Four-Dimensional Image-Guided Adaptive Brachytherapy for Cervical Cancer: A Systematic Review and Meta-Regression Analysis

Fei Li, Dan Shi, Mingwei Bu, Shuangchen Lu, Hongfu ZhaoFrontiers in Oncology 2022 · DOI

TL;DR

This systematic review and meta-regression analysis investigates the effectiveness of 4D image-guided adaptive brachytherapy (4D-IGABT) for cervical cancer. The study establishes a clear dose-response relationship between the radiation dose delivered to the high-risk clinical target volume (HR-CTV) and local tumor control, providing crucial insights for optimizing treatment planning and improving patient outcomes.

What Problem Does This Paper Address?

Cervical cancer treatment, particularly locally advanced stages, relies heavily on brachytherapy to deliver high doses of radiation directly to the tumor while sparing surrounding healthy tissues. Traditional brachytherapy methods often struggle to account for the dynamic changes that occur within the patient's body during treatment. These changes include tumor shrinkage (regression), movement of internal organs (like the bladder and rectum), and variations in organ filling. These uncertainties can lead to under- or over-dosing of the target volume and organs at risk, potentially compromising treatment efficacy and increasing the risk of side effects. The development of 4D-IGABT, which incorporates imaging and adaptive planning throughout the treatment course, aims to address these challenges. However, a clear understanding of the dose-response relationship for local control in the context of 4D-IGABT was needed to establish optimal planning aims.

Methods

The researchers conducted a comprehensive systematic review of existing literature, searching databases like PubMed, Web of Science, and Cochrane Library for studies published up to February 2022. They focused on studies that reported the minimum dose delivered to 90% of the high-risk clinical target volume (HR-CTV D90) and the local tumor control rate in cervical cancer patients treated with adaptive brachytherapy.

A key methodological component was the use of a probit model for meta-regression analysis. This statistical model is well-suited for analyzing the relationship between a continuous predictor variable (in this case, HR-CTV D90) and a binary outcome (local control, which is either achieved or not). By applying this model, the authors aimed to quantify the probability of local control at different dose levels and to identify the dose required to achieve specific control rates (e.g., 90% or 95%). They also performed a similar analysis for the intermediate-risk clinical target volume (IR-CTV).

Datasets

The study analyzed data from 19 individual studies that met their inclusion criteria, encompassing a total of 3,616 patients. These studies provided the necessary data on HR-CTV D90, IR-CTV D90, and local control rates. No new datasets were created or introduced by this paper; rather, it synthesized existing data from published research.

Key Insights and Ideas

  • Dose is King for Local Control: The most significant finding is the robust and statistically significant relationship between the HR-CTV D90 and local control. This reinforces the fundamental principle that delivering an adequate radiation dose to the tumor is paramount for eradicating it.
  • Quantifying the Target Dose: The study provides concrete dose targets for achieving high rates of local control. This moves beyond qualitative recommendations to quantitative benchmarks that clinicians can use in their planning.
  • Adaptive Planning Matters: The focus on 4D-IGABT highlights the importance of adapting treatment plans to account for anatomical and physiological changes during therapy. This adaptive approach is what allows for more precise dose delivery.
  • No Clear Dose-Toxicity Link (for Grade 3+): Interestingly, the probit model did not find a significant relationship between the dose to organs at risk (OARs) and the probability of severe (grade 3 and above) late gastrointestinal or genitourinary toxicity. This is a complex finding that warrants further investigation, as OAR dose constraints are critical for minimizing side effects.

Key Results

  • A significant relationship was found between the HR-CTV D90 and local control probability (P < 0.001).
  • The HR-CTV D90 required for a 90% local control probability was estimated at 79.1 GyEQD2,10.
  • A prescribed dose of 85 GyEQD2,10 for HR-CTV D90 was theoretically associated with a 92.1% local control rate.
  • For the intermediate-risk clinical target volume (IR-CTV), the IR-CTV D90 corresponding to a 90% local control probability was 66.5 GyEQD2,10.
  • A planning aim dose of 65 GyEQD2,10 for IR-CTV D90 was expected to yield an 87.2% local control probability.
  • The probability of grade 3 and above late gastrointestinal, genitourinary, and vaginal toxicity was generally low, ranging from 0.0% to 11.6%, with no significant dose-response relationship identified for OAR doses.

Limitations and Future Work

The authors acknowledge several limitations:

  • Heterogeneity of Data: The included studies exhibited considerable heterogeneity in patient populations, treatment protocols, and follow-up periods, which can impact the precision of the meta-regression.
  • Incomplete Reporting: Some studies did not report all necessary data (e.g., initial disease stage, tumor width), preventing a more detailed subgroup analysis.
  • Keyword Limitations: The search strategy might have missed relevant studies if "adaptive" or "IGABT" were not explicitly mentioned in the title or abstract.
  • OAR Dose-Toxicity Relationship: The lack of a significant relationship between OAR dose and severe toxicity was unexpected and may be due to simplifications in dose calculation, the way toxicity was reported, or the limited number of studies with detailed OAR dose data.

Future work should focus on developing more robust methods for dose accumulation, particularly for OARs, to reduce uncertainty. Further research is also needed to better understand the dose-toxicity relationship, potentially by incorporating more detailed patient data and advanced imaging techniques. The authors emphasize the need to strictly adhere to OAR dose constraints in IGABT, even in the absence of a statistically significant dose-response relationship in their current analysis.

Summary updated

Abstract

Four-dimensional brachytherapy takes into account uncertain factors such as tumour regression, internal organ motion and organ filling, and provides a more accurate and more therapeutic ratio delivery through adaptive delineation and replanning, replacement of the applicator, and the addition of interstitial needles. The dose volume effect relationship of four-dimensional brachytherapy between the HR-CTV D90 and the local control rate provides an objective planning aim dose.
BibTeX
@inproceedings{li2022four,
  author = {Fei Li and Dan Shi and Mingwei Bu and Shuangchen Lu and Hongfu Zhao},
  title = {{Four-Dimensional Image-Guided Adaptive Brachytherapy for Cervical Cancer: A Systematic Review and Meta-Regression Analysis}},
  booktitle = {Frontiers in Oncology},
  year = {2022},
  doi = {10.3389/fonc.2022.870570},
}
Figure