Therapy-related acute lymphoblastic leukemia (t-ALL) is a rare secondary leukemia following chemotherapy and/or radiotherapy for primary malignancies. Chromosomal 11q23 abnormality, frequently detected in therapy-related acute myeloid leukemia, is the most common cytogenetic alteration in t-ALL. However, t-ALL cases without 11q23 abnormality have been rarely described. We describe 6 adults with secondary t-ALL without 11q23 abnormalities following various treatment regimens for primary malignancies. We also reviewed 48 t-ALL cases, with complete chromosomal karyotyping, reported in the literature from 1992 to 2007. In the 48 cases, an 11q23 abnormality involving the MLL gene locus was the predominant chromosomal aberration (32 [67%]), followed by t(9;22) (6 [13%]) and a normal karyotype (4 [8%]). Compared with t-ALL cases with an 11q23 abnormality, cases without an 11q23 abnormality had a relatively longer latency period (median, 36 vs 19 months) and a different primary malignancy spectrum. No major difference was observed between groups in regard to age, sex, or receipt of a topoisomerase II inhibitor. The t(8;14)(q11.2;q32), a rare, nonrandom, balanced chromosomal translocation differing from the more common translocation involving c-MYC on chromosome 8q24, was seen in 1 adult t-ALL case, which may suggest another possible pathogenesis of this disease.
Therapy-related acute leukemia is a heterogeneous disease but can be broadly divided into 2 major types: alkylating agent/radiotherapy–related and topoisomerase II inhibitor–related.1 The alkylating agent–related subgroup is characterized by usually having antecedent myelodysplasia with a mean latency period of 5 to 7 years and frequently harboring complete or partial deletion of chromosome 5 or 7.2 These alkylating agent–related acute leukemias are mostly classified as acute myeloid leukemia (AML) M1 and M2, according to the widely used French-American-British classification.3 DNA topoisomerase II inhibitors, on the other hand, cause secondary leukemias with relatively short latent periods (1–5 years), and secondary acute leukemia usually develops without an intervening myelodysplastic phase. The most common characteristic chromosomal aberrations are translocations involving 11q23, the MLL gene locus.4 Most of the these leukemias are diagnosed as AML with monocytic differentiation (French-American-British subtypes M4 and M5).5
Therapy-related acute lymphoblastic leukemia (ALL), which represents approximately 12% of all therapy-related acute leukemias6 and 1.2% to 4%6,7 of adult ALLs, is seen much less frequently than therapy-related AML. While chromosomal 11q23 abnormalities are considered as the most common karyotypic alteration in therapy-related ALL, particularly in childhood therapy-related ALL, other cytogenetic abnormalities in therapy-related ALL are extremely rare but have been reported previously.
To have a better understanding of the clinical and biologic characteristics of these therapy-related ALLs without 11q23 abnormalities, we studied 6 adult therapy-related ALL cases lacking an 11q23 chromosome aberration and reviewed therapy-related ALL cases with detailed karyotype data published in the literature from 1992 to 2007.
Materials and Methods
Clinical Case Selection and Literature Review
Six cases were selected of therapy-related ALL, diagnosed at the City of Hope National Medical Center, Duarte, CA, and Duke University Medical Center, Durham, NC, with a documented history of primary malignancies and subsequent chemotherapy and/or radiotherapy. The pathologic diagnosis of the therapy-related ALL cases was made according to the current World Health Organization (WHO) classification.8 We also reviewed therapy-related ALL cases reported in the literature from 1992 to 2007. Therapy-related ALL cases without detailed karyotypes from the literature review were excluded.
Histologic and Cytologic Examination and Immunophenotypic Analysis
Peripheral blood and bone marrow aspirate smears were prepared with Wright-Giemsa staining for morphologic evaluation. The bone marrow trephine biopsy specimens were fixed in 10% neutral buffered formalin. Paraffin sections were stained with H&E for routine histologic examination. Immunophenotyping was primarily performed by 4-color flow cytometric analysis of bone marrow aspirate with a Coulter Epics XL cytometer (Beckman Coulter, Miami, FL), as previously described.9 Related flow cytometry panels were as follows: leukemia panel: CD1a, CD2, CD3, CD4, CD8, CD10, CD13, CD14, CD15, CD19, CD20, CD34, CD33, CD45, CD7, CD38, CD56, CD117, CD41, glycophorin A, HLA-DR, cytoplasmic terminal deoxynucleotidyl transferase (TdT), cytoplasmic CD22, and cytoplasmic CD3; and pre-B panel: CD19, CD34, CD33, and CD45 (Beckman Coulter). Immunohistochemical staining was performed on formalin-fixed, paraffin-embedded bone marrow biopsy sections using an automated immunostainer (DAKO, Carpinteria, CA) as described previously.10 The following antibodies were used: Pax5 (1:80 dilution; BD Biosciences, San Jose, CA), CD34 (1:5 dilution; DAKO), TDT (1:60 dilution; Supertechs, Rockville, MD), and myeloperoxidase (1:6,000 dilution; DAKO).
Cytogenetic Analysis and Target Fluorescence In Situ Hybridization
Cytogenetic analysis was performed on peripheral blood or bone marrow specimens using standard techniques. GTG banding was used to identify the individual chromosome. When available, at least 20 metaphases were examined from each case. Cytogenetic nomenclature followed standard 1995 International System for Human Cytogenetic Nomenclature criteria.11
Fluorescence in situ hybridization (FISH) studies for 11q23 translocation (MLL rearrangement) were performed as described previously.6 Briefly, slides were incubated in 2× saline sodium citrate (SSC) at 37°C for 30 minutes, denatured in 70% formamide/2× SSC (pH 7.0) at 72°C for 2 minutes, and hybridized using MLL break-apart single fusion probes (Vysis, Downers Grove, IL). Nonspecific probe binding was removed by a posthybridization wash with 0.4× SSC/0.3% Nonidet P40 (Roche Applied Science, Mannheim, Germany) (pH 7.0) at 72°C for 2 minutes, followed by 2× SSC/0.1% Nonidet P40 (pH 7.0) at room temperature for 1 minute. Cells were counterstained with 4060-diamino-2 phenyl indole dihydrochloride (Vysis). Lymphocytes from a phenotypically normal female were used as a negative control sample.
In all, 200 nuclei were scored for each patient. Samples were considered positive for an MLL rearrangement if greater than 3% of the interphase nuclei scored were positive (sensitivity, 100%; 95% confidence interval, 80.5%–100%; and specificity, 100%; 95% confidence interval, 71.5%–100%).
Statistical data were analyzed by using the χ2 test and the Student t test.
Clinical Features of Therapy-Related ALL Cases
There were 4 women and 2 men, with a median age of 60 years Table 1. Primary malignancies included breast adenocarcinoma (n = 2), multiple myeloma (n = 2), uterine leiomyosarcoma (n = 1), and Hodgkin lymphoma (n = 1). We recently described case 3.12 In all patients, precursor B ALL developed after chemotherapy and/or radiotherapy for the primary malignancies. Three patients received topoisomerase II inhibitors and 2 received local radiotherapy for stage I breast carcinoma without exposure to topoisomerase II inhibitors. The treatment regimen was not available for case 2 (Table 1). The latency period ranged from 48 to 156 months with a median of 72 months. Two patients were alive at 8 and 11 months’ follow-up; the other 4 patients died within 1 year after therapy-related ALL was diagnosed (range, 1–12 months).
Histologic, Cytologic, and Immunophenotypic Findings in Therapy-Related ALL Cases
The leukemic cells in peripheral blood and marrow aspirate smears and imprint preparations of each case showed typical lymphoblast morphologic features with small to medium-sized cells, with high nuclear/cytoplasmic ratios; round, oval, or slightly irregular nuclear contours; fine nuclear chromatin; indistinct nucleoli; and scant basophilic cytoplasm. In marrow biopsy specimens, the leukemic cells usually replaced bone marrow spaces with minimal residual hematopoiesis Image 1A and Image 1B. One of the therapy-related ALL cases had approximately 25% heavily granulated lymphoblasts with small to medium-sized, light pink to dark purple, compact cytoplasmic granules Image 2A.
Immunohistochemical tests showed the leukemic cells were positive for TdT, CD34 and B-cell lineage–specific markers (Pax5) Image 1C, Image 1D, Image 2B, and Image 2C and negative for myeloperoxidase Image 2D in all cases. Flow cytometric analysis demonstrated that all 6 cases had immunophenotypic profiles consistent with derivation of precursor B cells. No aberrant expression of myeloid or T lymphocyte–associated antigens was identified. CD10 expression was positive in 5 cases and negative in 1 case (Table 1). The morphologic features and immunophenotypic findings are diagnostic of precursor B ALL for all 6 cases according to the WHO criteria.
Cytogenetic Analysis and FISH in the Present Therapy-Related ALL Cases
Cytogenetic analysis and targeted FISH for 11q23 translocation were performed on all 6 cases. Four cases had normal karyotypes, and 1 (case 3) had t(8;14)(q11.2;q32), in addition to an extra X chromosome and partial deletion of chromosome 11 [del(11)(p11.2p13)]. Case 4 had del(9) (q13q32) in 3 of 20 scored cells. No 11q23-related translocations were detected by conventional karyotyping in any of the cases. FISH was performed on the first 3 cases and was negative for 11q23 rearrangement using break-apart probes (Table 1).
Biologic Features of Therapy-Related ALL Cases From the Literature Review and the Present Study
We identified 48 cases with detailed available karyotype data among therapy-related ALL cases reported in the literature from 1992 to 2007 and summarized in Table 2. As expected, the 11q23 abnormality involving the MLL gene locus was the predominant chromosomal aberration in therapy-related ALL (32 [67%]), followed by t(9;22) (6 [13%]) and a normal karyotype (4 [8%]). All 48 therapy-related ALL cases were further analyzed by comparing patient age, sex, primary malignancies, previous treatment, disease onset, and median latency period between the therapy-related ALL 11q23 group (cases with 11q23 abnormalities) and therapy-related ALL non-11q23 group (cases without 11q23 abnormalities, including normal karyotype) (Table 2). No significant differences were detected between these 2 groups in age or sex (P ≥ .05). In the 11q23 group, 21 (84%) of 25 had received topoisomerase II inhibitors, whereas in the non-11q23 group, 8 (67%) of 12 had received topoisomerase II inhibitor treatment. However, the difference was not statistically significant (P ≥ .05).
The distribution of primary malignancies between the 2 groups, however, appeared to be quite different. Hematologic malignancies (mainly Hodgkin lymphoma) were the most common primary neoplasm in the non-11q23 group (7/30 [44%]), but accounted for only 6 (19%) of 32 primary malignancies in the 11q23 group. On the other hand, breast cancer was the most frequent primary tumor in the 11q23 group (12/32 [38%]), but accounted for only 2 (13%) of 30 primary malignancies in the non-11q23 group. However, as a whole group of therapy-related ALL, there appeared to be no obvious association with any particular primary malignancy, which might be because only relatively small sample volumes were studied.
The latency period between primary malignancy and therapy-related ALL without 11q23 abnormalities was much longer than in therapy-related ALL with 11q23 abnormalities (median, 36 vs 19 months; P < .05). Generally speaking, the prognosis for patients with therapy-related ALL, regardless of the presence or absence of 11q23 abnormalities, was unfavorable (median survival, 2.55 months).
Compared with therapy-related AML, which is a well-recognized entity accepted by the WHO classification,1 therapy-related ALL is far less understood owing to its rarity. Only 21 therapy-related ALL cases were identified in a 901-case ALL study organized by the GIMEMA group in 1999, accounting for 2.3% of all ALL cases.7 Similar results were reported recently by Shivakumar et al,13 who reviewed 213 cases diagnosed as ALL from 1992 to 2007 and found only 7 therapy-related ALL cases (3.3%). In contrast, therapy-related AML represents 10% to 20% of all AML cases.14
Unlike de novo adult ALL, which occurs at a median age of 32 years (range, 16–65 years),15 our 6 therapy-related ALL cases obviously had a later onset with a median age of 60 years (range, 34–74 years). The cytomorphologic features of our precursor B therapy-related ALL cases also appeared to be no different from those of de novo adult ALL. Besides strongly expressing CD19, cytoplasmic CD22, and nuclear TdT in all of our 6 cases, most of our cases (5/6) also showed CD10 expression, consistent with the notion of correlation between CD10 expression in ALL and absence of MLL gene rearrangements in adult and infant ALL.16,17 It is noteworthy that 1 of our therapy-related ALL cases had heavily granulated lymphoblasts, which is extremely rare in adult ALL and may lead to misdiagnosis of acute leukemia with myeloid differentiation.18
To further understand the clinical and pathologic differences between therapy-related ALLs with 11q23 abnormalities and those without, we performed a literature review encompassing 48 therapy-related ALL cases with detailed cytogenetic data (Table 2). The latency period between primary malignancy and the onset of therapy-related ALL in the non-11q23 group was significantly longer than in the 11q23 group (36 vs 19 months; P < .05). It is interesting that the median latency period for our 6 non-11q23 cases was 72 months, which was considerably longer. Auxenfants et al19 in 1992 studied 9 cases with t(4;11)(q21;q23) and reported that the interval between treatment for the first cancer and diagnosis of therapy-related ALL was less than 24 months. Shivakumar et al13 pointed out that the interval for development of therapy-related ALL was significantly longer in therapy-related ALL cases with a complex karyotype (without 11q23 abnormalities) than in therapy-related ALL cases with all other aberrations. These findings suggest that the different karyotypes, or resultant different gene aberrations, may have a certain role in the development of therapy-related leukemia. In therapy-related AML, studies showed that polymorphisms that reduced the enzymatic activity of thiopurine methyltrans-ferase20 and polymorphisms in glutathione S-transferase P121 and NAD(P)H:quinone oxidoreductase (NQO1 gene)22 were associated with increased risk of the disease. Furthermore, increased susceptibility to therapy-related AML has also been linked to polymorphisms of DNA repair genes.23 However, as far as we know, no particular related mutations or polymorphisms were reported in the literature for therapy-related ALL cases.
It is well known that the most frequent chromosomal abnormalities in de novo ALL are t(12;21)(p13;q22) and hyperdiploidy of more than 50, which account for up to 29% and 25% of de novo ALL cases, respectively.8 However, in the review of therapy-related ALL cases, the 11q23 abnormality is the most frequent chromosomal abnormality. Prior topoisomerase II inhibitor treatments are highly associated with therapy-related AML cases harboring 11q23 abnormalities.24 Based on our review, such a relationship can also be seen in the majority of therapy-related ALL. Among the 25 therapy-related ALL cases with 11q23 abnormalities in our literature review, 21 cases (84%) had history of receiving topoisomerase II inhibitors. However, nearly one third (8/29 [28%]) of the patients who received topoisomerase II inhibitor treatment developed therapy-related ALL without 11q23 abnormalities, indicating mechanisms other than 11q23 aberration may also contribute to the therapy-related ALL induced by topoisomerase II inhibitors.25 We also noted that the overall median latency period of the 22 non-11q23 therapy-related ALL cases, including 16 reported cases from the literature and our 6 present cases, was approximately 48 months. This latency is similar to the latency period for therapy-related AML with exposure to alkylating agents or radiation (5–7 years) rather than exposure to topoisomerase II inhibitors.2 Because the latter is highly correlated with therapy-related AML with 11q23 abnormalities and with a shorter latency in therapy-related AML1 and ALL,19 we suggest that these cases of therapy-related ALL without 11q23 abnormalities may occur regardless of prior exposure to topoisomerase II inhibitors.
The 34-year-old man (case 3) had an unusual balanced translocation involving chromosomes 8 and 14: t(8;14) (q11.2;q32). The chromosomal abnormality was first reported in 1982.26 It is distinct from the more common translocation involving c-MYC on chromosome 8q24 [t(8;14)(q24;q32)], which is associated with classical Burkitt lymphoma with a mature B cell phenotype.27 In the literature, only 39 other cases of t(8;14)(q11.2;q32) were reported,7,26,28–39 and all of them were de novo ALLs and occurred mainly in children. According to the study, t(8;14)(q11.2;q32) defined a distinct cytogenetic entity of precursor B ALL in patients with a mean age of 14 years (median, 11.6 years), a high frequency of Down syndrome (trisomy 21),39 and a good rate of complete remission.38 The case we report herein suggests that t(8;14) (q11.2;q32) can also be seen in adult therapy-related ALL cases, which may suggest another possible pathogenesis of this disease.
The prognosis in de novo adult ALL is poor in comparison with about 80% disease-free survival for childhood ALL. Only one third of adults with ALL survive beyond 5 years.40,41 However, according to Moorman et al,42 the prognosis in therapy-related ALL with 11q23 abnormalities was even more unfavorable, with no patients surviving more than 5 years. In our review, however, there was no significant survival difference between the 11q23 group and non-11q23 group (median survival, 2.6 and 2.5 months, respectively), indicating that the overall prognosis in therapy-related ALL is poor, regardless of the presence or absence of 11q23 abnormalities.
- radiation therapy
- acute lymphocytic leukemia
- chromosome abnormality
- karyotype determination procedure
- chemotherapy regimen
- topoisomerase ii
- adult t-cell lymphoma/leukemia
- translocation (genetics)
- myeloid leukemia, acute, therapy related
- mll gene
- secondary leukemia
- t-cell leukemia, acute
- balanced chromosomal translocation