Low-Dose Involved-Field Radiation in the Treatment of Non-Hodgkin Lymphoma: Predictors of Response and Treatment Failure

International Journal of

Radiation Oncology biology

physics

www.redjournal.org

Clinical Investigation: Lymphoma

Low-Dose Involved-Field Radiation in the Treatment of Non-Hodgkin Lymphoma: Predictors of Response and Treatment Failure Andrea L. Russo, MD,* Yu-Hui Chen, MS, MPH,y Neil E. Martin, MD, MPH,z Anant Vinjamoori, BA,z Sarah K. Luthy, MD,z Arnold Freedman, MD,x Evan M. Michaelson, BA,z Barbara Silver, BA,z Peter M. Mauch, MD,z and Andrea K. Ng, MD, MPHz *Harvard Radiation Oncology Program, Boston, Massachusetts; yBiostatistics Core, Dana Farber Cancer Institute, Boston, Massachusetts; and Departments of zRadiation Oncology and xHematologic Oncology, Brigham and Women’s Hospital and Dana-Farber Cancer Institute, Boston, Massachusetts Received Oct 26, 2012, and in revised form Dec 6, 2012. Accepted for publication Dec 31, 2012

Summary Patients with indolent nonHodgkin lymphoma (NHL) often have frequent relapses. This study assessed the outcomes of treating NHL with low-dose involved-field radiation therapy of 4 Gy in 2 fractions. The overall response rate was 82%. Those with chronic lymphocytic leukemia (CLL) had an inferior response. Age 50 years at diagnosis, CLL, and mantle cell lymphoma predicted a shorter time to further therapy. Low-dose involvedfield radiation therapy should be a primary treatment in the palliation of NHL.

Purpose: To investigate clinical and pathologic factors significant in predicting local response and time to further treatment after low-dose involved-field radiation therapy (LD-IFRT) for non-Hodgkin lymphoma (NHL). Methods and Materials: Records of NHL patients treated at a single institution between April 2004 and September 2011 were retrospectively reviewed. Low-dose involved-field radiation therapy was given as 4 Gy in 2 fractions over 2 consecutive days. Treatment response and disease control were determined by radiographic studies and/or physical examination. A generalized estimating equation model was used to assess the effect of tumor and patient characteristics on disease response. A Cox proportional hazards regression model was used to assess time to further treatment. Results: We treated a total of 187 sites in 127 patients with LD-IFRT. Histologies included 66% follicular, 9% chronic lymphocytic leukemia (CLL)/small lymphocytic lymphoma, 10% marginal zone, 6% mantle cell lymphoma (MCL), and 8% other. Median follow-up time was 23.4 months (range, 0.0392.2 months). The complete response, partial response, and overall response rates were 57%, 25%, and 82%, respectively. A CLL histology was associated with a lower response rate (odds ratio 0.2, 95% confidence interval 0.1-0.5, PZ.02). Tumor size, site, age at diagnosis, and prior systemic therapy were not associated with response. The median time to first recurrence was 13.6 months. Those with CLL and age 50 years at diagnosis had a shorter time to further treatment for local failures (hazard ratio [HR] 3.63, PZ.01 and HR 5.50, PZ.02, respectively). Those with CLL and MCL had a shorter time to further treatment for distant failures (HR 11.1 and 16.3, respectively, P<.0001). Conclusions: High local response rates were achieved with LD-IFRTacross most histologies. Chronic lymphocytic leukemia and MCL histologies and age 50 years at diagnosis had a shorter time to further treatment after LD-IFRT. Ó 2013 Elsevier Inc.

Reprint requests to: Andrea L. Russo, MD, Brigham and Women’s Hospital, Department of Radiation Oncology, 75 Francis St, ASB1-L2, Boston, MA 02115. Tel: (617) 732-6310; E-mail: [email protected] Int J Radiation Oncol Biol Phys, Vol. 86, No. 1, pp. 121e127, 2013 0360-3016/$ - see front matter Ó 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ijrobp.2012.12.024

Presented in part at the 54th Annual Meeting of the American Society for Radiation Oncology, October 28-31, 2012, Boston, MA. Conflict of interest: none.

122

Russo et al.

Introduction There were approximately 66,360 new cases of non-Hodgkin lymphoma (NHL) in the United States in 2011 and 19,320 deaths (1). Indolent NHL, which includes follicular (FL), marginal zone (MZ), mantle cell (MCL), and chronic lymphocytic leukemia (CLL) lymphomas constitutes approximately 40% of all NHL (2). Although early-stage FL is potentially cured with definitive radiation therapy (RT), advanced-stage disease, which is diagnosed in more than two-thirds of patients at initial presentation (3), is presently not considered curable in the majority of patients. Patients with advanced-stage disease, however, often have a long median survival in excess of 10 years (2, 4) and tend to have frequent and late relapses, requiring treatment with various systemic agents, or occasionally are observed until symptoms develop (5). Similarly, those with nonfollicular NHLs are often treated initially with chemo-immunotherapy (6), but often will relapse after having an initial response. Given the long natural history of indolent NHLs, many individuals will require multiple types of treatment at many points during the disease course, and some may ultimately require stem cell transplantation (7). Low-dose involved-field RT (LD-IFRT) given in 2 daily fractions can offer durable local control and ideally delay the need for more potentially toxic systemic therapy. Others have previously reported on outcomes of patients with advanced indolent NHL treated with LD-IFRT and have shown response rates of 81-92% (8-14). These studies have found various inconsistent predictors of poor response, including older age at treatment, prior use of chemotherapy, and tumor size. We sought to evaluate our institution’s experience of the effectiveness of LD-IFRT in a large series of patients and to assess clinical and pathologic factors significant in predicting response to treatment.

Methods and Materials Patient population and treatment A retrospective review of medical records of patients with NHL treated between April 2004 and September 2011 was conducted. This study was approved by the Dana-Farber Cancer Institute institutional review board. Patients over the age of 18 years with advanced or recurrent NHL treated with LD-IFRT were included in the analysis. Patient, disease, and treatment characteristics including sex, age, date of diagnosis, stage at diagnosis, prior systemic or localized treatments, site of LD-IFRT, tumor size (cm) measured by largest dimension, indication for treatment, and treatment outcome including response to LD-IFRT, in-field and out-of-field progression, further treatments after LD-IFRT, time to further treatment, and status at last follow-up were reported. Tumor size >4 cm or 4 cm was used to assess response to treatment because >4 cm has previously been shown to be predictive of lack of response (8). A total of 187 sites in 127 patients were evaluated. Low-dose involved-field RT was given as 4 Gy in 2 fractions of 2 Gy per fraction over 2 consecutive days. The radiation treatment field was IFRT as per prior published guidelines, with inclusion of the involved nodal region in the treatment field (15). No other therapy was given concurrently with LD-IFRT. Before LD-IFRT

International Journal of Radiation Oncology  Biology  Physics patients were evaluated with physical examination and computed tomography scanning. Treatment response and disease control were determined by computed tomography scan and/or physical examination. Treatment response was analyzed by treatment site, not by individual patient, because several patients had multiple different sites requiring treatment at various time points. Complete response (CR) was defined as complete clinical or radiologic disappearance of disease. Complete response undocumented (Cru) was defined as >75% reduction in tumor size with residual mass. Partial response (PR) was defined as >50% reduction in tumor size with residual mass. Stable disease (SD) was defined as no change in tumor size. Progressive disease (PD) was defined as a 20% increase in tumor size. Overall response includes CR, Cru, and PR.

Statistical methods Of the 127 patients included in this analysis, 38 patients had multiple recurrence sites. The analysis unit used was recurrence site for both disease response and time to further treatment for failures. To assess the effect of tumor and patient characteristics on disease response, a generalized estimating equations model (16) with the exchangeable association structure was used to account for the correlation among disease sites within each patient. Times to further treatment for local failure and distant failure were defined as the time from the date of LD-IFRT to the date of further treatment administered for local failure and distant failure, respectively. Because the risk of failures changed owing to the administration of systemic treatment during the follow-up period, a Cox proportional hazards regression model with systemic treatment as a time-dependent covariate was used to evaluate the associations between patient/tumor characteristics and time to further treatment for failures. For this analysis, the marginal approach developed by Lee et al (17) was used to account for the dependence among disease sites within each patient. The patient/tumor characteristics of interest include tumor size (<4 cm vs 4 cm), site of RT (head and neck, supradiaphragmatic, infradiaphragmatic, pelvic, or cutaneous), histology (CLL, follicular, mucosa-associated lymphoid tissue/ MZ, MCL, or other), age at diagnosis (50 vs 51-70 vs >70 years), and prior systemic therapy. Kaplan-Meier curve estimates were generated to demonstrate time to further treatment for local and distant failures at 2 years.

Results Patient characteristics Patient and disease characteristics are listed in Table 1; treatment site characteristics are listed in Table 2. The median age at diagnosis was 54 years (range, 25-95 years). Fifty-four percent were male. Seventy percent of patients had 1 disease site, 19% had 2 sites, 7% had 3 sites, 3% had 4 sites, and 1% had 7 sites treated. The median follow-up time was 23.4 months (range, 0.03-92.2 months). The median follow-up time among survivors was 25.4 months. The median interval from diagnosis to the start of LDIFRT was 57.3 months (range, 0.3-284.5 months). Before receiving LD-IFRT, 52% of patients received systemic therapy, 27% received conventional-dose RT, and 7% underwent stem-cell transplant; 28% of patients received no prior treatment. The

Volume 86  Number 1  2013 Table 1

LD-IFRT for NHL Table 2

Patient and disease characteristics

All Age at diagnosis (y) Median Range 50 51-70 >70 Gender Female Male Stage at initial diagnosis IA IIA III IVA IVB Unknown Histology CLL Follicular MALT/MZ MCL Other Prior treatment Conventional RT Systemic therapy BMT/SCT None Interval from diagnosis to LD-IFRT (mo) Median Range

127 54 25-95 51 (40) 59 (46) 17 (13) 58 (46) 69 (54) 20 12 38 50 3

(16) (10) (31) (41) (2) 4

12 84 13 8 10

(9) (66) (10) (6) (8)

43 74 9 34

(34) (58) (7) (27)

57.3 0.3-284.5

Abbreviations: BMT Z bone marrow transplant; CLL Z chronic lymphocytic leukemia; LD-IFRT Z low-dose involved-field radiation therapy; MALT Z mucosa-associated lymphoid tissue; MCL Z mantle cell lymphoma; MZ Z marginal zone; RT Z radiation therapy; SCT Z stem cell transplant. Values are number (percentage) unless otherwise noted.

indications for treatment by site were cosmesis in 23%, pain in 29%, and bowel obstruction in 5%. For the remaining patients without symptoms, the main indication for LD-IFRT was significant local disease progression. In these cases, we did not encompass the entire disease burden but limited the RT field to the sites of local progression.

Response to LD-IFRT Patients were typically evaluated 8 to 12 weeks after treatment for response assessment. In this study patients were evaluated at a mean of 81 days after treatment. The absence or presence of a response, and if response, whether it was a CR or PR, were ascertained by retrospective review of clinical notes describing physical findings or radiographic reports. We found an overall response rate of 82%. The CR and PR rates were 57% and 25%, respectively. Six percent had SD, and 7% had either PD or were lost to follow-up. The response rates by patient characteristics and disease factors are listed in Table 3. The response rates were 82% for all factors considered, with the exception of supradiaphragmatic sites (75%), infradiaphragmatic sites (68%), CLL

123

Treatment-site characteristics

All Tumor size* (cm) Median Range <4 cm 4 cm Site of RT H&N Supradiaphragmatic (non-H&N) Infradiaphragmatic (nonpelvic) Pelvic Cutaneous Additional indications for treatment Pain Cosmesis Bowel obstruction

187 4 0.6-35 88 (50) 89 (50) 67 28 22 35 35

(36) (15) (12) (19) (19)

55 (29) 43 (23) 10 (5)

Abbreviation: H&N Z head and neck. Other abbreviation as in Table 1. Values are number of sites (percentage) unless otherwise noted. * Ten sites with tumor size missing.

histology (52%), and age at diagnosis 50 years (77%). The median time to local failure leading to reirradiation or initiation of systemic therapy was not reached. The median time to distant Table 3 factors

Response by patient characteristics and disease

Characteristic/factor

n*

No. with response (%)

Overall Tumor sizey (mm) <40 40 Site of RT H&N Supradiaphragmatic (non-H&N) Infradiaphragmatic (non-pelvic) Pelvic Cutaneous Histology CLL Follicular MALT/MZ MCL Otherz Age at diagnosis (y) 50 51-70 >70 Prior systemic therapy No Yes

187

154 (82)

88 89

72 (82) 74 (83)

67 28 22 35 35

59 21 15 30 29

(88) (75) (68) (86) (83)

21 125 18 10 13

11 106 16 9 12

(52) (85) (89) (90) (92)

75 83 29

58 (77) 68 (82) 28 (97)

80 107

66 (83) 88 (82)

Abbreviations as in Tables 1 and 2. * Number of sites. y Ten sites with tumor size unknown. z Other includes DLCL diffuse large cell lymphoma (nZ4), lymphoplasmacytic lymphoma (nZ2), cutaneous T-cell lymphoma (nZ4), nodular point differentiated lymphocytic lymphoma (nZ1), mixed small and large cell (nZ1), and B-cell unclassified (nZ1).

124

International Journal of Radiation Oncology  Biology  Physics

Russo et al.

Table 4 model

Factors associated with response using the GEE

Factor Overall Tumor sizek (mm) <40 40 Site of RT H&N Supradiaphragmatic (non-H&N) Infradiaphragmatic (nonpelvic) Pelvic Cutaneous Histology CLL Follicular MALT/MZ MCL Other Histology CLL Other Age at diagnosis (y) 50 51-70 >70 Prior systemic therapy No Yes

Odds n* ratioy 95% CIz Px 187

.77

88 89

1.0 1.1

0.5-2.4

67 28 22 35 35

3.1 1.3 1.0 2.4 2.8

1.2-8.1 0.5-3.8 0.8-7.6 0.6-12.6

.28

.19 21 125 18 10 13

1.0 4.9 6.9 7.3 9.4

1.8-13.4 1.3-36.7 0.8-69.8 1.0-87.9

21 166

0.2 1.0

0.1-0.5 -

75 83 29

1.0 1.3 6.1

0.5-3.0 0.7-50.7

.02

.07

.73 80 107

1.0 0.9

0.4-2.0

Abbreviations: GEE Z generalized estimating equations; CI Z confidence interval. Other abbreviations as in Tables 1 and 2. * Number of sites. y Ratio of the odds of response in a category versus the reference category; odds ratio of 1 is the reference category; odds ratios reported in this table have accounted for the correlation among disease sites within patient. z 95% confidence interval of odds ratio. x P value adjusted for the correlation among disease sites within patient. k Ten sites with tumor size unknown.

failure requiring irradiation or systemic therapy was 34.9 months.

Factors associated with response Table 4 demonstrates tumor response by patient characteristics and disease factors. A CLL histology compared with other histologies was the only factor significantly associated with response (odds ratio 0.2, 95% confidence interval 0.1-0.5, PZ.02). Tumor size, site, age at diagnosis, and the use of prior systemic therapy were not associated with response.

Disease progression Time to further treatment for local failure (TTFT-L) was analyzed by patient characteristics (Table 5). The 2-year freedom from further treatment for local failures was 74% (Fig. 1A). A CLL histology had a significantly shorter TTFT-L compared with other

Table 5 Time to further treatment for local (TTFT-L) and distant (TTFT-D) failures by patient characteristics TTFT-L Characteristic

N

Overall Tumor size (mm) <40 40 Unknowny Site of RT H&N Supradiaphragmatic (non-H&N) Infradiaphragmatic (nonpelvic) Pelvic Cutaneous Site of RT Cutaneous Other Histology CLL Follicular MALT/MZ MCL Other Histology CLL Other Age at diagnosis (y) 50 51-70 >70 Prior systemic therapy No Yes

187 88 89 10

HR*

TTFT-D

P*

HR

P

.08

-

.06

1.00 1.89 -

1.00 1.64 .07

67 28

1.28 2.52

1.86 1.09

22

3.66

1.12

35 35

2.07 1.00

2.26 1.00

35 152

1.00 2.04

21 125 18 10 13

10.72 3.33 2.82 1.78 1.00

21 166

3.63 1.00

75 83 29

5.50 2.83 1.00

80 107

1.00 1.69

.16

.16

.14 1.00 1.70 <.0001

.11 11.05 3.33 1.00 16.26 4.00

<.0001

.01 4.19 1.00 .02

.47 1.00 1.40 1.27

.16

.09 1.00 1.53

Abbreviation: HR Z hazard ratio. Other abbreviations as in Tables 1 and 2. * 25% HRs and P values after adjusting for the correlation among disease sites within patient as well as systemic treatment administration using a Cox proportional hazards regression model. y Unknown categories were not included in the analysis.

histologies (hazard ratio [HR] 3.63, PZ.01). Age at diagnosis was also associated with a shorter TTFT-L, with an HR of 5.50 for age 50 years (PZ.02). Tumor size, site of RT, and prior systemic therapy were not significant for TTFT-L. The 2-year freedom from further treatment for distant failures was 54% (Fig. 1B). Time to further treatment for distant failures (TTFT-D) analyzed by patient characteristics showed tumor size 4 cm to be of borderline significance (HR 1.64, PZ.06) (Table 5). Histologies of CLL and MCL were also associated with a shorter median TTFT-D (HR 11.1 and 16.3, respectively), when all 5 histologies were compared (P<.0001). Tumor site, age at diagnosis, and prior systemic therapy were not significant for TTFT-D.

Subsequent treatments Fifty-six patients (44%) did not receive any additional treatment up to the last follow-up. Thirty-one patients (24%) did not receive

Volume 86  Number 1  2013

LD-IFRT for NHL

125

Fig. 1. Kaplan-Meier curve estimates demonstrate time to further treatment for local (A) and distant (B) failures. The 2-year freedom from further treatment for local failures was 74% and for distant failures was 54%. any treatment for at least 1 year after LD-IFRT. For those patients who required treatment after LD-IFRT, 11% received subsequent LD-IFRT, 13% received conventional radiation, 40% received systemic therapy, and 10% underwent stem cell transplant. At last follow-up, 52% of patients were alive with disease, 27% were alive without disease, 20% were dead with disease, and 2% were dead without disease.

Discussion This study demonstrates the efficacy of LD-IFRT in terms of both treatment response and the durability of response in NHL. Lowdose involved-field RT can offer patients palliation from their disease with a well-tolerated, low-dose, short treatment course, thus delaying the need for more standard-dose radiation or systemic therapy. The majority of patients treated with LD-IFRT

(82%) had either a complete or partial response to treatment. Furthermore, all patients had advanced or recurrent disease at the time of treatment, and 57% had received prior systemic therapy or stem cell transplant, thereby demonstrating that LD-IFRT is efficacious even in patients with chemotherapy-refractory disease. This study included a range of histologies; however, 66% of patients had FL. The response rates remained high for all histologies, with the exception of CLL, which had an overall response rate of 52% and a CR of 4.8%. Patients with CLL also had a significantly shorter interval to further treatment for local and distant failures. Patients with MCL had a shorter time to further treatment for distant failures only. This highlights the radiosensitivity of MCL and its excellent response to and durable local control with LD-IFRT but also its tendency for early systemic progression. When considering other predictors of response, we did not find prior chemotherapy to be significant, as others have reported (10, 11). Table 6 summarizes prior studies of LD-IFRT

126

International Journal of Radiation Oncology  Biology  Physics

Russo et al.

Table 6

Prior studies of LD-IFRT in NHL patients

Study

n (patients/sites)

This study Rossier et al (11)

127/187 43

Chan et al (9) Girinsky et al (10)

54 48/135

Haas et al (12) Haas et al (18) Johannsson et al (13) Ng et al (19) Sawyer et al (14) Luthy et al (8)

109/304 Mostly FL 71/177 Mostly non-FL 22/31 10/14 11/16 33/43

Response rate (%) (overall/CR) 82/50 90 (included stable disease) 81/49 81/57

Predictors of poor response

92/61

CLL Age >73 y, prior chemotherapy, response to RT CLL, size >10 cm Age >65 y at time of treatment, no. of chemotherapy regimens, size >5 cm No. of sites, size >5 cm

87/48

None

82/55 90/70 94/38 95/84

NR NR NR Non-H&N, size <4 cm

Follow-up (mo)

Median time to progression (any/local; mo)

25.4 20

13.6 any 8/21

15.6 54

4/19 NR

7

25/42

9

12/22

8 6.5 NR 14

22/NR NR 7/NR 4/9

Abbreviations: CR Z complete response; NHL Z non-Hodgkin lymphoma; NR Z not reported. Other abbreviations as in Tables 1 and 2.

including predictors of poor response. We found high overall response rates (82%) in patients with large tumors (>4 cm), which has previously been shown to predict for poor response to treatment (8). Large tumor size was, however, possibly associated with a shorter time to further treatment for distant failures. We also found younger age was associated with shorter TTFT-L, which differs from what others have shown in terms of older age being associated with poorer response (10, 11). Thirty percent of patients had more than 1 site treated with LDIFRT. This is possible given the very low dose of 4 Gy and the limited treatment field size, which allows patients to be successfully palliated with multiple subsequent courses of LD-IFRT when their disease progresses. Low-dose involved-field RT is an effective approach to palliation that does not have systemic side effects and may be particularly appealing in patients whose systemic options are limited. The response rate in this study was similar to what others have found (Table 6). However, it is difficult to compare series because of the mix of histologies in all studies and the inability to control for patient selection. The 2 largest series by Haas et al showed response rates of 91% in their first study, composed primarily of patients with FL, and 87% in their subsequent study of patients with MCL, MZL, CLL, and diffuse large cell lymphoma DLCL (12, 18). It seems that an 80-90% response rate is consistent across most histologies, except perhaps CLL. Durability of response seems encouraging; the median time to first recurrence was 13.6 months, including both local and distant failures. However, 44% of patients did not require any additional treatment up to the last follow-up. This could be related to the relatively short median follow-up of 25.4 months in this study or could indicate a prolonged response. Another shortcoming of this study is its retrospective study design, and as such the findings may be limited by patient selection bias and the lack of a standardized response assessment. In addition, although we restricted the study population to patients with advanced or recurrent disease treated with palliative intent, patients with a variety of histologies were included. This heterogeneous population, however, did allow us to assess the differences in response, durability of local response, and time to distant failure by histologic subtypes.

To our knowledge, this is the largest series from North America reporting on the efficacy of LD-IFRT. Our results demonstrate the success of LD-IFRT in palliating NHL, including disease that has become refractory to chemotherapy. A randomized trial (CRUK-FoRT) recently conducted in the United Kingdom compared 24 Gy in 12 fractions versus 4 Gy in 2 fractions as either definitive or palliative treatment for FL or MZL (20). A significantly higher overall response rate (81% vs 74%) and 2-year local progression-free rate (94% vs 80%) were found in the 24-Gy arm, leading to the authors’ conclusion that although 4 Gy in 2 fractions can be effective in the palliative setting, it is significantly inferior to 24 Gy in 12 fractions, which should remain the schedule of choice for curative radiation therapy in FL or MZL. The overall response rate of 74% with 4 Gy in 2 fractions reported in the FoRT trial is lower than that reported by previous series, including the present study (Table 6), and this may reflect selection bias associated with retrospective study designs. The 2-year local progression-free rate of the FoRT study, however, compared favorably with that of the present study and may be due to their inclusion of FL and MZL histologies only. In conclusion, LD-IFRT yields high response rates and offers durable local control in patients with indolent NHL. It is effective in patients with chemotherapy-refractory disease and can be given multiple times to patients with recurrent disease. Histology seems to be most significant in predicting response to treatment and, along with young age at diagnosis, in predicting the need for subsequent treatments. Low-dose involved-field RT should be considered as one of the palliative treatment options in patients with NHL.

References 1. American Cancer Society. Cancer Facts & Figures 2011. Atlanta: American Cancer Society; 2011. 2. Landgren O, Tilly H. Epidemiology, pathology and treatment of nonfollicular indolent lymphomas. Leuk Lymphoma 2008;49(Suppl. 1): 35-42.

Volume 86  Number 1  2013 3. A clinical evaluation of the International Lymphoma Study Group classification of non-Hodgkin’s lymphoma. The Non-Hodgkin’s Lymphoma Classification Project. Blood 1997;89:3909-3918. 4. Horning SJ. Treatment approaches to the low-grade lymphomas. Blood 1994;83:881-884. 5. O’Brien ME, Easterbrook P, Powell J, et al. The natural history of low grade non-Hodgkin’s lymphoma and the impact of a no initial treatment policy on survival. Q J Med 1991;80:651-660. 6. National Comprehensive Cancer Network. NCCN Guidelines V3.2012 Non-Hodgkin’s Lymphomas. Fort Washington, PA: National Comprehensive Cancer Network; 2012. 7. Wrench D, Gribben JG. Stem cell transplantation for non-Hodgkin’s lymphoma. Hematol Oncol Clin North Am 2008;22:1051-1079, xi. 8. Luthy SK, Ng AK, Silver B, et al. Response to low-dose involved-field radiotherapy in patients with non-Hodgkin’s lymphoma. Ann Oncol 2008;19:2043-2047. 9. Chan EK, Fung S, Gospodarowicz M, et al. Palliation by low-dose local radiation therapy for indolent non-Hodgkin lymphoma. Int J Radiat Oncol Biol Phys 2011;81:e781-e786. 10. Girinsky T, Guillot-Vals D, Koscielny S, et al. A high and sustained response rate in refractory or relapsing low-grade lymphoma masses after low-dose radiation: Analysis of predictive parameters of response to treatment. Int J Radiat Oncol Biol Phys 2001;51:148-155. 11. Rossier C, Schick U, Miralbell R, et al. Low-dose radiotherapy in indolent lymphoma. Int J Radiat Oncol Biol Phys 2011;81:e1-e6. 12. Haas RL, Poortmans P, de Jong D, et al. High response rates and lasting remissions after low-dose involved field radiotherapy in indolent lymphomas. J Clin Oncol 2003;21:2474-2480.

LD-IFRT for NHL

127

13. Johannsson J, Specht L, Mejer J, et al. Phase II study of palliative lowdose local radiotherapy in disseminated indolent non-Hodgkin’s lymphoma and chronic lymphocytic leukemia. Int J Radiat Oncol Biol Phys 2002;54:1466-1470. 14. Sawyer EJ, Timothy AR. Low dose palliative radiotherapy in low grade non-Hodgkin’s lymphoma. Radiother Oncol 1997;42:49-51. 15. Yahalom J, Mauch P. The involved field is back: Issues in delineating the radiation field in Hodgkin’s disease. Ann Oncol 2002;13(Suppl. 1): 79-83. 16. Zeger SL, Liang KY. Longitudinal data analysis for discrete and continuous outcomes. Biometrics 1986;42:121-130. 17. Lee LJ, Wei EW, Amato DA. Cox-type regression analysis for large number of small groups of correlated failure time observations. In: Klein JP, Goel PK, editors. Survival Analysis: State of the Art. Dordrecht, The Netherlands: Kluwer Academic; 1992. p. 237-247. 18. Haas RL, Poortmans P, de Jong D, et al. Effective palliation by low dose local radiotherapy for recurrent and/or chemotherapy refractory non-follicular lymphoma patients. Eur J Cancer 2005; 41:1724-1730. 19. Ng M, Wirth A, Ryan G, et al. Value of low-dose 2 x 2 Gy palliative radiotherapy in advanced low-grade non-Hodgkin’s lymphoma. Australas Radiol 2006;50:222-227. 20. Hoskin P KA, Popova B, Brammers C, et al. FoRT: A phase III multicentre prospective randomised trial of low dose radiotherapy for follicular and marginal zone lymphoma. Presented at the Annual Meeting of the American Society for Radiation Oncology, October 2831, 2012, Boston, MA.