Stereotactic body radiation therapy for ≥ 5 cm node-negative non-small cell lung cancer : Survey of U . S . academic thoracic radiation oncologists

Mr. Post is a fourth-year medical student, Dr. Verma is chief resident, and Dr. Zhen is a professor, Department of Radiation Oncology, University of Nebraska Medical Center, Omaha, NE. Dr. Simone is an associate professor, Department of Radiation Oncology, University of Maryland Medical Center, Baltimore, MD. Disclosure: The authors have no conflicts of interest to disclose. None of the authors received outside funding for the production of this original manuscript and no part of this article has been previously published elsewhere. Abstract Purpose: Large (≥ 5 cm) node-negative non-small cell lung cancer (NSCLC) is relatively uncommon; efficacy and toxicities of stereotactic body radiation therapy (SBRT) in this unique population have been under-evaluated. Methods and Materials: We surveyed U.S. academic thoracic radiation oncologists regarding SBRT practice patterns in node-negative ≥ 5 cm NSCLC and assessed factors necessitating changes in SBRT management. A 25-question survey of demographics and practice patterns, including 5 clinical cases, was sent to 107 radiation oncologists who self-identified as thoracic/lung cancer specialists. Results: Response rate was 34% (36/107). Among respondents, two-thirds had at least 6 years of work experience following residency; 67% and 67% annually treated > 60 lung cancer and > 25 lung SBRT cases, respectively. Nearly all (97%) routinely offered SBRT for ≥ 5 cm NSCLC, and 55% used a SBRT treatment of 50-60 Gy in 5 fractions, with fractions delivered every other day in 60%. Dosing/fractionation were most commonly altered for central disease (77%). Sixty percent would offer additional chemotherapy; chemotherapy was strongly considered for patients with good performance status (74%), younger age (69%), and larger tumor size (68%). The 5 clinical cases revealed significant practice variability in dose, fractionation, treatment timing, and chemotherapy use. Conclusions: Practice patterns of SBRT for ≥ 5 cm NSCLC display substantial heterogeneity. Five-fraction regimens with biologically effective dose ≥100 Gy were most commonly performed, with common endorsement of every other day delivery and chemotherapy.

N on-small cell lung cancer (NSCLC) is the most common cause of cancer death in the world. 1,2][4][5][6][7][8][9][10][11][12] Initial studies of SBRT have demonstrated excellent local control rates of > 90%, but these large cohort studies have consisted primarily of small (≤ 4 cm) primary tumors, with node-negative NSCLC tumors ≥ 5 cm being vastly under-represented. 123][14][15] Hence, the National Comprehensive Cancer Network (NCCN) does not offer concrete guidelines on treatment of node-negative ≥ 5 cm NSCLC. 3ecause of this limited data and lack of consensus, there is great heterogeneity in how these cases are treated in clinical practice, and many questions remain regarding practicality of several SBRT schemes in this population.Hence, we

Methods and Materials
We asked 107 thoracic radiation oncologists from 71 U.S. academic institutions to participate in a 25-question survey.All invited participants self-identified as specializing in thoracic and/or lung radiation oncology.A single thoracic/ lung radiation oncologist was invited per institution in most cases; however, multiple radiation oncologists were invited for select larger institutions in which multiple providers specifically focus their clinical practice on lung cancer.The invitation contained instructions for participation and information regarding the study.The first invitation was sent on June 29, 2016.Participants who requested not to be contacted in the future were immediately removed from the database.The remaining respondents were contacted with a reminder email on July 12, 2016, to maximize response rate.No further communication with participants ensued.
Responses were anonymous and were recorded with Google (N = 34) or Word documents (N = 2).The complete survey (Supplemental Figure 1) was divided into demographic questions, clinical scenarios in which respondents commented on typical treatment preferences, and 5 clinical cases to assess dose/fractionation of SBRT and chemotherapy administration.Demographic questions addressed clinical experience, the nature of the clinician's practice, and patient volume.Next, preferences on mediastinal staging modalities, chemotherapy use and timing, and practical/technical aspects of SBRT were recorded.Subsequently, various clinical scenarios were presented to assess whether each respondent would change management.Respondents selected from a list of several potential reasons for adding chemotherapy in addition

Demographics
The overall response rate was 34% (36/107).Table 1 illustrates respondent demographics.Thirty-three percent had 0-5 years of work experience after residency, 19% had 6 to 10 years, 28% had 11 to 20 years, and 19% had > 20 years.Most respondents practiced in an urban location (78%), and they most commonly worked in the Northeast (33%) and Midwest (28%).Forty-seven percent were partners in a radiation oncology practice of 10 to 25 radiation oncologists, whereas 33% were in a practice of 2 to 9 physicians.
Lung cancer patients comprised over half of the practitioner's patient volume for approximately half (47%) of respondents, with lung cancer patients constituting 26% to 50% of the practice in an additional 31% of respondents.Half of the surveyed population saw > 90 lung cancer cases per year.Two-thirds of respondents (67%) delivered SBRT to at least 26 patients annually, with high volume providers (> 75 cases per year) accounting for 25% of total respondents.Most respondents (86.1%) had significant experience delivering SBRT to NSCLC tumors ≥ 5 cm, with 28% treating 1 to 2 cases per year, 28% treating 3 to 5 cases per year, 14% treating 6 to 10 cases per year, and 17% treating >10 cases per year.Of those surveyed, 94% participated in lung cancer cooperative group trials.

Practice Patterns
Table 2 highlights the collective responses to the survey's practice pattern questions.Eighty-nine percent used endobronchial ultrasound (EBUS) and/or mediastinoscopy in addition to positron emission tomography (PET) scanning as part of the initial staging workup.One respondent did not treat any NSCLC ≥ 5 cm with SBRT.Among respondents, 55% most typically treated ≥ 5 cm NSCLC with 50 to 60 Gy in 5 fractions, with 18% using 48 to 54 Gy in 3 fractions, and 8% each preferring 48 to 50 Gy in 4 fractions, 60 Gy in 8 fractions, and 70 Gy in 10 fractions.Sixty percent of respondents would deliver fractions every other day, whereas 40% would deliver fractions daily.
Sixty percent of respondents recommended chemotherapy use in ≥ 5 cm NSCLC patients being definitively treated with SBRT, with 81% and 19% preferring chemotherapy administration following and prior to SBRT, respectively.The factors most commonly reported as leading to consideration of chemotherapy included good performance status (74%), larger tumor size (69%), and younger age (69%).The responses to several other pertinent clinical factors influencing chemotherapy use are recorded in Table 2. Twenty-six percent would consider chemotherapy if no pathologic mediastinal staging was performed, and 20% would consider chemotherapy if there was visceral pleural involvement or adenocarcinoma histology.Five respondents (14%) would not consider chemotherapy regardless of any of the above-mentioned factors.

Cases
The results of the surveyed clinicians' recommended dosing and fractionation schemes in 5 clinical cases are shown in Table 4. Respondents offered SBRT for all cases with the exception of 2 respondents who refrained from using SBRT in case 2, the case in which the largest tumor size (7.5 cm) was depicted.

Discussion
Although ≥ 5 cm NSCLC cases are relatively uncommon thoracic malignancies, there is no consensus recommendation for this patient population. 3dditionally, in regard to the utility and efficacy of SBRT in large node-negative NSCLC, guidelines regarding dose and fractionation are lacking.As such, there is no consensus among providers regarding patient stratification and adjusting management accordingly based on various patient and tumor characteristics.Thus, our survey was designed to evaluate the diverse opinions of   [16][17][18][19][20] A vast majority (88%) of respondents preferred the addition of EBUS or mediastinoscopy in addition to PET scanning for staging, despite little evidence to sup-port that lymph node sampling improves outcomes in stage I-IIIA NSCLC. 21,22owever, it must be recognized that large tumors, especially central ones, have notably higher risks of occult nodal involvement, 23 likely explaining why respondents preferred lymph node sampling in this higher risk patient population.Despite this increased risk, a recent multi-institutional retrospective analysis revealed no improvement in tumor control (local, regional and distant) or survival with the addition of mediastinal lymph node sampling. 24Analysis to determine which subgroup(s) of patients with larger lesions that benefit the most from pathologic mediastinal evaluation is warranted.
7][18] Respondents also supported delivering treatments every other day (60%); however, there was considerable variation in this regard.Some studies have shown decreased toxicity with fractions delivered every other day, and that spacing out SBRT

STEREOTACTIC BODY RADIATION THERAPY NSCLC SURVEY
applied radiation oncology treatments in other neoplasms can also reduce toxicities. 25,26Decreased toxicity with every other day vs daily treatment has been reported for this patient population. 17Moreover, inverse planning with VMAT was preferred (63%).This might reflect the recent increased use of VMAT and its advantage of reducing treatment times and potentially improving conformity of dose coverage.However, there are conflicting dosimetric data comparing IMRT and VMAT as means for SBRT delivery, [27][28][29] and the significance of dynamic motion effects during VMAT is currently not well defined for tumors ≥ 5 cm. 30mong age, performance status and central tumor location, the latter was most commonly associated (71%) with a change in management by the surveyed population.Of the 24 respondents who would change management, 11 (46%) switched from a classic SBRT scheme of ≤ 5 fractions to > 5 fractions.Given that prior reports of SBRT for lesions < 5 cm have demonstrated increased toxicity when treating centrally located lesions, 31 and that treating larger tumors presumably has higher risks of toxicities than smaller tumors, this finding of switching fractionation schemes for central tumors is not unanticipated.Higher rates of toxicities have been reported for central lesions; 16 however, more recent data suggest no toxicity differences based on tumor location. 18dditional clinical outcomes data are needed to determine whether SBRT of larger tumors is associated with higher rates of toxicities than for < 5 cm tumors, and if toxicity rates are higher in central lesions despite more widespread adoption of modern SBRT techniques.
The addition of chemotherapy to SBRT was endorsed by 60% of respondents, of whom 81% preferred chemotherapy to be sequenced after SBRT.Despite this preference, only 2 studies have shown an overall survival (OS) improvement with the addition of adjuvant chemotherapy to SBRT. 32,33In the current survey, chemotherapy was more commonly considered in patients with good performance status (74%), younger age (69%), and larger tumor size (69%).These characteristics highlight that the perceived ability to tolerate chemotherapy, rather than specific tumor characteristics, is a common guiding rationale behind recommending chemotherapy in this high-risk population.Interestingly, 74% and 57% of respondents chose not to offer chemotherapy in cases 1 and 5, which depicted younger patients with a good performance score.Regardless, with distant failure occurring in 19% to 33% of patients, [16][17][18][19][20] studies that assess the exact clinical benefit of adjuvant chemotherapy are greatly needed, and novel approaches of trialing SBRT and immunotherapy for this patient population may also prove beneficial. 34esponses to the 5 clinical cases further identified which clinical parameters altered SBRT treatment regimens and chemotherapy usage.SBRT regimens > 5 fractions were prescribed most commonly in case 5 (46%) and case 2 (40%), which presented a 6.2-cm central tumor and a 7.5-cm peripheral tumor, respectively.Regarding treatment timing, although 60% advocated this in the initial question, in no clinical case did > 60% of respondents endorse every other day fractionation.Administration of SBRT fractions every other day was highest in case 1 (57%), presenting a 5.0-cm peripheral tumor.In fact, the 2 cases with central disease showed the lowest proportion of respondents recommending every other day fractionation (43% and 40%), although these were least likely to receive 5-fraction regimens to begin with.Of note, a 3-fraction regimen was most common in case 1 (29%), a patient with a 5.0-cm peripheral tumor, and in case 4 (20%), a patient with a 5.4-cm peripheral tumor.Case 4 also displayed the lowest rate of chemotherapy administration (20%).In contrast, chemotherapy was recommended most commonly in case 5 (43%), which depicted a 62-year-old patient with good performance status and a 6.3-cm moderately differentiated central lesion.The responses to the cases differed from the generic practice patterns questions, clarifying that each treatment plan was indeed created on a case-by-case basis.
SBRT for large node-negative NSCLC has many challenges, notably increased risks of toxicities and poorer tumor control, but its efficacy and toxicity have been reported in several recent studies.7][18][19][20] Despite the efficacy and safety of SBRT for large NSCLC, toxicity minimization is of the utmost importance in this population.The use of proton therapy could be a promising alternative to photon-based SBRT, wherein physical properties of the heavier proton particle that limits irradiation to normal adjacent tissues may translate into reduced toxicities to organs at risk, as well as potentially allow for dose escalation to improve local control.][37] Respiratory gating, which propagates radiation delivery only at designated phases of the respiratory cycle, most commonly at the end of expiration, can further reduce dose to OARs.Inverse plan optimization of gating using patient specific data (ie, 4-dimensional computed tomography [4D-CT] and individual breathing patterns), as compared to traditional gating methods, has been shown to significantly reduce irradiation doses to the heart, esophagus and spinal cord. 38Lastly, increased use of PET imaging for radiation treatment planning 39 and improvements in MRI-guided SBRT may allow for better delineation of the tumor from healthy tissue, leading to sharper planning treatment volumes. 40lthough this is the first survey of its kind assessing practice patterns for patients with large, node-negative non-small n APPLIED RADIATION ONCOLOGY www.appliedradiationoncology.com

March 2018
STEREOTACTIC BODY RADIATION THERAPY NSCLC SURVEY applied radiation oncology cell lung cancer, there are several limitations to this work.First, analysis is based on a limited number of respondents (n = 36).We limited the survey to academic thoracic radiation oncologists who self-identified as specialists in lung cancer to target a study population of providers who are most experienced in treating large node-negative NSCLC with SBRT.Fortunately, we do note a considerably high response rate among the total population surveyed (34%).Additionally, participation bias likely exists, as providers with more experience treating large tumors may have been more likely to complete the survey.As such, our results may not be representative of the practice patterns of SBRT in this unique patient population among the radiation oncology workforce outside of U.S. academia.Also, as in all surveys, wording of questions and limited space to offer a comprehensive clinical vignette or response options provided in the survey may have inappropriately simplified the complex nature of treatment planning in this challenging patient population.For instance, to simplify the wording of the survey, we did not acquire each respondent's dose/fractionation SBRT scheme simultaneously with dosing frequency, as we did for the clinical cases, and we instead used 2 separate questions to obtain this information.Lastly, when we assessed for chemotherapy usage in the cases, it was presented in a binary manner, which may have influenced respondents to not choose chemotherapy if they could not also dictate when it would be administered in relation to SBRT.

Conclusion
There are no current recommendations regarding SBRT for ≥ 5 cm node-negative NSCLC.Most commonly, respondents advocated treatment with 50 to 60 Gy in 5 fractions using VMAT, with fractions delivered every other day.However, substantial variability existed across treatment parameters.Central tumor location prompted most respondents to adjust their SBRT management, with roughly half adopting a > 5 fraction regimen.Chemotherapy was recommended more often in patients with good performance status, younger age and larger tumor size.

Supplemental Figure 1. Complete Survey Sent to Academic Thoracic Radiation Oncologists
Continued from the previous page

Table 4
continues on the next page n APPLIED RADIATION ONCOLOGY www.appliedradiationoncology.com
If yes, what would be your preferred dose and fractionation for patients with poor performance status? ____________ Gy in ______ fractions 9. Would a central location of the tumor lead you to change the dose/ n APPLIED RADIATION ONCOLOGY www.appliedradiationoncology.com