Combining deep-inspiration breath hold and intensity-modulated radiotherapy for gastric mucosa-associated lymphoid tissue lymphoma: Dosimetric evaluation using comprehensive plan quality indices

Mucosa-associated lymphoid tissue lymphoma (MALToma) accounts for approximately 19% of all non-Hodgkin lymphomas and can arise at any extranodal site. In Korea, however, at least half of all MALTomas present as primary gastric lymphomas [1, 2]. Several institutions have reported excellent disease control with radiation therapy alone, supporting the use of modest doses in field radiotherapy (30–40 Gy) for patients with stage I–II gastric MALToma [3‐8], although systemic therapy may also be needed in cases involving unsuccessful Helicobacter pylori eradication or H. pylori-negative disease, depending on the disease stage.

The recent National Comprehensive Cancer Network (NCCN) guidelines recommend 30 Gy involved-site radiation therapy for gastric MALToma [9]; this generally involves the entire stomach and adjacent perigastric lymph nodes (if involved). Although this involved-site radiation therapy uses low doses, the doses to the organs-at-risk (OARs) near the stomach, such as the kidneys, small bowel, or liver, must be considered. To date, several planning techniques involving anterior-posterior/posterior-anterior fields (AP/PA) and 3-dimensional conformal radiotherapy (3D-CRT) have been used to reduce radiation exposure to the OARs to within tolerance limits, and intensity-modulated radiotherapy (IMRT) techniques for gastric MALToma have recently been introduced [10‐15]. Despite the dosimetric advantages of these modalities in terms of the doses to the OARs, concerns regarding higher scattered doses have not been resolved [16]. Additionally, the effect of respiratory motion on the accuracy of the treatment has raised concerns, thus limiting the widespread use of IMRT.

Some institutions have implemented the deep-inspiration breath hold (DIBH) technique to address these concerns. However, only a few planning studies have included modern radiotherapy techniques when evaluating the efficacy of this technique [11‐13, 17]. Therefore, in the present study, we performed a dosimetric comparison of 3D-CRT and IMRT plans using DIBH or free-breathing (FB) techniques to determine the most optimal treatment plan for gastric MALToma. We additionally compared IMRT plans using various techniques (step-and-shoot IMRT [_SIMRT], volumetric-modulated arc therapy [VMAT], and tomotherapy) to identify the most dosimetrically optimal plan, using the plan quality metric (PQM) to ensure an objective assessment.

In this study, modern radiotherapy plans combining IMRT and DIBH (VMAT-DIBH, _SIMRT-DIBH) were found to be significantly superior to 3D plans for gastric MALToma. We further observed that tomotherapy plans had exceptionally high MU values and yielded no notable benefits in this context. Although it remains unclear whether DIBH or IMRT is more useful in a clinical setting, VMAT-FB seems to yield better dosimetric outcomes when compared to 3D-DIBH. One of the strengths of this study is that we performed a dosimetric comparison among various modern radiotherapy techniques combining DIBH and IMRT and suggested the most appropriate treatment combination strategy regarding target coverage, OAR doses, and comprehensive evaluation for gastric MALToma.

Since the Memorial Sloan-Kettering Cancer Center (MSKCC) first described the treatment of gastric MALToma with RT alone [3], subsequent studies have shown favorable results in various populations, including patients with H. pylori-independent disease [4‐8]. However, significant variations in stomach size and shape, digestive movement, and respiratory motion are known to cause uncertainty during the simulation and delivery of treatment to the stomach, and a safe, efficient irradiation technique that can optimally overcome these uncertainties with adequate margins has not been well established. The simplest approach to this issue involves the addition of an adequate margin to the CTV. Previous studies of RT planning techniques for gastric lymphoma have recommended a PTV comprising the CTV plus a 1.5–2-cm margin in all directions [10, 12] for the delivery of radiation in a FB state. Other studies have used respiratory synchronized 4D-CT images to provide information about respiratory-induced organ motion during treatment planning and minimize motion uncertainties [13, 23]. Specifically, Matoma et al. [17] compared the usefulness of 4D-CT vs. a uniform margin for the treatment planning of gastric MALToma, and found that the former yielded a significantly smaller mean PTV volume, with significantly lower mean doses to the liver and heart. According to the International Lymphoma Radiation Oncology Group (ILROG) guidelines for the CTV [24], the entire stomach should be considered to harbor disease even if the tumor appears confined to one area, and abnormal or suggestive perigastric lymph nodes can be included in the CTV. Respiratory motion-induced changes in stomach position should be detected using 4D-CT simulation or fluoroscopy when determining the ITV, and an additional margin of approximately 1 cm is often added to the CTV for this purpose. The PTV (normally ~ 1 cm) should account for setup variations. Moreover, radiotherapy planning based on 4D-positron-emission tomography (PET)-CT/4D-CT together with online cone-beam CT might be helpful to define PTV margins used for optimizing individual target coverage and estimating interfractional or intrafractional gastric movement [25].

DIBH has become a standard cardiac-sparing technique during the treatment of left-sided breast cancer [26‐29], and several experts in lymphoma treatment centers have recently reported the use of DIBH for mediastinal lymphoma [30‐33]. In most patients, DIBH reduces heart and lung doses by elongating the heart, resulting in greater separation from the target volume and increasing the lung volume. In DIBH, smaller PTV margins could be applied to further accentuate the organ-sparing benefit. Therefore, this respiratory technique is being considered for use during the treatment of other organs (i.e., liver and stomach) at various specialized institutions, including ours. More recently, Wang et al. demonstrated that there could exist substantial interfractional variation in stomach volume despite treatment with breath-hold and restriction of oral intake, and daily CT image guidance RT (CT-IGRT), in combination with a DIBH, enabled better target coverage with even smaller PTV margins (0.5–1.0 cm) while treating gastric MALToma [34].

Several investigators have also described the use of IMRT for the treatment of mediastinal lymphoma [32, 35, 36]. As expected, IMRT improves target conformity and reduces OAR doses. However, this technique also increases the volume of low-dose exposure in tissues such as the lungs, heart, and breasts. Notably, the use of the IMRT or DIBH technique in the treatment of Hodgkin lymphoma results in better protection of the heart and lungs [30, 31, 37]. Therefore, the combination of IMRT and DIBH would be expected to further reduce the doses to the OARs. In addition to the information on the breath-hold technique included in the ILROG guidelines [38], more number of clinical attempts is ongoing to combine IMRT and DIBH effectively for Hodgkin lymphoma.

In contrast to the situation with Hodgkin lymphoma, only a few studies have evaluated the use of modern radiotherapy techniques for gastric MALToma. In the first planning study for gastric MALToma, conducted by the MSKCC to determine the most advantageous technique [10], the PTV was defined as the CTV plus a 2-cm margin to account for respiratory-induced movement of the stomach during FB. In a comparison of AP/PA, 3D-CRT, and IMRT plans, 4-field 3D-CRT markedly decreased the dose to the kidneys when this organ overlapped with the PTV, and the findings with IMRT plans suggested that the kidney and liver doses could be incrementally improved in selected patients. Two relevant studies were also reported by Korean researchers. Lim et al. [12] retrospectively compared 2 different planning techniques (2D and 3D-CRT) for gastric MALToma based on CT with a FB status; the PTV was defined as the CTV plus a 1–1.5-cm margin, and an additional 1-cm margin was added in the craniocaudal direction to compensate for respiratory-induced stomach motion. For 3D-CRT plans, AP/PA or 3–4 non-coplanar fields were used according to the physician’s preference. Although that retrospective study compared treatment outcomes rather than dosimetric planning, the 3D-CRT plans yielded significantly better PTV coverage, conformity, and kidney doses on both sides when compared to 2D-RT plans, without compromising the oncologic outcomes. Furthermore, Bae at al. [13] compared 5 planning techniques (AP/PA, 4-field, 3D-CRT, IMRT with only coplanar beams, and IMRT with a few non-coplanar beams). 4D-CT was conducted under a FB status, ITV was defined as the sum total of the entire stomach at every respiratory stage, and CTV was defined as the ITV plus a 1-cm margin. The authors observed the highest mean kidney and liver doses with the AP/PA plan and a 4-field plan, respectively. As observed in our study, Bae and colleagues observed better conformity and hepatic toxicity with IMRT plans, but found no significant difference between the coplanar and non-coplanar IMRT plans. Unlike our study, however, these previous studies were limited by a lack of comparison with the DIBH technique and with more recent IMRT plans (e.g., VMAT and tomotherapy). In our institution, we defined CTVs as the GTV plus a 1.5-cm margin for FB plans, and GTV plus a 1.0-cm margin for DIBH plans, without 4D-CT scans. There was only a small difference in CTV margins compared to that in other planning studies. Moreover, considering that this study was a dosimetric study for comparing optimal radiotherapy planning techniques in patients, the conclusion would not change and this seems reasonable compared to other studies.

Similar planning studies have been conducted abroad, although neither modern DIBH nor various IMRT planning techniques have been evaluated. A similar planning study that compared four-field 3D-CRT, half-field RT, and IMRT was performed by Japanese researchers [11]. In that study, planning CT was performed during shallow exhale and inhale phases; the CTVs from both phases were fused, peristalsis margins were added to obtain the ITV, and the PTV was generated by expanding the ITV by 1 cm in all directions. The IMRT and half-beam methods were found to reduce the doses to the kidneys and liver, compared to 3D-CRT. In a retrospective Chinese study [23], the dosimetric superiority and efficacy, toxicity, and quality of life were investigated in patients with gastric diffuse large B-cell lymphoma who received IMRT. IMRT was performed with a FB status, and the PTV was defined as the CTV plus 1–2 cm in all directions. The 5-year overall survival, progression-free survival, and locoregional control rates were 80, 75, and 93%, respectively, with excellent target coverage and long-term global and functional quality of life scores.

The application of DIBH to gastric lymphoma might differ significantly with respect to the dose and target location. The use of DIBH requires 10–20 min of coaching during the treatment simulation and prior to administration of the first fraction. A patient is expected to hold their breath for 10–20 s per respiratory cycle, during which radiation must be delivered. This limitation prolongs the daily treatment time by a few minutes, and the addition of IMRT can further affect the treatment time. Additionally, each treatment requires a very sophisticated protocol. However, patients with gastric lymphoma account for only a small proportion of departmental workloads, and therefore, these limitations may be outweighed by the improved treatment accuracy and reduction in late adverse events.

This study had a few limitations. First, the relatively small sample size might have limited our ability to make firm recommendations regarding the usefulness of DIBH. Second, as this was a treatment planning study, we could not demonstrate the oncologic outcomes or prevalence of secondary malignancies after long-term follow-up in the same cohort. Still, we note that almost 2 years have passed since our institution actually used this technique, and no treatment failures have been reported, although a long follow-up duration would be needed given the late disease recurrence of this disease entity. Third, the plan IQ-based scoring method only calculates the target and OAR doses using CT density, but does not consider the delivery modality. In fact, no precise radiotherapy plan scoring system uses subjective criteria to evaluate each objective item. In other words, it would not be possible to define an absolute score for each treatment plan, although relative comparisons among different plans are possible. Our results should be interpreted in consideration of the aforementioned points, and they need to be verified through further clinical studies.

Our findings demonstrate that modern radiotherapy plans combining DIBH with VMAT or _SIMRT were significantly more beneficial than 3D plans for gastric MALToma by saving the OARs and enhancing conformity, regardless of concerns about increased MUs. Although whether DIBH, VMAT, or _SIMRT is more useful in the clinical setting is unclear, VMAT-FB seems to be better than 3D-DIBH regardless of target margins. Further studies are needed to confirm the relevance of the most practical plan considering clinical outcomes.