Fax: (507) 266-9277
Safety and efficacy of external beam radiation therapy for non-Hodgkin lymphoma in patients with prior 90Y-ibritumomab tiuxetan radioimmunotherapy†
Article first published online: 12 JUN 2006
Copyright © 2006 American Cancer Society
Volume 107, Issue 2, pages 433–438, 15 July 2006
How to Cite
Justice, T. E., Martenson, J. A., Wiseman, G. A. and Witzig, T. E. (2006), Safety and efficacy of external beam radiation therapy for non-Hodgkin lymphoma in patients with prior 90Y-ibritumomab tiuxetan radioimmunotherapy. Cancer, 107: 433–438. doi: 10.1002/cncr.21998
Presented at the 45th Annual Meeting of the American Society for Therapeutic Radiology and Oncology, Salt Lake City, UT, October 2003.
- Issue published online: 5 JUL 2006
- Article first published online: 12 JUN 2006
- Manuscript Accepted: 7 MAR 2006
- Manuscript Received: 5 JUL 2005
- National Cancer Institute. Grant Number: CA97274
- tumor response
Non-Hodgkin lymphomas (NHL) are radiosensitive tumors. Recent trials have demonstrated the effectiveness of beta irradiation delivered by monoclonal antibodies conjugated to radioisotopes such as yttrium-90 or iodine-131. Many patients eventually relapse after radioimmunotherapy (RIT) and require additional treatment. The goal of this study was to determine the safety and efficacy of external beam radiation therapy (EBRT) delivered after RIT.
We reviewed the records of 135 patients with relapsed B-cell NHL who had received RIT with 90Y-ibritumomab tiuxetan to identify patients who subsequently received EBRT. The response rates and radiation-induced toxicity from the EBRT were assessed.
Nineteen of 135 (14%) patients received EBRT to a total of 39 tumor sites. Complete radiation therapy records were available for 16 patients (36 tumor sites), which formed the basis for this analysis. The median EBRT dose was 28.5 Gy (range, 10–40 Gy), with a median fraction size of 2 Gy (range, 1–5 Gy). Response data were available for 29 treated sites. The overall response rate was 90% (26/29) with 12 complete responses (41%), 7 complete clinical responses (24%), 7 partial responses (24%), and 3 stable (10%). Toxicities were generally transient, reversible, and corresponded to the anatomic regions irradiated.
EBRT can produce tumor responses at sites of NHL that relapse after RIT with acceptable toxicity. Cancer 2006. © 2006 American Cancer Society.
The inherent radiosensitivity of most non-Hodgkin lymphomas (NHL),1–3 and the availability of a good antigenic target (CD20) and corresponding monoclonal antibody provided the rationale for the development of radioimmunotherapy (RIT) for the treatment of B-cell NHL.4, 5 RIT, which involves the administration of radionuclides linked to highly specific antibodies, can potentially treat systemic disease by delivering focused radiation to a targeted population of cells.6 In 2002, 90Y-ibritumomab tiuxetan (Zevalin™, BiogenIdec, San Diego, CA) became the first radiolabeled antibody approved by the FDA for the treatment of cancer. This was followed by the approval of 131I-tositumomab (Bexxar™, GlaxoSmithKline) in 2003. Initial phase I/II studies with single-agent 90Y-ibritumomab tiuxetan in relapsed patients produced response rates of 64%–74%, with higher rates for patients with low-grade disease.7, 8 In a subsequent randomized phase III study comparing 90Y-ibritumomab tiuxetan with the unlabeled anti-CD20 antibody rituximab, RIT was associated with improved response rates, prolonged treatment-free intervals, and greater improvement in cancer-related symptoms.9 Unfortunately most patients will eventually relapse and require additional therapy after RIT.10 In such situations, when complete eradication of disease is impossible, efforts must be directed toward slowing malignant progression while preserving or improving the patient's functional status.
Conventional external beam radiation therapy (EBRT) can be very effective in NHL as curative treatment in early stage of the disease or for palliation of unremitting tumor-related symptoms in patients with relapsed NHL. The known propensity for NHL to disseminate widely throughout the body has limited the use of EBRT in NHL. Chemotherapy has been demonstrated to be effective in patients who relapse after RIT11; however, the response rate to EBRT in RIT failures is unknown. The purpose of this retrospective study was to assess the safety and efficacy of EBRT after previous systemic treatment with RIT.
METHODS AND MATERIALS
Between 1996 and 2002, 135 patients with histologically confirmed, relapsed or refractory CD20+ B-cell NHL were treated with 90Y-ibritumomab tiuxetan RIT on clinical trials. All patients had previously consented to have their medical records reviewed for research purposes. The study was approved by the Mayo Institutional Review Board. We identified patients that underwent EBRT at some point in their treatment course after receiving 90Y-ibritumomab tiuxetan. The radiation therapy records were reviewed for field location and size, dose of radiation, toxicity incurred, and results of treatment. Patients who were treated at other institutions were contacted for approval to request their radiation records. Radiation treatment records and laboratory data were reviewed for evidence of toxicity occurring during and up to 3 mo after completion of EBRT. Toxicity from EBRT was defined as any documented unfavorable adverse event possibly, probably, or definitely attributable to EBRT. All reported toxicities were retrospectively graded using the Common Toxicity Criteria, version 2.0.
The NCI-International Working Group criteria were used to retrospectively assess tumor responses within the irradiated fields.12 Complete response (CR) was defined as disappearance of the irradiated lesion. An unconfirmed complete response (CRu) was defined as regression of the tumor by at least 75% with the persistence of some radiologic abnormalities. Regression in size by between 50 and 75% was termed a partial response (PR). Stable disease (SD) was defined as no change in the measurable dimensions of the irradiated lesion or regression by less than 50%.
Nineteen of the 135 patients (14%) treated with 90Y-ibritumomab tiuxetan subsequently received palliative EBRT for relapsed NHL. These 19 patients form the basis for this report (Table 1); the other 116 patients are not discussed further. Ninety-five percent (18/19) of the patients were diagnosed with CD20+ B-cell NHL at presentation. One patient initially diagnosed with peripheral T-cell NHL and treated with chemotherapy, later developed recurrent neck lymphadenopathy and was found to have transformed CD20+ diffuse large cell NHL. Prior to the treatment with 90Y-ibritumomab tiuxetan, all 19 patients had received combination chemotherapy with a median of 3 prior regimens (range, 1–9). In addition, 37% (7/19) had received previous involved field EBRT. Eight patients (42%) exhibited transformed histology prior to RIT administration. At relapse, each of the 19 patients enrolled on a clinical trial and received one course of standard outpatient 90Y-ibritumomab tiuxetan RIT. This consisted of rituximab 250 mg/m2 i.v. (Day 1 and Day 8), 111In-ibritumomab tiuxetan (Day 1), and 90Y-ibritumomab tiuxetan 0.4 mCi/kg (maximum 32 mCi) (Day 8). One patient received 0.3 mCi/kg because of mild thrombocytopenia. The RIT absorbed radiation doses to the normal tissue (except liver which accumulates the radiolabeled antibody) from the 90Y-ibritumomab tiuxetan were calculated in eight patients who had serial whole body images obtained. The mean normal tissue absorbed radiation dose was found to be 0.38 Gy (range 0.27–0.55 Gy), using the whole body remainder (total body dose minus the source organs) as the normal tissue absorbed dose. The whole body remainder absorbed doses were measured from the serial 111In-ibritumomab tiuxetan whole body images and calculated from the total body and normal source organ gamma camera activities using the MIRDOSE 3 computer software (Oakridge Associated Universities, Oakridge, TN).
|Gender, Male||13 (68)*|
|Median, 66 yr|
|Range, 27–78 yr|
|Large cell||4 (21)|
|Mantle cell||2 (11)|
|Small lymphocytic||1 (5)|
|Prior external beam radiotherapy||7 (37)|
|Response to 90Y-Ibritumomab tiuxetan|
|Complete response||1 (5)|
|Partial response||8 (42)|
|Mixed response||3 (16)|
The overall response rate (ORR) to 90Y-ibritumomab tiuxetan in these 19 patients was 47% (9/19) with eight PR and one CR. The remaining 10 patients had stable disease or progressed after 90Y-ibritumomab tiuxetan RIT. One patient relapsed in a site (mesentery) previously treated with EBRT and had a PR to 90Y-ibritumomab tiuxetan in that site.
Each of these 19 patients developed symptomatic tumor recurrences after RIT and received palliative EBRT. None of the relapses post-90Y-ibritumomab tiuxetan treated with EBRT were in sites that had received EBRT pre-RIT. The median elapsed time from RIT to EBRT administration was 8.2 mo (range, 1.2–63 mo). All patients had measurable, gross disease on clinical examination or radiographic imaging prior to EBRT. Complete treatment records were available for 36 treatment sites from 16 patients; the other 3 patients were deceased with no known relatives to authorize release of information. Thirteen received EBRT at other institutions. Tumor size was documented for 31 of 36 treated sites. Sixteen tumors (42%) measured ≥5 cm and were characterized as bulky. Sixty-one percent (22/36) of the irradiated sites were tumor-bearing lymph node regions. Extranodal lymphomas involving bone, soft tissues, and skin accounted for the majority of the remaining sites (Table 2). In all cases, involved-field EBRT was given in continuous courses (5 d/wk) with either electrons or megavoltage photons, using a variety of standard beam arrangements. The median dose of delivered radiation was 28.5 Gy (range, 10–40 Gy) with a median fraction size of 2 Gy (range, 1–5 Gy).
|Cervical lymph nodes||6|
|Extremity soft tissues||5|
|Axillary lymph nodes||4|
|Mesenteric lymph nodes||3|
|Retroperitoneal lymph nodes||2|
|Pelvic lymph nodes||2|
|Inguinal lymph nodes||2|
|Epitrochlear lymph nodes||2|
|Mediastinal lymph nodes||1|
Response data were available for 29 treated sites (Table 3; Fig. 1). There were no follow-up tumor measurements for the remaining 7 irradiated sites from 2 terminally ill patients. The ORR to EBRT in the treated lesions was 90% (26/29)—with 41% (12/29) CR, 24% (7/29) CRu, and 24% (7/29) PR. Three tumor sites (10%) remained clinically and radiographically stable. Among the 16 bulky tumor sites, the ORR was 87% (14/16), with 25% (4/16) CR, 31% (5/16) CRu, and 31% (5/16) PR. No tumor progressed during or immediately following EBRT. Eight patients who were stable or progressed after 90Y-ibritumomab tiuxetan had response data after EBRT. Seven of the 8 achieved a tumor response; one was stable by imaging but experienced significant pain relief.
|All Sites||N = 29 sites|
|Overall response||26 (90)†|
|Complete response||12 (41)|
|Complete response unconfirmed||7 (24)|
|Partial response||7 (24)|
|Bulky tumors (≥5 cm)||N = 16 sites|
|Overall response||14 (88)|
|Complete response||4 (25)|
|Complete response unconfirmed||5 (31)|
|Partial response||5 (31)|
|Stable disease||2 (13)|
Follow-up ranged from 1 to 23 mo, with a median of 3 mo. Six patients died from advanced, systemic NHL within 3 mo of the completion of EBRT. Failures in the site of EBRT were observed in only 7% (2/29) of assessable treated sites. One patient, who had received 3600 cGy in 20 fractions to bulky paraaortic adenopathy, failed locally 18 mo later in the setting of widespread disease progression. The second local failure occurred 4 mo after EBRT in a patient who had received 1200 cGy in 8 fractions for mesenteric adenopathy. A complete radiographic response was subsequently achieved after retreatment with 3000 cGy in 20 fractions. Local tumor control persisted until the time of death 2 mo later.
EBRT was generally well tolerated with incidences of grades 1–2 and 3–4 adverse events per total number of treated sites of 50% (18/36) and 8% (3/36), respectively (Table 4). Side effects and symptoms were generally mild and self-limited and varied between patients depending on what areas of the body that were being treated. One patient developed moderately severe cellulitis within the treated field that responded to parenteral antibiotics and a brief hospitalization. One additional patient with chronic renal insufficiency and bulky abdominal disease experienced tumor lysis syndrome while under treatment, resulting in acute renal failure. This was not felt to represent a direct nephrotoxic effect of the radiation. The creatinine and uric acid levels gradually returned to baseline with fluids and hemodialysis. This patient also developed Grade 3 anemia during hospitalization, requiring red blood cell transfusion. The patient ultimately succumbed to progressive NHL 1 month later.
|Grade 1–2||Grade 3–4|
|Any adverse event||18 (50)*||3 (8)|
Acquired radioresistance after chronic exposure to low-dose rate irradiation is a well-documented phenomenon in lymphoma cells.13, 14 In murine studies an increase in radioresistance in L5178Y murine lymphoma cells was observed more than 70 days after termination of protracted low dose radiation exposure.15 On the basis of these studies it was indeed conceivable that tumors that recurred after radiolabeled antibody therapy would be less responsive to subsequent EBRT. Despite this theoretical concern, this study demonstrates that EBRT can produce a very high ORR of 90% with 41% CR in tumor sites previously treated with RIT. Tumor shrinkage was quite marked in some cases (Fig. 1). Similarly high response rates with EBRT were observed in 16 patients with bulky tumors, with overall and complete response rates of 87% and 25%, respectively. Although tumor bulk predicts a lower ORR to RIT,7, 8, 16, 17 it did not adversely impact the effectiveness of EBRT in this study. Safety and tolerability data were gathered retrospectively, with obvious limitations. There was no evidence that prior systemic irradiation with standard dosages of RIT had a clinically significant impact on the tolerability of the involved-field EBRT that followed. In general, the frequency and severity of adverse events observed in this group were not more than what could reasonably be expected in patients without a history of RIT. A meaningful analysis of the relationship between the site irradiated, the radiation dose, and toxicity was not possible in view of the low rate of major toxicity (Table 4) and the small number of radiation courses administered to any given site (Table 2).
Historically, involved-field EBRT for indolent NHL in RIT-naive patients has been associated with greater than 90% long-term local control.18–20 Likewise, recent data regarding low-dose palliative EBRT for relapsed NHL have demonstrated excellent response rates (ORR 87%–94%), with median duration of response estimated at 22 mo.21, 22 Although follow-up of our patients was quite limited, given that many were terminally ill at the time of treatment, we observed only 2 local failures in the 29 tumor sites that could be assessed. One recurrence after suboptimal irradiation was successfully retreated with EBRT, resulting in local tumor control for the remainder of the patient's life.
To date, there has been little published information regarding the use of both external radiation and radiolabeled antibodies in humans. Macklis and White have examined the impact of prior EBRT on safety and efficacy of subsequent 90Y-ibritumomab tiuxetan RIT.23 In that study previous EBRT did not significantly predict the likelihood of response to future RIT, nor did it predispose to most types of organ-specific toxicity. There was a slight trend suggesting an association between prior EBRT and subsequent neurologic and pulmonary toxicity. The authors noted, however, that some patients who experienced these effects had actually never before been irradiated to these anatomic sites, calling into question a true cause and effect relationship. Seven of our patients had received one or more courses of EBRT preceding their RIT (Table 1), and we did not find excessive RIT-related toxicity. Specifically, there were no untoward neurologic or pulmonary adverse events recorded.
The excellent ORR of EBRT delivered after relapse to RIT, especially in patients with bulky disease, raises the question of whether the 2 modalities should be tested in combination in future trials. Indeed, the safety and efficacy of combining EBRT and RIT has been investigated in the laboratory using human colon cancer xenografts in a nude mouse model.24, 25 In this setting, multimodality therapy proved to be well-tolerated and produced greater tumor shrinkage and longer time to progression than either EBRT or RIT alone. Additionally, concurrent treatment with EBRT and RIT produced better results than sequential therapy. Studies in human colon carcinoma xenografts have been conducted to examine tumor uptake of 125I-labeled tumor-associated antibodies when administered alone or in combination with EBRT.26 Maximum radiolabeled antibody tumor uptake was achieved when the antibody was administered prior to external irradiation, which may have implications regarding optimal sequencing of therapies.
With respect to relapsed NHL, experience has shown that responses to 90Y-ibritumomab tiuxetan are variable and may depend on a number of tumor-related factors, including histology and tumor bulk.7, 8, 16, 17 Both nonfollicular histology and large tumor size (variably defined) have been implicated as negative predictive factors. Combining focal EBRT with RIT is conceptually appealing in that it may allow higher absorbed doses for relatively radioresistant cells, and might result in improved, more homogeneous, dose delivery to bulky tumors.6 The study of external beam radiation therapy and RIT in the context of prospective clinical trials will also allow for more refined analysis of the relationship between dose, site irradiated, effect of prior treatment, response, and toxicity.
- 9Randomized Controlled Trial of 90Y-Labeled Ibritumomab Tiuxetan (Zevalin') Radioimmunotherapy versus Rituximab Immunotherapy for Patients with Relapsed or Refractory Low-Grade, Follicular, or Transformed B-cell Non-Hodgkin's Lymphoma. J Clin Oncol. 2002; 20: 2453–2463., , , et al.
- 23Rationale and feasibility of combined radiotherapy and anti-CD20 immunotherapy in CD20+ B-cell NHL. Am Soc Ther Radiol Oncol (ASTRO). 2002; 54( Suppl): 141–142., .