Improved Survival in Patients With Early Stage Low-Grade Follicular Lymphoma Treated With Radiation

Original Article Improved Survival in Patients With Early Stage Low-Grade Follicular Lymphoma Treated With Radiation A Surveillance, Epidemiology, and End Results Database Analysis Thomas J. Pugh, MD; Ari Ballonoff, MD; Francis Newman, MS; and Rachel Rabinovitch, MD BACKGROUND: External beam radiation therapy (RT) is the standard treatment for stage I-II, grade 1-2 follicular lymphoma. Because of an indolent natural history, some advocate alternative management strategies, including watchful waiting for this disease. The relative improvement in outcomes for patients treated with and without RT has never been tested in randomized trials. METHODS: The Surveillance, Epidemiology, and End Results database was queried for adult patients with stage I-II, grade 1-2 follicular lymphoma diagnosed from 1973 to 2004. Retrievable patient data included age, sex, race, stage, extranodal disease, and treatment with RT within the first year after diagnosis. Actuarial overall survival (OS) and disease-specific survival (DSS) were analyzed. RESULTS: A total of 6568 patients were identified. DSS at 5, 10, 15, and 20 years in the RT group was 90%, 79%, 68%, and 63% versus 81%, 66%, 57%, and 51% in the no RT group (hazard ratio [HR], 0.61; 95% confidence interval [CI], 0.55-0.68; P <.0001). OS at 5, 10, 15, and 20 years in the RT group was 81%, 62%, 45%, and 35% versus 71%, 48%, 34%, and 23% in patients not receiving RT (HR, 0.68; 95% CI, 0.63-0.73; P <.0001). On multivariate analysis, upfront RT remained independently associated with improved DSS (P <.0001, Cox HR, 0.65; 95% CI, 0.57-0.72) and OS (P <.0001; Cox HR, 0.73; 95% CI, 0.67-0.79). Lymphoma was the most common cause of death (52%). Only 34% of patients received upfront RT. CONCLUSIONS: Upfront RT was associated with improved DSS and OS compared with alternate management approaches, a benefit that persisted over time. This benefit suggests that watchful waiting with administration of salvage therapies on progression/relapse do not compensate for inadequate initial definitive treatment. Although it is the standard of care for this disease, RT for early stage low-grade follicular lymphoma is greatly underused in the US population; increased use of upfront RT could prevent thousands of deaths from lymphoma in these patients. Cancer 2010;116:3843 51. VC 2010 American Cancer Society. KEYWORDS: follicular lymphoma, radiation therapy, indolent non-hodgkin lymphoma, watchful waiting. In 2008, an estimated 66,120 new cases of non-hodgkin lymphoma (NHL) were diagnosed in the United States. 1 Follicular B-cell lymphoma represents 30% of these cases, making it the most commonly diagnosed indolent subtype. 2 Whereas the majority of patients present with advanced disease (stage III-IV), approximately 30% of newly diagnosed cases of follicular lymphoma are early stage (stage I-II), or approximately 5950 patients annually in the United States. 2,3 Unlike advanced follicular lymphoma, for which there is no established curative therapy, early stage disease is potentially curable with regional or involved field radiation therapy (RT), long considered the standard treatment for patients with stage I-II disease. RT results in excellent complete response rates and long-term local control rates of >90%, documented in numerous single institution series. 4-10 Out-of-field recurrences are common, however; approximately half of all patients treated with RT alone will relapse within 10 years of treatment. 4-8 For the half of patients who remain disease-free Corresponding author: Rachel Rabinovitch, MD, Department of Radiation Oncology, University of Colorado Health Sciences Center, 1665 Aurora Court, Suite 1032, P.O. Box 6510, Mail Stop F-706, Aurora, CO 80045-0508; Fax: (720) 848-0222; Rachel.rabinovitch@ucdenver.edu Department of Radiation Oncology, University of Colorado Denver Comprehensive Cancer Center, Aurora, Colorado This article is the original work of the authors and was presented at the 50th Annual Meeting of the American Society of Therapeutic Radiology and Oncology, Boston, Massachusetts, September 22, 2008. We thank Jason Zhang for his assistance in data analysis. DOI: 10.1002/cncr.25149, Received: August 19, 2009; Revised: November 1, 2009; Accepted: November 10, 2009, Published online May 17, 2010 in Wiley Inter- Science (www.interscience.wiley.com) Cancer August 15, 2010 3843

Original Article for 10 years, first recurrences thereafter are infrequent, demonstrating that treated patients with stage I-II follicular lymphoma, in contrast to those patients presenting with advanced disease, can be cured by definitive RT. 4-5,11 Despite the long-documented curative potential of definitive RT for early stage disease, a contrasting approach of no therapy (watchful waiting) has also been proposed as reasonable, given the indolent natural history of the disease and the high likelihood of treatment failure. 9,12,13 The relative outcome differences between a watchful waiting approach and initial treatment with RT have never been tested in a prospective randomized trial for patients with early stage follicular lymphoma. We used the Surveillance, Epidemiology, and End Results (SEER) database to determine whether there are differences in disease-specific survival (DSS) or overall survival (OS) in patients treated with upfront RT compared with an alternative approach such as observation in patients with stage I-II, grade 1-2 follicular lymphoma. Patterns of RT utilization over time were also evaluated. METHODS AND MATERIALS The Surveillance, Epidemiology, and End Results (SEER) Program The SEER Program of the National Cancer Institute (NCI) collects and publishes information from population-based registries covering approximately 26% of the US population. 14 SEER routinely collects data regarding patient demographics, morphology, stage at diagnosis, first course of treatment, survival, and cause of death. Data from these registries are deidentified and electronically submitted to the NCI on a biannual basis. The data are subsequently made available to the general public. Study Population The SEER database (SEER 17 November 2006 data submission [1973-2004]) was queried for adult patients (18 years old) with microscopically confirmed stage I, IE, II, or IIE, grade 1 (follicular small cleaved cell) or grade 2 (follicular mixed) lymphoma diagnosed from 1973 to 2004 for whom outcome information was available. SEER*Stat 6.3.5 software was used to perform all queries. Stage designation was based on the SEER Program Coding and Staging Manual 2007, 15 which conforms to the Ann Arbor staging system and the American Joint Committee on Cancer sixth edition 16 staging for NHL. Patients coded as either receiving external beam radiation or not receiving radiation were included. Patients coded as not receiving radiation were defined by the following conditions: 1) there is no information in the patient s medical record about radiation, 2) multiple treatment options were offered and the patient selected treatment that did not include radiation, 3) the patient elected to pursue no treatment after the discussion of radiation treatment, and 4) a watchful waiting approach was the chosen treatment plan. Patients coded as receiving radiation are defined as patients who received external beam radiation to cancer tissue. The SEER database records first course of therapy data, which is defined as a treatment plan initiated within 12 months of diagnosis. Second-line therapies, including those initiated after a watchful waiting approach, are not recorded. Therefore, patients coded as receiving RT included only those patients who received radiation therapy as initial treatment soon after diagnosis, hereafter referred to as upfront RT. Data regarding patient age, sex, race, stage, grade, and extranodal disease were retrieved for all identified patients. Known prognostic factors for NHL, such as presenting lactate dehydrogenase (LDH), hemoglobin, and total number of involved lymph node regions, 17 are not recorded in SEER. Data were retrieved by 3-month intervals for a maximum follow-up of 360 months (30 years). Patients with multiple primary cancers, unknown age, or unknown ethnicity were excluded from analysis. Endpoints and Statistical Analysis DSS was determined from the time of diagnosis to the time of death from NHL. Death from lymphoma of any histology was chosen to account for patients initially diagnosed with early stage follicular lymphoma who experienced pathologic transformation to more aggressive NHL subtypes. OS was determined from the time of diagnosis to the time of death from any cause. Comparisons of characteristics between patients treated with upfront RT and those not receiving RT were made using the chi-square test. Actuarial DSS and OS analyses were illustrated using the Kaplan-Meier method 18 and compared using the logrank test. 19 Five-, 10-, 15-, and 20-year actuarial rates of DSS and OS are reported. Chi-square analyses between retrievable variables and treatment groups were conducted. Covariates analyzed included age (<60 vs 60 years), sex, race (white vs nonwhite), stage (I vs II), grade (1 vs 2), presence of extranodal disease, and treatment with RT. Multivariate analyses were performed to evaluate the influence of covariates on DSS and OS using Cox 3844 Cancer August 15, 2010

Radiation for Follicular Lymphoma/Pugh et al Table 1. Patient Characteristics Variable RT (%) No RT (%) P Total 2222 (34) 4346 (66) Age Median age, y 61 64 <.0001 <60 1050 (47) 1683 (39) 60 1172 (53) 2663 (61) Sex Women 1170 (53) 2268 (52).71 Men 1052 (47) 2078 (48) Race White 2074 (93) 4054 (93).93 Nonwhite 148 (7) 292 (7) No E 1263 (57) 2662 (61).0007 E 414 (19) 687 (16) Unspecified a 544 (24) 996 (23) Stage I 1717 (77) 2711 (62) <.0001 II 505 (23) 1635 (38) Grade 1 1202 (54) 2564 (59) <.0001 2 1020 (46) 1782 (41) RT¼radiation therapy; E, extranodal disease. P values were calculated using chi-square analysis. a Extranodal disease not coded prior to 1988. proportional hazards survival regression analysis. 20 Twosided P values and 95% confidence intervals (CIs) are reported. Statistical significance is defined as a P value.05. RESULTS A total of 6568 patients met the defined selection criteria. The median age of the entire cohort was 63 years. A total of 2222 (34%) patients were treated with upfront RT. A comparison of baseline demographic data and disease characteristics for patients in the RT and no RT groups is presented in Table 1. Patients in the RT group were significantly more likely to be younger, and to have stage I, grade 1, and no extranodal disease. A total of 3714 patients were alive at last follow-up (1429 patients died from NHL, 1343 died from other causes, and 82 patients had no cause of death recorded). Median follow-up time for all patients was 66 months (range, 3-360). Survival Outcomes Figure 1 shows the Kaplan-Meier DSS curves for all patients treated with and without RT. At 10 years, the DSS for RT and no RT groups were 79% and 65%, respectively (P <.0001). The absolute improvement in DSS for patients treated with RT at 5, 10, 15, and 20 Figure 1. Non-Hodgkin lymphoma-specific survival with or without upfront external beam radiation therapy (RT) is shown. HR indicates hazard ratio. years was 8.6%, 13.1%, 11.8%, and 11.6% (P <.0001; hazard ratio [HR], 0.61; 95% CI, 0.55-0.68). Factors associated with improved DSS for all patients on univariate analysis were age <60, stage I disease, and treatment with RT (Table 2). Sex, race, extranodal disease, and grade were not significantly associated with DSS. The effect of RT was analyzed within each patient subgroup, confirming an improved DSS associated with upfront RT in all subgroups. Figure 2 shows the Kaplan-Meier OS curves for all patients treated with or without RT. The absolute improvement in OS for patients treated with RT at 5, 10, 15, and 20 years was 10.3%, 14.2%, 11.0%, and 11.4% (P <.0001; HR, 0.68; 95% CI, 0.63-0.73). Factors strongly associated with improved OS on univariate analysis were younger age, stage I disease, and upfront RT (Table 3). Subgroup comparisons between treatment groups verified a consistent OS improvement associated with upfront RT. To account for the imbalance of characteristics in the RT and no RT groups, a multivariate analysis of all factors was performed (Table 4), demonstrating that upfront RT remained significantly and independently associated with both improved DSS (P < 0.0001; Cox HR, 0.65; 95% CI, 0.57-0.72) and improved OS (P < 0.0001; Cox HR, 0.73; 95% CI, 0.67-0.79). Other factors independently associated with improved DSS were age <60 years, stage I disease, and the absence of extranodal disease. Stage I disease and upfront treatment with RT were the only disease- or treatment-related factors independently associated with improved OS. RT Utilization A separate query of the database was conducted by time interval to determine treatment patterns during the Cancer August 15, 2010 3845

Table 2. Disease-Specific Survival Univariate Analyses Characteristic No. Actuarial 10-Year DSS, % P a HR 95% CI Age <60 2711 78 <.0001 0.51 0.46-0.56 60 3775 63 Sex Women 3395 70.68 1.02 0.92-1.13 Men 3091 70 Race Nonwhite 438 75.24 1.13 0.92-1.40 White 6048 70 Yes 1092 75.21 0.90 0.77-1.06 No 3883 73 Stage I 4366 73 <.0001 0.68 0.61-0.76 II 2120 65 Grade 1 3707 70.78 0.99 0.89-1.10 2 2779 70 Radiation RT 2206 79 <.0001 0.61 0.55-0.68 No RT 4280 65 RT effect by subgroup Age <60 years RT 1044 85 <.0001 0.61 0.51-0.73 No RT 1667 74 Age 60 years RT 1162 72 <.0001 0.65 0.57-0.74 No RT 2613 59 Men RT 917 78 <.0001 0.6 0.51-0.71 No RT 2044 65 Women RT 1159 79 <.0001 0.64 0.56-0.74 No RT 2366 66 White RT 2058 78 <.0001 0.61 0.54-0.67 No RT 3990 65 Nonwhite RT 148 80.21 0.75 0.48-1.18 No RT 290 71 RT 413 80.001 0.61 0.45-0.83 No RT 679 72 No RT 1253 81 <.0001 0.55 0.47-0.64 No RT 2630 69 Stage I RT 1702 81 <.0001 0.64 0.56-0.74 No RT 2664 72 Stage II RT 504 71.01 0.78 0.65-0.94 No RT 1616 63 Grade 1 RT 1189 78 <.0001 0.65 0.56-0.74 No RT 2518 66 Grade 2 RT 1017 80 <.0001 0.57 0.48-0.67 No RT 1762 64 DSS indicates disease-specific survival; HR, hazard ratio; CI, confidence interval; E, extranodal disease; RT, radiation. a P values calculated using log-rank test. 3846 Cancer August 15, 2010

Radiation for Follicular Lymphoma/Pugh et al Figure 2. Overall survival in patients with low-grade, stage I-II follicular lymphoma treated with or without upfront external beam radiation therapy (RT) is shown. HR indicates hazard ratio. analysis period (Fig. 3). Over the 30-year study period, approximately 1 = 3 of patients received upfront RT. The incidence of treatment utilization with upfront RT was stable when analyzed by decade: 36% (1973-1985), 35% (1986-1995), and 33% (1996-2004). DISCUSSION RT is the standard definitive treatment for early stage grade 1-2 follicular NHL based on single institution series reported over the past several decades documenting a proportion of long-term disease control. 4-8 Although no therapy (watchful waiting) has been advocated as an acceptable alternative strategy for these same patients, this approach has never been tested against RT in a randomized trial. Our SEER analysis was performed to evaluate the effect of RT on DSS and OS for low-grade, stage I-II follicular lymphoma by analyzing a large population database. As standard curative intervention for this group of patients consists of RT as the only treatment modality (surgery and chemotherapy are not standard treatments), the known limitations of SEER regarding absence of chemotherapy data are not relevant for this analysis, and SEER is a uniquely well-suited database for evaluating this question. To our knowledge, this is the largest reported analysis of early stage follicular lymphoma, and it significantly associates upfront RT with a clinically meaningful improvement in DSS (13% absolute improvement at 10 years). Evidence demonstrating prolonged disease-free survival in patients with stage I-II follicular lymphoma treated with RT was first published in the mid-1970s. 21,22 Investigators from Stanford University reported retrospective outcomes of 177 early stage patients treated with RT between 1961 and 1994. 4 The RT techniques were highly variable, with treatment volumes ranging from involved field RT, to extended field RT (mantle, inverted Y, whole abdomen), to subtotal nodal irradiation. Total radiation doses ranged from 35 to 50 gray (Gy), with most patients receiving 44 Gy. OS at 5, 10, 15, and 20 years was 82%, 64%, 44%, and 35%. Investigators from the Princess Margaret Hospital recently updated their series of 460 patients with stage I-II follicular lymphoma treated with involved field RT. 5 Ninety-eight percent of patients achieved a complete response, and only 5.5% relapsed within the radiation field. OS rates were very similar to the Stanford series: 79% and 62% at 5 and 10 years, respectively. The results of this SEER analysis in over 6500 patients confirm the findings of these and other investigators reporting prolonged survival for patients with early stage follicular lymphoma treated with RT. 4,5,8 The 10- year OS rate of 62% for patients treated with RT in the current study is virtually identical to the rates reported from Stanford and Princess Margaret. Other series reporting outcomes of early stage patients treated with RT describe 10-year OS rates ranging from 56% to 75%. 4-8 Table 5 summarizes the majority of published series and documents the similar outcomes of patients treated with RT from this SEER and other analyses. Deferred therapy in asymptomatic patients is a standard option in advanced (stage III-IV) follicular lymphoma, given the lack of efficacy of systemic chemotherapies and the indolent nature of the disease. This approach has been validated in several prospective randomized trials, confirming that immediate systemic treatment confers no survival benefit in patients with advanced disease. 9,23-25 The outcome of a watchful waiting strategy was subsequently applied to 43 select early stage patients at Stanford University diagnosed between 1969 and 2000. 12 Reasons for deferring treatment included physician or patient preference (40%), advanced patient age or comorbid disease (23%), and concern regarding potential complications of therapy (37%). Patients were observed until they displayed progressive disease or histologic transformation to a more aggressive NHL subtype. All patients included for analysis received no treatment for at least 3 months after diagnosis. The median survival in this cohort was 19 years, and 10-year OS was 85%; interestingly, these results were better than those for patients treated with upfront RT alone from the same institution. 4 Although it is difficult to make definitive conclusions from this small Cancer August 15, 2010 3847

Table 3. Overall Survival Univariate Analyses Characteristic No. Actuarial 10 y OS (%) P a HR 95% CI Age <60 years 2733 71 <.0001 0.37 0.35-0.40 60 years 3835 40 Sex Women 3438 54.08 0.94 0.87-1.01 Men 3130 51 Race Nonwhite 440 60.06 0.86 0.74-1.01 White 6128 52 Yes 1101 57.81 0.99 0.88-1.10 No 3925 56 Stage I 4428 55 <.0001 0.82 0.75-0.88 II 2140 48 Grade 1 3766 53.93 1.00 0.92-1.10 2 2802 52 Radiation RT 2222 61 <.0001 0.68 0.63-0.73 No RT 4346 48 RT effect by subgroup Age <60 RT 1050 79 <.0001 0.67 0.58-0.77 No RT 1683 65 Age 60 RT 1172 47 <.0001 0.73 0.67-0.80 No RT 2663 36 Male RT 1052 60 <.0001 0.65 0.59-0.73 No RT 2078 46 Female RT 1170 63 <.0001 0.70 0.63-0.78 No RT 2268 49 White RT 2074 62 <.0001 0.67 0.62-0.73 No RT 4054 47 Non-White RT 148 64.12 0.77 0.55-1.07 No RT 292 57 RT 414 61 <.0001 0.69 0.59-0.81 No RT 2229 47 No RT 1263 66 <.0001 0.59 0.53-0.66 No RT 2662 51 Stage I RT 1717 64 <.0001 0.63 0.58-0.70 No RT 2711 49 Stage II RT 505 55.01 0.82 0.71-0.95 No RT 1635 46 Grade 1 RT 1202 61 <.0001 0.70 0.63-0.77 No RT 2664 49 Grade 2 RT 1020 64 <.0001 0.64 0.57-0.72 No RT 1782 45 OS indicates overall survival; HR, hazard ratio; CI, confidence interval; E, extranodal disease; RT, radiation. a P values calculated using log-rank test. 3848 Cancer August 15, 2010

Radiation for Follicular Lymphoma/Pugh et al Table 4. Multivariate Analysis of All Variables in Patients With Localized Low-Grade Follicular Lymphoma Characteristic Disease-Specific Survival Overall Survival CHR a 95% CI P CHR a 95% CI P Age (<60 vs 60) 0.50 0.44-0.55 <.0001 0.35 0.32-0.38 <.0001 Sex (women vs men) 0.97 0.88-1.08.61 0.88 0.81-0.94.0003 Race (white vs nonwhite) 1.11 0.89-1.39.34 1.12 0.95-1.32.18 (no E vs E) 0.84 0.71-0.99.04 0.93 0.83-1.05.25 Stage (I vs II) 0.71 0.64-0.79 <.0001 0.81 0.75-0.87 <.0001 Histologic grade (1 vs 2) 0.96 0.87-1.06.46 0.97 0.90-1.04.42 Radiation (RT vs no RT) 0.65 0.57-0.72 <.0001 0.73 0.67-0.79 <.0001 CHR indicates Cox hazard ratio; CI, confidence interval; E, extranodal; RT, radiation therapy. P values were determined using log-rank test. a CHR <1 indicates reduced risk associated with first listed variable. Figure 3. Utilization of upfront external beam radiation therapy for localized low-grade follicular lymphoma in the United States is shown by decade. retrospective series, the results suggest that deferred therapy can be acceptable for some patients. However, a subset of patients for whom treatment was withheld experienced a less favorable disease course, with aggressive transformation occurring as early as 14 months after diagnosis and 14% (6 of 43) dying from progressive lymphoma. Despite prolonged DSS and competing causes of mortality, lymphoma remains the leading cause of death for patients with early stage, low-grade follicular lymphoma. MacManus and colleagues evaluated the impact of intercurrent deaths in these patients compared with age-matched controls. 4 Their study reported high rates of intercurrent deaths in patients aged 60 years or older, but suggested that the presence of follicular lymphoma negatively impacted survival for both younger and older patients. The Harvard Longwood area hospitals reported their experience in 106 patients with stage I-II, grade 1-2 follicular lymphoma diagnosed between June 1972 and February 2000. 8 Most patients received involved field RT (60%) as opposed to extended treatment fields with a median dose of 36.7 Gy. In a favorable group consisting predominantly of stage I patients, with median survival of 19 years, lymphoma was the predominant cause of mortality, accounting for 61% (20 of 33) of the observed deaths. The results of our analysis confirm these findings. Of the 2772 observed deaths in our cohort for which cause of death information was available, 52% (1429 of 2772) were attributed to NHL. The current analysis demonstrates that treatment with RT within 1 year of diagnosis significantly reduced the risk of dying from lymphoma, with an absolute risk reduction of 13.1% at 10 years. The DSS benefit appreciated with upfront RT was sustained throughout the 30- year analysis period, indicating that salvage therapy on progression does not compensate for deferred upfront definitive therapy for patients presenting with localized disease. Furthermore, the similar OS benefit associated with upfront RT documents the safety of this therapy. The similar magnitude of the DSS and OS benefit associated with RT delivery (HR, 0.61 and 0.68, respectively) indicates that upfront RT should not be withheld because of concerns of long-term increased mortality risks related to this therapy. In fact, other than age, RT delivery had the greatest impact on improved DSS and OS, as indicated by the HRs in the multivariate analyses (Table 4). The survival benefits of upfront therapy were consistent and significant across all patient subgroups, with the exception of nonwhite patients. This finding is explained by the small number of nonwhite patients included in the study cohort. With a total of 438 nonwhite patients accounting for just 7% of the total study population, there was inadequate power to reach statistical significance in this small subgroup, although outcomes were numerically improved for RT-treated patients. As in all SEER analyses, the following limitations must be considered: Cancer August 15, 2010 3849

Original Article Table 5. Stage I-II Follicular Lymphoma Treatment Results Study Institution Patient No. Treatment 10-Year FFR 10-Year OS 10-Year DSS Soubeyran 1988 7 Fondation Bergonié 103 RT Chemo 49% a 56% NA Kelsey 1994 11 British National Lymphoma 148 RT 33% 52% NA Investigation (RCT) RT þ Chemo 42% 42% NA Vaughan Hudson 1994 6 British National Lymphoma 208 RT 47% 64% 70% Investigation Pendlebury 1995 26 Royal Marsden, England 58 RT 43% b 79% NA MacManus & Hoppe 1996 4 Stanford 177 RT 44% 64% NA Seymour 2003 27 M. D. Anderson 83 RTþ Chemo 72% 80% NA Petersen 2004 5 Princess Margaret Hospital, 460 RT 51% 62% 79% Canada Advani 2004 12 Stanford 43 WW NA 85% NA Guadagnolo 2006 8 Joint Center 106 RT Chemo 46% 75% NA Current study NCI SEER 2222 RT NA 62% 79% 4346 No RT NA 48% 65% FFR indicates freedom from relapse; OS, overall survival; DSS, disease-specific survival; RT, radiotherapy; Chemo, chemotherapy; NA, not available; RCT, randomized control trial; WW, watchful waiting; NCI, National Cancer Institute; SEER, Surveillance, Epidemiology, and End Results. a Relapse-free survival rate. b Progression-free survival rate. Whereas well-known prognostic factors such as age and disease stage were obtainable, other validated prognostic factors for follicular lymphoma (LDH, number of involved nodal sites, bulky disease, B symptoms, and hemoglobin level) are not available through SEER. Therefore, we were unable to adjust for these factors in our analysis. Systemic therapy data and subsequent therapy after first course of treatment are not available through the SEER program. This latter consideration is unlikely to confound our results, as chemotherapy has no proven efficacy for early stage disease and is not standard therapy. Antibody and radiolabeled antibody therapies are currently under investigation for this patient population, but were not and are not approved by the US Food and Drug Administration or indicated for this group of patients. Although there has never been a prospective randomized trial comparing RT to no therapy for early stage grade I-II follicular lymphoma, our results in a population dataset involving >6500 patients provide the largest and strongest evidence to date of a significant correlation between upfront RT and improved DSS and OS. Treatment choices were made based on the patients clinical status combined with the treating physicians beliefs and judgment. In the absence of randomized data, the clinician is relegated to using retrospective and pooled analyses to guide clinical decision making. Given the lack of prospective randomized data and even with the likelihood of selection bias inherent in all retrospective analyses, we contend that the current study provides the highest level of evidence available for this group of patients supporting the initial use of definitive RT. Despite the historical acceptance of upfront RT as the only standard and potentially curative therapy for this patient population, only a minority of patients received it, an observation that was stable over the 30-year evaluation period. In our analysis, RT was associated with a 13% absolute risk reduction of death from NHL at 10 years (62% vs 49%; Fig. 2). In other words, for every 8 patients treated with upfront RT, 1 death from NHL was prevented at 10 years (number needed to treat ¼ 100/13 ¼ 8). If we apply these statistics to the US population, we can calculate the theoretical number of preventable NHLrelated deaths as a result of upfront RT underutilization. Considering the incidence of newly diagnosed early stage, low-grade, follicular lymphoma (5950 patients annually), the percentage of eligible patients who are not treated with upfront RT (66%), and the number needed to treat to prevent 1 death from NHL, 8 uniform use of upfront RT would prevent 491 deaths at 10 years (5950 patients 0.66 not treated/8). Over the 30-year time period analyzed, this translates into 14,430 (491 30) preventable deaths attributable to the withholding of standard definitive therapy. In conclusion, this population-based analysis, reflective of care and outcomes in >6500 patients treated in the United States over the past 30 years, demonstrates that upfront RT for patients with stage I-II, grade 1-2 follicular lymphoma is associated with improved DSS and OS compared with any alternative approach such as watchful waiting. These benefits persist with time, indicating that delayed or salvage strategies do not compensate for inadequate initial definitive treatment. Similar observed gains 3850 Cancer August 15, 2010

Radiation for Follicular Lymphoma/Pugh et al in both DSS and OS associated with RT administration document the safety of this therapy. Upfront RT should be offered as the standard therapy to all newly diagnosed early stage patients based on the potential to decrease death from lymphoma or any cause. Such treatment is dramatically underused in the US population assessed through SEER, suggesting the possibility of thousands of lives lost because of withholding of this therapy. These data strengthen the position that upfront RT is considered the standard against which all emerging treatment paradigms should be compared. Because of the indolent natural history, long-term follow-up will be required to test the survival effects of any new treatment approach. The relative rarity of localized disease and the necessity of long-term follow-up are significant barriers to completion of phase 3 trials in patients with low-grade, early stage follicular lymphoma. Therefore, multi-institutional collaboration will be imperative to prospectively confirm these findings and further define optimal management strategies for these patients. CONFLICT OF INTEREST DISCLOSURES The authors made no disclosures. REFERENCES 1. National Cancer Institute. Available at: http://www.cancer.gov/ cancertopics/types/non-hodgkin Accessed October 6, 2008. 2. 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. 3. Armitage J, Weisenburger D. New approach to classifying non-hodgkin s lymphomas: clinical features of the major histologic subtypes. J Clin Oncol. 1998;16:2780-2795. 4. MacManus MP, Hoppe RT. Is radiotherapy curative for stage I and II low-grade follicular lymphoma? Results of a long-term follow-up study of patients treated at Stanford University. J Clin Oncol. 1996;14:1282-1290. 5. Petersen PM, Gospodarowicz R, Tsang R, et al. Long term outcome in stage I and II follicular lymphoma following treatment with involved field radiation therapy alone. J Clin Oncol. 2004;22:6521. 6. Vaughan Hudson B, Vaughan Hudson G, MacLennan KA, Anderson L, Linch DC. Clinical stage 1 non-hodgkin s lymphoma: long-term follow-up of patients treated by the British National Lymphoma Investigation with radiotherapy alone as initial therapy. Br J Cancer. 1994;69:1088-1093. 7. Soubeyran P, Eghbali H, Bonichon F, et al. Localized follicular lymphomas: prognosis and survival of stages I and II in a retrospective series of 103 patients. Radiother Oncol. 1988; 13:91-98. 8. Guadagnolo BA, Sigui L, Neuberg D, et al. Long term outcome and mortality trends in early stage, grade 1-2 follicular lymphoma treated with radiation therapy. Int J Radiat Oncol Biol Phys. 2006;64:928-934. 9. National Comprehensive Cancer Network. Clinical Practice Guidelines in Oncology: Non-Hodgkin Lymphoma. Available at: http://www.nccn.org/professionals/physician_gls/pdf/nhl. pdf Accessed October 12, 2008. 10. Engelhard M, Stuschke M. Report on workshop: UICC workshop Therapy of NHL in early stages. Part 1: Follicular lymphoma. Ann Hematol. 2001;80:13-15. 11. Kelsey SM, Newland AC, Hudson GV, Jelliffe AM. A British National Lymphoma Investigation randomized trial of single agent chlorambucil plus radiotherapy versus radiotherapy alone in low grade, localized, non-hodgkins lymphoma. Med Oncol. 1994;11:19-25. 12. Advani R, Rosenberg SA, Horning SJ. Stage I and II follicular non-hodgkin s lymphoma: long-term follow-up of no initial therapy. J Clin Oncol. 2004;22:1454-1459. 13. Horning SJ, Rosenberg SA. The natural history of initially untreated low grade non-hodgkin s lymphomas. N Engl J Med. 1984;311:1471-1475. 14. Surveillance, Epidemiology, and End-Results Program. Available at: http://seer.cancer.gov/about Accessed October 12, 2008. 15. Johnson CH, Adamo M. The SEER Program Coding and Staging Manual 2007. National Cancer Institute, NIH Publication number 07-5581. Available at: http://seer.cancer.gov/ manuals/2007/spcsm_2007_maindoc.pdf Accessed October 20, 2008. 16. Greene FL, Page DL, Fleming ID, et al. AJCC Cancer Staging Handbook. 6th ed. New York, NY: Springer; 2002. 17. Solal-Celigny P, Roy P, Colombat P, et al. Follicular lymphoma international prognostic index. Blood. 2004;104:1258-1265. 18. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc. 1958;53:457-458. 19. Mantel N. Evaluation of survival data and 2 new rank order statistics arising in its consideration. Cancer Chemother Rep. 1966;50:163-170. 20. Cox DR. Regression models and life-tables. J R Stat Soc [B]. 1972;34:187-202. 21. Peters MV, Bush RS, Brown TC, et al. The place of radiotherapy in the control of non-hodgkin s lymphomata. Br J Cancer. 1975;31:386-401. 22. Chen MG, Prosnitz LR, Gonzalez-Serva A, Fischer DB. Results of radiotherapy in control of stage I and II non- Hodgkin s lymphoma. Cancer. 1979;43:1245-1254. 23. Brice P, Bastion Y, Lepage E, et al. Comparison in low-tumor-burden follicular lymphomas between an initial notreatment policy, prednimustine, or interferon alfa: a randomized study from the Groupe d Etude des Lymphomes Folliculaires. Groupe d Etude des Lymphomes de l Adulte. J Clin Oncol. 1997;15:1110-1117. 24. Ardeshna KM, Smith P, Norton A, et al. British National Lymphoma Investigation: long term effect of a watch and wait policy versus immediate systemic treatment for asymptomatic advanced stage non-hodkin lymphoma a randomized control trial. Lancet. 2003;362:516-522. 25. Young RC, Longo DL, Glatstein E, et al. The treatment of indolent lymphomas: watchful waiting versus aggressive combined modality treatment. Semin Hematol. 1988;25(supp 2):11-16. 26. Pendlebury S, el Awadi M, Ashley S, et al. Radiotherapy results in early stage low grade nodal non-hodgkin s lymphoma. Radiother Oncol. 1995;36:167-171. 27. Seymour JF, Pro B, Fuller LM, et al. Long-term follow-up of a prospective study of combined modality therapy for stage I-II indolent non-hodgkin s lymphoma. J Clin Oncol. 2003;21:2115-2122. Cancer August 15, 2010 3851