BID-ding wars: will dose-escalated twice daily radiotherapy for limited stage small cell lung cancer be the new standard of care?-critical analysis of the literature.

Introduction
About one third of patients with small cell lung cancer (SCLC) present with non-metastatic local-regional disease. This disease entity is labeled ‘limited stage’ (LS) SCLC (1). Treatment for LS-SCLC in fit patients should be given with curative intent, using a combination of platinum-based chemotherapy and thoracic radiotherapy (RT). Over 25 years ago, Turrisi and colleagues from the National Cancer Institute (NCI)-funded Eastern Cooperative Oncology Group (ECOG) and Radiation Therapy Oncology Group (RTOG) cooperative groups showed in a randomized trial improved 5-year survival with twice-daily (BID) radiotherapy (RT) over once-daily (QD) RT (2). The chemotherapy (cisplatin/etoposide) and cumulative RT dose (45 Gy) were the same in both arms.
Fast forwarding to today, we are very pleased to provide this commentary on an excellent randomized phase II trial by Grønberg et al. in patients with LS-SCLC that compared concurrent chemoradiation with 45 Gy in 30 BID fractions (the Turrisi regimen) to 60 Gy in 40 BID fractions. This study of 170 patients found that the dose-escalated cohort had numerically prolonged progression-free survival (PFS) and significantly prolonged overall survival (OS) (3). Eligibility was well controlled; all patients had performance status of 0–2 and positron emission tomography (PET) and brain magnetic resonance imaging (MRI) confirmed LS disease (4). Forced expiratory volume in 1 second needed to be greater than 1 L or 30% of predictive value and diffusion limit of carbon monoxide needed to be greater than 30% of predicted value. If pleural effusion was present and amenable to sampling, at least one negative cytology was required. Patients received standard chemotherapy with 4 cycles of cis/carboplatin and etoposide. Radiation started concurrent with cycle 2 of chemotherapy (20–28 days after cycle 1 initiation). For patients with stable disease or response after concurrent chemoradiation, prophylactic cranial irradiation (PCI) was recommended (25–30 Gy in 10–15 fractions).
The authors should be commended on the outstanding outcomes presented in their study as well as the rigorous evaluation of relapse patterns and their accompanying article analyzing detailed radiation plans (5). Their results at the very least justify the need for a confirmatory phase III study of 60 Gy BID RT in the modern era of adjuvant durvalumab and add to the growing body of data suggesting that BID radiotherapy is the preferred standard of care in LS-SCLC. The purpose of our article is to further probe into the strengths and limitations of the Grønberg study, with an emphasis on radiation technical factors and how that may affect outcomes—both efficacy and toxicity.

The nitty gritty—radiation volumes, coverage, and constraints
Although cross-trial comparison is inherently flawed, the benefit seen in this study warrants scrutiny of the radiation planning and delivery in the context of other contemporary studies. The rigorous comparison may help explain the results and help radiation oncologists who may wish to utilize this regimen in select patients off study.
The treatment volumes used on the Grønberg study were remarkably similar to that of other recent randomized trials. Specifically, after the radiation oncologist defined the gross tumor volume (GTV), and it was mandatory to add a 5-mm circumferential ‘clinical target volume’ (CTV) margin to account for microscopic disease. An internal target volume (ITV) was then created, using the CTV plus an internal margin (IM) for tumor/target respiratory motion, based either on 4-dimensional computed tomography (4D CT) scan or a generic 5–10 mm. Finally, planning target volume (PTV) was a final expansion on the ITV and was institutionally dependent for setup margin (SM) (usually another 5 mm circumferential margin). The PTV thus may be 2 to 3× as large as the GTV (Figure 1).
To limit the total body volume irradiated, no attempt is made to electively irradiate nodal regions that are clinically negative by PET/CT. These margins are relatively similar to the margins used in other contemporary SCLC studies such as Yu et al. (6), CALGB (7), CONVERT (8), and NRG LU-005 (9) where CTV margins were 0.5–0.8 cm and PTV margins were 0.5–1 cm and margins were slightly larger if no 4D CT was utilized. While target definitions are very similar among the clinical trials, significant differences arise, however, when looking at PTV target radiation dose coverage requirements and recommended/mandated organ at risk (OAR) constraints, as shown in Table 1.
Some of the OAR dose constraints used on the Grønberg study are considerably more liberal than prior studies of BID radiotherapy, which had a lower target prescription dose and lower number of total fractions. For example, the NRG LU-005 protocol mandated esophagus maximum dose of 51.75 Gy and spinal cord maximum dose of 41 Gy—compared with 60 and 54 Gy, respectively, for Grønberg et al. (Table 1).
When interpreting trial toxicity data, it is critical to study radiation dose actually delivered rather than protocol specifications. This was reported by Levin et al. (5) which further strengthens the Grønberg study. The radiation dose to OARs is reported in Table 2.
With the protocol defined OAR constraints and resultant organ dosimetry, the dose-escalated regimen was very well tolerated and had similar toxicity to the standard regimen (21% G3 esophagitis, 3% G3–4 pneumonitis, and 1% grade 5 pneumonitis). Long term adverse effects included 2% esophageal stenting, 6% swallowing dysfunction, and 1% cognitive impairment. The most critical point here is that esophagus maximum point doses were high in the 60 Gy arm, with median 59.5 Gy to 0.05 cc and interquartile range (IQR) of 50–60.2. In other words, at least 75% of patients on the dose-escalated arm had maximum dose to the esophagus that exceeded the recommended maximum dose constraint on NRG LU-005 [albeit with slight biologically effective dose (BED) differences given the more protracted regimen on the present study], and still esophagitis rates were acceptable, possibly due to the relatively favorable mean doses reported (20.1 Gy with IQR of 13.5–27.3) as mean dose is known to be associated with high grade esophagitis (10).
One limitation of the study was that PTV coverage was not specified in the protocol. The same dosimetric study from Levin reported median PTV median dose was 100% (IQR, 99.3–100.1%), median D98% was 93.2% (IQR, 91.4–94.2%) and median D2% was 103.6% (IQR, 102.9–104.6%).
In personal communication with the manuscript authors, they were gracious to report additional PTV coverage metrics that are typically recorded in United States radiation oncology clinics (Figure 2).
Figure 2 shows that in the Grønberg study, only about half of the PTV is fully covered by the prescription dose. This contrasts with the U.S.-led LU-005 study and Yu et al. which required that 95% of the PTV be covered by the prescription dose. In our personal communication, the study authors stated that greater priority was placed on ensuring coverage of the CTV/ITV rather than the larger PTV volume. Specifically, the authors note differential coverage metrics for PTV in lung tissue (V90% ≥99.5%) and for PTV in soft tissue (V95% dose ≥98%) while maintaining median dose to the CTV of approximately 100%. Evaluation of Figure 2 shows that a median of approximately 95% of the PTV was covered by 95% of the dose.
These details are very important as they highlight significant RT planning differences across countries that may inadvertently affect the safe and effective delivery of dose escalated RT. In the U.S., most radiation oncologists prioritize dose coverage of the PTV—for example 95% (not 50%) of the PTV receiving 100% of the prescribed RT dose. RT plans used in the Grønberg study would thus often be rejected in the U.S., with radiation oncologists then facing the decision of: (I) increasing the coverage of the RT fields, thus resulting in higher doses to OAR’s and potentially higher toxicity than seen in the Grønberg study; or (II) decreasing the prescription radiation dose in order to maintain safe OAR doses but potentially losing the PFS/survival benefit of the dose-escalated Grønberg regimen.
Notably the in Grønberg’s standard arm (BID RT to 45 Gy), survival was lower than expected: median survival was 22.5 months compared with 28.7–39.5 months in recent trials. While this may be due to many factors, is it possible that one factor might be PTV underdoing in this arm given that the PTV coverage was similarly lower than many conventional standards? There is no easy way to determine this, though it would perhaps be helpful if Grønberg performed a detailed radiomic analysis of the location(s) of the local-regional failures in their study.

Statistically significant improvement in overall but not PFS
A surprising result of the study is the statistically significant improvement in median OS [43.5 vs. 22.5 months, hazard ratio (HR) 0.68, 95% confidence interval (CI): 0.48–0.98, P=0.04] without a statistically significant improvement in median PFS (18.6 vs. 10.9 months, HR 0.76, 95% CI: 0.53–1.08, P=0.13). Though not statistically significant, the magnitude of the PFS difference is clinically meaningful, and with the current HR signal it is likely that the PFS difference would be significant with more statistical power by way of a confirmatory phase III study. The authors point out a very fair point that since there was no central radiological review, cases of radiation induced fibrosis (which may be more prevalent in the dose escalated arm) may have been marked as progression events, falsely lowering the PFS in the dose escalated arm. As such it is important that any confirmatory trial consider central radiology review with a dedicated thoracic radiologist with expertise in reviewing post radiotherapy imaging. Interestingly the authors note that, for unclear reasons, patients in the standard arm had higher rates of multiorgan relapse. We pose that metastatic failure may be due to either micrometastases that predated the initial chemoradiation regimen and/or subsequent distant dissemination in the setting of uncontrolled local disease. As such, it is plausible that the higher dose regimen with improved local control may positively affect the burden of distant metastasis, even though the incidence of distant metastasis is the same due to pre-existing micrometastatic disease. In line with this, there were large numerical differences in median post-progression survival between the dose-escalated and standard arms (22.9 vs. 11.1 months, HR 0.5, 95% CI: 0.21–1.16, P=0.11).
The difference in magnitude of the PFS and OS improvements do not significantly affect our interpretation of the study. As an important historical comparison, the practice-changing Turrisi trial showed significant improvement in OS (P=0.04) without a significant improvement in failure-free survival (P=0.10) (2).

Challenges in implementing dose-escalated twice daily radiotherapy in clinical practice?
Irrespective of the limitations thereof, there are now two positive randomized studies for BID dose escalation in the SCLC space, recognizing that neither was powered at the level of a phase III study. Additionally secondary analysis of the subgroups of patients on LU-005 and ADRIATIC receiving BID radiotherapy (45 Gy) had improved OS compared to ‘dose-escalated’ QD (60–66 Gy) radiotherapy (9,11). We contend that there is more impetus than ever on at least offering BID radiotherapy for patients with LS-SCLC. A discussion of the biological rationale for bid radiotherapy in SCLC, a rapidly growing cancer, is beyond the scope of this article; those interested may refer to the Hall et al. (12).
There still remains, however, considerable difficulties adopting BID treatment universally (13), and recent National Cancer Database analysis showed only half of United States radiation oncology clinics has offered twice daily radiotherapy from 2004–2019 (14). This may reflect an opinion in the community that the patients seen in the ‘real world’ are more frail than those who enroll in clinical trials and thus less likely to tolerate BID radiation. It is important to note, however, that randomized comparison of QD vs. BID radiotherapy in the CONVERT and CALGB studies show no significant difference in toxicity between study arms. It is also possible that a number of patients are offered BID RT but decline due to the social/logistical difficulties coming twice daily for RT. We are hopeful that adoption of BID radiotherapy will increase given the accumulation of data supporting these regimens. National organizations may need to consider supporting this endeavor through policy development with financial considerations to help adoption of BID radiotherapy in community clinics; for example patient assistance with transportation and/or lodging (13).

Dueling dose-escalated BID radiotherapy regimens
It is important to note that Grønberg study regimen involves a fourth week of BID treatment, and it is clear from the aforementioned data that even delivery of 3 weeks of BID treatment to 45 Gy in the United States is difficult to adopt. Due to these logistical difficulties, if dose-escalated BID radiotherapy gains further support, it is possible that shorter regimens like the Yu et al. regimen (bid RT to 54 Gy over 3 weeks using a complex RT simultaneous integrated boost technique) may be preferred.
A hypothetical advantage of a shorter regimen is a reduced risk of lymphopenia. While the Grønberg et al.’s and Yu et al.’s studies did not report hematologic toxicity, CALGB data showed that a more protracted QD regimen to 70 Gy (7 weeks) caused greater lymphopenia than 45 Gy BID (7). Similarly, a recent study of concurrent chemoradiation for SCLC found worse lymphopenia with 60 Gy QD over 6 weeks compared to 45 Gy QD over 3 weeks (Bi et al., World Conference on Lung Cancer, 2025). Collectively, these findings suggest that longer treatment courses and higher cumulative doses increase the risk of lymphopenia—a toxicity associated with worse survival in non-small cell lung cancer patients receiving adjuvant durvalumab (15). As adjuvant durvalumab is now standard for LS-SCLC based on the ADRIATIC trial (11), minimizing lymphopenia should be considered even more relevant.

Is dose-escalated twice daily radiotherapy ready for routine implementation?
The results of the Grønberg et al. are impressive though must be validated in a phase III setting in the adjuvant immunotherapy era (11) to become a gold standard. Given the differences in RT planning and coverage between modern studies, it would also be of value to the radiation oncology community for any future phase III study in this setting to reach an international consensus on RT coverage goals for translation into the clinic.
If a phase III study of bid dose escalation is not available, can the Grønberg regimen be offered ‘off study’? We suggest that the 60 Gy in 40 fraction regimen may be reasonable to consider in highly select patients with exceptional performance status, organ function, as well as target volumes and OAR anatomy that will result in very favorable radiation dosimetry. Careful attention must be placed on not exceeding safe and tolerable doses and volumes of OAR’s including the spinal cord, lung, heart and esophagus when treating this curable disease. It is crucial not to cause high grade, prolonged radiation toxicity that may delay or prevent the delivery of post-radiation systemic therapy, which has been unequivocally proven to prolong survival in this terrible disease. To best replicate the dose-escalated regimen, critical attention must be given to the nuances described herein with respect to target coverage and OAR constraints.

Supplementary
The article’s supplementary files as