Detection of 16 P Deletions by FISH in Patients with Inv(16) Or T(16;16) and Acute Myeloid Leukemia (AML)

Detection of 16 p deletions by FISH in patients with inv(16) or t(16;16) and acute myeloid leukemia (AML) D Martinet1,DMu¨hlematter1, M Leeman1, V Parlier1, U Hess2, J Gmu¨r3 and M Jotterand1

1Division Autonome de Ge´ne´tique Me´dicale, Centre Hospitalier Universitaire Vaudois (CHUV), Lausanne; 2on behalf of the SIAK/SAKK Leukemia Group, Klinik C fu¨r Innere Medizin, Kantonsspital, St-Gallen; and 3chairman of the SAKK/Hovon AML protocols, Abteilung Ha¨matologie, Departement Innere Medizin Universita¨tsspital, Zu¨ rich, Switzerland

Deletions of sequences centromeric to the p-arm breakpoint 16q22, a gene encoding for the ␤ subunit (CBFB) of the core have been described in a subset of patients with inv(16) and binding transcription factor (CBF) whose DNA target acute myeloid leukemia (AML) and reported to be associated sequences have been identified in the promoter–enhancer with a relatively good prognosis. We have investigated 16 p 8,14 deletions in a cohort of 15 patients with AML and inv(16) or regions of hematopoietic specific genes. Both rearrange- t(16;16) and compared non-deletion and deletion patients in ments result in the formation of two abnormal fusion genes, terms of clinical course. Patients were studied by fluorescence one consisting of the 5Ј end of CBFB and the 3Ј end of MYH11 in situ hybridization (FISH) using cosmid zit14 as a probe to (CBFB–MYH11), the other of the 5Ј end of MYH11 and the detect the presence of 16 p deletions in metaphase chromo- 3Ј end of CBFB (MYH11–CBFB). somes of leukemic cells. While seven patients (47%) revealed no evidence of a deletion, five patients (33%) presented 16 p A deletion of sequences centromeric to the p-arm break- deletions, thus bringing further support to the relatively fre- point has been observed in addition to inv(16) in 14 out of 77 quent occurrence of this event in inv(16) patients. Remarkably, patients studied.15–18 Deletions have been shown to involve a two patients with inv(16) and one patient with t(16;16) showed region of 160 kb to 350 kb and to affect the 5Ј part of the a mosaicism of deletion and non-deletion metaphases suggest- MYH11 gene.18 In the deletion cases studied, no MYH11– ing the presence of two distinct leukemic cell populations. CBFB transcript was identified by RT-PCR.17 As these patients Results let us assume that 16 p deletions are not restricted to inv(16) and may occur subsequently to inv(16) or t(16;16). The exhibit the same phenotype as non-deletion ones, it can be presence of a 16 p deletion in a case of inv(16) associated with concluded that it is the CBFB–MYH11 fusion product that is 18 CBFB–MYH11 transcript type E indicates that deletions are not critical for the inv(16)/t(16;16)-associated leukemogenesis. limited to CBFB–MYH11 transcript type A rearrangements. Sur- Deletions were reported to be associated with a significantly vival of deletion patients was compared with that of non- longer time from diagnosis until death or relapse.16,17 As deletion and mosaic ones. No significant differences were deletions were found to include at least the 5Ј part of the mul- observed. The advantage of FISH for enumerative and quantitative assessment of submicroscopic rearrangements of clini- tidrug resistance-associated protein gene (MRP), the authors cal significance is further emphasized. hypothesized that the partial loss of one MRP allele accounts Keywords: inv(16)(p13q22); t(16;16)(p13;q22); 16 p deletions; for the relatively favorable outcome in this group. However, FISH; acute myeloid leukemia; prognosis in another study, no difference in prognosis was observed between deletion and non-deletion patients.18 To assess further the incidence of the 16 p deletions and to Introduction improve the understanding of their prognostic significance, we studied 15 patients with AML and inv(16) or t(16;16) by flu- In 1983, Arthur and Bloomfield1,2 reported five patients with orescence in situ hybridization (FISH) using the zit14 cosmid acute myeloid leukemia (AML) of myelomonocytic type (M4), as a probe. Clinical, cytogenetic and FISH data are reported increased marrow eosinophils and a structurally abnormal here. chromosome 16. Soon after, a pericentric inversion of chromosome 16, inv(16)(p13q22), was described in 18 patients who had M4 leukemia and morphologically abnormal eosino- Patients, materials and methods phils with irregular, basophilic-staining granules and distinct cytochemical properties.3,4 A related chromosome abnor- Patients mality, t(16;16)(p13;q22), was reported in one patient with similar characteristics.5 Further studies have suggested that Fifteen patients with either inv(16) or t(16;16) were diagnosed these two abnormalities were uniquely associated with M4 at the divisions of hematology of the University Hospitals of and abnormal eosinophils and the correlation between the Zu¨rich (four patients), Basel (three), Bern (three), Lausanne karyotype and morphology has led to a subcategory in the (two) and of the Canton Hospitals of Sankt-Gallen (two) and FAB classification, namely M4Eo.6 However, both rearrange- Aarau (one) between 1992 and 1995 (Table 1). Each patient ments have been reported in other types of myeloid dis- was assigned a unique patient number for this and previous orders.7,8 In most studies, the presence of inv(16)/t(16;16) has publications.19–21 Additionally, 10 patients with AML (nine) or been associated with a high cure rate with standard chemo- acute lymphoblastic leukemia (ALL) (one) and normal karyo- therapy and a relatively favorable prognosis.9–13 types or aberrations other than inv(16) or t(16;16) were investi- Inv(16) and t(16;16) have been shown to involve, on 16p13, gated by FISH and served as controls. Informed consent was a smooth muscle myosin heavy chain gene (MYH11) and, on obtained from each patient. The diagnosis and classification of patients were based on morphologic and cytochemical examination of peripheral blood (PB) films, bone marrow (BM) aspirate and biopsy speci- Correspondence: M Jotterand, Division autonome de geńe´tique me´ dicale, Centre Hospitalier Universitaire Vaudois, 1011 Lausanne, mens obtained before therapy according to criteria proposed Switzerland by the FAB and the MIC cooperative study groups.3,4,6,22–24 Received 15 January 1997; accepted 3 March 1997 Patients were followed until death or to May 1996. Detection of 16 p deletions in inv(16)/t(16;16) patients D Martinet et al 965 Table 1 Clinical outcome, cytogenetic findings and FISH data in 15 patients with inv(16)/t(16;16) and AML

Clinical data Conventional cytogenetics FISH using zit 14 probe

Case Patient Age/Sex Diagnosis Survival Karyotype % of abnormal % of metaphases Patient No. FAB months metaphases with a single status [metaphases signalc analyzed] [metaphases scored]

1 2353/92 36/M M4Eo 40+ inv(16) 100 [25] 90 [10] del 2 1952/93 25/F M2Eo 8 inv(16)b 97 [31] 88 [17] del 3 52/95 30/M M4Eo 16+ inv(16) 100 [30] 94 [65] del 4 1473/95 55/M M4Eo 10+ inv(16)b 100 [25] 96 [26] del 5 1920/95 29/F M4Eo 7+ inv(16) 96 [23] 89 [27] del 6 1882/92 55/M M4Eo 13 t(16;16) 100 [19] 4 [26] non-del 7 568/93 25/M M5 38+ inv(16) 92 [13] 0 [3] non-del 8 1147/93 54/M M4Eo 2 inv(16) 100 [12] 0 [25] non-del 9 1552/93 44/M M4 5 days inv(16) 100 [10] 9 [11] non-del 10 270/94 31/M M4Eo 2 inv(16)b 100 [25] 4 [54] non-del 11 528/94 14/M AMLa 27+ inv(16) 75 [28] 9 [11] non-del 12 987/95 67/F M4Eo 12+ inv(16)b 96 [26] 10 [21] non-del 13 184/93 48/F M4Eo 39+ inv(16) 92 [24] 25 [16] mosaic 14 1427/93 40/M M4 32+ t(16;16) 100 [20] 35 [17] mosaic 15 434/94 23/F M4Eo 26+ inv(16)b 100 [28] 29 [14] mosaic aSecondary to T and B cell ALL. bPresence of additional chromosome abnormalities. cA signal is deﬁned as either one or two spot(s) on the same chromosome. del, deletion; survival +, still alive after x months survival.

Conventional cytogenetics 4 × SSC/0.05% Tween 20 for 3 min at room temperature (RT). After a preincubation in WB/5% non-fat dry milk (NFDM) at Chromosomes were prepared from PB or BM mononuclear RT for 10 min, slides were washed twice 5 min each in WB. cells and stained in G bands as described.21,25,26 The number Biotin-labeled probes were detected after an incubation with of metaphases analyzed fully varied between 10 and 31 6 ␮g/ml fluorescein isothiocyanate (FITC)-avidin (DCS, according to material available. Vector, Burlingame, CA, USA) and an amplification with 5.5 ␮g/ml biotinylated goat anti-avidin (Vector) followed by an incubation with FITC-avidin.28 All reagents were diluted in FISH 4 × SSC/5% NFDM. Each incubation was performed in a moist chamber at 37°C for 30 min and followed by three washes of Slides were prepared from methanol/acetic acid fixed cell pel- 5 min in WB. For dual-color experiments, signal revelation − ° lets, fresh or stored at 20 C for up to 3 years. To detect was performed as described by Dauwerse et al,29 except that deletions proximal of the p-arm breakpoint, one-color FISH FITC-avidin was used for detecting biotin-labeled probes and was performed on metaphase chromosomes with the pHuR 10 ␮g/ml sheep anti-mouse Cy3 (Sigma, Buchs, Switzerland) 195 probe specific for the heterochromatic region of chromo- for detecting mouse anti-DIG. DNA was counterstained with some 1627 and the cosmid probe zit14.15 To confirm results DAPI or propidium iodide and slides were mounted with an obtained in mosaic patients, the zit18 cosmid probe was used anti-fade solution (Vectashield; Vector). in addition to the pHuR 195 and zit14 probes in dual-color Hybridization signals were defined as follows. Either one or FISH. Both zit14 and zit18 probes, each spanning 40 kb, over- two spots on the same chromosome were considered as one lap and are specific for sequences located in the deletion signal. As a consequence, metaphases with two spots on both region, 50 kb centromeric to the 16 p-arm breakpoint. Zit14 chromosomes 16 (2/2), two spots on one chromosome 16 and hybridizes to sequences slightly proximal to zit18 target sequences. Probes were labeled by nick-translation with one spot on the other (2/1), or one spot on both chromosomes biotin-16-dUTP (pHuR 195 and zit14 probes) or digoxygenin 16 (1/1) were considered as harboring two signals. Meta- (DIG)-11-dUTP (zit18 probe). Slides were denatured at 80°C phases with two or one spot(s) on one chromosome 16 only in formamide 70%/2 × SSC for 5 min, dehydrated in ice-cold (2/0 and 1/0 respectively) were considered as presenting ethanol and air-dried. pHuR 195 (4 ng/␮l) and cosmid probes one signal. (20 ng/␮l) including 50 × human Cot 1 DNA were dissolved Hybridization signals were scored by at least two observers in 50% formamide/2 × SSC/10% dextran sulfate, denatured at on a Zeiss Axioplan fluorescence microscope (Zeiss, Zu¨rich, 80°C for 5 min. DNA was allowed to reanneal at 37°C for Switzerland) equipped with FITC and rhodamine filters, as 10 min and 15 ␮l hybridization mixture was applied under a well as with double (FITC/rhodamine) and triple 22 × 26 mm coverslip. After overnight hybridization at 37°C (FITC/rhodamine/DAPI) band pass filters. Photographs were in a moist chamber, slides were washed in 50% taken with a Kodak Ektachrome 320T film (Kodak, formamide/2 × SSC at 42°C, then in 1 × SSC at 60°C three Lausanne, Switzerland). times 5 min each and in the washing buffer (WB) Detection of 16 p deletions in inv(16)/t(16;16) patients D Martinet et al 966 Statistical analysis Conventional cytogenetics

Continuous variables were compared using the non-para- Thirteen patients presented with inv(16)(p13q22) and two with metric Wilcoxon rank-sum test. Comparisons of counts were t(16;16)(p13;q22) (Table 1). Ten patients had inv(16) or performed using the Fisher’s exact test. A P value Ͻ0.05 was t(16;16) as a single defect. The remaining patients had considered statistically signiﬁcant. Survival was calculated additional chromosome abnormalities. A clone with a normal from the date of diagnosis to the date of last follow-up or karyotype was observed in six patients. Cytogenetic analysis death. Survival curves were established by the Kaplan–Meier of BM obtained in patient 13 during complete remission (39 method and differences in the survival of subgroups assessed months after diagnosis) showed disappearance of the abnor- by log-rank test. All statistical computations were performed mal clone. Cytogenetic ﬁndings will be described in detail using STATA 4.0 (Stata Corporation, College Station, TX, USA) elsewhere (Jotterand et al, manuscript in preparation). for Apple Macintosh.

FISH Results Based on zit14 results, three distinct groups were observed Patients’ characteristics and clinical features among the inv(16)/t(16;16) patients (Figures 2 and 3; Tables 1 and 2). In the first group (five patients), the proportion of meta- Study patients were five women and 10 men (Table 1). Their phases with one signal ranged from 88.2 to 96.2% (mean age ranged from 14 to 67 years. Median age was 36 years. At 92.4%) and patients were considered as deletion patients; the diagnosis, 10 patients were categorized as M4Eo. Two 3.4% metaphases with signals on both chromosomes 16 are additional patients had the features of M4. Other FAB categor- probably due to the presence of some normal metaphases as ies observed included one patient with M5 and one patient demonstrated by conventional cytogenetics. In the second with M2Eo. One patient, a 14-year-old male, had an AML of group (seven patients), the proportion of metaphases with sig- probable M1 type secondary to an early B CD10+ ALL which nals on both chromosome 16 ranged from 90.5 to 100% occurred 2 years after a T cell ALL diagnosed in 1988; at the (mean 94.7%); as this proportion is similar to that observed time of ALL, this patient did not reveal any evidence of in the control group, patients were considered as non-deletion inv(16). Patients will be described in detail elsewhere patients (Table 2). Case 7 was included in this group as two (Jotterand et al, manuscript in preparation). signals were observed in all three metaphases available. In the Fourteen patients received standard chemotherapy accord- third group (three patients), the proportion of metaphases with ing to the Hovon 4A – SAKK 30/92 (12 patients, seven of them a single signal ranged from 25.0 to 35.3% (mean 29.8%) and included in the Dutch–Swiss study) and Hovon 29 – SAKK 30/95 (two patients, both of them included in the Dutch–Swiss study) protocols. The first cycle of remission induction treat- ment consisted of cytarabin for 7 days (200 mg/m2) and of either daunomycin (40–60 mg/m2) for 3 days (11 patients, 4 days in one elderly patient) or idarubicin for 3 days (12 mg/m2, two patients). The second cycle consisted of cytarabin for 6 days (1000 mg/m2) and amsacrine for 3 days (120 mg/m2). Patient 14 underwent allogeneic BM transplan- tation 5 months after diagnosis. The young patient with secondary AML (case 11) received an induction therapy compris- ing etoposide/cytarabin and etoposide/endoxan, alternatively. Median overall survival of the whole group has not been reached. Survival curves according to the presence/absence of the deletions were not significantly different (P = 0.17) (Figure 1). Median survival was 13.4 months in the seven non- deletion patients (four deceased patients); it was not reached in the five deletion patients (one deceased patient) and in the three mosaic cases (no deceased patient).

Figure 2 Schematic representation of p-arm and q-arm break- points, probes localization (zit14 and pHuR 195) and deletion region Figure 1 Kaplan–Meier survival curves in 15 inv(16)/t(16:16) in inv(16), inv(16)/deletion and t(16;16), t(16;16)/deletion situations patients. (drawings are not to scale). Detection of 16 p deletions in inv(16)/t(16;16) patients D Martinet et al 967

Figure 3 FISH analysis in control, inv(16), inv(16)/deletion and t(16;16)/deletion cells using the zit14 and pHuR 195 probes. (a) normal chromosomes 16; (b) metaphase cell from inv(16) patient showing signal on both chromosomes 16; (c–e) metaphase cells from inv(16)/deletion patients demonstrating a single signal from the normal chromosome 16; (f) metaphase cell from t(16;16)/deletion mosaic patient showing loss of signal from one rearranged chromosome 16 (16 p+). patients were considered as mosaic patients. In all inv(16) deletion and deletion patients, more than 50% metaphases do metaphases with a single signal, the signal was localized on not present any spot loss (mean values are equal to 58.3, 61.6 the normal chromosome 16, as assessed by pHuR195 localiz- and 71.0%, respectively). In these patients, the proportion of ation with regard to arm ratio. One mosaic patient (case 13) metaphases with one spot loss (2/1 or 1/0) amount to 35.4, was studied in complete remission; a signal was present on 25.2 and 21.4%, respectively, those of metaphases with two both chromosomes 16 in all metaphases analyzed. spots loss represent 6.3, 12.5 and 4.1%. In control and non- Spot loss was evaluated in each group of patients and the deletion patients, metaphases with 2/0 (1.6% and 4.6%) are proportion of metaphases with 2/1, 1/1, 2/0, 1/0 and 0/0 were less frequent than metaphases with 1/1 (4.7% and 7.9%). The compared to those of metaphases with optimal hybridization situation is reversed in the mosaic group, where metaphases (either 2/2 in control and non-deletion or 2/0 in deletion with 2/0 (21.3%) occur much more frequently than those with patients) (Table 2). Results indicate that, in control, non- 1/1 (10.6%). Detection of 16 p deletions in inv(16)/t(16;16) patients D Martinet et al 968

Figure 3 (Continued).

Table 2 zit14 spot distribution in inv(16)/t(16;16) and control patients

Signal numbera 210

Total Spot numberb 2/2 2/1 1/1 Total 2/0 1/0 Total 0/0

Deletion patients 1 [10]c 0.0d 10.0 [1] 0.0 10.0 [1] 70.0 [7] 20.0 [2] 90.0 [9] 0.0 2 [17] 0.0 0.0 0.0 0.0 47.1 [8] 41.2 [7] 88.2 [15] 11.8 [2] 3 [65] 1.5 [1]c 0.0 0.0 1.5 [1] 76.9 [50] 16.9 [11] 93.8 [61] 4.6 [3] 4 [26] 0.0 0.0 0.0 0.0 69.2 [18] 26.9 [7] 96.2 [25] 3.8 [1] 5 [27] 3.7 [1] 7.4 [2] 0.0 11.1 [3] 74.1 [20] 14.8 [4] 88.9 [24] 0.0 Total [145] 1.4 [2] 2.1 [3] 0.0 3.4 [5] 71.0 [103] 21.4 [31] 92.4 [134] 4.1 [6]

Non-deletion patients 6 [26] 53.8 [14] 34.6 [9] 3.8 [1] 92.3 [24] 3.8 [1] 0.0 3.8 [1] 3.8 [1] 7 [3] 66.7 [2] 33.3 [1] 0.0 100.0 [3] 0.0 0.0 0.0 0.0 8 [25] 80.0 [20] 20.0 [5] 0.0 100.0 [25] 0.0 0.0 0.0 0.0 9 [11] 54.5 [6] 27.3 [3] 9.1 [1] 90.9 [10] 9.1 [1] 0.0 9.1 [1] 0.0 10 [54] 64.8 [35] 24.1 [13] 7.4 [4] 96.3 [52] 3.7 [2] 0.0 3.7 [2] 0.0 11 [11] 36.4 [4] 18.2 [2] 36.4 [4] 90.9 [10] 9.1 [1] 0.0 9.1 [1] 0.0 12 [21] 57.1 [12] 23.8 [5] 9.5 [2] 90.5 [19] 9.5 [2] 0.0 9.5 [2] 0.0 Total [151] 61.6 [93] 25.2 [38] 7.9 [12] 94.7 [143] 4.6 [7] 0.0 4.6 [7] 0.7 [1]

Mosaic patients 13 [16] 50.0 [8] 18.8 [3] 6.3 [1] 75.0 [12] 18.8 [3] 6.3 [1] 25.0 [4] 0.0 14 [17] 11.8 [2] 35.3 [6] 17.6 [3] 64.7 [11] 29.4 [5] 5.9 [1] 35.3 [6] 0.0 15 [14] 35.7 [5] 28.6 [4] 7.1 [1] 71.4 [10] 14.3 [2] 14.3 [2] 28.6 [4] 0.0 Total [47] 31.9 [15] 27.7 [13] 10.6 [5] 70.2 [33] 21.3 [10] 8.5 [4] 29.8 [14] 0.0

Total control patients [254] 58.3 [148] 35.4 [90] 4.7 [12] 98.4 [250] 1.6 [4] 0.0 1.6 [4] 0.0

The number of metaphases scored in each case ranged between 15 and 32 in the control patients and between 10 and 65 in inv(16)/t(16;16) patients, except in patient 7, where only three metaphases could be analyzed. aA signal is defined as either one or two spot(s) on the same chromosome. bSpot distribution on both chromosomes 16. cMetaphase number. dProportion of metaphases with 2/2, 2/1, 1/1, 2/0, 1/0, and 0/0 spot(s). Detection of 16 p deletions in inv(16)/t(16;16) patients D Martinet et al 969 In mosaic patients, dual-color FISH with zit18 and zit14 To our knowledge, one case of inv(16) with a 16 p deletion cosmids was performed on the slides used for zit14 hybridiz- mosaicism has been reported.16 Subsequently however, this ation assessment. In all metaphases available, the presence or case was considered as non-deleted based on additional mol- absence of zit18 signals correlated with that of zit14 signals ecular data.17 In the other studies, the number of metaphases (data not shown). observed in each case and the distribution of FISH signals were not reported in detail; therefore, possible mosaic cases may have been considered as non-deleted when only a few Discussion metaphases were available. Fourteen of the 15 patients reported here have been studied We have investigated 16 p deletions in 13 patients with by RT-PCR.19,20 CBFB–MYH11 transcript type A (nomen- inv(16) and two patients with t(16;16). Seven patients (47%) clature after Liu et al8) was found in 12 cases (86%), transcript showed no evidence of a deletion, whereas five patients (33%) type E in two (14%) (cases 5 and 7). The presence of the presented 16 p deletions. The presence of the deletion was deletion in a patient with transcript type E indicates that the confirmed by Southern blot analysis with a MYH11 6.9 kb p-arm deletion is not restricted to CBFB–MYH11 transcript BamHI probe in three patients.20 Our rate of deletion patients type A and suggests that it can be associated with other CBFB– (33%) is close or superior to those in other reports.16–18 This MYH11 fusion types as well. finding provides further evidence of the relatively frequent Deletions in 16 p have been reported to be associated with occurrence of this event in inv(16) patients and to the pres- improved survival.16,17 As the deleted region was shown to ence of some degree of genomic instability in this region pre- include the promoter and the 5Ј part of the MRP gene located disposing to such molecular rearrangements.18 150 kb proximal to the p-arm breakpoint, Kuss et al suggested Remarkably, two inv(16) and one t(16;16) patients showed that MRP hemizygosity may be relevant in determining the a mosaic pattern suggesting the existence of two distinct leu- outcome of inv(16) AML patients. Although MRP status was kemic cell populations, one with the deletion and one with- not assessed specifically, part of the MRP gene was supposed out. It could be argued that, in these cases, the presence of to be codeleted in our deletion patients and these patients one and two signals in respectively 30 and 70% of meta- were compared with non-deleted and mosaic ones in terms phases (mean values), may be due to an admixture of normal of survival. No significant difference could be demonstrated; and abnormal cells. This seems unlikely according to the the result was unaltered when mosaic patients were grouped results of conventional cytogenetics (100% abnormal meta- with deletion or with non-deletion ones. This finding does not phases in two patients and 92% in the third one whatever the corroborate the formally reported positive correlation between culture condition) which suggest that the metaphases with two deletion and improved prognosis. However, although in signals do not all represent normal cells. A second argument accordance with other observations,18 our data have to be views the presence of 30% metaphases with a single signal as considered with caution as larger numbers of patients have to reflecting defective probe hybridization in these cases. There be longitudinally studied to draw any definitive conclusion are several reasons for not supporting this assumption, regarding prognosis. Moreover, multidrug resistance is a com- although it cannot be excluded. In control patients, more than plex and multifactorial phenomenon that involves different 96% metaphases present a double signal. Experiment varia- mechanisms and several distinct genes such as MRP and bility in the hybridization yield can be ruled out as each MDR1.30 If the MRP gene is expected to play a role in the analysis relied on several slides, all hybridized in separate clinical response to chemotherapy, follow-up studies should assays. Defective hybridization is expected to generate a be performed not only at the DNA level but also at the mRNA decrease in the proportion of metaphases with 2/2 (control and protein levels before any definitive statements concerning and non-deletion patients) or 2/0 (deletion patients) and an its role in leukemic drug resistance can be made. increase in metaphases with spot loss. To characterize better Our study reinforces the usefulness of FISH for assessing mosaic patients and to test if defective hybridization may submicroscopic rearrangements of possible diagnostic and account for the presence of more than 25% metaphases with prognostic significance that escape conventional cytogenetics a single signal in these cases, the distribution of fluorescent due to their small size. While molecular techniques such as spots was analyzed in detail. Whereas the proportion of meta- Southern blotting or microsatellite marker assessment may phases with three spots (2/1) is similar to those with 1 spot allow detection of the presence of the deletion in loss (2/1 or 1/0) in all other groups, the proportion of meta- inv(16)/t(16;16) patients, metaphase FISH with appropriate phases with two spots loss (2/0 or 1/1) is much higher and the probes has the advantage of being enumerative and quantitat- reduction of the number of 2/2 metaphases correlates with ive. Due to the existence of mosaic cases, the FISH approach a specific increase in 2/0 metaphases without any significant requires the observation of possibly a large number of meta- increase in the proportion of 1/1 metaphases. Moreover, in all phases for quantitative values to be significant. inv(16) metaphases with a single signal, the signal was recog- nized in the normal chromosome 16. The use of the zit18 probe in addition to zit14 (dual-color FISH) in a number of Acknowledgements metaphases from mosaic patients with two or one zit14 signal(s) did not reveal any discrepancy between zit14 and We are indebted to Dr A Tichelli, Prof Dr A Tobler, Prof Dr zit18 signal distribution. M Fey, Dr V Spataro, Dr PM Schmidt, Prof Dr A Hirt, Dr M Thus, if the mosaicism interpretation is correct, these results Wernli and the members of the Leukemia Group (Chairman, mean that, in a subset of patients, 16 p deletions may arise Dr U Hess) of the SIAK (Chairman Prof Dr H Wagner)/SAKK subsequently to inv(16)/t(16;16). Case 14 would represent the (Chairman Prof Dr A Goldhirsch) for providing patient first reported t(16;16) patient with a deletion detected by FISH. samples and clinical data. We thank Dr C Castagne´, O Bruz- Unfortunately, no more material was available in the mosaic zese, M Wicht and the other technicians who assisted in the patients for further study and additional patients have to be chromosome analysis. We are grateful to B van der Reijden tested so as to confirm definitively the mosaicism hypothesis. for helpful comments and interest in the project. We thank Dr Detection of 16 p deletions in inv(16)/t(16;16) patients D Martinet et al 970 M Breuning, J Dauwerse and R Giles for providing probes and 15 Dauwerse JG, Wessels JW, Giles RH, Wiegant J, van der Reijden for critical reading of the manuscript. Dr G van Melle is BA, Fugazza G, Jumelet EA, Smit E, Baas F, Raap AK, Hagemeijer acknowledged for advice in statistical analysis. This work was A, Beverstock GC, Van Ommen GJ, Breuning MH. Cloning the breakpoint cluster region of the inv(16) in acute nonlymphocytic supported by grants from the Ligue Suisse contre le Cancer leukemia M4Eo. Hum Mol Genet 1993; 2: 1527–1534. (FOR 209), the Ligue Argovienne contre le Cancer, the 16 Kuss BJ, Deeley RG, Cole SP, Willman CL, Kopecky KJ, Wolman Recherche Suisse contre le Cancer, the Cramer Foundation, SR, Eyre HJ, Lane SA, Nancarrow JK, Whitmore SA, Callen DF. the Ligues Vaudoise and Zougoise contre le Cancer, the Leu- Deletion of gene for multidrug resistance in acute myeloid leukemia group of the Swiss Institute for Applied Cancer kaemia with inversion in chromosome 16: prognostic impli- Research/Swiss Group for Clinical Cancer Research cations. Lancet 1994; 343: 1531–1534. 17 Kuss BJ, Deeley RG, Cole SP, Willman CL, Kopecky KJ, Wolman (SIAK/SAKK), the Swiss National Science Foundation (31- SR, Eyre HJ, Callen DF. The biological significance of the multi- 32319.91), the Stiftung zur Krebsbeka¨mpfung, the Sandoz, the drug resistance gene MRP in inversion 16 leukemias. Leuk Lym- SICPA Foundations and the Division autonome de geńe´tique phoma 1996; 20: 357–364. me´dicale (Prof G Pescia) to Martine Jotterand. 18 Marlton P, Claxton DF, Liu P, Estey EH, Beran M, Le Beau M, Testa JR, Collins FS, Rowley JD, Siciliano MJ. Molecular charac- terization of 16 p deletions associated with inversion 16 defines the critical fusion for leukemogenesis. Blood 1995; 85: 772–779. 19 Van der Reijden BA, Lombardo M, Dauwerse HG, Giles RH, References Mu¨hlematter D, Jotterand Bellomo M, Wessels HW, Beverstock GC, Van Ommen GJ, Hagemeijer A, Breuning MH. RT-PCR diag- 1 Arthur DC, Bloomfield CD. Partial deletion of the long arm of nosis of patients with acute nonlymphocytic leukemia and chromosome 16 and bone marrow eosinophilia in acute nonlym- inv(16)(p13q22) and identification of new alternative splicing in phocytic leukemia: a new association. Blood 1983; 61: 994–998. CBFB–MYH11 transcripts. Blood 1995; 86: 277–282. 2 Arthur DC, Bloomfield CD. Correspondence. Blood 1983; 62: 20 Van der Reijden BA, Martinet D, Dauwerse JG, Giles RH, Wessels 931. JW, Beverstock GC, Smit B, Mu¨hlematter D, Jotterand Bellomo M, 3 Le Beau MM, Larson RA, Bitter MA, Vardiman JW, Golomb HM, Gabert J, Lafage-Pochitaloff M, Reiffers J, Bilhou-Nabera C, Van Rowley JD. Association of an inversion of chromosome 16 with Ommen GJ, Hagemeijer A, Breuning MH. Simple method for abnormal marrow eosinophils in acute myelomonocytic leukemia. detection of MYH11 DNA rearrangements in patients with New Engl J Med 1983; 309: 630–636. inv(16)(p13q22) and acute myeloid leukemia. Leukemia 1996; 10: 4 Bitter MA, Le Beau MM, Larson RA, Rosner MC, Golomb HM, 1459–1462. Rowley JD, Vardiman JW. A morphologic and cytochemical study 21 Castagne´ C, Mu¨hlematter D, van Melle G, Gachoud V, Jotterand of acute myelomonocytic leukemia with abnormal marrow Bellomo M. Effect of conditioned media, nutritive elements and eosinophils associated with inv(16)(p13q22). Am J Clin Pathol mitotic synchronization on the accuracy of the cytogenetic analy- 1984; 81: 733–741. sis in acute nonlymphocytic leukemia patients presenting 5 Hogge DE, Misawa S, Parsa NZ, Pollak A, Testa JR. Abnormalities inv(16)/t(16;16) or t(15;17). Cancer Genet Cytogenet (in press). of chromosome 16 in association with acute myelomonocytic leu- 22 Bennett JM, Catovsky D, Daniel MT, Flandrin G, Galton DA, Gral- kemia and dysplastic bone marrow eosinophils. J Clin Oncol nick HR, Sultan C. Proposals for the classification of the acute 1984; 2: 550–557. leukaemias: French–American–British (FAB) Co-operative Group. 6 Bennett JM, Catovsky D, Daniel MT, Flandrin G, Galton DA, Gral- Br J Haematol 1976; 33: 451–458. nick HR, Sultan C. Proposed revised criteria for the classification 23 Bennett JM, Catovsky D, Daniel MT, Flandrin G, Galton DA, Gral- of acute myeloid leukemia. A report of the French–American–Brit- nick HR, Sultan C. Proposals for the classification of the myelo- ish Cooperative Group. Ann Intern Med 1985; 103: 620–625. dysplastic syndromes. Br J Haematol 1982; 51: 189–199. 7 Mitelman F. Catalog of Chromosome Aberrations in Cancer. 24 Second MIC Cooperative Study Group. Morphologic, immuno- Wiley-Liss: New York, 1994. logic and cytogenetic (MIC) working classification of the acute 8 Liu PP, Hajra A, Wijmenga C, Collins FS. Molecular pathogenesis myeloid leukaemias. Br J Haematol 1988; 68: 487–494. of the chromosome 16 inversion in the M4Eo subtype of acute 25 Jotterand Bellomo M, Parlier V, Schmidt PM, Beris P. Cytogenetic myeloid leukemia. Blood 1995; 85: 2289–2302. analysis of 54 cases of myelodysplastic syndrome. Cancer Genet 9 Larson RA, Williams SF, Le Beau MM, Bitter MA, Vardiman JW, Cytogenet 1990; 46: 157–172. Rowley JD. Acute myelomonocytic leukemia with abnormal 26 Jotterand Bellomo M, Parlier V, Mu¨hlematter D, Grob JP, Beris P. eosinophils and inv(16) or t(16;16) has a favorable prognosis. Three new cases of chromosome 3 rearrangement in bands q21 Blood 1986; 68: 1242–1249. and q26 with abnormal thrombopoiesis bring further evidence to 10 Schiffer CA, Lee EJ, Tomiyasu T, Wiernik PH, Testa JR. Prognostic the existence of a 3q21q26 syndrome. Cancer Genet Cytogenet impact of cytogenetic abnormalities in patients with de novo acute 1992; 59: 138–160. nonlymphocytic leukemia. Blood 1989; 73: 263–270. 27 Moyzis RK, Albright KL, Bartholdi MF, Cram LS, Deaven LL, Hilde- 11 Plantier I, Lai JL, Wattel E, Bauters F, Fenaux P. Inv(16) may be brand CE, Joste NE, Longmire JL, Meyne J, Schwarzacher-Robin- one of the only ‘favorable’ factors in acute myeloid leukemia: a son T. Human chromosome-specific repetitive DNA sequences: report on 19 cases with prolonged follow-up. Leukemia Res 1994; novel markers for genetic analysis. Chromosoma 1987; 95: 18: 885–888. 375–386. 12 Burnett AK. Karyotypically defined risk groups in acute myeloid 28 Pinkel D, Straume T, Gray JW. Cytogenetic analysis using quanti- leukaemia. Leukemia Res 1994; 18: 889–890. tative, high-sensitivity, fluorescence hybridization. Proc Natl Acad 13 Marlton P, Keating M, Kantarjian H, Pierce S, O’Brien S, Freireich Sci USA 1986; 83: 2934–2938. EJ, Estey E. Cytogenetic and clinical correlates in AML patients 29 Dauwerse JG, Wiegant J, Raap AK, Breuning MH, van Ommen GJ. with abnormalities of chromosome 16. Leukemia 1995; 9: 965– Multiple colors by fluoresence in situ hybridization using radio- 971. labelled DNA probes create a molecular karyotype. Hum Mol 14 Liu P, Tarle´ SA, Hajra A, Claxton DF, Marlton P, Freedman M, Genet 1992; 1: 593–598. Siciliano MJ, Collins FS. Fusion between transcription factor 30 Zo¨ chbauer S, Haas OA, Schwarzinger I, Lechner K, Pirker R. Mul- CBF␤/PEBP2␤ and a myosin heavy chain in acute myeloid leuke- tidrug resistance in acute myeloid leukaemia with inversion in mia. Science 1993; 261: 1041–1044. chromosome 16 of FAB M4Eo subtype. Lancet 1994; 344: 894.