Türk Çocuk Hematoloji Dergisi

Transkript

1 Türk Çocuk Hematoloji Dergisi Yay n Kurulu Editör Ömer Devecio lu Editör Yard mc lar Tiraje Celkan Duygu Uçkan Çetinkaya Kaan Kavakl Türkan Pat ro lu Sekreter Ferhat Karakafl TPHD YÖNET M KURULU Baflkan: Sema Anak II. Baflkan: Sabri Kemahl G. Sekreter: Gülyüz Öztürk Sayman: Gönül Aydo an Üye: Gülersu rken Üye: Mualla Çetin Üye: M. Akif Yeflilipek Sabri Acar Leyla A ao lu Mehmet Nejat Akar Saadet Akarsu Recep Akda Serap Aksoylar Davut Albayrak Canan Uçar Albayrak Sema Anak Ali Bülent Antmen Yusuf Ziya Aral Gönül Aydo an Hilmi Apak Avni Atay Yeflim Ayd nok Gönül Aydo an Can Balkan Su Gülsün Berrak Betül Biner Özcan Bör Duran Canatan Cengiz Canpolat Tiraje Celkan Ümran Çal flkan Mualla Çetin Nazan Çetingül Duygu Uçkun Çetinkaya Ayhan Da demir Ömer Devecio lu Feride Duru Erol Erduran Mehmet Ertem Tunç F flg n Hüseyin Gülen Fatma Gümrük Ünsal Günay Aytemiz Gürgey Türkiz Gürsel Gülersu rken Savafl Kansoy Mehmet Kantar Ceyda Karadeniz Zeynep Karakafl Kaan Kavakl Rejin Kebudi Sabri Kemahl Yurdanur K l nç Tezer Kutluk Emin Kürekci Ahmet Koç Ülker Koçak Adalet Meral Lale Olcay Nur Olgun Ahmet Faik Öner Hale Ören Nam k Yaflar Özbek Okan Özcan Mehmet Akif Özdemir Ünsal Özgen Hatice Emel Özyürek Gülyüz Öztürk Figen Pekün Türkan Pat ro lu Aziz Polat Nazan Sarper Betül Sevinir Tansu Sipahi Murat Soker Lebriz Yüksel Soycan lgen fiaflmaz Tülin Revüde fiayl Atilla Tanyeli A Murat Tuncer Bahattin Tunç Hayri Bozkurt Toksoy Canan Uçar Leyla Zümrüt Uysal Emel Ünal Selma Ünal Ayflegül Ünüvar Canan Vergin Nevin Yalman Ifl n Yaprak Nefle Yaral Nilgün Yar fl dil Yenicesu Mehmet Akif Yeflilipek nci Y ld z Y ld z Y ld rmak LET fi M ADRESLER ED TÖR Prof. Dr. ÖMER DEVEC O LU - Dergi editörü.ü..t.f. Çocuk Sa. ve Hast. ABD. Pediatrik Hematoloji/ Onkoloji BD fiehremini/çapa/ stanbul Tel: (0212) Fax: (0212) E-Posta: odeveci@istanbul.edu.tr TPHD OF S Türkoca Cad. stanbul Tabip Odas Binas No: 17/6 PK Ca alo lu/eminönü/ stanbul Tel: (0212) Fax: (0212) E-Posta: Tphdofis@yahoo.com Web Adres: Sahibi Türk Pediatrik Hematoloji Derne i nin resmi yay n organ Journal of Turkish Pediatric Hematology Haz rl k ve Bask 3 ayda bir yay mlan r. Y ld z Posta Cad. Sinan Apt. No. 36 D.66/ Gayrettepe- stanbul Tel: Fax: nternet: logos@logos.com.tr

2 YAZARLARA B LG Çocuk Hematoloji dergisi Türk Pediatrik Hematoloji Derne inin resmi yay n organ d r. Çocuk kan ve kanser hastal klar alan ndaki araflt rmalar, ilginç olgular ve davetli derlemeleri yay nlar. Derginin yay n dili Türkçe'dir. Yaz lar n dergide yer alabilmesi için daha önce baflka bir dergide yay nlanmam fl olmas ve yay nlanmak üzere gönderilmemifl olmas gereklidir. Sunulan yaz yay n kurulu taraf ndan belirlenen en az iki hakem taraf ndan de erlendirilir. Son karar dergi yay n kurulunundur. Yay n kurulu hakem kurulu taraf ndan yay n koflullar na uygun bulunmayan yaz lar yay nlamamak, düzeltmek üzere yazar na geri vermek, biçimce düzenlemek ve düzeltmek ya da k saltmak yetkisindedir. Yay n kurulu bilimsel aç dan yeterli görülen bir yaz n n yay nlanmas n etik nedenlerden dolay reddedebilir. Yaz kabul edildi i takdirde bütün bask haklar (copyright) dergiye geçmifl olur. nsan denekleri üzerindeki çal flmalar 1975 Helsinki Bildirgesinin 1983'de düzenlenmifl flekline uygun olmal, her denekten bilgilendirilmifl onay ve lokal Etik Kuruldan izin al nm fl olmal d r. Ayr ca çocuklarla ilgili çal flmalarda belirlenmifl kesin etik standartlar olmad ndan yazarlar Arch Dis Child 1978; 54: 441-2'de bildirilen politikay ve çal flmalar n hukuki yönü için Arch Dis Child 1978; 53: ve Lancet 1977; 2: 754-5'de yay nlanan kriterleri benimsemifltir. Yaz m koflullar için Uniform requirements for manuscripts submitted to biomedical journals bafll alt nda yay nlanan kurallar kabul edilmifltir. Yaz lar n Haz rlanmas Yaz lar orijinali ve 2 fotokopi olmak üzere 3 nüsha fleklinde gönderilmelidir. Beraberinde tüm yaz ve flekilleri içeren bir disket de bulunmal d r. Disket içeri inin yaz s Times New Roman karakteriyle ve 10 punto büyüklü üyle yaz lmal d r. Düzeltme istenen yaz larda, yaz n n düzeltilmifl halini içeren bir disket de tekrar ayr ca gönderilmelidir. Derlemeler en fazla 10, araflt rmalar ise 7 sayfa olmal, olgu bildirileri ise 5 sayfay geçmemelidir. Yaz lar bir baflvuru mektubu ile gönderilmeli ve bu mektubun sonunda tüm yazarlar n imzas bulunmal d r. Yaz lar n sorumlulu u yazarlara aittir. Yaz lar standart dosya ka d na daktilo veya bilgisayar ile, sayfan n sadece bir yüzüne olacak flekilde, iki aral kl olarak yaz lmal, sayfan n her iki kenar nda yaklafl k üçer santim boflluk b rak lmal d r. Yaz lar, bafll k sayfas ndan bafllamak üzere ve sa alt köflede olacak flekilde numaraland r lmal d r. Yaz larda bulunmas gereken k s mlar flunlard r: A) Araflt rmalar için: 1-Bafll k sayfas, 2-Türkçe özet, 3- ngilizce özet, 4-Girifl, 5-Gereç ve Yöntem, 6-Bulgular, 7-Tart flma, 8-Kaynaklar, 9-Tablolar, 10-fiekil ve Foto raflar B) Olgu bildirileri için: 1-Bafll k sayfas, 2-Türkçe özet, 3- ngilizce özet, 4-Girifl, 5-Olgu bildirisi, 6-Tart flma, 7-Kaynaklar, 8-Tablolar, 9-fiekil ve Foto raflar. Bafll k sayfas : Çal flman n ad Türkçe ve ngilizce, büyük harflerle olacak flekilde alt alta yaz lmal d r. Yazarlar n aç k ad (küçük harf)-soyad (büyük harf), ünvanlar, çal flt klar kurum ve çal flman n yap ld kurum belirtilmelidir. Yaz ile ilgili bilgi (kongrede sunulmufl olmas, herhangi bir kurum deste i, ald ödül vb) varsa belirtilmelidir. Yaz fl lacak yazar n posta adresi ile telefon, faks ve e-posta adresleri yaz lmal d r. Özet: Bafll k sayfas ndan sonra, ayr bir sayfada araflt rma ve derlemeler için en fazla 250 kelimeyi, olgu bildirileri için 100 kelimeyi aflmayan bir özet bulunmal ; özetin amaç, gereç-yöntem, bulgular ve ç kar mlar alt bafll klar olmal d r. Türkçe özetin alt nda ayn düzendeki ngilizce özet yer almal d r. ngilizce özette aim, methods, results, conclusion alt bafll klar olmal d r. Her yaz da özetlerin hemen alt nda Türkçe ve ngilizce 3-10 kelime aras nda anahtar kelime verilmeli, anahtar kelimeler Index Medicus'un konu bafll klar na uygun olmal ve ilk harflerine uygun alfabetik s ralanmal d r. Ana Metin: Girifl, gereç ve yöntem, bulgular ve tart flma bölümlerinden oluflur. Varsa teflekkür yaz s, kaynaklardan önce konulmal d r. Olgu sunumlar : 100 kelimeyi aflmayan bir Türkçe ve ngilizce özetten sonra, en az 3 kelimelik bir anahtar kelimeler k sm içermelidir. Ana metin, olgu/ olgular, tart flma, kaynaklar bölümlerinden oluflmal d r. fiekil ve Foto raflar: Siyah-beyaz ve parlak ka da bas lm fl olmal ve ayr bir zarf içinde herhangi bir ka da yap flt r lmadan gönderilmelidir. Grafikler ve teknik resimler çini mürekkebi ile ayd nger ka d na veya beyaz ka da çizilmeli, ya da bilgisayarda yap lmal d r. fiekil ve foto raflar da üç nüsha fleklinde gönderilmelidir. Her fleklin arkas na fleklin yönü, numaras ve ilk yazar n ismi kurflun kalemle yaz lmal d r. fiekillerin alt yaz lar ayr bir ka da, flekil numaras bildirilerek yaz lmal ve flekil numaralar metin içinde mutlaka belirtilmelidir. Mikroskopik resimlerde büyütülme oran ve boyama tekni i aç klanmal d r. Tablolar: Ayr bir ka da çift aral kl olarak yaz lmal, tablo içinde enine ve boyuna bölme çizgileri kullan lmamal d r. Her tablonun üzerine numara (romen rakkam ile) ve bafll k yaz lmal d r. Tablo numaralar metin içinde mutlaka kullan lmal - d r. Kaynaklar n yaz l fl : Kaynaklar yaz da geçifl s ras na göre numaraland r lmal ve numaralar parantez içinde olarak, cümle sonunda veya cümle içinde araflt r c ismi varsa bu isimden hemen sonra yaz lmal - d r. Dergi isimleri Index Medicus'a uygun olarak k salt lmal d r. Yazar isimleri alt y geçiyorsa, ilk üç isim yaz lmal ve sonunda Türkçe kaynaklarda "ve ark.", yabanc dildeki kaynaklarda "et al" kelimeleri eklenmelidir. Örnek: Pizzo PA, Rubin M, Freifeld A, Walsh TJ. Child with cancer and infection. II. Nonbacterial infections. J Pediatr 1991; 119: Yazar verilmemiflse kaynak afla daki gibi yaz lmal d r. Örnek: Coffee drinking and cancer of the pancreas (editorial). BMJ 1981; 283: 628.

3 Kaynak kitap ise afla daki flekilde yaz lmal d r. Örnek: Arceci RJ, Hann IM, Smith OP. Pediatic Hematology 3rd ed. Massachusetts: Blackwell, 2006: Kaynak kitap içinde bölüme ait ise afla daki flekilde yaz lmal d r. Örnek: Shahlaee AH, Arceci RJ. Histiocytic disorders In: Arceci RJ, Hann IM, Smith OP (eds). Pediatic Hematology third ed. Massachusetts:Blackwell, 2006: K saltmalar Bafll k ve özette k saltma kullan lmamal d r. Metin içinde k - saltma ilk kez kullan ld nda önüne kelime veya kelimeler grubunun tümü aç k olarak yaz lm fl bulunmal d r. Yaz flma adresi: Prof. Dr. Ömer Devecio lu.ü..t.f. Çocuk Sa. ve Hast. ABD. Pediatrik Hematoloji/Onkoloji BD fiehremini/çapa/ stanbul Tel: (0212) Fax: (0212) E-Posta: odeveci@istanbul.edu.tr Yazarlar için"kontrol Listesi": 1) Tüm yazarlar n imzalad baflvuru mektubu 2) Yaz n n 3 kopyas (bafll k sayfas ndan itibaren numaraland - r lmal d r) * Bafll k sayfas (Çal flman n Türkçe ve ngilizce ad, yazarlar n aç k ad, ünvanlar, çal flt klar kurum ve çal flman n yap ld kurum, yaz fl lacak yazar n posta adresi, telefon ve faks numaras, adresi) * Türkçe ve ngilizce özet (anahtar kelimeleri ile birlikte) * Girifl * Gereç ve Yöntemler * Bulgular * Tart flma * Kaynaklar * Tablolar * fiekil ve foto raflar 3) Yaz y flekil ve tablolar ile birlikte içeren bir disket (disket üzerinde çal flman n ad ve ilk yazar n ad bulunmal d r).

4 Türk Çocuk Hematoloji Dergisi Ç NDEK LER / 1 ALL de Relaps Sa alt m ve Yeni Yaklafl mlar Guenter HENZE... Adolesan ALL de Sa alt m David WEBB... Philadelphia Pozitif ALL Sa alt m nda Yenilikler Guenter HENZE... Çocukluk Ça AML Sa alt m ve UK-MRC Sonuçlar David WEBB... Demir fielasyon Tedavi Amaçlar ve Tedavi Etkinli inin zlenimi Chaim HERSKO... Immunfenotiplendirme Türkan PATIRO LU... Sitogenetik ve Moleküler Çal flmalar n Ak lc Planlanmas Tahsin YAKUT... Minimal Rezidüel Hastal k Araflt r lmas n n Lösemi zleminde Yeri Duygu Uçkan ÇET NKAYA... St. Jude ALL Sonuçlar Mualla ÇET N... COG-ALL Sonuçlar Ayflegül ÜNÜVAR... BFM-ALL Sonuçlar Hale ÖREN... Çocukluk Ça ALL Tedavisinde COG Protokolleri Leyla A AO LU... BFM ALL Protokolleri Lebriz YÜKSEL-SOYCAN Türk Pediatrik Hematoloji Dergisi üç ayda bir (y lda 4 say ) olarak yay nlan r. Bu dergide yay mlanan yaz lar n telif haklar Logos Yay nc l k Tic. A.fi. ye ait olup, yay nc n n yaz l izni olmadan hiçbir flekilde tümü veya herhangi bir bölümü kopya edilemez, herhangi bir dilde tamamen veya k smen yay nlanamaz. Bu dergi Acid Free (Alkali) ka da bas lmaktad r. / This journal is printed on Acid-Free paper

5 Türk Çocuk Hematoloji Dergisi Ç NDEK LER / 2 Lösemi Verileri ve KHT Endikasyonlar Akif YEfi L PEK... KHT Verileri Savafl KANSOY... BFM-AML Sa alt m Sonuçlar Hilmi APAK... MRC-AML Sa alt m Sonuçlar Adalet Meral GÜNEfi... Neonatal Hemofagositik Lenfohistiyositoz Aytemiz GÜRGEY... Langherhans Hücreli Histiyositoz ve Santral Sinir Sistemi Ömer DEVEC O LU... Yeni Demir fielatörü; ICL670 (Exjade) Yeflim AYDINOK... Medikal Sa alt m Zafer fialcio LU... Fizyopatoloji ve Klinik Yaklafl m Berna ATABAY... Radyoizotop Sinovektomi Kaan KAVAKLI... Konjenital Nötropeni Nam k ÖZBEK... Edinsel Nötropeni Tansu S PAH... Pediatrik Olgularda Sorunlu Transfüzyonlar Sabri KEMAHLI... Transfüzyon Yan t n n De erlendirilmesi Gülyüz ÖZTÜRK

6 ED TÖRDEN Sevgili Türk Pediatrik Hematoloji Derne i Üyeleri, lklerin her zaman yaflam m zda özel bir yeri vard r. Bir a lama ile bafllayan yaflam serüvenimizde hep ilkler sorgulanmaz m? lk a lamam z için saniyelerin bile önemi yok mudur? Annemizin ilk sütü çok özel de il midir? Hayata geldi imiz o ilk günü sonra ölene kadar do um günü olarak kutlam yor muyuz? lk diflin, ilk ad m n, ilk cümlenin kayd n hep bir yerlere yazmaz m y z? lk bisikletin, ilk ö retmenin, ergenli e girerken ilk sevgilinin, üniversitedeki ilk günün, ihtisastaki ilk nöbetin an lar m zdaki yeri hep çok farkl de il midir? Elinizdeki say iflte böyle bir ilk. Derne imiz yönetim kurulunun ayl k bir dergi ç karma karar n n ilk ürünü. Hemen burada bu zor ve onurlu görevi bana ve editör yard mc s arkadafllar - ma veren yönetim kurulumuza çok teflekkür etmek istiyorum. K sac k öykümüze gelince! Görevi ald ktan ve editör yard mc s arkadafllar m da ö rendikten sonra, Bursa kongresine yetiflebilmek için çok acele davranmam z gerekti ini düflündük. lk ifl olarak hakem kurulumuzu oluflturduk. Bunun için profesör, doçent, flef ve flef yard mc s arkadafllar m z kurula katt k. lk say y kongre konuflmalar yla, ikinci say y da kurs konuflmalar ile oluflturmay planlad k. Böylece 2007 y l n n iki say s tamamlanm fl oldu, di er say lar içinde yeterli zaman aral olufltu. Bu arada renk ve d fl kapak görünümü için hepimizin üyesi oldu u Türk Hematoloji Derne inin yay n organ ndan esinlendik. Kapak resmi Op. Dr. Bekir Tu cu ya aittir. Kendisine ayr ca teflekkür ediyoruz. Dergimiz üç ayda bir olmak üzere y lda dört adet ç kacakt r. Her say bir derleme, 3-4 adet araflt rma yaz s, 3-4 adet olgu sunumu, editöre mektuplar, tan n z nedir, kongre haberleri ve reklamlardan oluflacakt r. fiimdiden 2007 y l n n üçüncü ve dördüncü say lar için yaz lar n z bekliyoruz. Hepinize iyi kongreler dileklerimizle. Prof. Dr. Ömer Devecio lu

7 BAfiLARKEN Türk Pediatrik Hematoloji Derne i de erli üyeleri, de erli meslektafllar m, Türk Pediatrik Hematoloji Derne i kuruldu u 1999 y l ndan beri Pediatrik Hematoloji alan nda pek çok sorunun çözümünde giriflimlerde bulunmufl, pek çok bilimsel aktiviteye imza atm fl ve en önemlisi de e itim konusunda tüm Türkiye çap nda programlar n sürdürmüfltür. Bu süreçte görev alm fl tüm hocalar ma ve arkadafllar ma teflekkürü bir borç bilirim. Bu süren bayrak yar fl nda 2005 y l ndan beri görevini sürdüren yönetim kurulumuz da önce e itim ilkesinin yans malar olarak May s 2007'de Bursa'da yapaca m z VII. Ulusal Pediatrik Hematoloji Kongresi'nde pek çok ilki gerçeklefltirme arzusundad r. Bunlardan ilki, belki de en önemlisi, Pediatrik Hematoloji Dergisi dir. Bilimde üretkenli in zirveye ç kt 2000'li y llarda bu ilerlemeyi yans tman n en pozitif yolu yap lan çal flmalar n ka da dökülmesi veya yeni anlay flla bilgisayar ortam nda kayda geçirilmesidir. Bu amaçla kullan ma giren web sitemizden sonra siz üyelerimizin ve konuyla ilgili tüm bilim insanlar - n n çal flmalar n derlemelerini ve olgu sunular n yay nlayabilece i bir dergiyi ç karma karar ald k. lk 2 say s nda kongremizle ilgili konuflma metinlerinin ve kurs metinlerinin, ayr ca ek olarak bildiri özetlerinin yer alaca mecmuam z n h zla yap land r lmas nda ve ortaya ç - kar lmas nda büyük emekleri geçen editörümüz Prof. Dr. Ömer Devecio lu'na ve editör yard mc lar na, Logos Yay nevine ve eme i geçen tüm arkadafllar m za teflekkürü bir borç bilirim. Bilim dünyas na katk s n n olmas ümidiyle sayg lar m sunar m. Prof. Dr. S. Sema Anak Baflkan TPHD Yönetim Kurulu ad na

8 TREATMENT OF RELAPSED ALL AND RECENT APPROACHES Prof. Dr. Günter HENZE Berlin Charite University, Pediatric Hematology Department Despite high cure rates relapse of ALL is still a frequent diagnosis in pediatric oncology. Even with the best available treatment regimes relapse rates are still in the range of 15 to 20 %. Because of risk adapted front-line protocols relapses occur widely per random, but at a higher frequency in patients with high-risk for relapse features. Since stem cell transplantation (SCT) has increasingly become a regular part of the first-line treatment relapses occur also in patients who had already been treated with SCT, and these children are rather not eligible for retreatment according to common relapse protocols. These are the reasons why patient numbers in relapse trials have decreased over the past years. In the multicentric international ALL relapse trials of the BFM Relapse Study Group we have registered more than 2000 patients since The probability of event-free survival is about 33%, and the overall survival is about 40% for patients who have been treated within one of the 6 consecutive trials. These figures show that the prognosis and the chance of longterm survival are substantially reduced in case of relapse. The lower remission rates are due to a higher rate of induction deaths as a consequence of reduced tolerance to treatment after previous organ toxic frontline therapy. In addition, non-responses to treatment are common, since surviving blast cells had the chance to develop resistance towards commonly used antileukemic drugs. Even if a complete remission (CR) can be achieved, the rate of subsequent relapses is high. Time, site, and the immunological phenotype of the leukemic cells are the most important clinical predictors of prognosis. Early time point of relapse is associated with a worse prognosis; even if a CR can be achieved it cannot be maintained with chemotherapy, and effective salvage therapy has therefore to include SCT for the majority of children. Surprisingly, children with a combined bone marrow (BM) relapse have a better prognosis than the ones with an isolated BM relapse. Children with a relapse of B-cell precursor ALL have a better outcome as compared with T-cell leukemias. In addition to these clinical parameters, genetic and molecular genetic findings have been shown to have an important impact on outcome. Fusion genes can be found at about the same frequency as in front-line studies; however the percentage of Philadelphia chromosome (Ph1) positive ALL relapse has become less due to the fact that most of the Ph1-positive leukemias are diagnosed at first presentation, and stem cell transplantation is being performed in the majority of these children. Whereas, like during front-line therapy, the TEL-AML1 positive patients have a better prognosis also after relapse, prognosis of Ph1-positive ALL relapse is dismal. In recent years, another factor has gained increasing importance for prognostication: Like in 8

9 front-line trials response to therapy is one of the most significant single parameters determining the outcome. In the current BFM relapse trial ALL-REZ BFM 2002, we use response to therapy as measured by quantification of minimal residual disease based on molecular genetics for stratification of intermediate risk patients. Good responding patients will be maintained on chemotherapy whereas patients with a high MRD level at day 36 of treatment are being referred to SCT. Response to therapy determines also to a certain extent the choice of a stem cell donor. SCT from matched unrelated donors should only be performed in poor responders whereas transplants from siblings may also be performed in good responders because the transplant related mortality has become very low in this setting. Currently, we are considering to design targeted intensification therapies for patients with poor prognosis features. These are either patients with an anticipated low CR rate or children with a high MRD signal prior to SCT. Thus, remission rates should be improved by more effective induction regimens. Furthermore, attempts will be made to improve the pretransplant remission quality in order to reduce the relapse rate after SCT in children with a higher residual leukemic cell load prior to SCT. For these purposes, immunological appro aches (antibodies) or the use of new drugs, e.g., novel purine nucleoside analogues may be used. Such therapies should however be subject of controlled and prospective clinical trials. 9

10 TREATMENT OF ACUTE LYMPHOBLAST C LEUKAEM A IN ADOLESCENTS AND YOUNG ADULTS Prof. Dr. David KH WEBB Great Ormond Street Hospital for Sick Children Hematology Department, United Kingdom Age is an important prognostic factor in the outcome of treatment for acute lymphoblastic leukaemia (ALL). Current therapies have resulted in 80% event free survival (EFS) for children 1-10 years of age, whilst less than 40% of adults are cured. The success story of therapy in childhood ALL has not translated to adolescents, a population at present potentially eligible for treatment on either paediatric or adult protocols. Their outcome has continued to lag behind the paediatric subgroup with a 6-year EFS of 59% reported in Adolescents constitute only a small proportion of the population in both paediatric and adult ALL trials. Historically they have generally been treated in either a paediatric or an adult trial based on referral patterns rather than by selection on patient or disease characteristics. The studies analysing the outcome of adolescents treated on paediatric or adult trials have consistently shown better survival on paediatric trials. Despite a generally similar remission rate this has not translated into similar event-free survival (EFS) Paediatric trials have consistently shown a significantly better outcome for children 1-9 years of age than those over 10 years of age. In the last decade the 5 year EFS for the age group of over 10 years, was between 52.5% and 79%. There was a wide heterogeneity in treatment protocols, and in the age range of adolescent groups. Outcomes reported for groups aged under 20 years were 38% DFS at 8 years, 50% survival at 7 years, and 43% survival at 5 years. The consistent difference in the 5 year EFS (15-35%), favouring patients treated on paediatric trials, which is observed even with similarly matched major adverse prognostic factors, may be due to many factors. These include differing protocol design, drugs administered, the patterns of referral, compliance with treatment, and psychosocial and supportive care. In most paediatric trials adolescents are treated in the high risk group whereas in adult trials they are regarded as low risk. Overall the paediatric protocols use more vincristine, steroids and L - Asparaginase but lower amounts of alkylating agents, high dose cytarabine and anthracyclines than the adult trials. The paediatric trials use a BFM based model with intensive induction and consolidation phase followed by interim maintenance and 1 or 2 delayed intensifications. Adult trials are based on the perception that older patients have significantly less tolerance to corticosdteroids, asparaginase and vincristine compared to children. Whether or not this low tolerance is applicable to the adolescent patients who are being treated along with the remaining adults in the adult trials is a contentious issue; although some toxicities appear more common with increasing age, it is often appropriate to trade this against maximum efficacy. In the adult trials due to the poorer general outcome, allogeneic and autologous bone marrow transplantation is usually attempted in first remission, in spite of the associated increased toxicity compared with chemotherapy. Allogeneic transplantation 10

11 from matched sibling donors in first remission is of proven benefit only in a small subset of Ph+ ALL in the paediatric protocols. Whilst ALL constitutes the commonest malignancy in childhood, the incidence steadily declines with age and accounts for only 6% of adolescent cancers compared to 30% in the children younger than 15 years. Hence adolescents tend to be treated on multicentre trials in paediatric oncology centres; whilst in the United States only 25-40% of adolescents treated on adult trials are managed by community based adult oncologists. What may be more critical is that ALL is uncommon in adults and in some countries is often managed by physicians, who may have less well defined leukaemia programs. In the USA, more than 90% of children less than 15 years of age with cancer are managed at institutions participating in National Cancer Institute sponsored clinical trials. This contrasts with only about 20% of year olds with cancer seen in such institutions. In their National Cancer Database report, the year old patients treated in the CCG and POG sites with ALL and other solid tumours had a better 5-year survival rate than those treated elsewhere. A similar observation emerged when year olds in the Canadian Childhood Cancer Surveillance and Control program were evaluated Results from the studies that have evaluated treatment burden on the various age groups have shown a higher incidence of morbidity in induction and more post remission deaths in adolescents than in the 1-9 years age group of children and more also in adult trials than in the paediatric trials. The contemporary paediatric trials, MRC UKALL X, AIEOP, ALL-BFM 90, DFCI 87-01, 91-01, SJCRH XI-XIV, EORTC 58881, have reported deaths during induction and in first CR to be between % whilst the adult trials CALGB 9111, Hyper CVAD trials with more chemotherapy, have recorded 6-12% death rates. This counterbalanced any reduction of relapse rates which might have occurred with intensification of treatment. These deaths were predominantly infective in origin, especially fungal particularly related to dexamethasone, increased mortality with more anthracyclines and neutropenic episodes following high dose cytarabine in the adult trials attempting to intensify therapy. Changes such as the more potent form of schedules of asparaginase therapy and the substitution of dexamethasone for prednisolone contributed to a higher incidence of sepsis (6.5%) and death due to toxicity (11%) during induction therapy in two clinical trials. Adolescents have a higher incidence of steroid and L-asparaginase induced diabetes mellitus compared to younger children. Pancreatitis is also more common in adolescents. Avascular necrosis of bone (AVN) or osteonecrosis of bone (ON) is becoming an increasingly recognised toxicity in the patients more than 10 years of age. The overall 5 year cumulative incidence of ON was 1.8% for patients below 18 years of age, who were treated according to ALL- BFM95 protocol between 1996 and This incidence was significantly higher for patients over 10 years of age (0.2 % versus 8.9%) and over 15 years of age (16.7%, p=0.003). 11

12 TREATMENT OF PHILADELPHIA POSITIVE ALL AND RECENT DEVELOPMENTS Prof. Dr. Günter HENZE Berlin Charite University, Pediatric Hematology Department The observation of the Philadelphia chromosome as the first constant chromosomal aberration and its association with chronic myelogenous leukemia dates back to In 1973, it could be shown that the Philadelphia chromosome, a shortened chromosome 22, also referred to as Ph1 is the result of the balanced translocation t(9;22)(q34;q11). The translocation the human homologue of the Abelson murine leukemia virus protooncogene abl from chromosome 9 to chromosome 22 next to the bcr gene leads to the formation of the fusion gene BCR-ABL. BCR-ABL encodes a tyrosine kinase which is responsible for the leukemic transformation. The Philadelphia chromosome can also be found in acute lymphoblastic leukemias. In child hood ALL it occurs at a frequency of about 3% to 5%. The presence of the Ph1 is one of the most unfavorable prognostic factors in childhood ALL. Only about one third of the patients can be cured with chemotherapy, in particular those who have a good initial respose to treatment (good prednisone response). The prognosis is significantly better with allogeneic SCT. In a cooperative analysis of the BFM and the AIEOP (Italy) group, the event-free survival was over 80% if the children had received SCT from sibling donors. In recent years the outcome could be improved with trans plants from unrelated donors as well. Thus, in general SCT from a matched donor is indicated in patients with Ph1-positive ALL. The BCR-ABL encoded tyrosine kinase can be inhibited with a compound that has been designed as a specific inhibitor: STI 571, now better known as imatinib mesylate or Gleevec. Imatinib has been developed for chronic myelogenous leukemia with the 210kD fusion protein and has been found to be very effective for the affected patients. In particular, elderly patients who are no more eligible for transplants can be treated, and long-term control of the leukemia can be achieved while maintaining a very good quality of life. Imatinib is also active in Ph1-positive ALL in which the fusion protein is shorter (190kD). However, during recent years it turned out that the duration of the specific effect is limited because the leukemic cells develop resistance against the drug. This has led to the development of new tyrosine kinase inhibitors, like dasatinib, which are capable of inhibiting multiple tyrosine kinases, also receptor tyrosine kinases. The role of the tyrosine kinases for the treatment of childhood ALL has yet to be determined. A recent international trial is aiming at exploring the role of imatinib in context with BFM-type chemotherapy. Another approach to treat Ph1-positive leukemias could be immune therapy. In animal models the onset of leukemia could be delayed or even prevented by vaccination with gene mo- 12

13 dified (B7.1 or granulocyte macrophage colony-stimulating factor expressing) BCR-ABL positive leukemic cells. Thus, there is evidence that the Philadelphia chromosome can be used as an immunologic target in order to elimate residual leukemic cells or even to provide protection against this prognostic unfavorable subset of leukemia. 13

14 TREATMENT OF CHILDHOOD AML, AND RESULTS OF UNITED KINGDOM MEDICAL RESEARCH COUNCIL PROTOCOLS Prof. Dr. David KH WEBB Great Ormond Street Hospital for Sick Children Hematology Department, United Kingdom De novo AML comprises 5% of childhood cancer. The age-specific incidence is highest in young children, and 20% of cases occur in the first 5 years of life, after which the incidence is essentially stable throughout childhood. Although the aetiology of most cases is unknown, AML is more common in children with Down syndrome, Klinefelter syndrome, Bloom syndrome, Neurofibromatosis type I, ataxia telangectasia, and congenital bone marrow failure disorders (Fanconi anaemia, Severe Congenital Neutropenia, Shwachman syndrome, Diamond-Blackfan anaemia, Thrombocytopenia-absent radius syndrome). Secondary AML was first described in patients who had received alkylating agents, and more recently after highdose chemoradiotherapy and infusion of autologous bone marrow, but is rare in children. Alkylating agent-induced MDS/AML tends to occur after 4-5 years, is associated with a preceding phase of MDS and is characterized by deletions from chromosomes 5 and 7. This disease is usually refractory to chemotherapy. The second type is related to treatment with topoisomerase II inhibitors and is associated with abnormalities of chromosome 11q23, a shorter induction period and presentation as acute leukemia without a preceding MDS. More recently, a third group of therapy-related leukemias has been described in association with t(8;21), inv(16) and t(8;16) and t(15;17) after topoisomerase inhibitor therapy, alkylating agent therapy or anthracycline therapy. These do not usually have a dysplastic prodrome and the risk factors have not been so precisely defined, but there is a similar correlation between the cytogenetic findings, morphology and response to treatment, as in de novo AML, so that patients may respond to chemotherapy. The overall risk of leukaemia is increased fold in children with Down syndrome (DS), and is especially marked in the first 5 years of life, when it is 50 times that in normal children. The spectrum of myeloid disease includes a usually self-limiting myeloproliferative disorder of the newborn called transient abnormal myelopoiesis (TAM) or transient leukaemia (TL), myelodysplasia (MDS), and acute myeloid leukaemia (AML). MDS and AML in DS are closely linked with biological and clinical features distinct from the diseases in non - DS children, and are now recognised as a single specific entity, Myeloid leukaemia of Down syndrome (ML - DS). Although there are insufficient data to be precise regarding the incidence of TAM in DS, best estimates suggest 10-20% of all DS babies may be affected. The leukaemic blasts in TAM and ML-DS are typically, but not exclusively, classified as French American British (FAB) group M7. Recently, GATA-1 mutations have been described in DNA extracted from blast cells in DS children with TAM, and ML-DS, and are already considered pathognomonic. GATA-1 is a transcription factor necessary for the development of normal megakaryocytes, erythrocytes, eosinophils and mast cells, and in animal knockout models, absence of GATA-1 results in increased megakaryocyte proliferation with megakaryocyte 14

15 dysplasia, differentiation arrest and apoptosis in erythroid progenitors. As TAM is self-limiting, further unidentified genetic changes are considered necessary to produce the ML-DS phenotype. Gene dosage or disomy of a leukaemia predisposition gene or genes regulating haemopoiesis on chromosome 21 is a presumed mechanism of leukaemogenesis in DS, although there is a lack of clear evidence to favour this hypothesis. Several groups of children with AML are at high risk of early life threatening complications. Patients with acute promyelocytic leukaemia (APL) may have a severe coagulopathy,due to the release of cytoplasmic granules with procoagulant activity, with life threatening bleeding or thrombosis. High count AML,especially monocytic disease, carries risk of hyperviscosity syndrome, haemorrhage and thrombosis in the CNS, and pulmonary infiltrates and respiratory insufficiency. Release of lysozyme from monoblasts causes renal potassium loss, and hypokalaemia may be profound. Monocytic AML is also associated with tumour lysis syndrome and coagulopathy. As a result, treatment is a medical emergency, and rapid white cell count reduction is important in symptomatic high count patients. Anthracyclines are the most useful drugs in this regard and should be given on day 1 of therapy. At diagnosis, bone marrow cytogenetics should be obtained in every case as they provide major information on prognosis. In AML, favourable karyotypes comprise the t(15;17) characteristic of APL, t(8;21) and inversion of chromosome 16. Adverse karyotypes include monosomy 7, deletions of 5q, abnormal 3q, t(9;22), and complex karyotypes with 3 or more abnormalities. All remaining karyotypes are intermediate risk. Inversion 16 and t(8;21) involve genes encoding subunits of the core binding factor (CBF). The t(8;21) generates a novel fusion protein AML1/ETO, causing abnormal protein-protein signalling. The AML1/ETO interacts with CBF but recruits proteins involved in repression of transcription. Inversion 16 creates another fusion protein, CBF /MYHII which also disrupts transcription via CBF by inhibiting binding of AML1. Accordingly the two karyotypic changes affect a common transcriptional pathway, leading to their description as CBF leukaemias. The CBF pathway is under investigation for targets for new therapies. The match between FAB types and CBF leukaemias is not exact, but most children with t(8;21) have FAB type M1 or M2, whilst most with inversion 16 have M4 with eosinophilia (M4Eo). The t(15;17) is tightly associated with APL, and is discussed in detail in the APL section of this chapter. In AML, abnormalities of 11q with a breakpoint at 11q23, the site of the MLL gene, are particularly but not exclusively associated with monocytic leukaemia. These mutations disrupt the role of the MLL gene in transcriptional regulation. A large number of partner genes are observed, and it is possible that the different fusion genes have varying impact on prognosis - for example, several groups report a better outcome for children with the t(9;11). However, these findings are not consistent across studies, and require further elucidation, especially as the influence of prognostic factors may be protocol dependent. Adverse karyotypes are associated with a reduced complete remission rate and increased relapse risk, with overall survival of less than 50%. Monosomy 7 carries a poor prognosis in AML, and occurs as a single abnormality, or as part of a more complex karyotype. Abnormalities of 5q, 3q and the t(9;22) are adverse risk factors in adults, but are very rare in children with AML, and abnormal 3q is not adverse in children treated on AML 10 and AML 12. Molecular characterisation of AML has gained increasing importance, as a number of molecular markers may be used to refine the assessment of prognosis, and may be molecular targets. 15

16 Modern AML therapy is based on induction chemotherapy with an anthracycline combined with cytarabine, although schedules vary in choice of anthracycline (usually daunorubicin, idarubicin or mitozantrone), total doses and methods of administration. There is no convincing evidence for superiority of any anthracycline over daunorubicin. Most schedules include a third induction agent, although the superiority of additional agents compared with anthracycline and cytarabine alone has not been shown in a randomised study. With this combination 80-90% of children achieve complete remission after 2 courses. Induction death rates are in the region of 5%, with the remainder of induction failures due to resistant disease. Consolidation chemotherapy is necessary for cure, although the number of courses given varies between collaborative groups, and the optimum number within each regimen remains unknown. Consolidation courses usually contain high dose cytarabine, and it is unknown whether the addition of other drugs improves efficacy. This issue is being addressed in the MRC AML 15 study. Involvement of the central nervous system at diagnosis occurs in 2% of children, and is associated with monocytic leukaemia and AML in infants. The importance of CNS directed therapy for the majority of children is unclear as CNS relapse is uncommon, and affects around 3% of children. A limited number of intrathecal injections with combinations of methotrexate, cytarabine and hydrocortisone are generally administered routinely to children who are CNS negative at diagnosis. There is no role for prophylactic cranial radiation, although this is included in BFM protocols because of a reduction in bone marrow relapse observed in a subgroup of patients in BFM 87. AML chemotherapy is intensive, myelosuppressive, and associated with mucositis which may be severe. Treatment related death rates of up to 20% have been described, but are now typically under 10%. Accordingly, supportive care needs are high, and therapy should only be administered in experienced units. It is standard practice to keep children in hospital until bone marrow recovery occurs. Planned progressive therapy is required for febrile neutropenia, with early introduction of empiric antifungal drugs if fever persists. Since the early introduction of broad spectrum antibiotics for fever has become standard, gram negative sepsis has been better controlled and most infection related deaths are now fungal, although occasionally viral infections (cytomegalovirus, adenovirus, respiratory syncitial virus) have proved fatal. Stem cell transplant (SCT) now has a very limited role in first line AML therapy, although this area remains contentious. There is no evidence that SCT improves outcome for children with adverse cytogenetics, and SCT is contraindicated in those with favourable cytogenetics due to the high cure rate with chemotherapy, and the toxicity associated with SCT (up to 20% treatment related deaths in most series). The situation is more complex in patients with intermediate karyotypes; although SCT may reduce relapse, the treatment related death rate means that the overall survival benefit is small and European groups have generally discontinued SCT in first remission for these children. However, SCT continues to be considered in the United States for intermediate and poor risk cases. Prognostic factors are to an extent protocol dependant but include age (worse in adolescents and young adults), presenting white cell count (worse with high counts), presence of Down syndrome (better in DS), FAB type (M3 superior), cytogenetics, race (worse in African Americans treated in the United States), response to therapy (patients with more than 15% blasts in the bone marrow on recovery from course 1 have an adverse prognosis), performance score (worse with low score) and some molecular markers. Of these prognostic factors, cytogenetics, M3 FAB type, response to therapy, Down syndrome and internal tandem duplications 16

17 of the fms-like tyrosine kinase (FLT3-ITD) are the most important on multivariate analyses. Overall survival (OS) and event free survival (EFS) range from 25-66% and 47-56% respectively in major collaborative studies (table 1), with relapse rates of 30-50% the main cause of treatment failure. Although survival rates are broadly similar between leading collaborative groups, cumulative doses of anthracycline and cytarabine vary considerably (table 2).Salvage therapy cures around 25% of children who relapse, and the main prognostic factors at relapse are length of first remission and cytogenetics. Children with favourable cytogenetics usually have long first remissions and the best outcome with retreatment, whilst those with adverse cytogenetics relapse early and are very difficult to cure. A wide range of standard AML chemotherapy regimens are satisfactory for reinduction of remission, achieving remission rates of 60-70% overall and with no clear differences in efficacy between them. In the United Kingdom high dose cytarabine and fludarabine (FLA) has been favoured in recent years due to the high anthracycline dosage used in first line therapy. Consolidation is usually by SCT, either from a matched donor or using autologous cells. The survival following SCT is around 40% with no significant differences between the source of stem cells - matched unrelated donor SCT has a lower rate of relapse but more treatment related deaths than matched family donor SCT. Autologous SCT has the lowest treatment related mortality but also the highest rate of relapse. Haploidentical SCT has been successfully used in the absence of a matched donor, but has the highest treatment related mortality. Some children survive after chemotherapy alone in second remission., so this is an option if treatment toxicities preclude SCT. Medical Research Council AML 10 and AML 12 trials These two studies assessed different combinations of anthracycline and cytarabine based induction chemotherapy (ADE versus DAT, AML 10), and found no difference in efficacy whether thioguanine or etoposide was used as the third drug. If mitozantrone was substituted for daunorubicin (MAE versus ADE, AML 12), there was increased myelotoxicity with delayed recovery of the full blood count, but no difference in overall or event free survival. The addition of either autologous or allogeneic SCT to 4 courses of chemotherapy reduced relapse in AML 10, but allogeneic SCT was associated with an increased treatment-related death rate which largely offset the reduction in relapse. As the reduction in relapse observed with SCT in AML 10 may have been due to a fifth course of treatment rather than SCT per se, AML 12 randomised children between 4 or 5 courses of treatment in all (5th course of chemotherapy high dose cytarabine plus asparaginase). The addition of a fifth course was well tolerated but there was no improvement in overall or event free survival. Analyses of AML 10 demonstrated that karyotype was a prime determinant of prognosis, with t(8;21), t(15;17) and inversion 16 favourable, whilst monosomy 7, abnormal 5q,abnormal 3q, t(9;22) or complex karyotypes were adverse. All other karyotypes were intermediate risk. This risk score was adapted to include response to the first course of chemotherapy (more than 15% blasts adverse) to define 3 risk groups for AML 12. Acute Promyelocytic Leukaemia (APL) APL is uncommon, and comprises 8% of AML in children but has important biological and clinical characteristics. Around 95% of cases of APL are associated with a t(15;17) translocation, resulting in the PML/RAR ± fusion gene. In the remaining cases, RAR ± is fused to an alternative partner, in children most commonly the nucleophosmin gene (NPM1) in the t(5;17). RAR ± is a member of the RA nuclear receptor family that acts as a ligand inducible transcription factor. PML controls p53 dependent induction of apoptosis, growth suppressi- 17

18 on, and is required for transcriptional repression by other tumour suppressors. The PML- RAR ± fusion protein functions as an aberrant retinoid receptor, and is resistant to physiologic concentrations of retinoic acid. The block is overcome however, by therapeutic concentrations produced by all - trans retinoic acid (ATRA), and arsenic trioxide (ATO) degrades the protein whilst also inducing apoptosis via induction of the proenzymes of caspase 2 and 3 and activation of caspases 1 and 3. Besides the high risk of coagulopathy, APL has two unique features, namely a high sensitivity to anthracyclines, possibly due to low expression of P glycoprotein, and marked responsiveness to ATRA and ATO. ATRA as a single agent achieves a high CR rate, but relapse was common unless consolidation chemotherapy was administered. Several clinical trials demonstrated that ATRA administered prior to chemotherapy was associated with improved survival compared with chemotherapy alone due to reduced relapse, but these results were subsequently improved by the simultaneous administration of ATRA with chemotherapy, until CR was achieved. Although anthracyclines are key components of therapy, questions remain as to whether any one anthracycline is superior to the others, and whether the addition of other agents, especially cytarabine,improves outcome. Supportive care during induction therapy is of critical importance. ATRA usually leads to rapid resolution of DIC, but support of coagulation by transfusion of fresh frozen plasma, cryoprecipitate and platelets must be used as necessary until the coagulopathy resolves. A standard approach is maintenance of fibrinogen above 1g/l and platelets above 50 x 109 /l. A risk of both ATRA and ATO is the development of a differentiation syndrome characterised by fever, tachypnoea, hypoxia, pulmonary infiltrates, headaches and confusion in around 10% of patients. This complication is more common with presenting white cell count above 10 x 109 /l, and should be treated with dexamethasone 10mg/sq m twice daily. ATRA/ATO should be temporarily discontinued if the syndrome is severe, or fails to respond to dexamethasone. Up to 40% of children present with a white blood cell count over 10 x 109 /l, associated with M3 variant morphology, FLT3-ITD, and a higher risk of treatment failure due to both increased induction deaths and relapse. Evaluation of treatment response must be circumspect. Abnormal promyelocytes may persist for days, and CR is achieved in virtually all patients. Molecular and cytogenetic tests at the end of induction have no prognostic value, and must not be misinterpreted as indicating resistant disease. Anthracycline based consolidation therapy is associated with very high levels of molecular remission. The advantage of continued ATRA after remission has been achieved has not been proven in a randomised study. The role of cytarabine in consolidation is controversial. Variables confounding conclusions from the published series include differing combinations of anthracyclines, and total anthracycline doses. Cytarabine may be important in regimens with lower total anthracycline exposure. The role of ATO is still under evaluation. Several studies have demonstrated high CR rates with ATO alone or especially when combined with ATRA. ATO has now been included in phase III studies for newly diagnosed disease in both children and adults, and as salvage therapy for molecular or haematologic relapse. The role of haemopoietic stem cell transplant is very limited. Children with persistent molecular disease ( by end - stage polymerase chain reaction with low sensitivity of 10-3 ) at the end of consolidation, or following relapse, are considered for HSCT after salvage therapy with ATO and gemtuzamab. Autologous HSCT may be adequate in children who become molecularly negative, with allogeneic HSCT reser- 18

19 ved for persistent molecular positivity. Molecular monitoring is important, as a better outcome has been shown for patients diagnosed and treated for molecular relapse rather than at haematological relapse. Maintenance therapy with ATRA, mercaptopurine and methotrexate is of unproven benefit in APL. A study by the European APL group indicated that maintenance was beneficial, especially for patients with a high white cell count at presentation, whereas a study by the GI- MEMA group showed no benefit for maintenance (although the total dose of anthracycline was higher in this study). It is possible that efficacy is related to the intensity of initial therapy, that is maintenance is of benefit for patients receiving less intensive protocols with lower doses of anthracycline in induction and consolidation.i in the future, the use of molecular monitoring may allow selection of patients for maintenance therapy. Myeloid leukaemia of Down syndrome ( ML-DS) A remarkable feature of ML - DS is the sensitivity of the leukaemia to therapy with disease resistance and relapse uncommon. Conversely, DS children suffer high rates of treatment toxicity, with an increased risk of treatment related, primarily due to mucositis and infection. TAM also shows marked sensitivity to treatment, and short courses of low dose cytarabine have proved effective in reducing white blood cell count in selected children. A number of genes have been identified on chromosome 21 where gene dosage could influence the effectiveness of chemotherapy. Most particularly in regard to therapy for AML, overexpression of cystathionine _ synthase appears to confer increased sensitivity to cytarabine in the blasts of children with ML-DS. Assay of Ara-C triphosphate levels has shown a median five-fold increase to non-ds blasts. In vitro studies using the MTT assay indicate a 12 fold increase in sensitivity to cytarabine for ML-DS blasts compared to non-ds blasts. This study also demonstrated increased sensitivity to anthracyclines ( 2-7 fold), mitoxantrone (9 fold), and etoposide (20 fold) in ML-DS. Given the high potential for cure in ML-DS, but the risk of complications of therapy, the refinement of treatment to optimise cure rates whilst minimising toxicity is the major focus of modern treatment protocols. As studies using reduced doses of anthracycline have proved very effective and limit mucositis and infective deaths, contemporary regimens for ML-DS include reduced doses of anthracycline compared to those for other children with AML. Late effects of AML chemotherapy, unlike those of SCT, are limited, although there is considerable concern regarding the lifetime risk of cardiac failure due to anthracycline cardiomyopathy. The best cardioprotection strategy is reduced exposure, and so dose reduction is a prime issue in contemporary trials. The use of cardioprotectants remains limited, and the efficacy of liposomal daunorubicin is being studied by the BFM group. There is risk of otoxicity due to the routine use of aminoglycosides in the management of febrile neutropenia and this occurs in some children despite routine monitoring of serum levels of these drugs. Otoxicity is especially prevalent in rare children who carry the A1555G mitochondrial DNA mutation, a cause of familial deafness which confers exceptional sensitivity to aminoglycoside ototoxicity. This mutation should be screened for if there is a history of familial deafness. Total body irradiation (TBI) used in SCT is associated with additional complications and younger children are most vulnerable. Children who have received TBI have growth failure which is in part due to spinal shortening and partly to hypothalamic-pituitary failure; this will be exacerbated if there is chronic GVHD. Delayed puberty and gonadal failure are com- 19

20 mon and there is a variable risk of hypothyroidism, cataracts and learning problems. The combination of busulfan and cyclophosphamide has not been subjected to such rigorous long-term follow-up studies as TBI and it may induce sterility, but it is unlikely to have such a significant effect on skeletal growth and neuropsychologic development. Novel therapeutic approaches Gemtuzamab is a new agent with considerable promise in AML, both as a single drug, and as part of combination chemotherapy. The drug comprises a recombinant, humanised anti- CD33 antibody conjugated to the anti-tumour antibiotic calicheamicin, a DNA damaging agent. CD33 is expressed by myeloid precursors in bone marrow and over 80% of cases of AML. The drug is myelotoxic, but in particular is hepatotoxic causing elevation of bilirubin and transaminases and veno-occlusive disease of the liver in a minority of patients. Gemtuzamab shows good activity in relapsed AML, and has now been incorporated into phase II- I studies in de novo disease in adults and children. A variety of molecular changes have been identified that are associated with prognosis, and hold promise as a method of further refining prognosis or offering the prospect of targeted therapy. Internal tandem duplications of FLT3, are clearly a poor prognostic factor, particularly where the mutated to wild type allelic ratio is high, and FLT3 mutations have been shown as the strongest prognostic factor within cytogenetically normal AML. There is no clear evidence that treatment intensification improves prognosis, and inhibition of constitutively activated FLT3 is now the subject of phase III studies. Mutations of c-kit are found in up to 40% of patients with CBF leukaemia, and specific tyrosine kinase (TK) inhibitors have activity against specific c-kit mutations. Accordingly, phase III studies considering TK inhibition are of interest in these patients. Summary and future directions for management The prognosis of AML has steadily improved through a combination of intensive chemotherapy and better supportive care. Prognostic groups have been identified, largely based on cytogenetics and response to therapy. The role of SCT has been restricted. It is likely further refinement is possible by the incorporation of molecular markers. However, futher improvements to the control of disease by adjustments in conventional chemotherapy protocols are likely to be limited. This is especially true for children with poor risk disease. Increasingly strategies will be developed for subgroups of children; this has already proved necessary in APL and Down syndrome, and new targeted therapies, for example to FLT3-ITD or the CBF pathway are needed or are under evaluation. 20