Abstract
Purpose: $^{177}$Lu-DOTATATE therapy is an effective treatment for advanced neuroendocrine tumors (NETs), despite its dose-limiting hematotoxicity. The significance of off-target splenic irradiation remains largely unknown. This study aims to identify predictive markers of peptide receptor radionuclide therapy (PRRT)-induced leukopenia. Methods: We retrospectively analyzed blood counts and imaging data of 88 patients with histologically confirmed, unresectable metastatic NETs treated with $^{177}$Lu-DOTATATE. Automated deep-learning-based segmentation was used to calculate spleen volumes from routine imaging. Results: Absolute and relative white blood cell (WBC) counts and relative spleen volume reduction were independent predictors of radiation-induced leukopenia at 24 months. A 30% decline in spleen volume 12 months after treatment most accurately predicted leukopenia at 24 months (AUC 0.91, sensitivity 0.93, specificity 0.90), significantly outperforming baseline blood counts. Conclusion: Automated splenic volume assessment is a valuable, quantitative imaging-based biomarker for predicting radiation-induced leukopenia, offering potential for improved clinical risk assessment.
Introduction
Peptide Receptor Radionuclide Therapy (PRRT), specifically using $^{177}$Lu-DOTATATE, has revolutionized the treatment of advanced neuroendocrine tumors. However, the therapy's efficacy is often constrained by off-target radiation to healthy tissues, most notably the kidneys and bone marrow. Hematotoxicity, resulting in leukopenia (low white blood cell count), is a major dose-limiting side effect.
Currently, clinicians rely on baseline blood counts to estimate risk, but these metrics often fail to identify patients who will develop late-onset toxicity. Because the spleen expresses high levels of somatostatin receptor subtype 2A (SSTR2A), it accumulates significant radiation during PRRT. This study investigates whether the resulting splenic volume reduction can serve as a "canary in the coal mine" for the broader hematopoietic system.
Methods and Materials
The researchers conducted a retrospective study of 88 patients with metastatic NETs treated between 2009 and 2021 at the Technical University of Munich. Inclusion required at least 24 months of follow-up and a minimum of four treatment cycles. Patients receiving blood transfusions or additional therapies during follow-up were excluded to isolate the effects of PRRT.
To ensure objective and efficient measurement, the team employed TotalSegmentator, an open-source deep-learning model, to automatically segment and calculate spleen volumes from routine CT and MRI scans. These volumetric changes at 12 months were then correlated with hematological parameters at 12 and 24 months.
Results
The study revealed a strong and statistically significant correlation between early spleen shrinkage and late-onset white blood cell decline. Approximately 33% of the cohort developed leukopenia by the 24-month mark. Patients who developed leukopenia exhibited an average spleen volume reduction of 36%, compared to only 19% in those who maintained normal blood counts.
Key findings include:
- The 30% Threshold: A 30% decline in spleen volume at 12 months was identified as the optimal predictor for leukopenia at 24 months.
- Superior Accuracy: This imaging biomarker achieved an Area Under the Curve (AUC) of 0.91, far exceeding the 0.68 AUC achieved by baseline WBC counts.
- Predictive Power: The 30% cutoff demonstrated a sensitivity of 0.93 and a specificity of 0.90.
Discussion and Conclusion
Splenic contraction likely reflects lymphoid depletion and fibrosis caused by targeted radiation. This volume reduction acts as a measurable proxy for the total radiation burden on the immune system. Unlike static baseline blood tests, automated volumetry captures the dynamic physiological response to ongoing therapy.
While this retrospective, single-center study requires prospective validation, the results suggest that automated splenic volumetry could be seamlessly integrated into clinical workflows. By identifying high-risk patients early, clinicians may be able to adjust treatment intervals or initiate protective measures before severe hematotoxicity occurs. Ultimately, spleen volume reduction represents a robust, routinely available, and highly accurate biomarker for managing long-term toxicity in PRRT.
Summary updated