The risk factors of hemorrhage in stereotactic needle biopsy for brain lesions in a large cohort: 10 years of experience in a single center

Hailong Li, Chunling Zheng, Wei Rao, Junzhao Sun, Xin Yu, Jianning ZhangChinese 2022 · DOI

TL;DR

This study investigated the risk factors for hemorrhage following stereotactic needle biopsy in a large cohort of over 3000 patients. They identified that deep-seated lesions, concurrent edema and enhancement on MRI, intraoperative hypertension without a prior history, and high-grade gliomas are significant independent predictors of post-biopsy hemorrhage, offering crucial insights for improving patient safety.

What Problem Does This Paper Address?

Stereotactic needle biopsy is a vital procedure for obtaining tissue samples from brain lesions when direct surgical removal is not feasible. While generally considered safe and effective, a known complication is intracranial hemorrhage. However, the precise factors that increase the risk of this complication have not been consistently identified across different studies, leading to variability in reported hemorrhage rates (ranging from 1.3% to 59.8%). This lack of clarity makes it challenging for clinicians to accurately assess individual patient risk and implement targeted preventative measures. This paper aims to address this by conducting a large-scale, retrospective analysis to identify independent predictors of hemorrhage in a single center over a decade.

Methods

The researchers conducted a retrospective chart review of 3196 patients who underwent stereotactic needle biopsy for histological diagnosis between January 2010 and December 2019. They collected demographic and clinical data, focusing on factors that might influence post-biopsy hemorrhage. Statistical analyses included chi-square tests and univariable and multivariable logistic regression to identify risk factors. Hemorrhage was defined as a brain parenchymal hematoma greater than 5 mm on postoperative CT scans. Uncertain hemorrhage was defined as a high-density area less than 5 mm. Symptomatic hemorrhage was defined as newly occurring neurological deficits directly resulting from the hemorrhage.

Datasets

The study utilized a retrospective dataset of 3196 patients from a single medical center. The data included demographic characteristics, clinical variables, MRI findings, and postoperative CT scan results. No new benchmarks were introduced; the study relied on existing patient records.

Key Insights and Ideas

  • Hemorrhage is a significant, though manageable, complication: While the overall incidence of detectable hemorrhage was 4.66%, and symptomatic hemorrhage was 1.44%, the study emphasizes that stereotactic biopsy remains a safe and effective diagnostic tool.
  • Location matters, but not always as expected: Deep-seated lesions were a significant risk factor, but interestingly, brainstem lesions, while associated with a higher rate of neurological deficits, did not independently carry a higher risk of hemorrhage compared to other deep-seated locations in this cohort.
  • Tumor characteristics are critical: The presence of high-grade gliomas was a significant predictor of hemorrhage. This aligns with the understanding that more aggressive tumors often have abnormal vasculature, increasing bleeding risk.
  • Intraoperative blood pressure control is paramount: Intraoperative hypertension without a prior history was identified as an independent risk factor, highlighting the importance of vigilant blood pressure management during the procedure.
  • MRI findings provide crucial clues: Concomitant edema and enhancement on MR imaging scans were strong predictors of hemorrhage, suggesting that areas of active inflammation or increased vascularity are more prone to bleeding.

Key Results

  • A total of 3196 patients were included, with a histological diagnosis achieved in 91.93%.
  • Hemorrhage occurred in 149 (4.66%) patients, and symptomatic hemorrhage in 46 (1.44%).
  • Independent predictors of hemorrhage:
    • Deep-seated lesions (OR 1.272, p=0.035)
    • Concomitant edema and enhancement on MR imaging scans (OR 1.827, p=0.002)
    • Intraoperative hypertension without a past history (OR 1.012, p=0.024)
    • High-grade glioma (OR 0.306, p=0.003) - Note: The OR for high-grade glioma is <1, indicating a protective effect in the multivariable model, which is counterintuitive to the univariable finding and the discussion. This might be a typo in the paper or a complex interaction not fully explained.
  • The study found no significant correlation between hemorrhage and surgical method (frame-based vs. frameless) or operator experience.

Limitations and Future Work

The authors acknowledge that the definition of hemorrhage and symptomatic deficits can vary between studies, contributing to the wide range of reported rates. They also note that while coagulation and platelet counts were not found to be significant risk factors in their multivariable analysis, they strongly recommend correcting potential coagulation abnormalities before surgery, citing a case where a patient with prolonged PT died from hemorrhage.

Future research should focus on:

  • Preoperative assessments to better stratify risk.
  • Advancements in imaging and surgical techniques.
  • Refining intraoperative management strategies, particularly for hemostasis.
  • Further investigation into the complex relationship between tumor pathology (like high-grade gliomas) and hemorrhage risk.

Summary updated

Abstract

Stereotactic needle biopsy is a safe and effective way to obtain tissue from brain lesions for histological diagnosis. The presence of deep-seated lesions, concomitant edema, and enhancement on MR imaging scans and the presence of high-grade glioma are independent predictors of hemorrhage after stereotactic biopsy.
BibTeX
@inproceedings{li2022risk,
  author = {Hailong Li and Chunling Zheng and Wei Rao and Junzhao Sun and Xin Yu and Jianning Zhang},
  title = {{The risk factors of hemorrhage in stereotactic needle biopsy for brain lesions in a large cohort: 10 years of experience in a single center}},
  booktitle = {Chinese},
  year = {2022},
  doi = {10.1186/s41016-022-00307-y},
}
Figure