- Email: [email protected]
Rates and Durability of Response to Salvage Radiation Therapy Among Patients With Refractory or Relapsed Aggressive Non-Hodgkin Lymphoma
International Journal of
Radiation Oncology biology
physics
www.redjournal.org
Clinical Investigation
Rates and Durability of Response to Salvage Radiation Therapy Among Patients With Refractory or Relapsed Aggressive Non-Hodgkin Lymphoma Yolanda D. Tseng, MD,* Yu-Hui Chen, MS, MPH,y Paul J. Catalano, ScD,y,z and Andrea Ng, MD, MPHx *Department of Radiation Oncology, University of Washington, Seattle, Washington; yDepartment of Biostatistics and Computational Biology, Dana-Farber Cancer Institute, Boston, Massachusetts; z Department of Biostatistics, Harvard School of Public Health, Boston, Massachusetts; and x Department of Radiation Oncology, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts Received Apr 11, 2014, and in revised form Sep 8, 2014. Accepted for publication Sep 30, 2014.
Summary Salvage radiation therapy has been used to cytoreduce or consolidate patients with relapsed or refractory aggressive non-Hodgkin lymphoma. Within this single-institution experience, 86% of sites treated with curative intent responded to salvage radiation therapy. However, the durability of response was moderate (5-year local control, 66%). Patients with refractory disease or nonresponse to initial chemotherapy had shorter time to local recurrence, suggesting that these patients
Purpose: To evaluate the response rate (RR) and time to local recurrence (TTLR) among patients who received salvage radiation therapy for relapsed or refractory aggressive non-Hodgkin lymphoma (NHL) and investigate whether RR and TTLR differed according to disease characteristics. Methods and Materials: A retrospective review was performed for all patients who completed a course of salvage radiation therapy between January 2001 and May 2011 at Brigham and Women’s Hospital/Dana-Farber Cancer Institute. Separate analyses were conducted for patients treated with palliative and curative intent. Predictors of RR for each subgroup were assessed using a generalized estimating equation model. For patients treated with curative intent, local control (LC) and progression-free survival were estimated with the Kaplan-Meier method; predictors for TTLR were evaluated using a Cox proportional hazards regression model. Results: Salvage radiation therapy was used to treat 110 patients to 121 sites (76 curative, 45 palliative). Salvage radiation therapy was given as part of consolidation in 18% of patients treated with curative intent. Median dose was 37.8 Gy, with 58% and 36% of curative and palliative patients, respectively, receiving 39.6 Gy or higher. The RR was high (86% curative, 84% palliative). With a median follow-up of 4.8 years among living patients, 5-year LC and progression-free survival for curative patients were 66% and 34%, respectively. Refractory disease (hazard ratio 3.3; PZ.024) and
Reprint requests to: Yolanda D. Tseng, MD, Department of Radiation Oncology, University of Washington, 1959 NE Pacific St, Box 356049, Seattle, WA 98195. Tel: 206-598-4100; E-mail: [email protected] Presented in part at the 55th Annual Meeting of the American Society for Radiation Oncology, September 22-25, 2013, Atlanta, GA. Int J Radiation Oncol Biol Phys, Vol. 91, No. 1, pp. 223e231, 2015 0360-3016/$ - see front matter Ó 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ijrobp.2014.09.041
Conflict of interest: none. AcknowledgmentsdThe authors thank Dr Peter M. Mauch for his valuable input, and all the lymphoma patients he has treated and diligently followed, which allowed this study to be conducted.
224
International Journal of Radiation Oncology Biology Physics
Tseng et al.
may benefit from dose escalation or addition of radiosensitizers.
lack of response to initial chemotherapy (hazard ratio 4.3; PZ.007) but not dose (PZ.93) were associated with shorter TTLR. Despite doses of 39.6 Gy or higher, 2-year LC was only 61% for definitive patients with refractory disease or disease that did not respond to initial chemotherapy. Conclusions: Relapsed or refractory aggressive NHL is responsive to salvage radiation therapy, and durable LC can be achieved in some cases. However, refractory disease is associated with a shorter TTLR, suggesting that radiation dose escalation, addition of radiosensitizers, or a combination of both may be indicated in these patients. Ó 2015 Elsevier Inc.
Introduction Definitive treatment of aggressive non-Hodgkin lymphoma (NHL) is associated with 5-year progression-free survival rates of 60% to 80% (1). In patients who relapse or have refractory disease after chemotherapy, prognosis is worse, with event-free survival ranging between 20% and 50% (2). Salvage radiation therapy may play an important role in these patients to cytoreduce and achieve minimal residual disease before high-dose therapy or as part of consolidation. Alternatively, salvage radiation therapy may provide local symptom control in the palliative setting. Prior retrospective studies and anecdotal experience suggest that relapsed or refractory aggressive NHL tend to be radio-resistant (3), with a local failure rate of >50%. Alternative strategies have been explored to intensify radiation therapy, including accelerated hyperfractionation and concurrent chemotherapy and radiation therapy (4, 5). However, most available series were based on a relatively small number of patients treated in the pre-rituximab era. In a randomized trial comparing 30 Gy versus 40 to 45 Gy for patients with aggressive NHL, there was no difference in overall response rate or local control (LC) between the standard and lower-dose arms (6), although subgroup analysis was not performed among the small group of patients with refractory or relapsed disease (17% of cohort). In this study, we aimed to assess the response rate and time to local recurrence (TTLR) of a modern cohort of relapsed or refractory aggressive NHL patients treated with salvage radiation therapy either with definitive or palliative intent. In addition, we sought to identify clinical factors that predict improved rates and durability of response to radiation therapy.
Methods and Materials Patient cohort and eligibility We retrospectively reviewed the records of 110 consecutive patients who completed a course of salvage radiation therapy to 121 sites for refractory or relapsed aggressive NHL between January 1, 2001, and May 31, 2011, at Brigham and Women’s Hospital/Dana-Farber Cancer Institute in accordance with the Dana-Farber Cancer Institute institutional
review board. Eligible patients were adults aged 18 years with aggressive NHL. Both de novo and histologic transformation of indolent to aggressive NHL (eg from follicular to diffuse large B-cell lymphoma [DLBCL]) were included. Patients were excluded if they received salvage radiation therapy for treatment of primary cutaneous lymphoma or disease involving the skin or the central nervous system. For patients who received more than 1 course of salvage radiation therapy, only the first course between January 2001 and May 2011 was included in this analysis. We extracted from medical records the following variables that could influence radiation therapy response and TTLR: age, gender, histology, transformed versus de novo disease, radiation therapy dose, once-versus twice-daily treatment, response to initial chemotherapy, and whether disease was relapsed or refractory. Refractory disease was defined as a <50% tumor response to the chemotherapy regimen given immediately before salvage radiation therapy (7). Because many patients received more than one chemotherapy regimen before salvage radiation therapy, some patients defined as having refractory disease may have initially responded to induction chemotherapy, whereas others never responded and may have even progressed on chemotherapy. Relapsed disease was new disease at any site on imaging or biopsy after an initial complete response (CR) to chemotherapy. A patient responded to chemotherapy if he or she achieved a CR or partial response (PR).
Salvage radiation therapy treatment Computed tomographyebased simulations were performed on all patients. When available, positron emission tomography fusion was used to delineate the target volume. Three-dimensional conformal techniques were primarily used, although intensity modulated radiation therapy was occasionally used if appropriate. In some urgent cases, patients were initially treated once daily but subsequently were switched after 2 to 3 fractions to twice-daily treatment; for the purposes of our analysis these patients were categorized as receiving twice-daily treatment. In addition to the dose delivered, the biological effective dose (BED) was calculated using an a/b ratio of 10 for each treated site. Treatment intent was decided by the treating radiation oncologist at the time of consultation and planning. Cases
Volume 91 Number 1 2015
in which salvage radiation therapy was given for consolidation after transplantation or after salvage chemotherapy or for cytoreduction with the intention of proceeding with transplantation were typically deemed as cases treated with curative intent. Patients who failed after transplantation or who were not transplantation candidates were treated with palliative intent, with radiation dose-fractionation schemes that tended to include larger fraction size and a lower total dose. For curative cases, only sites of relapsed or refractory disease were included in the clinical target volume. Other initially involved sites that achieved a CR with chemotherapy were generally not included, unless they were immediately adjacent to the relapsed/refractory site and their inclusion would result in limited toxicity. For palliative cases, only sites associated with active or impending symptoms were treated. Compared with curative patients, a smaller proportion of palliative patients (47% vs 93%) were irradiated to the only site(s) of active disease.
Assessment of response, disease recurrence/ progression, and death Response to salvage radiation therapy and chemotherapy was based on diagnostic gallium (10%), positron emission tomography/CT (66%), or CT scans (65%) usually within 3 months of treatment. Among symptomatic palliative patients, response was based on physician-reported outcomes and imaging, if available. Response at each treated site was graded as a CR, PR, stable disease (SD), or progressive disease (PD) according to the Revised International Response Criteria definition (8). Curative patients who responded (ie CR or PR) to salvage radiation therapy were then followed for LC. Local control was defined as maintenance of a local CR or the lack of progression of local disease in those with a PR. In general, patients were followed every 2 to 3 months for the first 1 to 2 years, then twice a year to 5 years, and annually thereafter. Death was a high competing risk among palliative patients, and 24% of these patients did not have imaging follow-up. Therefore, within this subgroup the analysis was limited to response rate and survival. If applicable, a patient’s date and cause of death was extracted from his or her medical record or the National Death Index.
Salvage RT for relapsed/refractory NHL
225
without LR were censored at date of last disease assessment. Progression-free survival was calculated from date of radiation therapy completion to date of any relapse or progression or death, whichever occurred first; patients were censored at the last date of disease assessment if they were alive without recurrence or if they died without documented recurrence but death occurred beyond the typical follow-up interval after the date of last disease assessment. Eleven patients who were treated to 2 separate sites were excluded from this analysis because PFS was analyzed by patient (vs site). Overall survival was defined as the time from radiation therapy completion to the date of death. Response to radiation therapy and TTLR were analyzed by site, whereas OS and PFS were analyzed by patient. Separate analyses were performed for sites that were treated with curative versus palliative intent, given the potentially confounding effects of the patient’s disease status and treatment. To assess the effect of tumor and treatment characteristics on response to radiation therapy, a univariate generalized estimating equations estimation model (9) was used with an unstructured correlation matrix to account for the correlation among disease sites within a patient. A univariate Cox proportional hazards regression model was used to evaluate the associations of tumor and treatment characteristics with TTLR and PFS. The following tumor and treatment characteristics were evaluated: refractory versus relapsed disease, transformed versus de novo disease, radiation therapy dose (<39.6 Gy vs 39.6 Gy), frequency of treatment (daily vs twice daily), and response to initial chemotherapy (responder vs nonresponder). Because administration of systemic therapy after salvage radiation therapy could alter the risk of LR, a timedependent covariate for systemic treatment was included in the Cox regression model. The marginal approach developed by Lee et al (10) was also used to account for the dependence among treated sites within a patient in the time-to-event models. Local control, PFS, and OS rates at 2 and 5 years were estimated using the Kaplan-Meier method. All analyses were conducted using SAS (version 9.3; SAS Institute, Cary, NC). P values of <.05 were considered significant. All tests were two-sided.
Results
Statistical analysis
Patients treated with curative intent
A total of 110 patients were treated to 121 sites during their first course of salvage radiation therapy and formed the cohort for this analysis. The primary endpoint was response to salvage radiation therapy, defined as either a CR or PR; secondary endpoints included TTLR and progression-free survival (PFS) among definitive patients and overall survival (OS). Time to local recurrence was defined as the time from radiation therapy completion to the date of local recurrence (LR) for sites that achieved a CR or PR. Sites
Seventy-two patients were treated with curative intent to 76 sites. The median age at diagnosis was 54 years, with 61% of patients being male, 78% with DLBCL, and 63% with refractory disease (Table 1). Most patients (63%) received a median of 6 cycles of rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP) initially (Table 2), with 56% achieving a response (ie responder to initial chemotherapy). Most patients (72%) at one point in their treatment course
226
International Journal of Radiation Oncology Biology Physics
Tseng et al.
Table 1 Patient demographic and disease characteristics among patients treated with curative (nZ72, 76 sites) and palliative intent (nZ38, 45 sites) Treatment intent Curative Characteristic Age at diagnosis (y)* (median and range) Refractory/relapsed Refractory Relapsed Gender Female Male Histology Diffuse large B-cell lymphoma Grade 3 follicular lymphoma T cell lymphoma Othery Transformed disease No Yes Time from diagnosis to transformation (mo) (median and range) Pathologically confirmed among transformed patients Histology before transformation G1-2 follicular Chronic lymphocytic leukemia Other
n
Palliative
Total
%
n
%
n
%
54
19-88
62.5
32-86
55
19-88
45 27
63 38
30 8
79 21
75 35
68 32
28 44
39 61
16 22
42 58
44 66
40 60
56 5 3 8
78 7 4 11
29 2 2 5
76 5 6 13
85 7 5 13
77 6 5 12
58 42 0-235.6
78 32 33.9
71 29 0-235.6
56 16 14.0
78 22 0-151.4
22 16 47.8
14
88
12
75
26
81
11 3 2
69 19 13
8 6 2
50 38 13
19 9 4
59 28 13
* Age at diagnosis for de novo disease patients and age at transformation for patients with transformed disease. y Including clinical suspicion of high grade non-Hodgkin lymphoma (NHL) (3), follicular versus large B-cell lymphoma (1), high-grade NHL not otherwise specified (1), large-cell lymphoma (2), mantle cell blastic variant (1), mixed follicular and large B-cell lymphoma (1), primary effusion lymphoma (1), T-cell rich B-cell NHL (2), and undifferentiated high grade B-cell NHL (1).
underwent high-dose chemotherapy with stem cell transplant (SCT; Table 2); 42 of these patients underwent SCT either immediately before or after salvage radiation therapy. Eighteen received salvage radiation therapy after SCT for consolidation (nZ8), cytoreduction (nZ6), or unknown disease status (nZ4), and 20 received salvage radiation therapy before SCT for residual disease (ie cytoreduction). After salvage chemotherapy, 5 patients received salvage radiation therapy as consolidation in lieu of SCT. Median dose to the 76 treated sites was 40 Gy, with 20% of sites being treated twice daily. The most commonly used twicedaily regimen was 1.4 Gy per fraction for 24 to 26 fractions.
Patients treated with palliative intent The majority of palliative patients (20 of 38, 53%) were symptomatic at the time of salvage radiation therapy, with pain (61%), obstructive symptoms (11%), cord compression (11%), bleeding (5%), brachial plexopathy (5%), and eye irritation (5%). Similar to patients treated with curative intent, the majority of these patients were male (58%), had DLBCL (76%), and refractory disease (79%). However, patients treated with palliative intent were older at
diagnosis (median, 62.5 years; Table 1) and were treated to a lower median salvage radiation therapy dose (36 Gy; Table 3).
Salvage radiation therapy response, LC, and PFS among patients treated with curative intent The majority (86%) of patients achieved a CR or PR to salvage radiation therapy (Table 4), and all patients with relapsed disease responded. With a median-follow-up of 4.8 years (range, 0.2-11.3 years) among living patients, the 2-year and 5-year LC rates were 72% and 66%. The majority of LRs occurred within the first 2 years after radiation therapy treatment (Fig. 1). Presence of de novo (vs transformed disease) was associated with a significantly higher odds (odds ratio [OR] 5.2; 95% confidence interval [CI] 1.3-20.5; PZ.02) of achieving a response to salvage radiation therapy. In contrast, presence of refractory disease (hazard ratio [HR] 3.3; PZ.024) and nonresponse to initial chemotherapy (HR 4.3; PZ.007) were associated with significantly shorter TTLR (Table 5). Among patients whose disease was either refractory or did not respond to initial chemotherapy and received 39.6 Gy (nZ27), the 2-year
Volume 91 Number 1 2015
227
Salvage RT for relapsed/refractory NHL
LC was only 61%. In the subgroup of patients treated definitively with doses of >50 Gy or with twice-daily treatments, 2-year LC was 67%. Progression-free survival at 2 and 5 years was 46% and 34%, respectively. There were no significant predictors for improved PFS (Table 5).
therapy dose was also paradoxically associated with a higher response rate (OR 6.2; 95% CI 1.1-36.1; PZ.044), which may be related to selection bias (discussed below).
Salvage radiation therapy response among patients treated with palliative intent
The median survival among patients who received curative and palliative radiation therapy was 4.7 and 0.4 years, respectively. Among the 74 patients (42 curative, 32 palliative) who have died, the majority (31 curative, 29 palliative) succumbed to their aggressive NHL. The 2- and 5-year OS rates among patients treated with curative intent were 60% and 43%, respectively (Fig. 1); rates were 19% and 8% among palliatively treated patients.
Despite the poor prognosis of palliative patients, 84% achieved a response (clinical or imaging) to salvage radiation therapy (Table 4). Among the 20 patients who were symptomatic at the time of salvage radiation therapy, 16 (80%) had symptomatic improvement during or immediately after treatment. In 3 patients, the clinical response was unknown, and there was no response in 1 patient. Response to initial chemotherapy was associated with a significantly higher response rate (OR 10.3; 95% CI 1.1-95.2; PZ.04). A lower radiation
Overall survival
Treatment toxicity Salvage radiation therapy was generally well tolerated, with 8% and 5% of curative and palliative patients experiencing grade 3
Table 2 Initial chemotherapy treatments received before salvage radiation therapy and rates and timing of high-dose chemotherapy with stem cell transplantation Treatment intent Curative
Palliative
Total
Parameter
n
%
n
%
n
%
No. of chemotherapy regimens before first salvage RT (median and range) Initial chemotherapy Regimen CHOP R-CHOP Other* No. of cyclesy (median and range) Response after initial chemotherapy CR PR SD PD Unknown Rates of any SCT (nZ67) SCT type Autologous SCT Allogeneic SCT >1 SCT SCT immediately before or after salvage RT (nZ42) SCT / consolidative RT SCT / cytoreductive RT SCT / RT (disease state unknown) Cytoreductive RT / SCT
2
1-4
2
1-7
2
1-7
20 45 7 6
28 63 10 1-11
11 23 4 6
29 61 11 1-8
31 68 11 6
28 62 10 1-11
23 16 14 17 2
33 23 20 24
17 4 4 12 1
46 11 11 32
40 20 18 29 3
37 19 17 27
46 6 6
64 8 8
8 7 1
21 18 3
54 13 7
49 12 6
8 6 4 20
11 8 6 28
0 1 0 3
0 3 0 8
8 7 4 23
7 6 4 21
Abbreviations: CR Z complete response; PD Z progressive disease; PR Z partial response; RT Z radiation therapy; SD Z stable disease; SCT Z stem cell transplantation; CHOP Z cyclophosphamide, doxorubicin, vincristine, prednisone; R-CHOP Z rituximab, cyclophosphamide, doxorubicin, vincristine, prednisone. * Including CVP (cyclophosphamide, vincristine, prednisone), chlorambucil, HAM (high-dose cytosine arabinoside, mitoxantrone), ProMACECytaBOM (prednisone, doxorubicincyclophosphamide, etoposide, cytarabine, bleomycin, vincristine, methotrexate, leucovorin), R-CVP (rituximab, cyclophosphamide, vincristine, prednisone), REPOCH (rituximab, etoposide, prednisone, vincristine, cyclophosphamide, doxorubicin), RESHAP (rituximab, etoposide, methylprednisolone, cytarabine, platinum agent), RICE (rituximab, ifosfamide, carboplatin, etoposide), and rituximab. y For 2 patients, the number of initial chemotherapy cycles was unknown.
228
International Journal of Radiation Oncology Biology Physics
Tseng et al.
Table 3 Details of salvage radiation therapy and highest acute toxicity by patient (nZ110). Location and salvage radiation therapy dose by site (nZ121) Curative Parameter No. of sites treated during first course 1 2 Total no. of RT courses (median and range) Time from diagnosis of de novo disease or transformation to salvage RT (mo) (median and range) Acute toxicity during RT course Grade 0 Grade 1 Grade 2 Grade 3 Unknown Location Head and neck Axilla Mediastinum or thorax Extranodal, GI tract Abdominal/pelvic lymph nodes Bone Paraspinal Soft tissue (extremity) Other* Unknown Radiation dose (Gy) (median and range) BED10 (Gy10) (median and range) 20 20.01-30 30.01-40 40.01-50 50.01-60 >60 Fractionation frequency Once daily Twice dailyy
Palliative
Total
n
%
n
%
n
%
68 4 1 10.6
94 6 1-7 0.6-219.6
31 7 1 10.2
82 18 1-6 0.2-93.6
99 11 1 10.5
90 10 1-7 0.2-219.6
17 42 6 6 1
24 59 8 8
15 18 3 2 0
39 47 8 5
32 60 9 8 1
29 55 8 7
12 4 17 3 31 4 0 2 2 1 40.0 47.2 1 0 13 38 22 2
16 5 23 4 41 5 0 3 3
14 2 1 3 12 4 4 4 1 0 36.0 43.2 0 2 10 27 5 1
31 4 2 7 27 9 9 9 2
26 6 18 6 43 8 4 6 3 1 37.8 43.8 1 2 23 65 27 3
22 5 15 5 36 7 3 5 3
61 15
16.5-55.7 19.0-64.8 1 0 17 50 29 3 80 20
35 10
13.0-54.0 21.5-64.8 0 4 22 60 11 2 78 22
96 25
13.0-55.7 19.0-64.8 1 2 19 54 22 2 79 21
Abbreviations: BED10 Z biological effective dose; GI Z gastrointestinal. Other abbreviations as in Table 2. * Including head and neck with mediastinum (1), soft tissue and spleen (1), and mediastinum and bone (1). y Includes patients who were initially started with once daily but received twice daily treatment for the majority of their radiation therapy course.
acute toxicity, respectively; no grade 4 acute toxicity was noted. Five definitively treated patients died from myelodysplastic syndrome (nZ3) or acute myeloid leukemia (nZ2).
Discussion In this relatively large, single-institution experience of salvage radiation therapy for refractory or relapsed aggressive NHL, response to radiation therapy was high, but durability of response varied among patients treated with curative intent for refractory versus relapsed disease (2-year LC 56% vs 92%). There was no evidence of a doseeresponse relationship with LC detected, and intensification of radiation therapy (>50 Gy or twice-daily treatment) was associated with an LC rate of only 67% at
2 years. These results suggest that the biology underlying relapsed and refractory aggressive NHL differs, at least with respect to durability of response to salvage radiation therapy, and that there remain subgroups of patients who may benefit from treatment intensification. Prior single-institution studies on salvage radiation therapy for relapsed or refractory lymphoma reported LC rates ranging from 47% to 73% at 2 years (3-5). In a retrospective, single-institution experience from Wayne State University, 35 patients (69% with refractory disease) were treated with salvage radiation therapy, which yielded a 2-year local failure rate of 54%. Furthermore, none of the 5 patients who progressed through chemotherapy achieved LC with salvage radiation therapy. However, there was a trend toward improved LC (66% vs 39% at 2 years; PZ.06) with radiation therapy doses of 39.6 Gy.
Volume 91 Number 1 2015 Table 4
Salvage RT for relapsed/refractory NHL
229
Response to salvage radiation therapy in disease sites treated with curative versus palliative intent Treatment intent Curative
Parameter
n
No. w/response (%)
Overall 76 Disease type Relapsed 27 Refractory 49 Transformed disease No 58 Yes 18 Dose (Gy) High (39.6) 42 Low (<39.6) 34 Frequency Twice daily 15 Once daily 61 Response to initial chemotherapy Response 39 Nonresponse 37 Number of prior chemotherapy 1-2 57 >2 19
Palliative 95% CI
n
No. w/response (%)
OR
95% CI
65 (86)
d
OR
d
45
38 (84)
d
d
27 (100) 38 (78)
d 1.0
d
8 37
8 (100) 30 (81)
d 1.0
d
53 (91) 12 (67)
5.2* 1.0
1.3-20.5*
27 18
23 (85) 15 (83)
1.2 1.0
0.2-6.0
35 (83) 30 (88)
1.0 1.5
0.4-5.0
16 29
11 (69) 27 (93)
1.0 6.2*
12 (80) 53 (87)
1.0 1.7
0.4-6.3
10 35
10 (100) 28 (80)
d 1.0
d
36 (92) 29 (78)
2.7 1.0
0.6-11.3
25 20
24 (96) 14 (70)
10.3* 1.0
1.1-95.2*
49 (86) 16 (84)
1.0 1.6
27 18
23 (85) 15 (83)
1.0 0.9
0.3-8.9
1.1-36.1*
0.2-4.4
Abbreviations: CI Z confidence interval; OR Z odds ratio. * Significant predictor for response to salvage radiation therapy.
In studies in which salvage radiation therapy was intensified, LC seemed to be higher. Among 34 patients (85% with refractory disease) who were treated at Princess Margaret Hospital with twice-daily radiation therapy (39.940.5 Gy in 30 fractions), a 2-year LC rate of 73% was
Fig. 1. Cumulative incidence of local control (LC), progression free survival (PFS), and overall survival (OS) among patients treated with curative intent. Note that the number at risk differs for each endpoint given the subgroups of the cohort examined. OS includes all patients treated with curative-intent RT; PFS only includes patients treated to a single site; and LC includes those patients who achieved a complete or partial response to radiation therapy.
achieved, with the majority of local and distant recurrences occurring within the first year after radiation treatment (5). A phase 2 study from the Institut Gustave Roussy reported an LC rate of 63% among 21 patients with bulky relapsed or refractory lymphoma (71% with refractory disease) treated with split course radiation therapy to a median dose of 40 Gy concurrent with chemotherapy (76% cisplatin with etoposide). However, in this study 70% of patients experienced grade 3 to 4 hematologic toxicity (4). In the present study, the 2-year LC rate of 72% among patients treated with curative intent compared favorably with patients from prior studies that used intensified treatment, but this may reflect a more favorable prognostic case mix, modern radiation therapy techniques, more effective systemic therapy, and inclusion of patients who received salvage radiation therapy as part of consolidation. This single-institution study is limited by its retrospective study design, and treatment received was subject to selection bias and likely correlated with aggressiveness of disease. As such, we were unable to detect a doseeresponse relationship for LC. Indeed, LC was higher among patients treated with low (<39.6 Gy) versus high (39.6 Gy) radiation therapy doses and daily versus twice-daily treatment, regardless of treatment intent. We also obtained similar results regardless of whether dose was dichotomized by either the median BED10 or median prescribed dose (data not shown). However, our study did identify subgroups of patients with significantly inferior LC with salvage radiation therapy, namely patients who failed to respond to initial chemotherapy and patients with
230
International Journal of Radiation Oncology Biology Physics
Tseng et al.
Table 5
Time to local recurrence in disease sites treated with curative versus palliative intent Treatment intent Curative
Parameter
N
Overall 65 Disease type Relapsed 27 Refractory 38 Transformed disease No 53 Yes 12 Dose (Gy) High (39.6) 35 Low (<39.6) 30 Frequency Twice daily 12 Once daily 53 Response to initial chemotherapy Response 36 Nonresponse 29
2-y LC (%)
Palliative P
n
2-y LC (%)
HR
P
72
d
HR
d
38
43
d
d
92 56
1.0 3.3*
.024
8 30
80 35
1.0 3.1
.22
69 88
4.1 1.0
.16
23 15
33 82
1.3 1.0
.75
67 79
1.1 1.0
.93
11 27
0 56
2.5 1.0
.23
60 73
1.2 1.0
.79
10 28
dy 34
dz
88 51
1.0 4.3*
.007
24 14
28 56
1.0 1.1
d
.90
Abbreviation: LC Z local control. Other abbreviations as in Table 4. * Significant predictor for shorter time to local recurrence. y All palliative sites treated twice daily were censored before 2 years. z Hazard ratio cannot be estimated because no local recurrences occurred among palliative sites treated twice daily.
chemotherapy-refractory disease at the time of radiation therapy. Despite radiation dose escalation in these patients, approximately one-third relapsed locally after salvage radiation therapy. The biology driving these observations is unknown, but they suggest that chemotherapy-resistant NHL may have lower intrinsic radiosensitivity, a high rate of proliferation, or both. Double hit pathology with 2 translocations involving c-MYC and BCL2 or BCL6 is associated with biologically aggressive lymphomas, but it is unknown what proportion of this cohort had these translocations because this was not routinely obtained. Additional studies are required to explore whether the double hit pathology is enriched among refractory or relapsed NHL patients and to determine whether novel approaches such as hyperfractionation or addition of radiosensitizers (11, 12) may benefit chemotherapy-resistant patients. Finally, this cohort represents a heterogeneous group of patients who received radiation therapy alone, in conjunction with SCT, or after salvage chemotherapy. The relatively high PFS and OS observed among curative patients likely reflect the use of tandem systemic therapy. Death was a significant competing risk, in particular among palliative patients. These patients were deemed to have poor prognosis and/or be ineligible for transplantation at the time of consultation. Overall survival at 2 years was 19%, and almost one-third of these palliative patients were unable to be assessed for relapse before death. Despite the poor prognosis, however, the majority of symptomatic, palliative patients had a clinical response to salvage radiation therapy, and the response rate among all palliative patients (with or without symptoms) was >80%. Therefore
salvage radiation therapy should be considered in symptomatic refractory or relapsed NHL patients, with the hope to improve quality of life. Similarly, it is remarkable that even among patients who have chemotherapy-refractory or relapsed NHL and who are treated with curative intent, the overall response rate to salvage radiation therapy was 86%. Given these findings, there may be a role to introduce salvage radiation therapy earlier in the course of a patient’s relapsed or refractory disease. Future studies are needed to examine whether, instead of using multiple chemotherapy regimens first, early intervention with salvage radiation therapy may be associated with improved LC and/or allow potentially a lower dose of radiation therapy to be used. To our knowledge, our study is among the first to report on the efficacy of salvage radiation therapy among patients whose aggressive NHL transformed from low-grade disease. Although we found that patients with transformed disease had a lower initial response rate to radiation therapy, those who achieved a response had significantly more durable LC than patients with de novo disease. These results suggest that patients with transformed disease fare as well as, if not better than, patients with de novo disease, in terms of durability of LC with salvage radiation therapy. In summary, our study showed that overall, relapsed or refractory aggressive NHL is responsive to salvage radiation therapy, and durable LC can be achieved in some cases. However, patients with refractory or nonresponse to initial chemotherapy had less durable LC, despite relatively high doses of radiation. The optimal salvage radiation therapy dose and fractionation scheme in addition to the
Volume 91 Number 1 2015
role for concurrent radiosensitizing agents remain to be determined and can be considered as part of future, prospective studies.
References 1. Miller TP, Dahlberg S, Cassady JR, et al. Chemotherapy alone compared with chemotherapy plus radiotherapy for localized intermediate- and high-grade non-Hodgkin’s lymphoma. N Engl J Med 1998;339:21-26. 2. Philip T, Guglielmi C, Hagenbeek A, et al. Autologous bone marrow transplantation as compared with salvage chemotherapy in relapses of chemotherapy-sensitive non-Hodgkin’s lymphoma. N Engl J Med 1995;333:1540-1545. 3. Aref A, Narayan S, Tekyi-Mensah S, et al. Value of radiation therapy in the management of chemoresistant intermediate grade non-Hodgkin’s lymphoma. Radiat Oncol Invest 1999;7:186-191. 4. Girinsky T, Lapusan S, Ribrag V, et al. Phase II study of concomitant chemoradiotherapy in bulky refractory or chemoresistant relapsed lymphomas. Int J Radiat Oncol Biol Phys 2005; 61:476-479.
Salvage RT for relapsed/refractory NHL
231
5. Martens C, Hodgson DC, Wells WA, et al. Outcome of hyperfractionated radiotherapy in chemotherapy-resistant non-Hodgkin’s lymphoma. Int J Radiat Oncol Biol Phys 2006;64:1183-1187. 6. Lowry L, Smith P, Qian W, et al. Reduced dose radiotherapy for local control in non-Hodgkin lymphoma: A randomised phase III trial. Radiother Oncol 2011;100:86-92. 7. Friedberg JW. Relapsed/refractory diffuse large B-cell lymphoma. Hematology Am Soc Hematol Educ Program 2011;2011:498-505. 8. Cheson BD, Pfistner B, Juweid ME, et al. Revised response criteria for malignant lymphoma. J Clin Oncol 2007;25:579-586. 9. Zeger SL, Liang KY. Longitudinal data analysis for discrete and continuous outcomes. Biometrics 1986;42:121-130. 10. Lee E, Wei LJ, Amato D, et al. Cox-type regression analysis for large numbers of small groups of correlated failure time observations. In: Klein J, Goel P, editors. Survival Analysis: State of the Art. Dordrecht, The Netherlands: Springer Netherlands; 1992:237-247. 11. Kapadia NS, Engles JM, Wahl RL. In vitro evaluation of radioprotective and radiosensitizing effects of rituximab. J Nucl Med 2008;49:674-678. 12. Elstrom RL, Leonard JP, Christos P, et al. Phase 1 study of radiosensitization using bortezomib in patients with relapsed non-Hodgkin’s lymphoma receiving radioimmunotherapy. Presented at the American Society of Hematology 54th Annual Meeting. December 811, 2012; Atlanta, GA.
Radiation Oncology biology
physics
www.redjournal.org
Clinical Investigation
Rates and Durability of Response to Salvage Radiation Therapy Among Patients With Refractory or Relapsed Aggressive Non-Hodgkin Lymphoma Yolanda D. Tseng, MD,* Yu-Hui Chen, MS, MPH,y Paul J. Catalano, ScD,y,z and Andrea Ng, MD, MPHx *Department of Radiation Oncology, University of Washington, Seattle, Washington; yDepartment of Biostatistics and Computational Biology, Dana-Farber Cancer Institute, Boston, Massachusetts; z Department of Biostatistics, Harvard School of Public Health, Boston, Massachusetts; and x Department of Radiation Oncology, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts Received Apr 11, 2014, and in revised form Sep 8, 2014. Accepted for publication Sep 30, 2014.
Summary Salvage radiation therapy has been used to cytoreduce or consolidate patients with relapsed or refractory aggressive non-Hodgkin lymphoma. Within this single-institution experience, 86% of sites treated with curative intent responded to salvage radiation therapy. However, the durability of response was moderate (5-year local control, 66%). Patients with refractory disease or nonresponse to initial chemotherapy had shorter time to local recurrence, suggesting that these patients
Purpose: To evaluate the response rate (RR) and time to local recurrence (TTLR) among patients who received salvage radiation therapy for relapsed or refractory aggressive non-Hodgkin lymphoma (NHL) and investigate whether RR and TTLR differed according to disease characteristics. Methods and Materials: A retrospective review was performed for all patients who completed a course of salvage radiation therapy between January 2001 and May 2011 at Brigham and Women’s Hospital/Dana-Farber Cancer Institute. Separate analyses were conducted for patients treated with palliative and curative intent. Predictors of RR for each subgroup were assessed using a generalized estimating equation model. For patients treated with curative intent, local control (LC) and progression-free survival were estimated with the Kaplan-Meier method; predictors for TTLR were evaluated using a Cox proportional hazards regression model. Results: Salvage radiation therapy was used to treat 110 patients to 121 sites (76 curative, 45 palliative). Salvage radiation therapy was given as part of consolidation in 18% of patients treated with curative intent. Median dose was 37.8 Gy, with 58% and 36% of curative and palliative patients, respectively, receiving 39.6 Gy or higher. The RR was high (86% curative, 84% palliative). With a median follow-up of 4.8 years among living patients, 5-year LC and progression-free survival for curative patients were 66% and 34%, respectively. Refractory disease (hazard ratio 3.3; PZ.024) and
Reprint requests to: Yolanda D. Tseng, MD, Department of Radiation Oncology, University of Washington, 1959 NE Pacific St, Box 356049, Seattle, WA 98195. Tel: 206-598-4100; E-mail: [email protected] Presented in part at the 55th Annual Meeting of the American Society for Radiation Oncology, September 22-25, 2013, Atlanta, GA. Int J Radiation Oncol Biol Phys, Vol. 91, No. 1, pp. 223e231, 2015 0360-3016/$ - see front matter Ó 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ijrobp.2014.09.041
Conflict of interest: none. AcknowledgmentsdThe authors thank Dr Peter M. Mauch for his valuable input, and all the lymphoma patients he has treated and diligently followed, which allowed this study to be conducted.
224
International Journal of Radiation Oncology Biology Physics
Tseng et al.
may benefit from dose escalation or addition of radiosensitizers.
lack of response to initial chemotherapy (hazard ratio 4.3; PZ.007) but not dose (PZ.93) were associated with shorter TTLR. Despite doses of 39.6 Gy or higher, 2-year LC was only 61% for definitive patients with refractory disease or disease that did not respond to initial chemotherapy. Conclusions: Relapsed or refractory aggressive NHL is responsive to salvage radiation therapy, and durable LC can be achieved in some cases. However, refractory disease is associated with a shorter TTLR, suggesting that radiation dose escalation, addition of radiosensitizers, or a combination of both may be indicated in these patients. Ó 2015 Elsevier Inc.
Introduction Definitive treatment of aggressive non-Hodgkin lymphoma (NHL) is associated with 5-year progression-free survival rates of 60% to 80% (1). In patients who relapse or have refractory disease after chemotherapy, prognosis is worse, with event-free survival ranging between 20% and 50% (2). Salvage radiation therapy may play an important role in these patients to cytoreduce and achieve minimal residual disease before high-dose therapy or as part of consolidation. Alternatively, salvage radiation therapy may provide local symptom control in the palliative setting. Prior retrospective studies and anecdotal experience suggest that relapsed or refractory aggressive NHL tend to be radio-resistant (3), with a local failure rate of >50%. Alternative strategies have been explored to intensify radiation therapy, including accelerated hyperfractionation and concurrent chemotherapy and radiation therapy (4, 5). However, most available series were based on a relatively small number of patients treated in the pre-rituximab era. In a randomized trial comparing 30 Gy versus 40 to 45 Gy for patients with aggressive NHL, there was no difference in overall response rate or local control (LC) between the standard and lower-dose arms (6), although subgroup analysis was not performed among the small group of patients with refractory or relapsed disease (17% of cohort). In this study, we aimed to assess the response rate and time to local recurrence (TTLR) of a modern cohort of relapsed or refractory aggressive NHL patients treated with salvage radiation therapy either with definitive or palliative intent. In addition, we sought to identify clinical factors that predict improved rates and durability of response to radiation therapy.
Methods and Materials Patient cohort and eligibility We retrospectively reviewed the records of 110 consecutive patients who completed a course of salvage radiation therapy to 121 sites for refractory or relapsed aggressive NHL between January 1, 2001, and May 31, 2011, at Brigham and Women’s Hospital/Dana-Farber Cancer Institute in accordance with the Dana-Farber Cancer Institute institutional
review board. Eligible patients were adults aged 18 years with aggressive NHL. Both de novo and histologic transformation of indolent to aggressive NHL (eg from follicular to diffuse large B-cell lymphoma [DLBCL]) were included. Patients were excluded if they received salvage radiation therapy for treatment of primary cutaneous lymphoma or disease involving the skin or the central nervous system. For patients who received more than 1 course of salvage radiation therapy, only the first course between January 2001 and May 2011 was included in this analysis. We extracted from medical records the following variables that could influence radiation therapy response and TTLR: age, gender, histology, transformed versus de novo disease, radiation therapy dose, once-versus twice-daily treatment, response to initial chemotherapy, and whether disease was relapsed or refractory. Refractory disease was defined as a <50% tumor response to the chemotherapy regimen given immediately before salvage radiation therapy (7). Because many patients received more than one chemotherapy regimen before salvage radiation therapy, some patients defined as having refractory disease may have initially responded to induction chemotherapy, whereas others never responded and may have even progressed on chemotherapy. Relapsed disease was new disease at any site on imaging or biopsy after an initial complete response (CR) to chemotherapy. A patient responded to chemotherapy if he or she achieved a CR or partial response (PR).
Salvage radiation therapy treatment Computed tomographyebased simulations were performed on all patients. When available, positron emission tomography fusion was used to delineate the target volume. Three-dimensional conformal techniques were primarily used, although intensity modulated radiation therapy was occasionally used if appropriate. In some urgent cases, patients were initially treated once daily but subsequently were switched after 2 to 3 fractions to twice-daily treatment; for the purposes of our analysis these patients were categorized as receiving twice-daily treatment. In addition to the dose delivered, the biological effective dose (BED) was calculated using an a/b ratio of 10 for each treated site. Treatment intent was decided by the treating radiation oncologist at the time of consultation and planning. Cases
Volume 91 Number 1 2015
in which salvage radiation therapy was given for consolidation after transplantation or after salvage chemotherapy or for cytoreduction with the intention of proceeding with transplantation were typically deemed as cases treated with curative intent. Patients who failed after transplantation or who were not transplantation candidates were treated with palliative intent, with radiation dose-fractionation schemes that tended to include larger fraction size and a lower total dose. For curative cases, only sites of relapsed or refractory disease were included in the clinical target volume. Other initially involved sites that achieved a CR with chemotherapy were generally not included, unless they were immediately adjacent to the relapsed/refractory site and their inclusion would result in limited toxicity. For palliative cases, only sites associated with active or impending symptoms were treated. Compared with curative patients, a smaller proportion of palliative patients (47% vs 93%) were irradiated to the only site(s) of active disease.
Assessment of response, disease recurrence/ progression, and death Response to salvage radiation therapy and chemotherapy was based on diagnostic gallium (10%), positron emission tomography/CT (66%), or CT scans (65%) usually within 3 months of treatment. Among symptomatic palliative patients, response was based on physician-reported outcomes and imaging, if available. Response at each treated site was graded as a CR, PR, stable disease (SD), or progressive disease (PD) according to the Revised International Response Criteria definition (8). Curative patients who responded (ie CR or PR) to salvage radiation therapy were then followed for LC. Local control was defined as maintenance of a local CR or the lack of progression of local disease in those with a PR. In general, patients were followed every 2 to 3 months for the first 1 to 2 years, then twice a year to 5 years, and annually thereafter. Death was a high competing risk among palliative patients, and 24% of these patients did not have imaging follow-up. Therefore, within this subgroup the analysis was limited to response rate and survival. If applicable, a patient’s date and cause of death was extracted from his or her medical record or the National Death Index.
Salvage RT for relapsed/refractory NHL
225
without LR were censored at date of last disease assessment. Progression-free survival was calculated from date of radiation therapy completion to date of any relapse or progression or death, whichever occurred first; patients were censored at the last date of disease assessment if they were alive without recurrence or if they died without documented recurrence but death occurred beyond the typical follow-up interval after the date of last disease assessment. Eleven patients who were treated to 2 separate sites were excluded from this analysis because PFS was analyzed by patient (vs site). Overall survival was defined as the time from radiation therapy completion to the date of death. Response to radiation therapy and TTLR were analyzed by site, whereas OS and PFS were analyzed by patient. Separate analyses were performed for sites that were treated with curative versus palliative intent, given the potentially confounding effects of the patient’s disease status and treatment. To assess the effect of tumor and treatment characteristics on response to radiation therapy, a univariate generalized estimating equations estimation model (9) was used with an unstructured correlation matrix to account for the correlation among disease sites within a patient. A univariate Cox proportional hazards regression model was used to evaluate the associations of tumor and treatment characteristics with TTLR and PFS. The following tumor and treatment characteristics were evaluated: refractory versus relapsed disease, transformed versus de novo disease, radiation therapy dose (<39.6 Gy vs 39.6 Gy), frequency of treatment (daily vs twice daily), and response to initial chemotherapy (responder vs nonresponder). Because administration of systemic therapy after salvage radiation therapy could alter the risk of LR, a timedependent covariate for systemic treatment was included in the Cox regression model. The marginal approach developed by Lee et al (10) was also used to account for the dependence among treated sites within a patient in the time-to-event models. Local control, PFS, and OS rates at 2 and 5 years were estimated using the Kaplan-Meier method. All analyses were conducted using SAS (version 9.3; SAS Institute, Cary, NC). P values of <.05 were considered significant. All tests were two-sided.
Results
Statistical analysis
Patients treated with curative intent
A total of 110 patients were treated to 121 sites during their first course of salvage radiation therapy and formed the cohort for this analysis. The primary endpoint was response to salvage radiation therapy, defined as either a CR or PR; secondary endpoints included TTLR and progression-free survival (PFS) among definitive patients and overall survival (OS). Time to local recurrence was defined as the time from radiation therapy completion to the date of local recurrence (LR) for sites that achieved a CR or PR. Sites
Seventy-two patients were treated with curative intent to 76 sites. The median age at diagnosis was 54 years, with 61% of patients being male, 78% with DLBCL, and 63% with refractory disease (Table 1). Most patients (63%) received a median of 6 cycles of rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP) initially (Table 2), with 56% achieving a response (ie responder to initial chemotherapy). Most patients (72%) at one point in their treatment course
226
International Journal of Radiation Oncology Biology Physics
Tseng et al.
Table 1 Patient demographic and disease characteristics among patients treated with curative (nZ72, 76 sites) and palliative intent (nZ38, 45 sites) Treatment intent Curative Characteristic Age at diagnosis (y)* (median and range) Refractory/relapsed Refractory Relapsed Gender Female Male Histology Diffuse large B-cell lymphoma Grade 3 follicular lymphoma T cell lymphoma Othery Transformed disease No Yes Time from diagnosis to transformation (mo) (median and range) Pathologically confirmed among transformed patients Histology before transformation G1-2 follicular Chronic lymphocytic leukemia Other
n
Palliative
Total
%
n
%
n
%
54
19-88
62.5
32-86
55
19-88
45 27
63 38
30 8
79 21
75 35
68 32
28 44
39 61
16 22
42 58
44 66
40 60
56 5 3 8
78 7 4 11
29 2 2 5
76 5 6 13
85 7 5 13
77 6 5 12
58 42 0-235.6
78 32 33.9
71 29 0-235.6
56 16 14.0
78 22 0-151.4
22 16 47.8
14
88
12
75
26
81
11 3 2
69 19 13
8 6 2
50 38 13
19 9 4
59 28 13
* Age at diagnosis for de novo disease patients and age at transformation for patients with transformed disease. y Including clinical suspicion of high grade non-Hodgkin lymphoma (NHL) (3), follicular versus large B-cell lymphoma (1), high-grade NHL not otherwise specified (1), large-cell lymphoma (2), mantle cell blastic variant (1), mixed follicular and large B-cell lymphoma (1), primary effusion lymphoma (1), T-cell rich B-cell NHL (2), and undifferentiated high grade B-cell NHL (1).
underwent high-dose chemotherapy with stem cell transplant (SCT; Table 2); 42 of these patients underwent SCT either immediately before or after salvage radiation therapy. Eighteen received salvage radiation therapy after SCT for consolidation (nZ8), cytoreduction (nZ6), or unknown disease status (nZ4), and 20 received salvage radiation therapy before SCT for residual disease (ie cytoreduction). After salvage chemotherapy, 5 patients received salvage radiation therapy as consolidation in lieu of SCT. Median dose to the 76 treated sites was 40 Gy, with 20% of sites being treated twice daily. The most commonly used twicedaily regimen was 1.4 Gy per fraction for 24 to 26 fractions.
Patients treated with palliative intent The majority of palliative patients (20 of 38, 53%) were symptomatic at the time of salvage radiation therapy, with pain (61%), obstructive symptoms (11%), cord compression (11%), bleeding (5%), brachial plexopathy (5%), and eye irritation (5%). Similar to patients treated with curative intent, the majority of these patients were male (58%), had DLBCL (76%), and refractory disease (79%). However, patients treated with palliative intent were older at
diagnosis (median, 62.5 years; Table 1) and were treated to a lower median salvage radiation therapy dose (36 Gy; Table 3).
Salvage radiation therapy response, LC, and PFS among patients treated with curative intent The majority (86%) of patients achieved a CR or PR to salvage radiation therapy (Table 4), and all patients with relapsed disease responded. With a median-follow-up of 4.8 years (range, 0.2-11.3 years) among living patients, the 2-year and 5-year LC rates were 72% and 66%. The majority of LRs occurred within the first 2 years after radiation therapy treatment (Fig. 1). Presence of de novo (vs transformed disease) was associated with a significantly higher odds (odds ratio [OR] 5.2; 95% confidence interval [CI] 1.3-20.5; PZ.02) of achieving a response to salvage radiation therapy. In contrast, presence of refractory disease (hazard ratio [HR] 3.3; PZ.024) and nonresponse to initial chemotherapy (HR 4.3; PZ.007) were associated with significantly shorter TTLR (Table 5). Among patients whose disease was either refractory or did not respond to initial chemotherapy and received 39.6 Gy (nZ27), the 2-year
Volume 91 Number 1 2015
227
Salvage RT for relapsed/refractory NHL
LC was only 61%. In the subgroup of patients treated definitively with doses of >50 Gy or with twice-daily treatments, 2-year LC was 67%. Progression-free survival at 2 and 5 years was 46% and 34%, respectively. There were no significant predictors for improved PFS (Table 5).
therapy dose was also paradoxically associated with a higher response rate (OR 6.2; 95% CI 1.1-36.1; PZ.044), which may be related to selection bias (discussed below).
Salvage radiation therapy response among patients treated with palliative intent
The median survival among patients who received curative and palliative radiation therapy was 4.7 and 0.4 years, respectively. Among the 74 patients (42 curative, 32 palliative) who have died, the majority (31 curative, 29 palliative) succumbed to their aggressive NHL. The 2- and 5-year OS rates among patients treated with curative intent were 60% and 43%, respectively (Fig. 1); rates were 19% and 8% among palliatively treated patients.
Despite the poor prognosis of palliative patients, 84% achieved a response (clinical or imaging) to salvage radiation therapy (Table 4). Among the 20 patients who were symptomatic at the time of salvage radiation therapy, 16 (80%) had symptomatic improvement during or immediately after treatment. In 3 patients, the clinical response was unknown, and there was no response in 1 patient. Response to initial chemotherapy was associated with a significantly higher response rate (OR 10.3; 95% CI 1.1-95.2; PZ.04). A lower radiation
Overall survival
Treatment toxicity Salvage radiation therapy was generally well tolerated, with 8% and 5% of curative and palliative patients experiencing grade 3
Table 2 Initial chemotherapy treatments received before salvage radiation therapy and rates and timing of high-dose chemotherapy with stem cell transplantation Treatment intent Curative
Palliative
Total
Parameter
n
%
n
%
n
%
No. of chemotherapy regimens before first salvage RT (median and range) Initial chemotherapy Regimen CHOP R-CHOP Other* No. of cyclesy (median and range) Response after initial chemotherapy CR PR SD PD Unknown Rates of any SCT (nZ67) SCT type Autologous SCT Allogeneic SCT >1 SCT SCT immediately before or after salvage RT (nZ42) SCT / consolidative RT SCT / cytoreductive RT SCT / RT (disease state unknown) Cytoreductive RT / SCT
2
1-4
2
1-7
2
1-7
20 45 7 6
28 63 10 1-11
11 23 4 6
29 61 11 1-8
31 68 11 6
28 62 10 1-11
23 16 14 17 2
33 23 20 24
17 4 4 12 1
46 11 11 32
40 20 18 29 3
37 19 17 27
46 6 6
64 8 8
8 7 1
21 18 3
54 13 7
49 12 6
8 6 4 20
11 8 6 28
0 1 0 3
0 3 0 8
8 7 4 23
7 6 4 21
Abbreviations: CR Z complete response; PD Z progressive disease; PR Z partial response; RT Z radiation therapy; SD Z stable disease; SCT Z stem cell transplantation; CHOP Z cyclophosphamide, doxorubicin, vincristine, prednisone; R-CHOP Z rituximab, cyclophosphamide, doxorubicin, vincristine, prednisone. * Including CVP (cyclophosphamide, vincristine, prednisone), chlorambucil, HAM (high-dose cytosine arabinoside, mitoxantrone), ProMACECytaBOM (prednisone, doxorubicincyclophosphamide, etoposide, cytarabine, bleomycin, vincristine, methotrexate, leucovorin), R-CVP (rituximab, cyclophosphamide, vincristine, prednisone), REPOCH (rituximab, etoposide, prednisone, vincristine, cyclophosphamide, doxorubicin), RESHAP (rituximab, etoposide, methylprednisolone, cytarabine, platinum agent), RICE (rituximab, ifosfamide, carboplatin, etoposide), and rituximab. y For 2 patients, the number of initial chemotherapy cycles was unknown.
228
International Journal of Radiation Oncology Biology Physics
Tseng et al.
Table 3 Details of salvage radiation therapy and highest acute toxicity by patient (nZ110). Location and salvage radiation therapy dose by site (nZ121) Curative Parameter No. of sites treated during first course 1 2 Total no. of RT courses (median and range) Time from diagnosis of de novo disease or transformation to salvage RT (mo) (median and range) Acute toxicity during RT course Grade 0 Grade 1 Grade 2 Grade 3 Unknown Location Head and neck Axilla Mediastinum or thorax Extranodal, GI tract Abdominal/pelvic lymph nodes Bone Paraspinal Soft tissue (extremity) Other* Unknown Radiation dose (Gy) (median and range) BED10 (Gy10) (median and range) 20 20.01-30 30.01-40 40.01-50 50.01-60 >60 Fractionation frequency Once daily Twice dailyy
Palliative
Total
n
%
n
%
n
%
68 4 1 10.6
94 6 1-7 0.6-219.6
31 7 1 10.2
82 18 1-6 0.2-93.6
99 11 1 10.5
90 10 1-7 0.2-219.6
17 42 6 6 1
24 59 8 8
15 18 3 2 0
39 47 8 5
32 60 9 8 1
29 55 8 7
12 4 17 3 31 4 0 2 2 1 40.0 47.2 1 0 13 38 22 2
16 5 23 4 41 5 0 3 3
14 2 1 3 12 4 4 4 1 0 36.0 43.2 0 2 10 27 5 1
31 4 2 7 27 9 9 9 2
26 6 18 6 43 8 4 6 3 1 37.8 43.8 1 2 23 65 27 3
22 5 15 5 36 7 3 5 3
61 15
16.5-55.7 19.0-64.8 1 0 17 50 29 3 80 20
35 10
13.0-54.0 21.5-64.8 0 4 22 60 11 2 78 22
96 25
13.0-55.7 19.0-64.8 1 2 19 54 22 2 79 21
Abbreviations: BED10 Z biological effective dose; GI Z gastrointestinal. Other abbreviations as in Table 2. * Including head and neck with mediastinum (1), soft tissue and spleen (1), and mediastinum and bone (1). y Includes patients who were initially started with once daily but received twice daily treatment for the majority of their radiation therapy course.
acute toxicity, respectively; no grade 4 acute toxicity was noted. Five definitively treated patients died from myelodysplastic syndrome (nZ3) or acute myeloid leukemia (nZ2).
Discussion In this relatively large, single-institution experience of salvage radiation therapy for refractory or relapsed aggressive NHL, response to radiation therapy was high, but durability of response varied among patients treated with curative intent for refractory versus relapsed disease (2-year LC 56% vs 92%). There was no evidence of a doseeresponse relationship with LC detected, and intensification of radiation therapy (>50 Gy or twice-daily treatment) was associated with an LC rate of only 67% at
2 years. These results suggest that the biology underlying relapsed and refractory aggressive NHL differs, at least with respect to durability of response to salvage radiation therapy, and that there remain subgroups of patients who may benefit from treatment intensification. Prior single-institution studies on salvage radiation therapy for relapsed or refractory lymphoma reported LC rates ranging from 47% to 73% at 2 years (3-5). In a retrospective, single-institution experience from Wayne State University, 35 patients (69% with refractory disease) were treated with salvage radiation therapy, which yielded a 2-year local failure rate of 54%. Furthermore, none of the 5 patients who progressed through chemotherapy achieved LC with salvage radiation therapy. However, there was a trend toward improved LC (66% vs 39% at 2 years; PZ.06) with radiation therapy doses of 39.6 Gy.
Volume 91 Number 1 2015 Table 4
Salvage RT for relapsed/refractory NHL
229
Response to salvage radiation therapy in disease sites treated with curative versus palliative intent Treatment intent Curative
Parameter
n
No. w/response (%)
Overall 76 Disease type Relapsed 27 Refractory 49 Transformed disease No 58 Yes 18 Dose (Gy) High (39.6) 42 Low (<39.6) 34 Frequency Twice daily 15 Once daily 61 Response to initial chemotherapy Response 39 Nonresponse 37 Number of prior chemotherapy 1-2 57 >2 19
Palliative 95% CI
n
No. w/response (%)
OR
95% CI
65 (86)
d
OR
d
45
38 (84)
d
d
27 (100) 38 (78)
d 1.0
d
8 37
8 (100) 30 (81)
d 1.0
d
53 (91) 12 (67)
5.2* 1.0
1.3-20.5*
27 18
23 (85) 15 (83)
1.2 1.0
0.2-6.0
35 (83) 30 (88)
1.0 1.5
0.4-5.0
16 29
11 (69) 27 (93)
1.0 6.2*
12 (80) 53 (87)
1.0 1.7
0.4-6.3
10 35
10 (100) 28 (80)
d 1.0
d
36 (92) 29 (78)
2.7 1.0
0.6-11.3
25 20
24 (96) 14 (70)
10.3* 1.0
1.1-95.2*
49 (86) 16 (84)
1.0 1.6
27 18
23 (85) 15 (83)
1.0 0.9
0.3-8.9
1.1-36.1*
0.2-4.4
Abbreviations: CI Z confidence interval; OR Z odds ratio. * Significant predictor for response to salvage radiation therapy.
In studies in which salvage radiation therapy was intensified, LC seemed to be higher. Among 34 patients (85% with refractory disease) who were treated at Princess Margaret Hospital with twice-daily radiation therapy (39.940.5 Gy in 30 fractions), a 2-year LC rate of 73% was
Fig. 1. Cumulative incidence of local control (LC), progression free survival (PFS), and overall survival (OS) among patients treated with curative intent. Note that the number at risk differs for each endpoint given the subgroups of the cohort examined. OS includes all patients treated with curative-intent RT; PFS only includes patients treated to a single site; and LC includes those patients who achieved a complete or partial response to radiation therapy.
achieved, with the majority of local and distant recurrences occurring within the first year after radiation treatment (5). A phase 2 study from the Institut Gustave Roussy reported an LC rate of 63% among 21 patients with bulky relapsed or refractory lymphoma (71% with refractory disease) treated with split course radiation therapy to a median dose of 40 Gy concurrent with chemotherapy (76% cisplatin with etoposide). However, in this study 70% of patients experienced grade 3 to 4 hematologic toxicity (4). In the present study, the 2-year LC rate of 72% among patients treated with curative intent compared favorably with patients from prior studies that used intensified treatment, but this may reflect a more favorable prognostic case mix, modern radiation therapy techniques, more effective systemic therapy, and inclusion of patients who received salvage radiation therapy as part of consolidation. This single-institution study is limited by its retrospective study design, and treatment received was subject to selection bias and likely correlated with aggressiveness of disease. As such, we were unable to detect a doseeresponse relationship for LC. Indeed, LC was higher among patients treated with low (<39.6 Gy) versus high (39.6 Gy) radiation therapy doses and daily versus twice-daily treatment, regardless of treatment intent. We also obtained similar results regardless of whether dose was dichotomized by either the median BED10 or median prescribed dose (data not shown). However, our study did identify subgroups of patients with significantly inferior LC with salvage radiation therapy, namely patients who failed to respond to initial chemotherapy and patients with
230
International Journal of Radiation Oncology Biology Physics
Tseng et al.
Table 5
Time to local recurrence in disease sites treated with curative versus palliative intent Treatment intent Curative
Parameter
N
Overall 65 Disease type Relapsed 27 Refractory 38 Transformed disease No 53 Yes 12 Dose (Gy) High (39.6) 35 Low (<39.6) 30 Frequency Twice daily 12 Once daily 53 Response to initial chemotherapy Response 36 Nonresponse 29
2-y LC (%)
Palliative P
n
2-y LC (%)
HR
P
72
d
HR
d
38
43
d
d
92 56
1.0 3.3*
.024
8 30
80 35
1.0 3.1
.22
69 88
4.1 1.0
.16
23 15
33 82
1.3 1.0
.75
67 79
1.1 1.0
.93
11 27
0 56
2.5 1.0
.23
60 73
1.2 1.0
.79
10 28
dy 34
dz
88 51
1.0 4.3*
.007
24 14
28 56
1.0 1.1
d
.90
Abbreviation: LC Z local control. Other abbreviations as in Table 4. * Significant predictor for shorter time to local recurrence. y All palliative sites treated twice daily were censored before 2 years. z Hazard ratio cannot be estimated because no local recurrences occurred among palliative sites treated twice daily.
chemotherapy-refractory disease at the time of radiation therapy. Despite radiation dose escalation in these patients, approximately one-third relapsed locally after salvage radiation therapy. The biology driving these observations is unknown, but they suggest that chemotherapy-resistant NHL may have lower intrinsic radiosensitivity, a high rate of proliferation, or both. Double hit pathology with 2 translocations involving c-MYC and BCL2 or BCL6 is associated with biologically aggressive lymphomas, but it is unknown what proportion of this cohort had these translocations because this was not routinely obtained. Additional studies are required to explore whether the double hit pathology is enriched among refractory or relapsed NHL patients and to determine whether novel approaches such as hyperfractionation or addition of radiosensitizers (11, 12) may benefit chemotherapy-resistant patients. Finally, this cohort represents a heterogeneous group of patients who received radiation therapy alone, in conjunction with SCT, or after salvage chemotherapy. The relatively high PFS and OS observed among curative patients likely reflect the use of tandem systemic therapy. Death was a significant competing risk, in particular among palliative patients. These patients were deemed to have poor prognosis and/or be ineligible for transplantation at the time of consultation. Overall survival at 2 years was 19%, and almost one-third of these palliative patients were unable to be assessed for relapse before death. Despite the poor prognosis, however, the majority of symptomatic, palliative patients had a clinical response to salvage radiation therapy, and the response rate among all palliative patients (with or without symptoms) was >80%. Therefore
salvage radiation therapy should be considered in symptomatic refractory or relapsed NHL patients, with the hope to improve quality of life. Similarly, it is remarkable that even among patients who have chemotherapy-refractory or relapsed NHL and who are treated with curative intent, the overall response rate to salvage radiation therapy was 86%. Given these findings, there may be a role to introduce salvage radiation therapy earlier in the course of a patient’s relapsed or refractory disease. Future studies are needed to examine whether, instead of using multiple chemotherapy regimens first, early intervention with salvage radiation therapy may be associated with improved LC and/or allow potentially a lower dose of radiation therapy to be used. To our knowledge, our study is among the first to report on the efficacy of salvage radiation therapy among patients whose aggressive NHL transformed from low-grade disease. Although we found that patients with transformed disease had a lower initial response rate to radiation therapy, those who achieved a response had significantly more durable LC than patients with de novo disease. These results suggest that patients with transformed disease fare as well as, if not better than, patients with de novo disease, in terms of durability of LC with salvage radiation therapy. In summary, our study showed that overall, relapsed or refractory aggressive NHL is responsive to salvage radiation therapy, and durable LC can be achieved in some cases. However, patients with refractory or nonresponse to initial chemotherapy had less durable LC, despite relatively high doses of radiation. The optimal salvage radiation therapy dose and fractionation scheme in addition to the
Volume 91 Number 1 2015
role for concurrent radiosensitizing agents remain to be determined and can be considered as part of future, prospective studies.
References 1. Miller TP, Dahlberg S, Cassady JR, et al. Chemotherapy alone compared with chemotherapy plus radiotherapy for localized intermediate- and high-grade non-Hodgkin’s lymphoma. N Engl J Med 1998;339:21-26. 2. Philip T, Guglielmi C, Hagenbeek A, et al. Autologous bone marrow transplantation as compared with salvage chemotherapy in relapses of chemotherapy-sensitive non-Hodgkin’s lymphoma. N Engl J Med 1995;333:1540-1545. 3. Aref A, Narayan S, Tekyi-Mensah S, et al. Value of radiation therapy in the management of chemoresistant intermediate grade non-Hodgkin’s lymphoma. Radiat Oncol Invest 1999;7:186-191. 4. Girinsky T, Lapusan S, Ribrag V, et al. Phase II study of concomitant chemoradiotherapy in bulky refractory or chemoresistant relapsed lymphomas. Int J Radiat Oncol Biol Phys 2005; 61:476-479.
Salvage RT for relapsed/refractory NHL
231
5. Martens C, Hodgson DC, Wells WA, et al. Outcome of hyperfractionated radiotherapy in chemotherapy-resistant non-Hodgkin’s lymphoma. Int J Radiat Oncol Biol Phys 2006;64:1183-1187. 6. Lowry L, Smith P, Qian W, et al. Reduced dose radiotherapy for local control in non-Hodgkin lymphoma: A randomised phase III trial. Radiother Oncol 2011;100:86-92. 7. Friedberg JW. Relapsed/refractory diffuse large B-cell lymphoma. Hematology Am Soc Hematol Educ Program 2011;2011:498-505. 8. Cheson BD, Pfistner B, Juweid ME, et al. Revised response criteria for malignant lymphoma. J Clin Oncol 2007;25:579-586. 9. Zeger SL, Liang KY. Longitudinal data analysis for discrete and continuous outcomes. Biometrics 1986;42:121-130. 10. Lee E, Wei LJ, Amato D, et al. Cox-type regression analysis for large numbers of small groups of correlated failure time observations. In: Klein J, Goel P, editors. Survival Analysis: State of the Art. Dordrecht, The Netherlands: Springer Netherlands; 1992:237-247. 11. Kapadia NS, Engles JM, Wahl RL. In vitro evaluation of radioprotective and radiosensitizing effects of rituximab. J Nucl Med 2008;49:674-678. 12. Elstrom RL, Leonard JP, Christos P, et al. Phase 1 study of radiosensitization using bortezomib in patients with relapsed non-Hodgkin’s lymphoma receiving radioimmunotherapy. Presented at the American Society of Hematology 54th Annual Meeting. December 811, 2012; Atlanta, GA.