An Evaluation of Total Internal Motions of Locally Advanced Pancreatic Cancer during SABR Using Calypso^® Extracranial Tracking, and Its Possible Clinical Impact on Motion Management

Round 1

Reviewer 1 Report

The authors carefully addressed all comments and suggestions raised, which significantly improved the manuscript, which is now acceptable for publication in the journal.

Reviewer 2 Report

The authors have addressed and corrected all my previous comments adequately

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

Major Comments

The manuscript addresses an essential clinical topic and is interesting for the field, and should be published in the journal after some revisions suggested within this review.

There are multiple language improvements required and indicated in the Minor Comments below. Also, clarifications need to be made for the reader to improve the ease of reading. These clarifications are listed in the Specific Comments below.

The authors claim (Lines 88 and 89) that the most accurate form of intrafraction motion management is achieved with implanted fiducials. This claim is incorrect as today, MR-Linac devices are in clinical operation that can monitor soft tissue during dose delivery. The authors are encouraged to weaken their statement and recognize the intrafraction motion monitoring capability of these MR-Linace devices.

Specific Comments

• Line 19: fiducial tumor motion tracking – consider writing fiducial-based tumor motion tracking
• Line 48: plays a crucial role – instead of plays crucial role
• Line 50: and it is critical – instead of and is critical
• Line 61: which also complicates image guidance – instead of which also complicates image guiding
• Line 63: reliable image guidance – instead of reliable image guiding
• Line 63/64: dose to the target – instead of dose to target
• Line 70: is also a possible – instead of is also possible
• Line 92: and time delay: it is not evident for the reader what is meant by time delay. In line 90, real-time tracking is mentioned using kilovoltage imaging. Why there is then a time delay. Please clarify.
• Line 129: scan instead of scans
• Line 133: T1 or T2 images – instead of phase
• Line 153: treatment instead of treatments
• Lines 153 – 156: Please specify for which OARs these constraints are used. For all OARs mentioned in Lines 152/153?
• Line 170: is a real – instead of is real
• Line 176: The system – instead of System
• Lines 203 – 205: This sentence is not clear for the reader. What you probably did is using the retrospective 4D-CT and breath-hold CT for analyses. First, it is not mandatory to talk about retrospective 4D-CT, as a typical 4D-CT is assumed to be retrospectively sorted according to the recorded breathing phase. You only need to mention this if you were also using prospective gating, which is not the case in this study as far as I can tell. In addition, there is not prospective and retrospective planning within Eclipse. However, you used Eclipse to identify the GTV motion using 4D-CT, created the ITV, and with the help of an Eclipse tool, the ITV based on the Calypso measurements (vITV). However, (dose/treatment)planning was not done, and therefore the reader might have difficulties understanding this sentence. Please clarify and rewrite accordingly.
• Line 209 – 210: t-MIP images: Typically, MIP is used for ITV generation. Not sure if t-MIP, which is not a well-known expression in 4D-CT for radiation therapy, is well recognized by the reader. t-MIP is a more often used expression in radiological procedures.
• Figure 3: Please also explain additional details for readers not experienced in reading these Calypso plots. Please consider addressing the gating threshold and the beam on indicators.
• Line 236: Virtual ITV: Consider calling it Calypso-based ITV comparable to your notification Calypso (or better Calypso-based) AR/CC/LR
• Figure 4: Please use a thicker line style for the contours. The contours are not well visible in the figure.
• Figures 5 and 6: use mean instead of average to be consistent in the figure and in the figure caption. In addition, write: with range max/min instead of with range.
• Line 285: Insert additional space
• Table 1 caption: cLR instead of lateralcLR
• Table 2: increase space between the last two columns
• Table 2 caption: left – right instead of left-right lateral
• Table 3: 0.05 instead of 0,05 and line break after from
• Line 323: in the work – instead of in work
• Lines 326-327: This sentence needs to be connected to the following sentence more clearly. Please rewrite.
• Line 330: In a recent – instead of In recent
• Line 331: 3 mm: was this an average of the surface, a region of interest, or a single point? Please clarify and rewrite.
• Lines 341 and 343: what does 100% motion mean? Please clarify and rewrite.
• Lines 354 – 355: This is unclear and different from the aim of the study written at the beginning of the abstract. Please consider rewriting.
• Lines 364-365: Does the 210% refer to the volume change ITV to vITV? What is a volume excursion? I guess you are referring to volume increase.
• Lines 366-367: Why a change of the breathing pattern (frequency, amplitude), stomach and intestine filling, and drift are not mentioned? Why is heartbeat mentioned? Is there a change from 4D-CT to the treatment phase?
• Line 372: the full – instead of full
• Line 378: at least at some time during treatment – instead of at least in some parts of the time during treatment
• Line 388: to all alternative – instead of to all; continuous tumor tracking (either with fiducials, Calypso, or MRI during treatment) is a powerful motion management tool. Therefore, one needs to differentiate this from other methods of motion management.

Reviewer 2 Report

Specific comments:

Introduction

„According to stages of the disease in total, resected patients have overall survival median of 12.6 months, unre-sected 3.5 months, and all patients 4.4 months, respectively [1].

This seems not a good reference for the overall outcome of pancreatic cancer patients. Although poor, it is not that poor. Resected patients usually have median OS times of 24 months in modern randomized trials and even metastatic patients have longer medial survival times if they can receive adaequate systemic therapy. Please cite modern data on this subject.

„Stereotactic ablative radiotherapy 42 (SABR) has emerged as an effective and safe form of local treatment for patients with 43 LAPC in recent years [4-8].“

This is generally true, but it should be mentioned that similar to chemoradiation the value of SABR in terms of prognosis is unclear, too.

„This technique is generally unpleasant, often painful for patients, and also pushes the OARs closer to the target volume.“

This statement is simply not true, abdominal compression is well tolerated without pain in the vast majority of patients, please refine the sentence

„The most accurate management currently available is intrafractional, 88 achieved with fiducials implanted into or nearby the tumor“

This sentence does not make sense to me, it seems that somethig is missig, please check

The introduction should include a paragraph with a description about the used fiducials (Calypso markers), which are different from usual fiducials. Their special features should be described at least briefly, because not all readers will be aware of them (especially as they are usually used in prostate cancer and in their anchored form in lung cancer) and their theoretical advantages should be mentioned in the introduction to describe the rationale for the study.

„Materials and Methods“

Please name the section „Patients and Methods“

„approved by the Institution’s Tumor board“ Does this mean that all patients were discussed in a multidiciplinary tumor board prior to SABR or does it mean the study was approved by the institutional review board ?

„LAPC was defined as unresectable, histologically proven pan-116 creatic adenocarcinoma or islet cell carcinoma“

How was unresectable defined ? According to prespecified criteria for example according to vessel involvement ? If so please state the definitions or an appropriate reference, if not (for example unresectability was simply defined by the surgeon in the board), please state for exampel that unresectability was defined by the surgeon based on his experience.

The paragraph on target volume definition and dose prescription is confusing. The authors mention that CTV was equal to GTV, which was defined on MRI and „further corrected“ on a contrast-free CT scan (including free breathing, deep breathhold and 4D-CT phases) . They further stated that „The plans were normalized to the mean dose to PTV“. Several coresponding questions:

I suppose that the MRI was not a 4D-MRI (otherwise it should be stated). Was it a free breathing MRI or a gated MRI (for example in breath hold technique) ? If so please describe, otherwise please explain what „correction“ of the GTV means. Did you create an ITV on the 4D-CT scans ? if so, how did you do that without contrast enhanced CT ? How was the PTV created ? Which margin was added, if any ? Please explain. How was dose exactly prescribed ? From the sentence above it seems that the dose was prescribed to the mean of the PTV, but later on the authors stated that the mean dose to the PTV was higher than the prescription dose ??

The authors stated their dose constraints for stomach, small intestine and duodenum and referenced two papers. First of all, the referenced evidence seems somewhat outdated as newer literature integrating the former papers exists (for example Hanna et al.). Second, the mentioned dose constraints do not fit the constraints from the references (for example ref 22 stomach max critical volume above treshold 10 cc, 1 fraction treshold dose D11.2 Gy, maximum point dose 12.4 Gy vs manuscript: Dmax V(0.03cc) < 23 Gy. Please explain why those dose constraints have been used and/or how they have been created.

Results and discussion

The authors created GTVs and ITVs on 4D-Cts and compared them with virtual ITVs derived from the intrafractional calypso measurements. I generally would question the possibility of an accurate delineation of pancreatic cancer target volumes on contrast-free CT scans, especially on reconstructed 4D-CT scans (see for example ESTRO ACROP guideline on target delineation in pancreatic cancer which clearly recommends contrast-enhanced CTs for target volume delineation). I further wonder why the authors have not simply measured the motion of the transponders themselves on the 4D-CT scans and compared those measurements to the „real“ motion measured by the Calypso system during RT. Please explain, comment and mention in the (missing) paragraph about the limitations of the study.

I do further question the assumption that the larger movements based on Calypso compared to the 4D CT scans are due to peristaltic movements or heartbeat induced. It seems more likely that the 4d-ct scans simply underestimated the real extent of the movement either because they represent only a one day image and respiratory motion is not that uniform as supposed to be or because of an insufficient ability (of the method) to correctly outline the GTV. Please comment and at least mention this possibility it in the limitatios paragraph.

I agree with the authors that tracking based on implanted fiduials may be (at least theoretically) the best method of motion management, however I do not feel that their statements regarding a possible benefit of single fraction treatments are supported by the presented data (especially regarding safety as no toxicity data is given at all, neither regarding CT-based placement of the fiducials nor SABR). Moreover, without efficacy and toxicity data, any conclusions regarding „safe“ dose escalation should be drawn more cautiously.

By the way: have fiducials been lost or moved during RT ?