Abstract
Background/Aim: A continued increase in the incidence of therapy-related myeloid neoplasms (t-MN) is expected due to the improvement of chemotherapeutic treatments for solid and haematological malignancies. The use of 5-azacytidine (AZA) is emerging in these patients. We, therefore, analyzed the outcome of patients with t-MN ineligible for intensive chemotherapy treated in the front-line with AZA. Patients and Methods: We retrospectively collected clinical data from consecutive patients with t-MN treated in the front-line with AZA at five Haematology Centers. Response to therapy, overall survival (OS) and safety were considered. Results: The overall response rate was of 35.7% with a median OS of 9.6 months. Patients who were heavily pre-treated for their primary malignancy (more than 3 lines of chemotherapy) presented a significant inferior OS (4.9 months). The principal reported toxicity was haematological with severe infections occurring in a minority of patients. Fatigue was the most common extra-haematological toxicity. Conclusion: New aspects emerged on the management of t-MN. AZA may represent a reasonable choice for patients ineligible for intensive treatment, with the exception of heavily pre-treated patients who presented –anyway- a worse outcome.
The improvement of chemotherapeutic treatments for solid and haematological malignancies is leading to an ever-increasing number of late complications. Among these, a higher incidence of therapy-related myeloid neoplasms (t-MN) is expected (1). T-MN account for 10-20% of all cases of myeloid neoplasms and comprise therapy-related myelodysplastic syndromes (t-MDS), acute myeloid leukaemias (t-AML) and myelodysplastic/myeloproliferative neoplasms (2).
T-MN are, by definition, a secondary event due to chemotherapy, radiotherapy or immunosuppressive treatment (2). The two most commonly recognised subsets occur: i) after exposure to alkylating agents and/or ionizing radiation; this form encompasses 70-80% of cases and occurs with a latency of 5-10 years manifesting as t-MDS, often with mutations of chromosomes 5 and 7; ii) after exposure to topoisomerase II inhibitors (20-30%), with a latency of 1-5 years, it manifests mainly as t-AML and may present balanced chromosomal rearrangements, e.g. involving 11q23 and 21q22 (1-3).
In addition to chromosomal alterations, several other forms of molecular damages induced by anti-neoplastic agents have been documented. A modification of the epigenetic regulation by DNA methylation has been described in t-MN and affects mainly promoter regions, which regulate genes involved in the malignant transformation of progenitor cells, angiogenesis, interaction with microenvironment, apoptosis and xenobiotic detoxification (4-6). Other alterations included the expression of genes and proteins leading to impaired mitochondrial function and accelerated telomere shortening, which could favour the pathogenesis of t-MDS/AML in the context of autologous hematopoietic stem cell transplant (SCT) (7-9). T-MDS/AML occur indeed in about 5-10% of patients who undergo autologous SCT for lymphoma (7, 9). Furthermore, the role of individual susceptibility in the development of a t-MN is recently emerging from some studies that have specific polymorphisms of genes involved in the DNA repair and regulation of apoptosis (10-12). Globally, these studies show a picture of genetic and epigenetic alterations, which account for the biological complexity of the disease.
In most cases, disease remission is indeed difficult to reach and prognosis remains poor (1,13-15). As in de novo counterparts, treatment strategies and prognosis depend on patients' characteristics and cytogenetics and range from supportive care to intensive chemotherapy and/or allogeneic SCT (1). The rationale for the use of 5-azacytidine (AZA), a cytidine analogue with hypomethylating activity, in t-MN lies in the aetiopathogenetic mechanisms mentioned above and its efficacy has been recently reported in a few retrospective series (14-18). An overall response rate (ORR) of 28-42% has been documented but data on overall survival (OS) are contradictory, probably due to the retrospective nature of the studies and to different inclusion criteria. Few data are available on the safety of AZA in this subset of patients (18).
In the present study, we analyzed the outcome and the toxicity of front-line AZA in a retrospective cohort of patients with t-MN ineligible for intensive chemotherapy.
Patients and Methods
We retrospectively collected clinical data from consecutive patients with t-MN treated front-line with AZA at 5 Haematology Centers. AZA was administered subcutaneously at a dose of 75 mg/m2/day during 7 days every 28 days. Patients received at least 2 cycles of AZA. Prophylactic anti-emetic treatment with 5-HT3 serotonin receptor inhibitor was performed.
Data collected at the time of treatment commencement included: age, ECOG performance status (PS), diagnosis according to WHO classification, complete blood count, percentage of marrow blasts, karyotype, and transfusion-dependency. Cytogenetics were categorized as good (normal, -Y, del(5), del(20)), poor (complex (3 or more abnormalities), chromosome 7 abnormalities), or intermediate (all others). The newly-proposed risk model from MD Anderson was applied to calculate patient risk only in the subset of MDS. Patients were assigned one point for each of the following characteristics: over 65 years of age, ECOG performance status (PS) higher than 1, monosomy 7 or complex karyotype, WHO MDS subtype of refractory anaemia with ring sideroblasts (RARS) or refractory anaemia with excess blasts-1 or -2 (RAEB-1 or -2), haemoglobin less than 11 g/dl, platelets less than 50x109/l, and transfusion-dependency. Patients were categorized as low risk (0–2 factors), intermediate risk (3-4 factors) and high risk (5 or more factors) (19).
Data on the previous diagnosis of malignancy or immune disease were collected and regarded according to type and time of chemotherapy, radiotherapy or immunosuppressive drug, latency from primary treatment to the onset of t-MN and outcome of primary malignancy.
Response was assessed according to the International Working Group (IWG-2006) criteria (20).
The safety assessment was performed from cycle 1 to 12 of AZA and included haematological and non-haematological toxicity at each cycle. Abnormal changes from baseline measures, not imputable to disease activity or progression, were reported as adverse events (AEs). AEs were graded according to the National Cancer Institute Common Toxicity Criteria (CTC-NCI, version 4.0).
Associations between patient characteristics and outcome were analyzed using the Fisher's exact test. The Wilcoxon rank-sum (Mann-Whitney) test was used to compare the distribution of baseline blood cell count and response. A p-value <0.05 was considered significant. OS was calculated from the start of AZA treatment to the date of the last follow-up or death. Survival curves were estimated using the Kaplan-Meier method, while the Log-rank test was applied to study survival differences.
The study was approved by the local Ethical Committee and patients gave written informed consent for the collection of clinical data in accordance with the Declaration of Helsinki.
Results
Eighteen consecutive patients (M/F 7/11; median age=65; range=49-78) were retrospectively selected from August 2006. None of the patients was eligible for intensive chemotherapy and allogeneic SCT due to age, comorbidity or persistence of the primary malignancy.
Eleven patients had a previous diagnosis of solid malignancy, six of haematological malignancy and one of primary amyloidosis. Overall, the median number of chemotherapy lines was 1.5 (range=1-6). Primary disease was still active in 5 patients when AZA was started (2 metastatic ovarian cancer, bladder cancer, chronic lymphocytic leukaemia and primary amyloidosis). The median interval from the treatment for primary disease to the diagnosis of t-MN was 5 years (range=1-18) (Table I).
According to the WHO classification, the diagnosis was MDS for 14 patients and AML for 4 patients. MDS could be further sub-divided in: refractory cytopenia with multilineage dysplasia (RCMD) (n=1), RAEB-1 (n=3), RAEB-2 (n=9) and myelodysplastic syndrome, unclassified (MDS-U) (n=1). For AML, the median blast count was 41.5% (range=28-90).
A cytogenetic evaluation was available for 14 patients (77.7%). The cytogenetic risk was: good for 6 (42.8%), intermediate for 1 (7.1%) and poor for 7 (50%), with 4 patients (28.5%) presenting a complex karyotype. Four patients (28.5%) presented abnormalities of chromosome 7 (1 patient del(7q) and 3 patients monosomy of chromosome 7); all of them had received cyclophosphamide and 3 also an autologous SCT. One patient presented a del(5) within a complex karyotype.
The ECOG PS was 0 for 2 patients, 1 for 8 patients and 2 for 8 patients. According to the MD Anderson Risk score calculated for MDS with available karyotype (n=12), 1 patient presented a low risk, 5 patients an intermediate risk and 6 patients a high risk.
At baseline we observed: median haemoglobin=8.8 g/dl (range=6.3-12.3), median WBC=2.87×109/l (range=1.34-43.69), median ANC=0.98×109/l (range=0.10-2.01) and median platelets=30.5×109/l (range=8-130). Ten patients (55.5%) were transfusion-dependent.
Characteristics of 18 t-MN patients.
Treatment. AZA was started after a median of 1 month from diagnosis (range, 0-6). The median number of administered cycles was 6 (range=2-22). AZA was discontinued due to progression of the solid tumour (n=2, 11.1%), progressive disease (n=7, 38.8%), AE (infection) (n=3, 16.6%), death due to ischemic events (n=2, 11.1%) or respiratory failure (n=1, 5.5%). AZA was ongoing in 3 patients (16.6%) at the time of the analysis.
Treatment outcome and response to AZA.
Response profile. After 4-6 months from the start of AZA, response assessment was possible in 14 patients (77.7%). Of the remaining 4 patients, after a median of 3 courses, 2 discontinued treatment for respiratory failure and 2 for pneumonia not attributable to the drug.
The ORR was of 35.7% with 4 complete remissions (28.5%), 1 partial remission (7.1%) and a median duration of response of 12 months (range=5-23). Four (28.5%) patients presented a stable disease and 5 (35.7%) a progressive disease (Table II).
The median OS time of the entire cohort from the start of AZA treatment was 9.6 months (95% confidence interval (CI)=4.9-14.4) (Figure 1A). The one-year OS was 42.4%.
As expected, patients carrying alterations of chromosome 7 or alterations of chromosome 7 associated to complex karyotype had an inferior OS compared to the remaining ones (p=0.015 and p=0.023, respectively). With the limitation of the small cohort, we analyzed how the previous treatment could have influenced the outcome, observing that patients who had received 3 or more than 3 lines of chemotherapy presented an inferior OS (median OS=4.9 months) in comparison to the ones treated with 1 or 2 lines of chemotherapy (median OS=13.7 months, p=0.0046) (Figure 1B). The occurrence of an AE of any grade did not impair the OS.
Safety profile. Haemoglobin reduction ≥1g/dl from baseline and not imputable to disease activity was reported in 10 patients (55.5%), grade 3 in 7 patients (38.8%) (Table III). It was observed mainly in the first cycles (median, 2; range, 1-4) and was managed with transfusions (n=7, 70%), erythropoiesis-stimulating agents (n=2, 20%) or delay of next cycle administration (n=1, 10%). Anaemia regressed in 50% of patients (n=5). ANC reduction was present in 14 patients (77.7%); grade 3 in 5 patients (27.7%) and grade 4 in the remaining 9 (50%). Ten patients (55.5%) experienced fever, 3 of whom (16.6%) presented a documented infection, a leading cause of death for two (11.1%) patients. Neutropenia occurred prevalently in the first cycles (median=2; range=1-6) and was managed in 12 cases (87.7%) with antimicrobial prophylaxis and in 2 (14.2%) with delay of the next cycle administration. Fever was managed in 3 cases with G-CSF administration. Neutropenia regressed in 4 patients (28.5%). All the patients with basal ECOG PS ≥2 presented a neutropenia of grade 3-4, while a PS ≥2 conferred a risk for the development of hyperpyrexia (87.5% of patients) in comparison to PS 0-1 (30% of patients) (p=0.046). Moreover, all patients with a high MD Anderson risk score for MDS (n=6/12) had a hyperpyrexia (p=0.030). Platelet reduction was observed in 9 patients (50%); thrombocytopenia was of grade 3 in 2 patients (11.1%) and grade 4 in 6 patients (33.3%). This AE occurred from cycle 1 to 7 (median=3) was managed with transfusions (n=4, 44.4%) and delay of therapy (n=1, 11.1%), while it regressed in 3 patients (33.3%). A bleeding of grade 2 occurred in 2 patients (11.1%), 1 presenting nosebleed and 1 mucosal bleeding. Fatigue was observed in 7 (38.8%) patients and was of grade 1-2. Gastrointestinal toxicity occurred in 4 patients (22.2%) who manifested constipation grade 2 (n=2) and diarrhea grade 1-2 (n=2). They occurred from cycle 2 to 5, were managed with symptomatic therapy in 3 cases and antibiotics in 1 case; the patients recovered promptly. Two patients (11.1%) reported a grade 2 injection site reaction (erythema) successfully treated with local corticosteroids. The median days of delay of AZA administration for toxicity was 13 (range, 7-14) and there were no dose reductions.
Overall survival for (A) the entire cohort of patients, 9.6 months (95% CI 4.9-14.4) and (B) patients who had received 3 or more than 3 lines of chemotherapy for the previous malignancy, median OS=4.9 months, in comparison to those treated with 1 or 2 lines of therapy, median OS=13.7 months (p=0.0046).
Toxicity recorded during front-line AZA treatment in t-MN.
Discussion
T-MN is a group of diseases generally with a poor outcome, due to both the complex biology and to the limited therapeutic chances. It manifests prevalently in advanced age and in patients with poor PS. In some patients, the t-MN is diagnosed when the primary haematological or solid tumour is still active. Furthermore, patients could present cumulative toxicity, immunosuppressive status and depletion of bone marrow hematopoietic stem cell reserve due to prior chemoradiotherapy (12-15, 21).
Allogeneic SCT represents the treatment with a potential curative intent for this subset of patients (13, 22) but the majority are not eligible for this approach (1, 13, 23). Recent retrospective studies have documented the efficacy of AZA in patients with t-MN (14-18). This drug has been mainly employed in patients ineligible for allogeneic SCT and in some cases as a bridge to it (14, 23).
In this study, we observed a prevalence of t-MN after treatment for breast cancer and non-Hodgkin lymphomas. The majority of our patients were heavily pre-treated with either chemotherapy and radiotherapy or autologous SCT. Some of them had the primary malignancy still in active phase. The biological characteristics at baseline were unfavourable for approximately half of the patients with aberrant karyotype, including chromosome 7 alterations, and adverse prognostic risk score. All patients were treated at front-line and we also included those with more than 30% of marrow blasts, which were not considered in the main reported series (14-18). The ORR to AZA in our population was comparable to that reported in previous studies (14-18). About 7-25% of patients in the previous series could further receive an allogeneic SCT (14-18), while our patients were ineligible. This aspect, together with the baseline unfavourable disease characteristics, could have influenced the OS. We reported a median OS of 9.6 months in comparison to a range of 10-21 months in the other series (14-18). Evidently, patients who were heavily pre-treated for the primary malignancy presented a significant inferior OS (4.9 months), in comparison to the ones treated with 1 or 2 lines of chemotherapy (median OS, 13.7 months).
Data on the safety profile of AZA in t-MN have been reported only in a single study (18). The principal toxicity registered in our population was haematological and was observed mainly in first cycles. Despite the high percentage of neutropenia (77.7%), severe infections developed in a minority of patients (16.6%). A poor ECOG PS and a high risk, according to MD Anderson score, were associated with a superior incidence of neutropenia and fever. Life-threatening infections mostly occurred in patients who were heavily pre-treated for the primary malignancy. Fatigue was the most frequent extra-haematological toxicity. The frequency of AE reported in our study results compared to that described for primary MDS in randomized phase III trials (24-26). With respect to the unique study on t-MN in which a grade 3-4 haematological toxicity was documented in 32% of patients, we observed a higher incidence. The frequency of infective events was comparable (16% vs. 16.6%) with an occurrence of fatal events higher in our series (4% vs. 11.1%). We did not document any severe extra-haematological toxicity, while a grade 3-4 injection-site reaction was present in 4% and grade 3-4 constipation in 4% of patients in the study of Fianchi et al. (18). It was not possible to compare our data on fatigue due to the lack of this evaluation in the previous study.
In conclusion, our study reports data on the efficacy of AZA front-line in t-MN, documenting new aspects on the subset of heavily pre-treated patients for their primary malignancy, who presented –anyway–a worse outcome. Furthermore, our study first describes a close examination of the safety profile of AZA in t-MN.
Acknowledgements
The Authors would like to thank C. Oakley (National Cancer Research Centre “Giovanni Paolo II”, Bari, Italy) for linguistic revision of the manuscript and R. Bonaduce (Haematology Unit, Department of Medical and Experimental Oncology, National Cancer Research Centre “Giovanni Paolo II”, Bari, Italy) for graphical support.
Footnotes
-
Conflicts of Interest
The Authors state no conflict of interest and have received no payment for the preparation of this manuscript.
-
Disclosures
Off-label use: Azacytidine in t-AML with more than 30% of marrow blasts.
- Received September 4, 2014.
- Revision received October 1, 2014.
- Accepted October 8, 2014.
- Copyright© 2015 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.