Persistent genomic instability in peripheral blood lymphocytes from hodgkin lymphoma survivors

Authors


Abstract

Advances in cancer treatment have led to an increase in patient survival. However, exposure to genotoxic chemotherapeutic agents and ionizing radiation may induce persistent genetic damage in cancer survivors. In this study, we detected genomic instability in chromosomes of peripheral blood lymphocytes from Hodgkin lymphoma patients, 2–17 years after MOPP (nitrogen mustard, Oncovin, procarbazine, and prednisone) chemotherapy with or without radiotherapy. Samples were obtained from 11 healthy individuals, 5 pretreatment patients, and 20 posttreatment patients. Cytogenetic analysis with GTG banding was performed on 1,000 lymphocyte metaphases per donor to identify genomic instability, including numerical and structural chromosomal aberrations, at a resolution of 10 Mb across the entire genome. Our results showed that anticancer treatment did not induce significant differences in the frequency of aneuploidy among the three study groups. However, 1 of the 11 healthy individuals, and 13 of the 20 posttreatment patients had a high frequency of chromosomal breaks and gross chromosomal rearrangements. The types of aberrations observed were random and complex, consistent with persistent genomic instability that was induced by cancer treatment. Clonal expansion of cells with chromosomal lesions was observed in one posttreatment patient only. These findings show that anticancer treatments induce persistent genomic instability, but not aneuploidy. Chemotherapy may affect genes with a role in DNA damage surveillance or repair, which in turn allows the accumulation of nontargeted structural chromosomal damage in future generations of cells. This genomic instability may facilitate the development of second malignancies in Hodgkin lymphoma survivors. Environ. Mol. Mutagen. 2012. © 2012 Wiley Periodicals, Inc.

INTRODUCTION

Hodgkin lymphoma (HL) is a malignant neoplasm affecting the lymphoid system and accounts for ∼30–40% of all malignant lymphomas [Magrath and Johnson,1993; Santoro and Viviani,1993; Diehl et al.,2003]. Worldwide, HL is the third most common pediatric cancer; in Mexico, HL ranks third among malignant neoplasms [Magrath and Johnson,1993; Rivera Luna,2006]. Various drug combinations have been used to treat HL, including MOPP (nitrogen mustard, Oncovin, procarbazine, and prednisone), ABVD/P (adriamycin, bleomycin, vinblastine, and dacarbazine or prednisone), and combinations of these schemes with or without radiotherapy [Hudson and Donaldson,1997; Thomson and Wallace,2002]. These therapeutic regimens result in 10 years disease-free survival in 80% of HL patients [Santoro and Viviani,1993; Muwakkit et al.,1999; Sieniawski et al.,2007; Constine et al.,2008]; currently, there are 12 million cancer survivors in the US, representing ∼4% of the total population [Arden-Close et al.,2010; Fairley et al.,2010].

Anticancer regimens are known to induce adverse side effects in HL survivors, including germ cell cytotoxicity, manifested as transitory or persistent azoospermia, and oligospermia, [Meistrich,1993]. In addition, second cancers have been detected in 20% of HL survivors [Friedman et al.,2010], including leukemia in 25%, non-Hodgkin lymphoma in 17% and solid tumors in 58% of this population [Jenkin et al.,1996; Hudson and Donaldson,1999; Linch et al.,2000; Foss Abrahamsen et al.,2002; Thomson and Wallace,2002]. The second cancer appears about 10 years after treatment with radiotherapy or chemotherapy alone. Chemotherapy generally confers greater risk, especially when treatment includes nitrogen mustard and procarbazine [Leone et al.,2010], which induce genomic instability in animal models, in both somatic and germ cells [Balis et al.,1993; Witt and Bishop,1996; Frias, 2002; Gobbi et al.,2006; Worrilow et al.,2008].

Genomic instability (GIN) refers to a range of genomic abnormalities, from point mutations to chromosomal rearrangements. These abnormalities can be classified based on the type of genomic defect. Chromosomal instability (CIN) refers to aneuploidy, which is the gain or loss of chromosomes. Micro- and mini-satellite instability (MIN) refers to expansions and contractions of repetitive DNA sequences. Gross chromosomal rearrangement (GCR) refers to events that involve a defect in the processing of DNA fragments and a change in the genetic linkage of two DNA fragments, leading to chromosomal deletions, duplications, inversions, translocations, and other structural abnormalities [Aguilera and Gomez-Gonzalez,2008].

GIN is a well-documented consequence of exposure to ionizing radiation and higher frequency of cells with persistent GIN has been observed in cells exposed to chemical agents or radiation in vitro [Huang et al.,2003; Little,2003; Lorimore et al.,2003; Preston,2005]. However, little work has been performed to investigate GIN as part of the long-term effects of cancer treatments including MOPP on cells from human patients. The purpose of this study was to: (a) investigate the frequency and types of GIN present in the peripheral blood lymphocytes (PBLs) from HL patients 2–17 years after MOPP combination chemotherapy with or without radiotherapy; and, (b) investigate whether cells that were directly damaged by anticancer treatment, generated cells with the same aberration or whether treatment generated random aberrations that are consistent with persistent nontargeted effects of chemotherapy treatment.

MATERIALS AND METHODS

Study Population

Peripheral blood samples were obtained from 11 healthy individuals (Group I) and 25 HL patients. Only men were included because we were also interested in the cytotoxic effects of chemotherapy on male germ cells. Among the HL patients, 5 HL patients were sampled before treatment with anticancer agents (Group II), and 20 patients were sampled 2–17 years after MOPP chemotherapy with or without radiotherapy (Group III). Among Group III patients, 10 men produced sperm (Group IIIa), and 10 were azoospermic (Group IIIb; Table I). The study participants were patients at the Cancer Survivor Clinics of the Oncology Service of the Instituto Nacional de Pediatria, Instituto Nacional de Cancerologia and Hospital General de Mexico. The Research and Ethics Committees approved the study. Both healthy individuals and patients signed informed consent forms and answered questionnaires regarding reproductive and medical history and lifestyle habits.

Table I. Characteristics of the Study Groups
CodeHL stageAge at blood samplingChemotherapy/cyclesRadiotherapy/GyaYears after treatment
  • a

    All irradiated patients had gonadal protection with lead plates. SD, supradiaphragmatic; PA, paraaortic.

Healthy individuals (Group I)
NL4-01 20   
NL7-01 20   
NL8-01 20   
NL9-02 20   
NL10-02 20   
NL11-03 18   
NL12-03 21   
NL13-03 20   
NL14-03 24   
NL16-03 25   
NL18-03 28   
Pretreatment HL patients (Group II)
Pre-Tx35-04IVB23   
Pre-Tx39-04IVB28   
Pre-Tx45-04IIB52   
Pre-Tx51-04IIA20   
Pre-Tx52-04IIA16   
MOPP-treated HL patients (Group III) with sperm (Group IIIa)
MOPP5-02III20MOPP/8 ABVP/5SD/3513
MOPP7-02III20MOPP/5 ABVP/4SD/42 PA/2512
MOPP12-02III18MOPP/5 ABVP/4SD/60 PA/2015
MOPP19-02IIIB27MOPP/6 ABVD/6 13
MOPP21-02IIIB20MOPP/6 ABVP/6PA/3013
MOPP23-02I19MOPP/5 ABVP/1SD/2013
MOPP24-03I15MOPP/10 ABVP/5SD/2015
MOPP29-03III19MOPP/1 COPP/1 ABVD/4SD/21.6 PA/21.68
MOPP31-03IIA22MOPP/4 ABVD/4SD/3517
MOPP36-04III18MOPP/2 ABVD/8SD/21.6 P A/21.67
Azoospermic (Group IIIb)
MOPP1-01IV19MOPP/7 COPP/3 9
MOPP9-02IV20MOPP/8 ABVP/8 12
MOPP13-01II16MOPP/6 COPP/1SD/3610
MOPP15-01IV17MOPP/8 COPP/1SD/40 PA/4010
MOPP22-02III20MOPP/4 ABVP/5SD/45 PA/3011
MOPP26-03IIIA16MOPP/4 ABVD/3SD/25.22
MOPP28-03IIA17MOPP/3 ABVD/3 9
MOPP30-03IIB38MOPP/4 ABVD/3 7
MOPP41-04IVB24MOPP/9 16
MOPP43-04IIIB20MOPP/6 ABVD/6SD/25 PA/21.610

Group I participants that smoked ten or more cigarettes per day, had consumed drugs, or had reproductive, hormonal, urogenital diseases, or cancer were excluded. Patients were chosen according to the following inclusion criteria: (a) HL was diagnosed following strict medical and laboratory criteria, (b) MOPP/ABVD/P treatment was administered at least 2 years prior to the initiation of the study, and (c) the patients were 18-years old or older. Patient exclusion criteria included the following: (a) the patient smoked ten or more cigarettes per day, (b) the patient had received a blood transfusion within the last 6 months, and (c) the patient had a history of a cancer relapse or secondary malignancy.

Treatment Schemes

Patients included in this study were treated with: MOPP (Nitrogen mustard 6 mg m−2 days 1 and 8, Oncovin 1.5 mg m−2 days 1 and 8, procarbazine 100 mg m−2 days 1–14, and prednisone 40 mg m−2 days 1–14. Repeated every 28 days), or combined MOPP/ABVD/P (Nitrogen mustard 6 mg m−2 day 1, Oncovin 1.5 mg m−2 day 1, procarbazine 100 mg m−2 days 1–7, and prednisone 40 mg m−2 days 1–7; Adriamycin 25 mg m−2 day 15, bleomycin 10 U m−2 day 15, vinblastine 6 mg m−2 day 15, and dacarbazine 375 mg m−2 day 15 or prednisone 40 mg m−2 day 15. Repeated every 28 days) or COPP regimen (Cyclophosphamide 600 mg m−2 days 1 and 8, Oncovin 1.5 mg m−2 days 1 and 8, procarbazine 100 mg m−2 days 1–14, and prednisone 40 mg m−2 days 1–14. Repeat every 28 days). Recruited patients received one of the following five MOPP chemotherapy regimens: (1) MOPP (nine cycles), (2) combined MOPP (four to eight cycles)/ABVP (one to eight cycles), (3) combined MOPP (two to sixcycles)/ABVD (three to eight cycles), (4) combined MOPP (six to eight cycles)/COPP (one to three cycles), and (5) combined MOPP (one cycle)/COPP (one cycle)/ABVD (four cycles). Irradiation treatment consisted of 20–60 Gy with Co60 to the supradiaphragmatic and/or paraaortic areas (Table I).

Lymphocyte Culture

Heparinized peripheral blood samples were cultured in 10 ml of RPMI 1640 medium (Gibco, USA) supplemented with 10% bovine fetal serum (Gibco, USA) and 0.2 ml of phytohemagglutinin (Gibco, USA). Cultures were incubated at 37°C and harvested after 48 hr. Prior to harvesting, 0.2 ml of colchicine (1 mg ml−1; Sigma, USA) was added to the cultures for 3 hr to arrest cells in mitosis. The cells were then treated with a hypotonic KCl solution (0.075 M; Sigma) at 37°C for 25 min, followed by fixation in methanol/acetic acid (v/v, 3:1; Merck, Germany). Fixed cells were dropped onto cold slides and air-dried. For GTG banding (Giemsa banding with trypsin), the slides were treated with a solution containing 0.015 g trypsin (Gibco, USA) and 0.01 g EDTA (Sigma, USA) diluted in Sörensen buffer (KH2PO4, Na2HPO4; Sigma). Chromosomes were stained with 5% Giemsa (Gibco, USA). Each slide was coded by a person not involved in the scoring and scored by a second person who was blind to the origin of the slide.

Chromosome Analysis

In this study, GTG banding was used for cytogenetic analysis due to its ability to detect numerical abnormalities and structural chromosomal aberrations at a resolution of 10 Mb across the entire genome. In addition, this technique enabled us to identify cells derived from a clonal population and carrying the same cytogenetic aberration.

For each sample, 15 cells were karyotyped to ensure that every individual included in this study had a normal constitutional karyotype. Cytogenetic analysis to detect GIN was performed by one expert cytogeneticist scorer, without computer assistance, by pairing-up homologs visually. A total of 1,000 GTG-metaphase cells were analyzed per sample. For each metaphase, 450–500 bands were resolved at 1,000 × to detect numerical and structural chromosomal abnormalities. To avoid false positives, the following analysis criteria were applied to each sample: (a) metaphase cells must have had adequate chromosomal dispersion, (b) only nonadjacent cells were analyzed, and (c) when a metaphase cell with aneuploidy was identified, all metaphase cells within a distance of 10 times that of the field were analyzed for the presence of the chromosome that was missing or extra in the original cell. If this chromosome was found in any of the nearby cells, neither of the two cells was considered aneuploid. This latter criterion was designed to avoid artifacts that may have occurred during chromosome preparation. Chromosomal structural changes were detected by analyzing the G banding pattern of each chromosome; we quantified unstable chromosomal damage as chromatid breaks, chromosome breaks, and acentric fragments and GCRs as translocations, deletions, duplications, inversions, marker chromosomes, and complex figures; the last two abnormalities may represent complex chromosomal rearrangements involving three or more breaks.

Statistical Analysis

Calculations were performed with the statistical analysis system software SPSS 16.0 (SPSS, Chicago, IL). To compare the frequencies of aneuploidy, breaks and GCR between pair of groups, we used the nonparametric Mann–Whitney-U test and we used the Kruskal–Wallis test for comparison among groups. Regression analyses were done for patient's age, chemotherapy cycles, HL stage, years after treatment and GCR. All tests were considered statistically significant when P < 0.05.

RESULTS

Study Population

Table I outlines patient characteristics for each group. Group I consisted of healthy donors, with a mean age of 22.0 ± 7.8 (mean ± SD) years (range, 18–28 years). The pretreatment group (Group II) included patients with a mean age of 27.8 ± 12.7 years (range, 16–52 years). The posttreatment groups, Group IIIa and IIIb, included patients with mean ages of 20.3 ± 2.8 years (range, 15–27 years) and 20.7 ± 6.2 years (range, 6–38 years), respectively. Mean time between the end of treatment and this study was 10.8 years (range, 2–17 years). Chemotherapy included MOPP treatment in all patients. Fifteen patients received MOPP/ABVP(D) with a variable number of cycles (1–10). In three patients, chemotherapy consisted of one to eight cycles of MOPP/COPP. One patient received one to four cycles of MOPP/COPP/ABVD, and one patient was treated only with nine cycles of MOPP. Six of the 20 patients did not receive radiotherapy, as noted in Table I.

None of the patients had been previously exposed to clastogenic or aneugenic agents, either via previous medical treatment or personal habits. The mean ages of the patients were similar among the groups, and there were only two patients over 35 years in age, one in Group II and one in Group IIIb.

Aneuploidies in PBLs of MOPP Patients

We found aneuploidies or CIN in all studied groups. The mean numbers of aneuploidies per 1,000 cells were as follows: Group I, 20.3; Group II, 11.6; and Group III, 22.2 (Table II). There were no statistically significant differences in the distributions or average frequencies of CIN among the three studied groups (Table III). Chromosomes 21 and Y were the most frequently involved in trisomies, followed by chromosomes 5 and X. Chromosomes 21, 22 and Y were the most frequently lost, followed by 12, 18, 19, and 20. In general, chromosome Groups A, B, and C and chromosome X were rarely lost in the individuals studied, while acrocentric Groups D and G and chromosome Y were lost at higher frequencies (data not shown).

Table II. Average Frequency of Aneuploidies/1,000 Cells in Healthy Individuals and HL Patients
CodeHypodiploidies mean (95% CI)Hyperdiploidies mean (95% CI)Total aneuploidy mean (95% CI)
  1. CI, confidence intervals.

Group I19.8 (10.9–28.6)0.55 (0–1.2)20.3 (11.5–29.1)
Group II11.2 (0.18–22.2)0.4 (0–1.5)11.6 (0.47–22.7)
Group III21.7 (13.2–30.2)0.5 (0.11–0.89)22.2 (13.6–30.8)
Table III. Average Frequencies of Numerical and Structural Chromosomal Aberrations per Groupa
Type of alterationGroup I; healthy individualsGroup II; pretreatment HL patientsGroup III; posttreatment HL patientsP valueb
  • a

    Frequencies per 1,000 cells.

  • b

    P value obtained by Kruskal–Wallis test, when comparing average of the frequency of aberrations/individual among the three studied groups.

  • c

    Chromosome and chromatid breaks.

  • d

    Number of events/cell = total number of chromosome breaks per cell. In the case of chromosome markers without a specific description, we assumed that there were two breaks per marker; CIN, chromosomal instability (aneuploidy); CI, confidence intervals.

Aneuploidies (CIN) mean (95% CI)20.3 (11.5–29.1)11.6 (0.47–22.7)22.2 (13.6–30.8)0.55
Structural aberrations (Breaksc) mean (95% CI)0.73 (0–1.47)0.2 (0–0.76)1.95 (0.8–3.1)0.13
Gross chromosomal rearrangements (GCR) mean (95% CI)0.18 (0–0.59)06 (1.37–10.6)0.002
Cells with GCR mean (95% CI)0.18 (0–0.59)02.05 (0.9–3.2)0.002
Chromosome breaks involved in GCRd mean (95% CI)0.36 (0–1.17)011.9 (3–20.9)0.002

Breaks and Gross Chromosomal Rearrangements in PBLs of HL Patients and Healthy Individuals

Breaks were detected in 4 out 11 individuals in Group I, 1 out 5 individuals in Group II, and 12 out 20 individuals in Group III (Tables III and IV). GCR were present in 1/11 healthy Individuals and in 13/20 posttreatment patients, and were not observed in noncancerous PBL pretreatment HL patients. The frequencies of individuals with GCR were significantly different among studied groups (P = 0.005), as well as the frequencies of GCR/1,000 cells among groups (P = 0.002) (Tables III and IV). One of the two abnormalities found in the healthy individual is a translocation between chromosomes 7 and 14, which is found in PBL of normal people with a frequency of 1/144 [Hecht et al.,1979].

Thirteen of the MOPP-treated patients in Group III had levels of GCR that ranged from one to eight cells. Four of these patients presented with only one aberrant cell each, nine patients had multiple aberrant cells with different chromosomes involved, and only one patient (MOPP28) showed a clonal event in which three cells contained the same deletion del(17)(p11.2) (Table IV). These findings are consistent with a broad range of DNA breakage events, ranging from 2 to 38 events per cell. The variation in aberration types among cells is consistent with GIN persistence, which differs from the clonal expansion of cells that are directly damaged during MOPP therapy.

Table IV. Frequency of Chromosomal Structural Alterations per 1,000 Cells
CodeBreaksaGross chromosomal rearrangements (GCR)
ChrChtGCR unbalancedGCR balancedACbEvc
  • a

    Breaks = Chromosome (Chr) and chromatid (Cht) breaks.

  • b

    Number of abnormal cells with the same structural chromosome alteration.

  • c

    Number of events/cell = total number of chromosome breaks per cell. In the case of chromosome markers without a specific description, we assumed that there were two breaks per marker.

  • d

    Patients Pre-Tx35-04 and Pre-Tx39-04 had the Hodgkin's lymphoma-related alterations del(7)(q36)[3] and del(21)(q22)[15], respectively (Shaffer et al.,2009).

  • e

    Patients treated only with chemotherapy.

Group I (healthy individuals)
NL4-0100  00
NL7-0121  00
NL8-0101  00
NL9-0211  00
NL10-0200  00
NL11-0300  00
NL12-0300  00
NL13-030246, XY, del(17)(p?) 12
 46, XY, t(7;14)(q34;q11)12
NL14-0300  00
NL16-0300  00
NL18-0300  00
Frequency/1,000 cells0.730.180.180.36
Group II (pretreatment patients)
Pre-Tx35-04d00  00
Pre-Tx39-04d00  00
Pre-Tx45-0400  00
Pre-Tx51-0410  00
Pre-Tx52-0400  00
Frequency/1,000 cells0.2000
Group IIIa (posttreatment HL patients/with sperm)
MOPP5-0223  00
MOPP7-020046, XY, add(20)(p?) 12
MOPP12-022046, XY, add(1)(p?)t(2;9)(?;?) 14
MOPP19-02e2045, multiple alterations 138
  47, XY, +mar 24
   45, XY, inv(1)(p22p36.3), -2112
MOPP21-0241 46, XY, inv(20) (p11.2p12)12
MOPP23-020046, XY, -20, del(7)(q35), +mar 14
  47, XY, +mar 12
MOPP24-034245, XY, t(9;17) 12
  47, XY, +mar 12
  46, XY, del(6)(q24?) 12
  46, XY, multiple alterations 14
  46, XY, del(17)(q25) 12
  46, XY, multiple alterations 14
   46, XY, inv(11)12
   46, XY, t(4;9)(p16;q32)12
MOPP29-0300 46, XY, inv(12)(p11.2q24.3)12
  48, XY, t(1;2)(q21;q35),48, XY, t(1;2)(q21;q35),18
  del(21)(q11.2q21),+mar1, +mar2del(21)(q11.2q21),+mar1, +mar2  
MOPP31-0300  00
MOPP36-040148, multiple alterations 132
  46, XY, del(X)(q28) 12
   46, XY, inv(14)(q11.2q32)12
Subtotal group IIIa; Frequency/1,000 cells26.22.212.4
Group IIIb (posttreatment HL patients/azoospermic 
MOPP1-01e10  00
MOPP9-02e10  00
MOPP13-0234  00
MOPP15-0243  00
MOPP22-021045, XY, pbe. t(2;12) 18
  (q37;q13), -12, -19,   
  −19, -20, +mar1,+mar2,   
  +mar3   
  44, multiple alterations 19
  45, X, multiple alterations 111
  45, XY, -8, -10, +mar 12
MOPP26-032046, XY, -12, +mar 12
  44, Y, multiple alterations 113
  46, XY, multiple alterations 111
  48, Y, multiple alterations 118
  48, XY, +mar1, +mar2 14
  46, Y, -17, +mar1, +mar2 14
  47, XY, multiple alterations 110
  46, XY, multiple alterations 19
MOPP28-03e0046, XY, del(17)(p11.2p11.2) 36
MOPP30-03e00 46, XY, t(5;5)(q21;q31)12
   46, XY, t(1;3)(p10;q10)12
   46, XY, t(6;19)(q15;p13.3)12
MOPP41-04e00  00
MOPP43-040047, XY, +mar 12
Subtotal group IIIb; frequency/1,000 cells1.9 5.81.911.5
Total group III; frequency/1,000 cells1.95 62.0511.9

We did not find significant differences in breaks or GCR between normospermic and azoospermic MOPP patients or when MOPP treatment was combined with radiotherapy. When the period of time after treatment was considered, 0.016 events per cell were observed in patients treated 2–11 years before the study and 0.0078 events per cell, in patients treated 12–17 years before the study. However, regression analysis showed that time after treatment alone did not significantly affect the frequencies of GCRs (P = 0.382).

DISCUSSION

We found that patients who had been treated up to 17 years earlier with MOPP, either with or without radiotherapy, presented GIN as demonstrated by the occurrence of structural chromosomal damage or GCR. However, we found no evidence of increased chromosomal number aberrations or CIN with respect to controls. The types of aberrations observed were random and complex, consistent with persistent genomic instability events that were induced by chemotherapy and radiotherapy treatment in HL patients. These results suggest that anticancer treatment may affect genes with a role in DNA surveillance or repair, which, in turn, allows the accumulation of nontargeted structural chromosomal damage in future generations of cells.

The Anticancer Treatment MOPP Does Not Alter the Frequency of Aneuploidy

To our knowledge, there are currently no published reports on the effects of MOPP chemotherapy (with or without radiotherapy) on the frequency of aneuploidy in human lymphocytes. In this study, the total aneuploidy frequencies were 2.0% for healthy individuals, 1.1% for pretreatment HL patients and 2.2% for posttreatment HL patients; no significant differences were found in CIN among the three groups studied (Tables II and III). This result is consistent with the lack of persistence of aneuploidy in germ cells after anticancer treatment [Robbins et al.,1997; Frias et al.,2003].

Hyperdiploidy was the least frequent aneuploidy found among the three groups studied (Table II), and its frequency in healthy individuals was up to eight times lower than those reported previously (0.1–4.17%) [Juberg et al.,1985; Leopardi et al.,2002]. The frequency of hypodiploidy in the lymphocytes of healthy individuals observed in this study (Table II), was also four times lower than those found in other studies [Cimino et al.,1986; Eastmond and Pinkel,1990]. These differences are most likely the result of the very strict criteria employed during analysis to avoid cells to be mistakenly identified as aneuploid due to technical artifacts.

Although some components of the chemotherapy studied here are able to induce aneuploidy, we did not find CIN in Group III patients. Even though aneuploidy cells may have been generated immediately (or during) treatment, they may have disappeared from the population over the time, presumably because the genetic imbalances created by aneuploidy imposed a selective disadvantage [Yuen and Desai,2008].

Complex Structural Chromosome Damage is Increased in HL Survivors

Cells with one chromatid or chromosome break, considered as unstable damage, were observed in all three study groups (Tables III and IV). We found a higher frequency of chromosome breaks in PBLs from HL patients up to 17 years after treatment. Complex GCR was the most frequently type of aberrations found in HL survivors and it was significantly higher in Group III with respect to the other groups (Tables III and IV). All GCR found in this study were stable aberrations [Shaffer et al.,2009] such as translocations, deletions, inversions, marker chromosomes and complex rearrangements, while asymmetrical exchanges (dicentrics, acentrics, and ring chromosomes) were not found in any of the studied groups.

A high frequency of chromosomal aberrations in pretreatment HL patients has been reported [M'Kacher et al.,2003], suggesting that these patients harbor basal chromosome instability. In our study, we detected cells with GCR in two of five pretreatment patients that consisted of clones of 3 and 15 cells with del(7)(q36) and del(21)(q22), respectively (Table IV). These chromosomal abnormalities have been related to cytogenetic changes in cancerous Reed-Sternberg Hodgkin cells [Barrios et al.,1988; M'Kacher et al.,2003]. Interestingly, these cells may be related to viral infections which even 2 years after treatment, may generate chromosomal instability in the peripheral blood of patients [M'Kacher et al.,2010]. In this study, these rearrangements were considered to be aberrations related to the disease and were discarded as a possible GIN in noncancerous lymphocytes [Nagai et al.,1986; Dennis et al.,1989].

GCRs can be transmitted through mitosis of hematopoietic stem cells (HSCs); these types of events, therefore, may be detected many years after exposure as clonal events. In our study, only one clone with the same gross chromosomal aberration was found, (1 of 13 individuals with GCR in Group III), whereas nonclonal GCRs were found in 12 of 20 individuals within the same group. All of these 12 individuals displayed chromosomal structural aberrations 2–17 years after treatment, consistent with persistent nontargeted effects of anticancer treatment. Because of the time that elapsed between treatment and the initiation of the study, it is unlikely that the high incidence of breaks and their rejoined product GCR was due to direct chromosomal damage produced by the treatment.

These results suggest that the origin of the observed GIN is due to mutations in long-term self-renewal HSCs, that affects one or more of the main mechanisms that contribute to maintain genomic stability: (a) high-fidelity of DNA replication in S-phase, (b) accurate distribution of chromosomes among daughter cells during mitosis, (c) error-free repair of DNA damage throughout the cell cycle, and (d) cell cycle progression and checkpoint control [Wright,1999; Shen,2011]. If the anticancer treatment induced mutations affecting any of these mechanisms, HSCs with a mutator phenotype could arise in bone marrow and be maintained for long periods of time; these abnormal HSCs may constantly generate blood cells with a heterogeneous spectrum of chromosomal abnormalities. In this study, only one patient showed a clone with the same chromosomal aberration del(17)(p). This clone may be originated by additional genomic changes that conferred a selective advantage that permitted the mutated HSCs to evolve as a lymphocyte clone.

We found a large variation in the number of events (total number of chromosome breaks per cell) that lead to GCR among patients. Specifically, 7 of 20 patients did not have gross chromosomal damage, whereas 13 of 20 displayed GCR and there was no correlation with personal habits, age, medical records or time after treatment among these patients. Consistent with our findings, M'Kacher et al. [2003] analyzing eight patients treated with MOPP/ABV found no correlation between the frequency of GCR and the patient's age, time after treatment, disease stage, or chemotherapy dose, indicating that there are clear individual variability in the response to anticancer treatment [M'Kacher et al.,2003]. Individual variance is most likely caused by differences in DNA-metabolizing enzymes or DNA repair capabilities [M'Kacher et al.,2003; Bonassi et al.,2004; Hagmar et al.,2004; Allan and Travis,2005].

We found a significantly higher frequency of GCRs in posttreatment patients than the healthy and pretreatment group (Table IV) suggesting that radiation and chemotherapy caused permanent damage to HSCs in these patients. Studies have shown that cells exposed to radiation may develop mutations and chromosomal aberrations that persist for several cell generations (over a year) [Wright,1999]. In this study, GCRs were found in 10 of 14 patients treated with both chemotherapy and radiotherapy, and in 3 of 6 treated only with chemotherapy, without significant difference in the frequency of GCR between the two groups. The immediate damage to cells may be greater due to the combined treatment; however, in survivors treated a long time before analysis (12–24 years), both radiotherapy and chemotherapy might affect genes related to DNA surveillance or repair, and once these genes are mutated, new cells with nontargeted chromosomal damage are continuously produced in sensitive patients [Wright,1999].

The development of secondary neoplasms as a result of the treatment of primary malignant tumors is frequent; the highest excess risk is reported for HL patients among survivors of childhood cancer [Friedman et al.,2010]. Cancer risk is associated with many factors, one of the most important being chromosomal damage. Several studies have indicated a clear association between cancer risk and the frequency of GCRs in PBLs, suggesting that GCRs in PBLs can be used as a biomarker of cancer risk [Blanco et al.,2001; Heng et al.,2006; Norppa et al.,2006; Boffetta et al.,2007]. In our patients, the majority of GCRs were observed in only one cell (nonclonal); however, GCRs are the primary source of genomic variation that can lead to unstable cell populations and, ultimately, initiate the conversion of a healthy cell into a cancer cell [Heng et al.,2006]. In our study, we found 6 GCRs per 1,000 lymphocytes (Table IV) or 6,000 GCRs in 106 cells. If the chromosomal damage found in PBLs is similar to the damage that occurs in other tissues, the frequency of GCRs may be an indication of cancer risk in any tissue [Norppa et al.,2006; Boffetta et al.,2007].

In the post treatment group, we found three chromosomal alterations that have been identified in patients with hematological neoplasias, or in patients with diseases associated with a high risk of developing cancer like myelodysplastic syndromes (MDS). The del(7)(q35) is present in MDS and acute myeloid leukemia (AML). The inv(14)(q11q32.1) is present in patients with T-cell leukemia/lymphoma, and the del(17)(p11), which was the only clonal alteration detected in this study, has been detected in MDS, blastic crisis of chronic granulocytic leukemia, de novo AML, or t-AML associated with alkylating agents, as well as in patients with polycythemia vera or essential thrombocytopenia treated with hydroxyurea [Mitelman,1991]. As a consequence of GCR, it is likely that oncogenes and tumor suppressor genes may be misregulated and facilitate the development of second malignancies [Smith et al.,1992; Bilban-Jakopin and Bilban,2001].

Previous studies that analyzed chromosomal abnormalities using nonbanded chromosomes or FISH, have found a significantly increased frequency of structural chromosomal aberrations in HL patients treated with MOPP chemotherapy in combination with radiotherapy [Papa et al.,1984; Smith et al.,1992; Bilban-Jakopin and Bilban,2001; M'Kacher et al.,2003; Ryabchenko et al.,2003]. Smith et al. [1992] analyzed five HL patients 12–21 years after treatment with chemotherapy/radiotherapy, using chromosome 4 painting probe and found a high frequency of chromosomal translocations (mean = 0.016). Similarly, M'Kacher et al. [2003] studied 50–200 cells from eight patients treated with chemotherapy including MOPP/ABV, either with or without radiotherapy using FISH painting probes for chromosomes 1, 3, and 4, and found a twofold higher frequency of chromosomal aberrations per cell in HL patients 2 years after treatment, as compared with pretreatment group. In our study, we analyzed all 46 chromosomes in 1,000 cells per patient, and found a 30-fold higher frequency in posttreatment patients as compared with healthy individuals and pretreatment group. These results show that the use of GTG banding by allowing the identification of each of the 46 chromosomes permits the detection of not only more cytogenetic alterations, but also the specific chromosomes involved in structural aberrations and the detection of balanced rearrangements.

In conclusion, we found genomic instability in lymphocytes of HL patients treated with MOPP many years after the end of chemotherapy. As these patients are very susceptible to developing secondary cancers [Van der Velden et al.,2003], our results suggest that genomic instability may affect the function of genes important for preventing cancer development. In particular, balanced exchanges that may modify the expression of oncogenes, or tumor suppressor genes but allow the survival of carrier cells without lethal chromosomal damage (i.e., unbalanced exchanges such as dicentrics) may represent a mechanism by which secondary cancers arise in these patients. Real-time PCR studies to detect populations of cells with chimeric proteins that may confer an increased risk of developing secondary cancers after HL treatment is an avenue for future research to confirm this hypothesis.

Acknowledgements

The authors thank Drs. Andy Wyrobek and Francesco Marchetti for the valuable discussion on the contents of this article.

Ancillary