was important to differentiate between 'missing at random' and 'not missing at random'.

If there had been missing data, we would have contacted the original investigator to request missing data. If data had still been missing, we would have made explicit assumptions of any methods used; for example, that the data were assumed missing at random or that missing values were assumed to have a particular value, such as a poor outcome.

In addition, we would have performed sensitivity analysis to assess how sensitive results were to reasonable changes in the assumptions that we made. We addressed the potential impact of missing data on the finding of the review in the Discussion.

Assessment of heterogeneity

We pooled all results by applying meta-analysis using a fixed-effect model, because of the small number of trials in the analysis we could not evaluate heterogeneity. If more trials had been available, we would have assessed heterogeneity of treatment effects between trials by using a Chi2 test with a significance level at P value < 0.1. We would have used the I2 statistic to quantify possible heterogeneity (I2 > 30% moderate heterogeneity, I2 > 75% considerable heterogeneity) (Deeks 2011). We would have explored potential causes of heterogeneity by sensitivity and subgroup analysis.

Assessment of reporting biases

In a meta-analysis with at least 10 trials, we would have explored potential reporting bias by generating a funnel plot and statistically testing this by conducting a linear regression test (Sterne 2011). We would have considered a P value less than 0.1 to be significant. However, we only included three trials so this test was not conducted (Sterne 2011).

Data synthesis

We performed analyses according to the recommendations of the Cochrane Handbook for Systematic Reviews and Interventions (Deeks 2011), and used aggregated data and the Cochrane statistical package Review Manager 5 for analyses (RevMan 2011). One review author (MR) entered data into the software and a second review author (NS) checked it for accuracy. We performed meta-analyses using a fixed-effect model (e.g. the generic inverse variance method for survival data outcomes and Mantel-Haenszel method for dichotomous data outcomes).

We used the software GRADEpro 3.2 to create 'Summary of findings' tables as suggested in the Cochrane Handbook for Systematic Reviews and Interventions (Schuenemann 2011). The main outcomes in the 'Summary of findings' were OS, PFS, AEs, TRM and QoL.

Subgroup analysis and investigation of heterogeneity

Due to the small number of studies in the analysis and the missing subgroup data, it was not possible to evaluate heterogeneity. We would have considered the following parameters for subgroup analysis, but the study authors did not provide study data:

  • age (e.g. adults < 65 years versus adults ≥ 65 years);

  • stage of disease (e.g. Ann-Arbor stages I to IV);

  • different types of HDCT (e.g. aggressive therapy or less aggressive);

  • different durations of follow-up;

  • different risk factors (e.g. time to relapse, response to salvage chemotherapy, B-symptoms, presence of anaemia at relapse).

Sensitivity analysis

We did not perform a sensitivity analysis because of the low number of included studies (N = 2) in the meta-analysis.

Results

Description of studies

Results of the search

We identified 1663 potentially relevant references. Of these, we excluded 1638 at the initial stage after screening the titles, because they did not fulfil our predefined inclusion criteria or were duplicates. We retrieved the remaining 25 publications as full-text publications and, if not possible, as abstract publications, for a detailed evaluation. Of these, we excluded a further 11 publications. Finally, we included 14 publications of three trials with 398 patients in this systematic review. The overall number of references screened, identified, selected, excluded and included are documented in the PRISMA flow diagram (Figure 1) (Moher 2009).

We conducted a meta-analysis with two trials (BNLI; HDR1). We did not include the third trial in the meta-analysis, as it assessed ASCT in both treatment arms (HDR2).

Included studies

Three trials published in 14 publications with 398 patients (range 40 to 241) fulfilled our inclusion criteria (see Characteristics of included studies for further details) (BNLI; HDR1; HDR2).

For one trial (BNLI), no dates on trial recruitment were provided, the earliest trial recruited from 1993 to 1997 (HDR1), and the latest from 2000 to 2006 (HDR2).

Design

All three included trials were two-armed RCTs.

Sample sizes

The smallest trial (BNLI) randomised 40 patients and the largest trial (HDR2) randomised 241 patients .

Location

BNLI did not report the countries were the trial was conducted. The other two trials were conducted in different European countries (HDR1; HDR2).

Participants

A total amount of 490 male and female patients with relapsed or refractory HL. Of these, we evaluated 398 randomised patients who received different chemotherapy regimens with or without ASCTfor the primary outcome.

Patients in the BNLI trial had a median age of 26 years (range 18 to 40) with 70% males in the HDCT plus ASCT arm and 32 years of age (range 17 to 49) with 50% males in the mini-BEAM arm. In the HDR1 trial, patients in the HDCT plus ASCT arm had a median age of 30 years (range 21 to 57) with 64% males, and those in the conventional chemotherapy arm had a median age of 34 years (range 16 to 55) with 64% males. In the HDR2 trial, patients in both treatment arms were 35 years old. The standard HDCT arm included 62% males and the SHDCT arm 68% males.

Interventions

In two trials, HDCT (BEAM (carmustine, etoposide, cytarabine, and melphalan)) followed by ASCT (bone marrow transplantation (BNLI) or either bone marrow or peripheral blood progenitor cells (HDR1)) was compared with chemotherapy alone (mini-BEAM (BNLI) or Dexa-BEAM (HDR1)). The third trial evaluated two cycles of DHAP plus SHDCT (high-dose cyclophosphamide, high-dose methotrexate plus vincristine, and high-dose etoposide) followed by BEAM plus ASCT from peripheral blood stem cells versus two cycles of DHAP plus BEAM and ASCT (HDR2).

Outcomes
Primary outcome measure

OS was analysed in all three trials, but only as a primary outcome in the BNLI trial with a median follow-up of 34 months, and was reported in the other trials as secondary outcome only (HDR1; HDR2).

Secondary outcome measures

Only one trial reported data on PFS (HDR2). For the HDR1 trial, we contacted the authors and received further data on PFS. BNLI reported cumulative progression rate, which can be converted to PFS. Response rates and TRM were analysed in all three trials, and AEs in two trials (HDR1; HDR2). Additional analysis included EFS in BNLI as secondary outcome. In addition, freedom from treatment failure was analysed in HDR1 and HDR2 as primary outcome measures with a median follow-up of 83 months (HDR1), and 42 months (HDR2).

Conflict of interest

One trial did not provide conflict of interest statements (BNLI). The authors of the other two trials stated no potential conflicts of interest (HDR1; HDR2). Two trials were supported by the German Cancer Aid (Deutsche Krebshilfe) (HDR1; HDR2), and one trial was also funded by the Mildred Scheel Stiftung (HDR1).

Excluded studies

After a detailed evaluation of the retrieved full-text publications, we excluded 11 trials, of which seven were not RCTs (Avigdor 2000; Evens 2007; Ferme 2002; Gutierrez-Delgado 2003; Kuruvilla 2004; Morschhauser 2008; Sweetenham 1997), three did not assess patients with relapsed HL separately (Bolanos-Meade 2007; Vellenga 2001; Weaver 1996), and one was a cost analysis (Van Agthoven 2001) (see Characteristics of excluded studies table).

Risk of bias in included studies

Overall, the quality of included trials was moderate. For more information, see 'Risk of bias' tables in the Characteristics of included studies table, and Figure 2 and Figure 3. The 'Risk of bias' graph illustrates the proportion of studies with each of the judgements 'low risk', 'high risk', and 'unclear risk' (Figure 2). The 'Risk of bias' summary figure presents all judgements in a cross-tabulation of study by entry (Figure 3).

Figure 2.

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Figure 3.

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Allocation

Random sequence generation was adequate in all three trials. The BNLI trial used a central sealed card system for randomisation. Patients in the HDR1 trial were randomly assigned by computer, and in the HDR2 trial allocation of patients was conducted by the co-ordinating centre of the trial. No information was available for allocation concealment in the other two trials (BNLI; HDR2).

Blinding

No trial reported information about blinding of patients and physicians. Therefore, we judged blinding of patients and physicians as 'high risk'. Usually trials evaluating the role of stem cell transplantation are not blinded, which leads to a potential high risk of bias. We judged the outcome assessor as 'unclear risk' of bias for two trials (BNLI; HDR1), as it was not reported, and 'high risk' for one trial that was not masked (HDR2).

Incomplete outcome data

Two of the three trials described missing outcome data in detail and stated that they had performed analyses according to the intention-to-treat principle (BNLI; HDR2), we therefore judged 'low risk' of attrition bias for these trials. In the HDR1 trial, we judged the risk of incomplete outcome data as 'high' because analyses were performed according to a per-protocol analysis.

Selective reporting

For two trials, there were no study protocols available at the meta-register of clinical trials (www.controlled-trials.com/mrct/). Therefore, we judged the potential risk of reporting bias as 'unclear' (BNLI; HDR1). The HDR2 trial study protocol was available and all outcomes were reported adequately, therefore, the risk of bias for this study was judged as 'low'.

Other potential sources of bias

Two trials were stopped before reaching the final trial size (BNLI; HDR1). In the BNLI trial, patients refused randomisation and demanded ASCT, and in the HDR1 trial, the scientific committee decided to stop the study early because of the low accrual of patients. Therefore, we judged both trials as 'unclear' for risk of other bias. HDR2 appeared to be free of other potential sources of bias, therefore, we judged it as 'low' risk of bias.

Effects of interventions

See: Summary of findings for the main comparison High-dose chemotherapy followed by autologous stem cell transplantation for patients with relapsed/refractory Hodgkin lymphoma

High-dose chemotherapy plus autologous stem cell transplantation versus conventional chemotherapy without autologous stem cell transplantation

Two RCTs (N = 157) evaluated the efficacy and safety of HDCT followed by ASCT compared with conventional chemotherapy alone (BNLI; HDR1). BNLI compared conventional chemotherapy (mini-BEAM (carmustine plus etoposide plus cytarabine plus melphalan)) versus high-dose chemotherapy (BEAM) followed by bone marrow ASCT. The HDR1 trial compared chemosensitive patients (after two cycles of salvage Dexa-BEAM) who received either two further cycles of Dexa-BEAM or high-dose BEAM followed by ASCT (transplantation of haematopoietic stem cells).

Primary outcome
Overall survival

Both trials reported data for OS in 157 relapsed HL patients (BNLI; HDR1). OS was not statistically significantly different between patients treated with HDCT plus ASCT or conventional chemotherapy alone (HR 0.67; 95% CI 0.41 to 1.07; P value = 0.10) (Figure 4).

Figure 4.

Forest plot of comparison: 1 High-dose chemotherapy before autologous stem cell transplantation, outcome: 1.1 Overall survival - overall analysis.

Secondary outcomes
Progression-free survival

We contacted Dr. Horst Müller, the statistician of the HDR1 trial, to request additional data on PFS. With his help, we were able to analyse PFS for 157 patients from both studies in the meta-analysis (BNLI; HDR1). Patients treated with HDCT followed by ASCT had a statistically significantly increased PFS compared with patients treated with conventional chemotherapy alone (HR 0.55; 95% CI 0.35 to 0.86; P value = 0.009) (Figure 5).

Figure 5.

Forest plot of comparison: 1 High-dose chemotherapy before autologous stem cell transplantation, outcome: 1.5 Progression-free survival.

Response rates
Complete response rate

Both trials with 157 relapsed or refractory patients analysed CRR (BNLI; HDR1). The CRR was statistically significantly higher in the HDCT plus ASCT arm (RR 1.32; 95% CI 1.07 to 1.64; P value = 0.01).

Overall response rate

The meta-analysis of ORR included both trials with 157 patients; there was no statistically significant difference between treatment with HDCT plus ASCT or conventional chemotherapy (RR 1.15; 95% CI 0.99 to 1.34; P value = 0.08) (BNLI; HDR1).

Treatment-related mortality

There was a trend to reduced TRM with HDCT plus ASCT (BNLI: two HDCT plus ASCT versus zero conventional chemotherapy; HDR1: one HDCT plus ASCT versus five conventional chemotherapy); however, this trend was not statistically significant between the treatment arms (RR 0.61; 95% CI 0.16 to 2.22; P value = 0.45) (Figure 6).

Figure 6.

Forest plot of comparison: 1 High-dose chemotherapy before autologous stem cell transplantation, outcome: 1.4 Treatment-related mortality - overall analysis.

Adverse events

AEs were reported in only one trial and are presented in Table 1 (HDR1). SAEs occurred in 51 (83.6%) patients treated with HDCT plus ASCT and in 49 (87.5%) patients treated with conventional chemotherapy. Overall, more AEs occurred in the conventional chemotherapy arm. However, the frequency of toxic effects did not differ significantly between treatment arms.

Table 1. Adverse events in HDR1
  1. ASCT: autologous stem cell transplantation; HDCT: high-dose chemotherapy; WHO: World Health Organization.

    WHO grades 3 and 4 toxic effects per treatment group

WHO grade 3 and 4 toxicity

Conventional chemotherapy

(N = 49)

HDCT plus ASCT

(N = 51)

P value
Infection24 (49%)24 (47%)0.848
Oral (mucositis)10 (24%)19 (37%)0.168
Gastrointestinal10 (20%)7 (14%)0.374
Pulmonary or respiratory tract6 (12%)2 (4%)0.125
Cardiac3 (6%)1 (2%)0.288
Neurological2 (4%)1 (2%)0.534
Hepatic2 (4%)00.145
Renal1 (2%)00.305
Quality of life

Neither study reported on QoL.

Additional sequential high-dose chemotherapy versus high-dose chemotherapy before autologous stem cell transplantation

One RCT (N = 241) evaluated the efficacy and safety of SHDCT before ASCT and compared it with standard HDCT before ASCT in patients with histologically confirmed relapsed or refractory HL (HDR2). Both treatment arms initially received two cycles of DHAP and were then randomised to the SHDCT arm before BEAM and ASCT or to the HDCT arm (with high-dose BEAM) before ASCT.

Overall survival

OS was analysed as secondary endpoint in the HDR2 trial. There was no statistically significant difference between SHDCT and HDCT. The three-year OS rate was 80% in the SHDCT arm compared with 87% in the HDCT arm (HR 0.93; 95% CI 0.5 to 1.74; P value = 0.816).

Progression-free survival

PFS was not statistically significantly different between the two treatment groups. The PFS rates at three years were 67% for treatment with SHDCT and 72% for treatment with HDCT (HR 0.87; 95% CI 0.58 to 1.30; P value = 0.505).

Response rates
Complete response rate

At the final evaluation, 101 (84%) patients in the SHDCT arm and 99 (83%) patients in the HDCT arm reached complete response/complete response, unconfirmed. We found no statistically significant difference in terms of CRR between SHDCT and HDCT (RR 1.00; 95% CI 0.89 to 1.12; P value = 0.93).

Overall response rate

We found no statistically significant differences between treatment arms with regard to ORR. ORR was 89% (N = 108) in the SHDCT arm and 89% (N = 106) in the HDCT arm (RR 0.99; 0.91 to 1.09; P value = 0.89).

Treatment-related mortality

TRM after the pre-phase lead to seven deaths in the SHDCT arm and one in the HDCT arm, a further five patients died owing to the toxicity of additional salvage treatment (three patients in the SHDCT arm versus two patients in the HDCT arm), and five as a result of infections/sepsis (one patient in the SHDCT arm versus four patients in the HDCT arm).

Adverse events

All AEs reported in the HDR2 trial are shown in Table 2.

Table 2. Adverse events in HDR2
  1. WHO: World Health Organization; WHO grade 3/4 toxic effects per treatment group.

WHO grade 3 and 4 toxicity

HDCT

(N = 113)

SHDCT

(N = 110)

P value
Anaemia59 (52%)77 (70%)0.007
Thrombopenia100 (89%)102 (93%)0.28
Leukopenia98 (87%)98 (89%)0.59
Infection37 (33%)53 (48%)0.02
Nausea40 (35%)50 (46%)0.13
Mucositis64 (57%)74 (67%)0.10
Respiratory7 (6%)11 (10%)0.30

The SHDCT arm resulted in higher toxicity compared with the HDCT arm (N = 223). After two cycles of DHAP, 97 (88%) patients in the SHDCT arm and 51 (45%) in the HDCT arm experienced WHO grade 4 toxicity (P value < 0.00001). Significantly more WHO grade 3/4 infections (48% SHDCT versus 33% HDCT; P value = 0.02) and WHO grade 3/4 anaemia (70% SHDCT versus 52% HDCT; P value = 0.007) occurred with the intensified treatment regimen. Thrombopenia (93% SHDCT versus 89% HDCT; P value = 0.2792) and leukopenia (89% SHDCT versus 87% HDCT; P value = 0.59) did not differ statistically significantly.

Quality of life

The trial did not assess QoL.

Discussion

Summary of main results

The following findings emerge from this Cochrane Review and meta-analysis evaluating HDCT followed by ASCT compared with conventional chemotherapy without ASCT (two trials, 157 patients):

  • based on current available research, HDCT followed by ASCT shows a potential advantage in OS as compared to conventional chemotherapy alone in patients with relapsed or refractory HL, but, because of the small number of patients in the trials, this advantage does not translate to statistical significance;

  • PFS and CRR are significantly improved in patients receiving HDCT plus ASCT compared with those treated with conventional chemotherapy;

  • AEs and TRM are not statistically significantly different in patients treated with HDCT plus ASCT compared with conventional chemotherapy alone;

  • None of the trials reported QoL.

For the analysis of additional SHDCT before HDCT followed by ASCT compared with standard HDCT followed by ASCT, we found one trial with 241 patients. We found:

  • there is no evidence that additional SHDCT before ASCT compared with standard HDCT before ASCT improves OS;

  • PFS was not significantly different, but there was a positive trend after treatment with HDCT;

  • AEs were experienced significantly more often by patients in the additional SHDCT before ASCT than by patients in the standard HDCT before ASCT arm;

  • TRM was not influenced by the increased AEs, and was similar in both arms;

  • QoL was not reported in this trial.

Overall completeness and applicability of evidence

This Cochrane Review and meta-analysis included two trials evaluating the efficacy and safety of HDCT followed by ASCT compared with conventional chemotherapy alone, and one trial assessing additional SHDCT followed by HDCT plus ASCT compared with HDCT followed by ASCT in patients with relapsed or refractory HL. Because of the clinical homogeneity of two included trials, we were able to pool the outcomes in a meta-analysis. The third trial analysed ASCT in both treatment arms, therefore, we could not include it in the meta-analysis.

Quality of the evidence

Overall, the quality of the three included trials (398 patients) was moderate. All trials were reported as randomised and open-label studies and only one trial reported allocation concealment. The open-label design and unclear allocation concealment could lead to selection, performance or detection bias. Blinding of outcome assessor would have increased the quality of the study and would have been feasible, but was not performed. Study protocol was not available for two trials, therefore, we judged selective reporting as unclear for two, and low risk for one trial. The premature closure of two trials due to patients refusal and low accrual could have introduced other biases. Patients in the HDCT followed by ASCT arm in the BNLI trial were younger (median 26 years old) than patients in the HDCT only arm (median 32 years old). For younger patients stem cell transplantation is more compatible. In addition, potential sources of bias may produce uncertainties in the HR calculations. In one trial, the HRs for OS and PFS were calculated from survival curves with a constant censoring as described by Tierney 2007. Because of the small number of trials and patients included in the main analysis, obtaining reliable information from subgroup and sensitivity analyses was not possible. Dr. Horst Müller from the GHSG provided additional data on PFS for the HDR1 trial, so that we were able to pool results. In the HDR1 trial, patients with chemosensitive disease only proceeded to the intervention and control arm, and only those were included in our meta-analysis.

In the HDR2 trial, all recruited patients received two cycles of DHAP induction treatment, and only patients who were not refractory to chemotherapy were then randomised to the intervention or control arms (possible occurrence of recruitment bias).

Potential biases in the review process

We tried to avoid bias by performing all relevant processes in duplicate. We are not aware of any obvious flaws in our review process. However, one aspect of the internal validity of the results should be mentioned. The significant results for PFS regarding HDCT followed by ASCT were based on only two RCTs assessing 157 patients. Results from one trial were from a per-protocol analysis (HDR1). These statistically significant results might be changed into non-significant by the inclusion of one further trial with an HR close to 1. In addition, the small number of trials included in this analysis may lead to publication bias, as they are not evaluable formally with a funnel plot as more trials would be necessary. However, it is important to note that publication bias could have an impact on a meta-analysis with only two trials. Therefore, further evidence regarding OS and particularly PFS from RCTs is needed.

Agreements and disagreements with other studies or reviews

To our knowledge, this is the first Cochrane Review with meta-analysis focusing on patients with first relapsed or primary refractory HL after first-line therapy comparing HDCT followed by ASCT versus conventional chemotherapy without ASCT. We included a second evaluation of additional SHDCT before HDCT followed by ASCT compared with HDCT before ASCT, which unfortunately did not fit into the meta-analysis and was described separately.

Comparing our results with results from other trials is challenging, because we have included all available RCTs in our Cochrane Review (BNLI; HDR1; HDR2). However, our results are in accordance with the results from other non-randomised trials in patients with relapsed HL. For example, Morschhauser et al. analysed risk-adapted salvage treatment with single or tandem ASCT after HDCT in patients with HL and first relapse and showed that intermediate-risk patients can achieve good treatment results with single ASCT (Morschhauser 2008).

The question of the optimal salvage therapy remains unclear. There are several options for salvage therapy that were used in different randomised and non-randomised trials: Dexa-BEAM, mini-BEAM, DHAP, ICE, GDP (gemcitabine, dexamethasone, cisplatin), GVD (gemcitabine, vinorelbine, doxorubicin), IEV (ifosfamide, etoposide, vinorelbine), MINE (mitoguazone, ifosfamide, vinorelbine, etoposide), and IV (ifosfamide, vinorelbine), but evidence to prove superiority was not found (Kuruvilla 2011). With the HDR1 trial of the GHSG, an ORR of 81% occurred after Dexa-BEAM. Unfortunately, this salvage therapy was associated with an increased TRM of 5%, while other trials reported a TRM range of 0% to 5% (Kuruvilla 2011). The acceptable toxicity of two cycles of DHAP before HDCT plus ASCT in the HDR2 trial and high rates of overall responses (89%) indicate the effectiveness of this regimen; therefore, it has become the standard for the GHSG and other co-operative groups. Additional analyses of the HDR2 trial revealed that a dose-density approach of DHAP statistically significant increases in PFS and OS are possible and, therefore, recommended (Sasse 2012).

There are some ongoing trials assessing newer agents and different approaches for patients with relapsed/refractory HL. Maintenance therapy is in focus in the ongoing AETHERA trial, assessing the effect of brentuximab vedotin compared with placebo. One ongoing randomised phase II trial will assess efficacy and safety of everolimus in combination with DHAP as induction therapy for patients with relapsed or refractory HL (HD-R3i).

However, some major topics still have limited or no evidence from RCTs available, such as: the role of ASCT in relapsed patients after HDCT followed by ASCT and direct comparisons of different HDCT regimens, so that it is not possible to make a final conclusion of optimal regimens. However, evidence from this Cochrane Review shows an increased PFS and acceptable safety with HDCT followed by ASCT as compared with conventional chemotherapy alone.

Authors' conclusions

Implications for practice

There is currently not enough evidence from randomised controlled trials (RCTs) to demonstrate that high-dose chemotherapy (HDCT) followed by autologous stem cell transplantation (ASCT) significantly improves overall survival (OS) in patients with relapsed or refractory Hodgkin lymphoma (HL). An advantage in OS is indicated in the current studies, but this did not translate to significance, which might be due to the small number of included patients. Progression-free survival is increased if HDCT followed by ASCT is compared with conventional chemotherapy alone. There is no evidence that additional sequential HDCT before HDCT followed by ASCT compared with HDCT followed by ASCT improves efficacy, but results in significantly more World Health Organization (WHO) grade 3 or 4 adverse events, including infections.

Implications for research

More RCTs with OS as the primary outcome are needed to assess the effect of HDCT followed by ASCT. This may be particularly important when progression-free survival is improved to evaluate if this effect also translates into improvement in OS. In addition, it will become important to assess the effects of newer substances (such as brentuximab vedotin) compared with HDCT.

Acknowledgements

We are grateful to the following people for their comments and for improving the protocol and review:

Prof. Lena Specht and Dr. Guido Schwarzer (editors of the Cochrane Haematological Malignancies Group), Sabine Kluge and Andrea Will (members of the editorial team), Céline Fournier (consumer editor of the Cochrane Haematological Malignancies Group), Heather Maxwell and Anne Lawson for copy editing. We are also very grateful to Dr. Horst Müller from the GHSG for providing additional data for the analysis of HDR1 trial.

Data and analyses

Download statistical data

Comparison 1. High-dose chemotherapy (HDCT) before autologous stem cell transplantation (ASCT)
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Overall survival - overall analysis2157Hazard Ratio (Fixed, 95% CI)0.67 [0.41, 1.07]
2 Overall response rate - overall analysis2157Risk Ratio (M-H, Fixed, 95% CI)1.15 [0.99, 1.34]
3 Complete response rate - overall analysis2157Risk Ratio (M-H, Fixed, 95% CI)1.32 [1.07, 1.64]
4 Treatment-related mortality - overall analysis2157Risk Ratio (M-H, Fixed, 95% CI)0.61 [0.16, 2.22]
5 Progression-free survival2157Hazard Ratio (Fixed, 95% CI)0.55 [0.35, 0.86]
Analysis 1.1.

Comparison 1 High-dose chemotherapy (HDCT) before autologous stem cell transplantation (ASCT), Outcome 1 Overall survival - overall analysis.

Analysis 1.2.

Comparison 1 High-dose chemotherapy (HDCT) before autologous stem cell transplantation (ASCT), Outcome 2 Overall response rate - overall analysis.

Analysis 1.3.

Comparison 1 High-dose chemotherapy (HDCT) before autologous stem cell transplantation (ASCT), Outcome 3 Complete response rate - overall analysis.

Analysis 1.4.

Comparison 1 High-dose chemotherapy (HDCT) before autologous stem cell transplantation (ASCT), Outcome 4 Treatment-related mortality - overall analysis.

Analysis 1.5.

Comparison 1 High-dose chemotherapy (HDCT) before autologous stem cell transplantation (ASCT), Outcome 5 Progression-free survival.

Appendices

Appendix 1. CENTRAL search strategy

#Searches
#1MeSH descriptor Lymphoma explode all trees
#2MeSH descriptor Hodgkin Disease explode all trees
#3(germinoblastom*)
#4(reticulolymphosarcom*)
#5(Hodgkin*)
#6(malignan* near/2 lymphogranulom*)
#7(malignan* near/2 granulom*)
#8(#1 OR #2 OR #3 OR #4 OR #5 OR #6 OR #7)
#9rezidiv*
#10relap*
#11refractor*
#12pregredient*
#13progredient*
#14(#9 OR #10 OR #11 OR #12 OR #13)
#15(#8 AND #14)
#16"accession number" near PubMed
#17(#15 AND NOT #16)

Appendix 2. MEDLINE search strategy

#Searches
#1LYMPHOMA/
#2exp HODGKIN DISEASE/
#3Germinoblastom$.tw,kf,ot.
#4Reticulolymphosarcom$.tw,kf,ot.
#5Hodgkin$.tw,kf,ot.
#6(malignan$ adj2 (lymphogranulom$ or granulom$)).tw,kf,ot.
#7or/1-6
#8rezidiv$.tw,kf,ot.
#9relap$.tw,kf,ot.
#10refractor$.tw,kf,ot.
#11pregredient$.tw,kf,ot.
#12progredient$.tw,kf,ot.
#13or/8-12
#14randomized controlled trial.pt.
#15controlled clinical trial.pt.
#16randomized.ab.
#17placebo.ab.
#18clinical trials as topic.sh.
#19randomly.ab.
#20trial.ti.
#21or/14-20
#22humans.sh.
#2321 and 22
#247 and 13
#2523 and 24

Contributions of authors

Michaela Rancea (MR): conception of the review, writing of the protocol and the review.

Dr. Nicole Skoetz (NS): methodological expertise and advice, proof read.

Dr. Bastian von Tresckow (BT): clinical expertise, advice and proof read.

Ina Monsef (IM): development of the search strategies, ran the database searches and provided reference databases.

Prof. Andreas Engert (AE): clinical expertise and advice.

All authors have read and accepted the final version of the review.

Declarations of interest

MR, NS, BT and IM have no known conflicts of interest. Andreas Engert is chair of the German Hodgkin Study Group and conducted two of the three included studies.

Sources of support

Internal sources

  • Department I of Internal Medicine, University Hospital of Cologne, Cologne, Germany.

External sources

  • No sources of support supplied

Differences between protocol and review

None.

Characteristics of studies

Characteristics of included studies [ordered by study ID]

BNLI

Methods

Randomisation

  • 2 arms: 1 week BEAM plus ABMT on day 7 versus mini-BEAM up to 3 cycles (after full haematological recovery)

Recruitment period

  • Not reported

Median follow-up time:

  • 34 months (range not reported)

Participants

Patients randomised (N = 40)

  • BEAM plus ABMT (N = 20); (N = 1 withdrew before informed consent was given, and was included in analyses of overall survival and event-free survival as a non-responder, but was excluded from the progression-free survival analysis)

  • Mini-BEAM (N = 20)

Eligibility criteria

  • Histologically confirmed HL

  • Patients with active resistant or relapsed disease at trial entry

  • Patients with high-risk HL: who did not achieve a CR with MOPP or a similar regimen; who had high-grade histology or an erythrocyte sedimentation rate above 59 mm/h at presentation; who did not attain a CR or who relapsed within 1 year of an initial alternating or hybrid regimen

  • Patients for whom 2 or more chemotherapy regimens had failed

Mean age

  • BEAM plus ABMT: 26 years (range 18 to 40); mini-BEAM: 32 years (range 17 to 49)

Gender (male)

  • BEAM plus ABMT: 14 (70%); mini-BEAM: 10 (50%)

Stage of disease (by BNLI)

  • Grade I (nodular sclerosis type I or lymphocyte predominant)

  • Grade II (nodular sclerosis type II, mixed cellularity or lymphocyte depleted)

Country

  • Not reported

Interventions

Previous treatment:

  • 1 regimen of previous chemotherapy was administered to 6 patients from BEAM plus ASCT arm and 5 patients from the mini-BEAM arm

  • 2 regimens of previous chemotherapy (including alternating regimens as MOPP and ABVD) was administered to 11 patients in the BEAM plus ASCT arm and 13 patients in the mini-BEAM arm

  • 3 regimens of pervious chemotherapy was administered to 3 patients in the BEAM plus ASCT arm and 4 patients in the mini-BEAM arm

  • Previous RT was conducted in 11 patients in the BEAM-ASCT arm and four patients from the mini-BEAM arm

BEAM plus ASCT

  • Patients received carmustine (300 mg/m2), etoposide (800 mg/m2), cytarabine (1600 mg/m2) and melphalan (140 mg/m2) for 1 week. On day 7, at least 10 cryopreserved autologous nucleated bone marrow cells per kilogram were re-infused. When a residual mass persisted after chemotherapy, RT was permitted.

Mini-BEAM

  • Patients received carmustine (60 mg/m2), etoposide (300 mg/m2), cytarabine (800 mg/m2) and melphalan (30 mg/m2) for 3 weeks. Patients who responded to a cycle, were given a second cycle after full haematological recovery. A third cycle could be given, if there was a further, but incomplete, response. When a residual mass persisted after chemotherapy, RT was permitted

  • Patients who did not respond were taken off protocol and received further therapy at their physician's direction

Outcomes

Reported

  • Primary outcome

    • Overall survival

  • Secondary outcome

    • Progression-free survival

    • Event-free survival

    • Response rate

Not reported (relevant for this review)

  • Adverse events

NotesThere were no conflicts of interest stated. The trial was closed early, before the target accrual was achieved (40 patients instead of 66 patients), because patients refused randomisation and demanded ABMT. No information about the funding of this trial is provided
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskComment: the authors describe the sequence generation process by "central sealed card system"
Allocation concealment (selection bias)Unclear riskNo information provided
Blinding of participants and personnel (performance bias)
All outcomes
High riskNo information provided. Trials evaluating stem cell transplantation are usually not blinded
Blinding of outcome assessment (detection bias)
All outcomes
Unclear riskNo information provided. Trials evaluating stem cell transplantation are usually not blinded
Incomplete outcome data (attrition bias)
All outcomes
Low risk

ITT analysis.

Quote: "1 of the 20 patients assigned BEAM treatment withdrew before it was given. This patient was included in analysis of overall survival and event-free survival as a non-responder, but was excluded from the progression-free analysis since no response assessment was possible"

Selective reporting (reporting bias)Unclear riskNo protocol available
Other biasUnclear risk

Early closing of the trial, before accrual was met

Quote: "Recruitment to the trial became increasingly difficult because patients refused randomisation and requested ABMT. It was therefore closed early (40 patients rather than 66 intended)"

HDR1

Methods

Randomisation

  • 2 arms: all patients received 2 cycles of Dexa-BEAM; BEAM plus ASCT (autologous haemopoietic stem cell transplantation) versus additional 2 cycles of Dexa-BEAM

Recruitment period

  • February 1993 to September 1997

Median follow-up time

  • 83 months (range not reported)

Participants

Patients enrolled and randomised to pretreatment (N = 161)

  • BEAM plus ASCT (N = 88); (27 were excluded after randomisation and pretreatment: 6 did not qualify, 1 incomplete data, 6 died, 4 severe infection, 2 protocol violation, 1 patient refused, 7 did not achieve PR)

  • Dexa-BEAM (N = 73); (17 were excluded after randomisation and pretreatment: 7 did not qualify, 3 incomplete data, 2 died, 1 severe infection, 1 protocol violation, 2 patients refused, 1 did not achieve PR)

Chemosensitive patients randomised (N = 117)

  • BEAM plus ASCT (N = 61)

  • Dexa-BEAM (N =56)

Eligibility criteria

  • Patients with advanced HL aged 16-60 years with a biopsy-confirmed relapse

  • Patients who had received primary multiple drug chemotherapy with or without RT

  • Patients with stage I or II at relapse with additional risk factors (bulky mediastinum, involvement of 3 or more lymph-node regions, extranodal disease, or erythrocyte sedimentation rate > 30 mm) or stage III or IV early in the course of disease

  • Karnofsky performance score > 70% and stable cardiac, pulmonary, renal and liver function

  • Only chemosensitive patients (in CR or PR) after 2 courses of Dexa-BEAM were randomised to further therapy

Mean age (117 chemosensitive patients reported)

  • BEAM plus ASCT: 30 years (range 21-57 years); Dexa-BEAM: 34 years (range 16-55 years)

Gender (male) (117 chemosensitive patients reported)

  • BEAM plus ASCT: 39 (64%); Dexa-BEAM 36 (64%)

Stage of disease

  • Stage IA to IIB: BEAM plus ASCT: 18 (30%); Dexa-BEAM: 23 (41%)

  • Stage IIIA to IVB: BEAM plus ASCT: 39 (64%); Dexa-BEAM: 30 (54%)

  • Missing: BEAM plus ASCT: 4 (7%); Dexa-BEAM: 3 (5%)

Stratum

  • Early first relapse: BEAM plus ASCT: 21 (34%); Dexa-BEAM 17 (30%)

  • Late first relapse: BEAM plus ASCT: 29 (48%); Dexa-BEAM 26 (46%)

  • Multiple relapses: BEAM plus ASCT: 11 (18%); Dexa-BEAM 13 (23%)

Countries

  • 58 participating centres from: Czech Republic, Norway, Sweden, Finland, Germany, Switzerland, Italy, and the UK

Interventions

Previous treatment:

  • All patients randomised to BEAM plus ASCT or Dexa-BEAM received prior treatment consisting of 2 cycles of Dexa-BEAM (8 mg/m2 dexamethasone, 60 mg/m2 carmustine, 250 mg/m2 etoposide, 100 mg/m2 cytarabine, and 20 mg/m2 melphalan)

  • All patients received G-CSF on day 8 of Dexa-BEAM, which was continued until leukocyte recovery or until last day of stem-cell harvesting if blood-progenitor cells were gathered

  • Chemosensitive patients in CR or PR after 2 cycles of Dexa-BEAM were randomised further

BEAM plus ASCT

  • BEAM started 4 weeks after WBC count returned to normal after Dexa-BEAM, and consisted of carmustine (300 mg/m2 intravenously on day -7), etoposide (150 mg/m2 intravenously, every 12 h from day -7 to day -4), cytarabine (200 mg/m2, every 12 h from day -7 to day -4), and melphalan (140 mg/m2, on day -3)

  • Bone marrow or progenitor cells were harvested after the second cycle of Dexa-BEAM and chemosensitive patients received transplantation of either autologous bone marrow or peripheral-blood-progenitor cells after high-dose chemotherapy

  • Cryopreserved bone-marrow or peripheral-blood-progenitor cells were infused on day 0 followed by G-CSF until leukocyte recovery

Dexa-BEAM

  • Chemosensitive patients, after the first 2 cycles of Dexa-BEAM, received the third cycle as soon as restaging showed at least partial remission and haemological recovery

  • The fourth cycle Dexa-BEAM was given after haemological recovery from cycle 3

Outcomes

Reported

  • Primary outcome

    • Freedom from treatment failure

  • Secondary outcomes

    • Response rate

    • Complete remission

    • Overall survival

Notes

The scientific committee decided to stop the study because of low accrual of patients after 117 chemosensitive patients had been recruited

There were no conflicts of interest declared

Funding: Deutsche Krebshilfe e.V. and Mildreed Scheel Stiftung

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk

Randomised controlled trial

Quote: "Data managers randomly assigned eligible patients..."

Allocation concealment (selection bias)Low risk

Information about allocation concealment provided

Quote: "Data managers randomly assigned eligible patients at the GHSG trial office by computer before any therapeutic intervention"

Blinding of participants and personnel (performance bias)
All outcomes
High riskNo information provided. Trials evaluating stem cell transplantation are usually not blinded
Blinding of outcome assessment (detection bias)
All outcomes
Unclear riskNo information provided. No information provided.Trials evaluating stem cell transplantation are usually not blinded
Incomplete outcome data (attrition bias)
All outcomes
High riskAnalysis was conducted per protocol; 23% mission data in Dexa-BEAM group and 31% missing data in BEAM plus ASCT group
Selective reporting (reporting bias)Unclear riskNo protocol available
Other biasUnclear riskEarly closure of study because of low accrual of patients after 117 chemosensitive patients had been recruited

HDR2

  1. a

    ABMT: autologous bone marrow transplant; ABVD: doxorubicin plus bleomycin plus vinblastine plus dacarbazine; BEAM; carmustine plus etoposide plus cytarabine plus melphalan: BNLI: British National Lymphoma Investigation; CR: complete response; dexa-BEAM: dexa-carmustine plus etoposide plus cytarabine plus melphalan; G-CSF: granulocyte colony-stimulating factor; GHSG: German Hodgkin Study Group; HL: Hodgkin's lymphoma; ITT: intention to treat; MOPP: mechlorethamine plus vincristine plus procarbazine plus prednisone; PBSCT: peripheral blood stem cell transplantation; PR: partial response; RT: radiotherapy; SHDCT: sequential high-dose chemotherapy; WBC: white blood cell; WHO: World Health Organization.

Methods

Randomisation

  • 2 arms: pre-phase with 2 cycles DHAP for all patients, then random assignment to BEAM and ASCT (autologous transplantation with peripheral blood stem cells) versus SHDCT followed by BEAM plus ASCT (autologous transplantation with peripheral blood stem cells)

Recruitment period

  • December 2000 to December 2006

Median follow-up time

  • 42 months (range not reported)

Participants

Patients randomised (N = 241)

  • 2 cycles of DHAP for all recruited patients (N = 284)

  • BEAM plus ASCT (N = 119); (7 were excluded from study after randomisation: 3 disease progression, 1 patient's wish, 1 toxicity and 2 other reasons)

  • SHDCT plus BEAM plus ASCT (N = 122); (11 were excluded from study after randomisation: 7 patient's wish and 4 other reasons)

Eligibility criteria

  • Patients with histologically confirmed early or late relapsed HL or multiple relapses and no prior ASCT

  • Patients aged 16-60 years

  • Eastern Cooperative Oncology Group performance status ≤ 2

  • Patients who had received primary multiple drug chemotherapy with or without RT

Mean age

  • BEAM plus ASCT: 35.3 years (range not reported); SHDCT plus BEAM plus PBSCT: 35.4 years (range not reported)

Gender (male)

  • BEAM plus PBSCT: 74 (62%); SHDCT plus BEAM plus PBSCT: 83 (68%)

Stage of disease (Ann Arbor stage)

  • Stage IA: BEAM plus ASCT: 17 (14%); SHDCT plus BEAM plus PBSCT: 17 (14%)

  • Stage IB: BEAM plus ASCT: 3 (3%); SHDCT plus BEAM plus PBSCT: 2 (2%)

  • Stage IIA: BEAM plus ASCT: 29 (24%); SHDCT plus BEAM plus PBSCT: 38 (31%)

  • Stage IIB: BEAM plus ASCT: 8 (7%); SHDCT plus BEAM plus PBSCT: 5 (4%)

  • Stage IIIA: BEAM plus ASCT: 21 (18%); SHDCT plus BEAM plus PBSCT: 15 (12%)

  • Stage IIIB: BEAM plus ASCT: 11 (9%); SHDCT plus BEAM plus PBSCT: 14 (11%)

  • Stage IVA: BEAM plus ASCT: 17 (14%); SHDCT plus BEAM plus PBSCT: 17 (14%)

  • Stage IVB: BEAM plus ASCT: 13 (11%); SHDCT plus BEAM plus PBSCT: 13 (11%)

Stratum

  • Early relapse: BEAM plus ASCT: 27 (23%); SHDCT plus BEAM plus PBSCT: 34 (28%)

  • Late relapse: BEAM plus ASCT: 74 (62%); SHDCT plus BEAM plus PBSCT: 68 (56%)

  • Multiple relapses: BEAM plus ASCT: 18 (15%); SHDCT plus BEAM plus PBSCT: 19 (16%)

Countries

  • Multicentre study from: Belgium, Croatia, Denmark, Germany, the Netherlands, Poland, Portugal and Switzerland

Interventions

Previous treatment:

  • All patients received 2 cycles of DHAP (dexamethasone (40 mg intravenously, days 1-4), cytarabine (2 x 2000 mg/m2 intravenously over 3 hours, day 3), cisplatin (100 mg/m2 continuous intravenously over 24 hours, day 1) followed by 10 µg/kg of daily G-CSF until the end of peripheral blood stem cell apheresis)

BEAM plus ASCT:

  • BEAM: carmustine (300 mg/m2 intravenously, over 2 hours, day 37), etoposide (2 x 150 mg/m2 intravenously over 30 minutes, days 37-40), cytarabine (2 x 200 mg/m2 intravenously over 30 minutes, days 37-40), melphalan (140 mg/m2 intravenously over 30 minutes, day 37) followed by ASCT with at least 2 x 106 of CD 34+ peripheral blood stem cells per kilogram body weight on day 42 and G-CSF 5 µg/kg subcutaneously twice a day from day 41 until WBC count was ≥ 3000/µL for 3 days

SHDCT plus BEAM plus PBSCT:

  • SHDCT: cyclophosphamide (4000 mg/m2 intravenously over 8 hours, day 37), prophylactic uromitexane (4000 mg/m2 continuously intravenously days 37-39) followed by G-CSF 5 µg/kg subcutaneously by day 38 until WBC count > 3000/µL for 3 days, followed by high-dose methotrexate (8000 mg/m2 intravenously for 6 hours, day 51) and vincristine (1.4 mg/m2, maximum 2 mg, intravenously, day 51) followed by high-dose etoposide (500 mg/m2 intravenously for 8 hours, days 58-61 with G-CSF 5 µg/kg subcutaneously from day 62 until WBC > 3000/µL for 3 days) followed by BEAM (at same dosages as the other treatment arm) plus ASCT (peripheral blood stem cells were re-infused and G-CSF were administered until haematological recovery)

  • Patients with residual lymphoma (> 1.5 cm on computed tomography scan), received additionally 30 Gy involved-field RT

Outcomes
  • Primary outcome

    • Freedom from treatment failure

  • Secondary outcomes

    • CR rate 100 days after PBSCT

    • Progression-free survival

    • Overall survival

    • WHO grade 3/4 toxicity

    • Secondary neoplasia

Notes

No conflicts of interest were reported

Funding: German Cancer Aid

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk

Randomised controlled trial

Quote: "...patients were centrally randomly assigned..."

Allocation concealment (selection bias)Unclear risk

No information about allocation concealment provided

Quote: "...patients were randomly assigned in the GHSG central trial office..."

Blinding of participants and personnel (performance bias)
All outcomes
High riskNo information provided. Trials evaluating stem cell transplantation are usually not blinded
Blinding of outcome assessment (detection bias)
All outcomes
High riskQuote: "...participating centres were informed without masking..."
Incomplete outcome data (attrition bias)
All outcomes
Low riskAnalyses were performed according to ITT principle
Selective reporting (reporting bias)Low riskProtocol available at clinicaltrials.gov/ct2/show/NCT00025636
Other biasLow riskThe trial seems free of other biases

Characteristics of excluded studies [ordered by study ID]

StudyReason for exclusion
  1. a

    HL: Hodgkin's lymphoma; NHL: non-Hodgkin's lymphoma.

Avigdor 2000Not a randomised controlled trial
Bolanos-Meade 2007No differentiation between HL and NHL
Evens 2007Not a randomised controlled trial
Ferme 2002Not a randomised controlled trial
Gutierrez-Delgado 2003Not a randomised controlled trial
Kuruvilla 2004Not a randomised controlled trial
Morschhauser 2008Not a randomised controlled trial
Sweetenham 1997Not a randomised controlled trial
Van Agthoven 2001Cost analysis
Vellenga 2001No differentiation between HL and NHL
Weaver 1996No differentiation between HL and NHL in outcomes

Ancillary