Rodrigo Ruiz-Soto is the recipient of the “FUNSALUD-AMEH-291” and the “Excelencia Academica Aventis” scholarships in Mexico.
Estimating late adverse events using competing risks after autologous stem-cell transplantation in aggressive non-Hodgkin lymphoma patients†
Article first published online: 13 NOV 2005
Copyright © 2005 American Cancer Society
Volume 104, Issue 12, pages 2735–2742, 15 December 2005
How to Cite
Ruiz-Soto, R., Sergent, G., Gisselbrecht, C., Larghero, J., Ertault, M., Hennequin, C., Manson, J., de Kerviler, E., Briere, J. and Mounier, N. (2005), Estimating late adverse events using competing risks after autologous stem-cell transplantation in aggressive non-Hodgkin lymphoma patients. Cancer, 104: 2735–2742. doi: 10.1002/cncr.21492
This study was presented in part at the American Society of Hematology 45th annual meeting, December 6–9, 2003, San Diego, California, USA.
- Issue published online: 8 DEC 2005
- Article first published online: 13 NOV 2005
- Manuscript Accepted: 13 JUL 2005
- Manuscript Revised: 15 JUN 2005
- Manuscript Received: 19 APR 2005
- adult lymphoma;
- regression analysis;
- agents/adverse effects;
- second primary
Consolidative autologous stem-cell transplantation (ASCT) is a valuable option in high-risk or disease recurrence large-cell non-Hodgkin lymphoma patients (NHL); however, its long-term toxicity must still be assessed.
Among the 439 lymphoma patients transplanted at our institution from January 1, 1993, to January 1, 2002, 158 exhibited aggressive NHL. The median age of the patients was 46 years (range, 18–69), 98 males and 60 females. Ninety (57%) patients received first-line ASCT. The median number of prior chemotherapy regimens was 2 (range, 1–10). Thirty-eight (24%) patients received total body irradiation conditioning. Here we report the adverse events which occurred at least 30 days after ASCT and before disease recurrence.
After a median follow-up of 3 years, the overall and disease-free survival rates were 61% and 55%, respectively. Sixty-eight late adverse events affected 43 (27%) patients, leading to a cumulative incidence of 34% at 3 years. Infections were the most frequent adverse events (n = 13), followed by neurologic (n = 12), pulmonary (n = 6), or cardiovascular (n = 4). Eight malignancies were diagnosed (six solid, two hematologic), leading to a cumulative incidence of 3.7% at 3 years. Taking into account the competing risks, multivariate analysis revealed that the number of progressions (relative risk [RR] = 2.68) and a mitoxantrone-containing conditioning regimen (RR = 2.98) significantly increased the incidence of late toxicity.
ASCT is effective in patients with aggressive NHL with a poor prognosis. However, careful long-term follow-up of survivors is recommended because of the increase in malignant and nonmalignant toxicities. Cancer 2005. © 2005 American Cancer Society.
Aggressive non-Hodgkin lymphomas (NHLs) are highly chemotherapy-sensitive malignancies. Treatment with conventional combination chemotherapy produces complete remission rates of 50–70% and disease-free survival (DFS) rates of approximately 50%.1 Advances in the definition of risk groups and in therapies have improved the prognosis of the disease and recent strategies, including the use of monoclonal antibodies, have improved these results.2, 3 In addition, high-dose therapy (HDT) followed by autologous stem-cell transplantation (ASCT) given as a first-line treatment for high-risk patients or after disease recurrence allows a subset of those patients with a poor prognosis to be cured.4–6
Despite the low early treatment-related mortality (1–2%), ASCT may be related to late adverse events due to treatment toxicity such as secondary cancers and major organ dysfunction.7 Within this framework, the incidence of secondary leukemia and myelodysplastic syndrome have been extensively studied.8–15 However, information regarding other late toxicities after ASCT is scarce and does not always take into account the presence of competing risks between death, disease recurrence, and late toxicities.16
The aim of the present study was to assess late toxicities after ASCT in a cohort of aggressive NHL patients treated at our institution.
MATERIALS AND METHODS
Patient Selection and Staging
The patients transplanted at our institution were identified with the database of the Societe Francaise de Greffe de Moelle. Files were reviewed and data were collected by means of a standard form. Between January 1, 1993, and January 1, 2002, 439 consecutive lymphoma patients were consolidated by HDT and ASCT.
The stage of the disease was evaluated by physical examination, a computerized tomography (CT) scan of the chest and abdomen, cerebrospinal fluid examination, bone marrow biopsy, and other investigations depending on the clinical symptoms. Patients were staged according to the Ann Arbor classification. The Eastern Cooperative Oncology Group (ECOG) performance status (PS) was assessed and lactate dehydrogenase (LDH) was expressed as the maximum/normal value ratio.
Tumor responses were assessed after the four cycles of induction chemotherapy were classified according to the International Workshop criteria17 as follows. Complete response (CR): the disappearance of all lesions and radiologic or biologic abnormalities observed at diagnosis and the absence of new lesions. In particular, the spleen must not be palpable and the bone marrow biopsy must be cleared. Unconfirmed complete response (CRu): CR but the persistence of some radiologic abnormalities (i.e., a residual lymph node mass greater than 1.5 cm in greatest transverse diameter), whose size had regressed by at least 75%. Individual nodes that were previously confluent must have regressed by more than 75% in their sum of the products of the greatest diameters (SPD). Partial response (PR): the regression of all measurable lesions by more than 50%, the disappearance of nonmeasurable lesions, the absence of new lesion, and no increase in the size of other lymph nodes, liver, or spleen. Stable disease: less than PR but not progressive disease. Progressive disease: the appearance of new lesions, or growth of any previously identified abnormal lymph node by more than 50% from nadir in the SPD.
All histologic slides were reviewed by an expert pathologist (J.B.), and lymphomas were reclassified according to the World Health Organization classification.18
In the present analysis, we included only the 158 patients who represented aggressive NHL.
Stem Cell Collection
Peripheral blood stem cells (PBSC) were collected from patients by leukapheresis during the hematologic recovery phase after the last inductive cycle. At the time of hematologic recovery, peripheral blood CD34+ cell counts were determined daily and leukapheresis was started when counts exceeded 10/μL. Leukapheresis procedures were continued for 2 to 4 consecutive days, depending on the number of CD34+ cells harvested. The targeted number of CD34+ cells was 3.106/kg, but the minimum required was fixed at 2.106/kg body weight. Marrow cells were harvested by multiple aspirations from the iliac crests under general anesthesia into 600-mL bags (Baxter, France) containing acid citrate dextrose (ACD) (Maco Pharma, France) and heparin (Sanofi, France). A minimum of 2.108 nucleated cells per kg body weight was required.
Cryopreservation and Thawing Procedures
Excess plasma was removed from PBSC components by centrifugation in 600-mL blood transfer bags at 650g for 10 minutes. The volume of the residual cell pellet was adjusted with human serum albumin between 1993 and 1998; with Elohes (Fresenius Kabi, France) between 1998 and 2000; and with voluven (Fresenius Kabi, France) since 2000. Precooled freezing solution was added slowly to the cell components with continual mixing to achieve a final concentration of 10% dimethylsulfoxide (DMSO). Cell concentrations before freezing were less than 3.108 per mL. Cells were cooled at 1 °C per minute to −40 °C with compensation for heat from superfusion, and then at 10 °C per minute to −100 °C, using a rate-controlled freezer (Nicool, Air Liquide, France). Frozen cells were stored in liquid nitrogen until reinfusion. Before reinfusion, hematopoietic progenitor cell (HPC) bags were thawed in a 37 °C water bath, washed twice in a glucose solution (B-Braun, France), and finally suspended in albumin. The cell suspension was then rapidly infused into the patient through a central venous catheter at about 20 ml per minute.
Patients were eligible for ASCT if they were responding after induction treatment and had reached CR or PR and after stem cells had been collected. All patients undergoing ASCT were required to have adequate pulmonary, liver, and renal functions. All patients had an indwelling central venous catheter and were housed in laminar air-flow rooms for the duration of aplasia. All received the entire collected dose of CD34+ cells and were given G-CSF (5 μg/kg/day), starting 1 day after stem cell reinfusion and acyclovir prophylaxis. Prevention of veno-occlusive disease with heparin (1 mg/kg) was used for patients receiving total body irradiation (TBI) conditioning. Patients who developed fever and whose absolute granulocyte count was less than 0.5 G/L were treated with intravenous broad-spectrum antibiotics. All patients received irradiated packed red blood cells and platelet products to maintain a hemoglobin level >8 g/dL and a platelet count >20 G/L.
Follow-up and Late Toxicity
All patients were followed in outpatient settings each month during the first 3 months after ASCT. Follow-up procedures included physical examination and complete blood count every 3 months for the first 2 years, then every 6 months for 3 years, and eventually once on a yearly basis. All patients had their annual follow-up visit in our institution. Computerized tomography (CT) scans of the chest and abdomen were performed every 6 months during the first 2 years, then at the discretion of the treating physician. In addition, before ASCT and after 1 and 2 years, patients underwent cardiac left ventricular ejection fraction evaluation (LVEF) by gammagraphy and lung function tests.
The primary endpoint of the study was the occurrence of late toxicity, defined by any adverse event that presented itself after Day 30 of ASCT (outpatient period) and before disease recurrence, when it occurred. Late toxicities were classified in accordance with the oncology section of the International Classification of Diseases.19 This endpoint was calculated as a cumulative probability and the effects of potential risk factors were examined using the Cox proportional-hazards model.20, 21
Demographic and baseline laboratory data were compared among subpopulations by Fisher exact test for nominal variables and Wilcoxon rank test for continuous variables. Overall survival (OS) was measured from the date of graft to death from any cause. Survival functions were estimated by the Kaplan–Meier method and compared by the log rank test.
Differences between the results of comparative tests were considered significant if the two-sided P-value was less than 0.05. All statistical analyses were performed using SAS 9.1 software (SAS Institute, Cary, NC) and Splus 6.2 software (MathSoft, Cambridge, MA).
Patient characteristics at diagnosis are listed in Table 1. Ninety-eight (62%) patients were male and the median age was 46 years (range, 18–69). The most frequent histologic subtype was diffuse large B-cell lymphoma in 101 (64%) patients. Of the remaining patients, 3 (2%) exhibited Burkitt-like cells, 14 (9%) mantle cells, 13 (8%) lymphoblastic cells, 10 (6%) peripheral T-cells (PTCL), and 5 (3%) anaplastic cells.18 Only a small fraction of 9 (6%) patients were in transformation from low-grade type NHL. In 3 (2%) cases large cell NHL could not otherwise be classified, due to the absence of an immunohistochemical study. At diagnosis, 116 (70%) patients were Stage III–IV and the majority presented an ECOG scale of 0–2 without B symptoms (85% and 57%, respectively). Marrow was involved in 37 (20%) patients. The score for the International Prognostic Index (aa-IPI) factors was 0 in 21 (13%) patients, 1 in 39 (25%), 2 in 59 (37%), and 3 in 19 (12%).
|Characteristics||At diagnosis (%)||At transplantation (%)|
|I–II||42 (26)||35 (22)|
|III–IV||116 (74)||123 (78)|
|Performance status (WHO grading)|
|0-2||135 (85)||125 (80)|
|>2||23 (15)||23 (14)|
|< Normal||48 (30)||88 (56)|
|> Normal||90 (57)||56 (35)|
|Missing||20 (13)||14 (9)|
|0||21 (13)||19 (12)|
|1||39 (25)||46 (29)|
|2||59 (37)||60 (38)|
|3||19 (12)||18 (11)|
|Missing||20 (13)||15 (10)|
Ninety (57%) patients received ASCT as first-line consolidation (IPI 2–3, 76%): 81 patients received only one induction treatment in order to achieve remission, 9 patients needed a second-line treatment regimen. Sixty-eight (43%) patients received ASCT as consolidation after chemosensitive disease recurrence (IPI 2–3, 25%). Sixty patients were in first disease recurrence. The induction chemotherapy regimens are listed in Table 2. The preferred regimen was an intensive regimen given at 2-week interval ACVBP (doxorubicin 75 mg/m2, cyclophosphamide 1200 mg/m2 Day 1; vindesine 2 mg/m2, bleomycin 10 mg Days 1 and 5; prednisone 60 mg/m2 Days 1–5) in 77 (49%) patients, followed by classic CHOP (doxorubicin 50 mg/m2, cyclophosphamide 750 mg/m2, vincristine 1.4 mg/m2 Day 1; prednisone 60 mg/m2 Days 1–5) in 41 (26%). No patients were treated with rituximab and chemotherapy.
|Type of treatment||Patients||%|
|At diagnosis (n = 158)|
|At relapse (n = 68)|
Table 3 lists the characteristics of transplantation. Overall, the median number of previous chemotherapy regimens was 2 (range, 1–10) and the median number of prior chemotherapy courses was 8 (range, 3–42). In all, 115 (73%) patients who received HDT were in complete remission, 35 (22%) were in partial remission, and only 8 (5%) were in stable disease. A total of 154 (97%) patients had no marrow involvement at the time of graft harvest. BEAM (carmustine 300 mg/m2, etoposide 800 mg/m2, aracytine 800 mg/m2, and melphalan 140 mg/m2) was used as a conditioning regimen in 86 (54%) patients and CBV-N (mitoxantrone 40 mg/m2, carmustine 300 mg/m2, cyclophosphamide 6000 mg/m2, and etoposide 1000 mg/m2) in 24 (15%). Fractionated 12 Gy TBI was part of the conditioning regimen in 38 patients (24%); of those, 76% were given cyclophosphamide and VP-16. The source of stem cells was PBSC in the majority (n = 149). The median number of CD34+ cells per kg infused was 5.0 106 (range, 2.2–55.0]. Purging was used in only 11 (7%) patients and G-CSF was given to 92% of the patients. Involved field radiotherapy was administered after ASCT to 31 (20%) patients.
|Number of previous relapses||Patients (n = 158)||%|
|Median (range)||0 (0–3)|
|Number of prior chemotherapy regimens|
|Median [range]||2 (1–10)|
|Number of prior chemotherapy courses|
|Median (range)||8 (3–42)|
|Status of disease|
|TBI + Chemotherapy||38||24|
|Involved field radiation therapy|
Survival and Late Toxicities
With a median follow-up of 3 years after ASCT (range, 0.2–10) and 4.5 years after diagnosis (range, 0.5–12), 102 patients were alive and 63 with disease recurrence. The 3-year OS was 61 ± 9% and the DFS was 55 ± 8%. Figure 1 compares OS between first-line consolidation and after disease recurrence.
Sixty-eight nonmalignant late adverse events occurred in 43 (27%) patients. Twenty-five patients had two late adverse events (n = 16) or three late adverse events (n = 9). The most frequent of these were infections (see Table 4). In 33 of the 43 (20%) patients, late adverse events were retrospectively graded as severe (Grade 3–4), but none of the patients died. In 10 patients they were associated with infectious episodes (two pneumocystis, one viral B hepatitis, five pulmonary infections, two sinusitis), six with neurologic (peripheral neuropathy), four with pulmonary (three fibrosis, one embolism), four with cardiovascular (three heart failure, one angina), four with digestive (three gastric ulcers, one pancreatitis), one with hemorrhagic cystic, and two with musculoskeletal problems; two more were associated with cataracts.
|Organ||Events N = 68||%||Diagnosis (median time in months from autologous stem-cell transplantation)|
Eight cancers occurred. Six were solid tumors, one myelodysplastic syndrome and one acute myeloid leukemia (after 30 and 3 months, respectively). Details are given in Table 5.
|Malignancy N = 8||Diagnosis (in mos from autologous stem-cell transplantation)||Status (in mos from malignancy)|
|Gastric cancer||10||Dead (5)|
|Colonic villous tumor||79||Alive (44)|
|Acute myeloid leukemia||3||Dead (19)|
|Myelodysplastic syndrome||30||Dead (50)|
|Breast cancer||21||Dead (20)|
|Cervix cancer||118||Alive (6)|
Competing Risks Analysis
Figure 2 shows the cumulative incidence for competing risks: the incidence at 3 years was 34% for nonmalignant toxicities (25% after exclusion of infections) and 3.7% for malignancies.
In order to account for competing risks, the age, sex, IPI, type of conditioning regimen (mitoxantrone-containing regimen vs. BEAM), TBI, source of stem cells, time between diagnosis and ASCT, number of disease recurrences, number of regimens, and number of chemotherapies before ASCT were included in regression analysis. The only prognostic factor for death was the number of disease recurrences before ASCT (relative risk [RR] = 1.26, P = 0.04). Age was significantly associated with disease recurrence after ASCT (RR = 1.3 for 10-year increment, P = 0.02) and development of secondary cancer (RR = 1.8 for 10-year increment, P = 0.05). The number of disease recurrences before ASCT was associated with late toxicity (RR = 2.68, P = 0.01) and secondary cancer (RR = 5.68, P = 0.02). The use of mitoxantrone in the conditioning regimen (CBV-N) for 24 patients was related to development of 7 late toxicities (RR = 2.9, P = 0.03 in comparison with BEAM regimen) including: neuropathy (2), infectious pneumopathy (2), cardiac failure (1), hypogodanism (1), and pancreatitis (1). Note that there was no significant risk for the TBI conditioning regimen (vs. the chemotherapy-only conditioning regimen), which led to 10 adverse events for 38 patients: tendonitis (1), neuropathy (2), lung fibrosis (1), pneumocystis (1), cataract (2), heart failure (1), and gastric ulcers (2).
This study demonstrates a 61% survival rate in 158 poor prognosis aggressive NHL patients treated with ASCT either in first-line therapy (n = 90) or after disease recurrence (n = 68). In order to estimate the risk of late toxicity, we used the cumulative incidence, which differs from the Kaplan–Meier estimate in that patients must both survive competing risks and be diagnosed with late toxicity. The cumulative incidence found at 3 years was 34%, but does not constitute an important risk of death (none of our patients died as a consequence of it). However, due to the retrospective design of the present study we were not able to give that estimate for the remaining subset of 281 patients transplanted within the same period at our institution, which was a very heterogeneous population (i.e., follicular lymphoma, chronic leukemia, and Hodgkin disease).
As anthracycline-based chemotherapy is fundamental in the treatment of NHL, we tried to emphasize its toxicity. Nine (6%) patients presented LVEF dysfunction but only 3 (2%) had clinical symptoms and required treatment. The cumulative incidence of cardiac late toxicity found was 9%; the crude incidence has been reported between 0–12% with predictive factors for heart failure post-ASCT (higher risk with LVEF less than 50%,22 older patients,23 and use of high doses of cyclophosphamide in conditioning regimens24), where we did not find any correlation. TBI was not found to increase cardiac toxicity. More late toxicities were noted when mitoxantrone was used as part of the conditioning regimen and exclusively after first-line therapy. So far, this bears no relation to an increased risk of cardiac failure or development of second malignancies.
The development of secondary solid tumors in relation to ASCT has been reported in the central nervous system, kidney, bladder, lung, stomach, and melanoma,25 which are not different from those found in the present study. We reported a total of 8 secondary malignancies with a 3-year cumulative incidence of 3.7%. Incidences described in the literature vary between 3–18 years post-ASCT. Pavlovsky et al.26 transplanted 279 lymphoma patients (including Hodgkin disease and low-grade NHL) with an incidence of 2.86 second malignancies at 10 years after ASCT. In older NHL patients (> 60 years) treated with ASCT, an incidence of 7% secondary cancers was found, with a median survival time of 12 months (range, 0.6–79).27 In a large cohort of 605 NHL patients treated by ASCT (TBI and cyclophosphamide conditioning), there was a 10-year OS rate of 44% and an estimated cumulative incidence of 18% of second tumors, with no difference between patients transplanted in first remission or after therapy for disease recurrence.28 By contrast, in our experience previous disease recurrence, prior lines of chemotherapy before ASCT, and age were found to be risk factors for developing a secondary tumor.
The mutagenic risk of conventional chemotherapy was described in a series of 2837 NHL patients treated with an ACBVP regimen, with a cumulative incidence of 3.5% (81 cancers, of those 17 hematologic), which was not different from malignancies affecting the general population.29 As the present study population shares the same characteristics of age and gender as the ACVBP study population, we consider that the risk is of the same magnitude, but a larger follow-up is needed before doing a matched analysis.
Lastly, in our experience, out of eight secondary malignancies two corresponded to sAML/myelo-dysplastic syndrome (MDS) (one acute myeloid leukemia (AML) and one MDS) and almost all (94%) received PBSC infusion (with a median age of 46 years, and none of them presented a poor stem cell harvest). TBI has also been noted as an important factor in the etiology of sAML/MDS.16 Some studies indicate that there is a dose-dependent relation with the risk of sAML/MDS with TBI that diminishes when TBI doses are 12 Gy or less.30 The latter corresponds to the treatment protocol used in our institution and we did not find an increased risk of second malignancies.31
The results of this cohort study confirmed the efficacy of ASCT. In the meantime, a careful continuous follow-up of lymphoma patients is advisable. An approach to minimize the risks can include the use of ASCT in first CR in poor prognosis patients at high risk of disease recurrence to reduce prior exposure to mutagenic agents. Long-term nonmalignant toxicities are frequent, although rarely severe, but should be evaluated prospectively with the introduction of new agents such as monoclonal anti-CD20.
The authors thank Olive Eccleston for editing the English and Sylvie Corre for secretarial assistance. They also thank the editor and referees for helpful comments and suggestions.
- 11Secondary leukaemia and myelodysplasia after autografting for lymphoma: results from the EBMT. EBMT Lymphoma and Late Effects Working Parties. European Group for Blood and Marrow Transplantation. Br J Haematol. 1999; 106: 1020–1026., , , et al.
- 19World Health Organization. Manual of the international statistical classification of disease, injuries and causes of death: based on the recommendations of the Ninth Revision Conference, 1975, and adopted by the Twenty-ninth World Health Assembly, vol. I. Geneva: World Health Organization, 1977.
- 20Modeling survival data: extending the Cox model. New York: Springer, 2000., .
- 26Ten years of autologous peripheral stem cell transplantation with the same conditioning regimen (CBV) in 279 consecutive lymphoma patients done in a single center. Evaluation of prognostic factors affecting outcome. Proc Am Soc Clin Oncol. 2003; 22: 607 (Abstract 2441)., , , , , .
- 27High dose therapy with autologous stem cell transplantation (ASCT) for non-Hodgkin's lymphoma (NHL) in elderly patients. A single institution retrospective study. Blood. 2003; 102 (Abstr 2719)., , , , .