TL;DR
This study investigates whether omitting axillary lymph node dissection (cALND) for breast cancer patients with sentinel node micro-metastases is safe. Using a large retrospective cohort and advanced statistical methods, they found that cALND omission is associated with an increased risk of recurrence and death, particularly in patients treated with breast-conserving surgery (BCS). This suggests that current guidelines may need refinement to better identify patients who can safely avoid cALND.
What Problem Does This Paper Address?
Sentinel lymph node (SLN) biopsy is a standard procedure to determine if breast cancer has spread to the lymph nodes. If the SLN is positive for cancer cells (micro-metastases or isolated tumor cells), the traditional approach has been to perform a completion axillary lymph node dissection (cALND) to remove all remaining lymph nodes in the armpit. However, cALND can lead to significant side effects like lymphedema and nerve damage.
Several clinical trials (like NSABP B-32, ACOSOG Z0011, and IBCSG 23-01) have explored whether cALND can be safely omitted in certain patient groups, particularly those with limited SLN involvement (isolated tumor cells or micro-metastases) and who are undergoing breast-conserving surgery (BCS) with adjuvant therapy. While some studies suggest non-inferiority (meaning the outcomes are not worse), the data is considered insufficient by current guidelines to universally avoid cALND for all patients with SLN micro-metastases. This paper aims to provide more robust evidence by analyzing a larger, multicenter retrospective cohort to better understand the impact of cALND omission on long-term survival outcomes.
Methods
The researchers employed a retrospective, multicenter cohort study design. To address the inherent biases in observational studies (where treatment assignment is not randomized), they utilized Inverse Probability of Treatment Weighting (IPTW). This statistical technique creates a "pseudo-population" where the treatment groups (cALND vs. no cALND) are balanced on observed confounding factors. This allows for a more robust estimation of the causal effect of cALND omission on survival outcomes.
Specifically, they:
- Derived individual propensity scores using logistic regression, predicting the probability of receiving cALND based on various prognostic and treatment characteristics.
- Applied stabilized IPTW (sIPTW) to create weights for each patient, ensuring that the weighted groups were comparable.
- Calculated adjusted Kaplan-Meier (AKM) survival curves using these weights to estimate disease-free survival (DFS) and overall survival (OS).
- Used weighted Cox proportional hazards regression (sIPTW-Cox) to quantify the magnitude of the causal effect of cALND omission.
- Conducted subgroup analyses based on propensity scores to explore differential effects in specific patient groups.
Datasets
The study utilized a multicentric retrospective database comprising 23,134 patients who underwent breast cancer surgery across 15 French academic centers. For this specific analysis, they included 1451 patients who were younger than 75 years old, underwent breast-conserving surgery (BCS), had sentinel lymph node micro-metastases (defined as >0.2mm to ≤2mm), and did not receive neoadjuvant therapy. Patients treated with mastectomy or those with macro-metastases in the SLN were excluded.
Key Insights and Ideas
- The "good prognosis" caveat: While previous studies suggested non-inferiority for cALND omission, this study highlights that "good prognosis" might not be a universal shield against negative outcomes when cALND is omitted, especially for BCS patients.
- IPTW as a powerful tool: The use of IPTW is crucial here, moving beyond simple comparisons to estimate the causal impact of cALND omission, accounting for complex confounding factors in real-world data.
- Time-dependent effects: The divergence in survival curves suggests that the negative impact of cALND omission might become more pronounced over longer follow-up periods.
- Subgroup vulnerability: The most significant negative impact of cALND omission was observed in a specific subgroup of women who were less likely to miss cALND (lower propensity scores), suggesting these patients might be more susceptible to adverse outcomes.
Key Results
- Overall Survival (OS): cALND omission was associated with a significant increase in the risk of death (HR = 2.77, CI95% = 1.36-5.66) in the overall cohort. Adjusted Kaplan-Meier curves showed a significant difference in OS (p=0.002).
- Disease-Free Survival (DFS): A borderline significant association was found between cALND omission and an increased risk of recurrence (p=0.090). Adjusted Kaplan-Meier curves showed a borderline significant difference in DFS (p=0.090).
- Subgroup Analysis: In the subgroup of patients with lower propensity scores (ps ≤ 0.13), cALND omission was associated with significant differences in DFS (p=0.031) and OS (p=0.004). The hazard ratios for DFS and OS were 2.09 and 3.23, respectively, in this subgroup.
Limitations and Future Work
- Retrospective Design: The study is retrospective, meaning it relies on existing data, which can have limitations in data completeness and potential biases. The authors acknowledge the presence of missing data, particularly for HER2 status, and that 24% of the initial cohort was excluded due to missing data for the propensity score model.
- Generalizability: While multicenter, the study focused on patients <75 years old treated with BCS and SLN micro-metastases. Results may not directly apply to older patients, those treated with mastectomy, or those with SLN macro-metastases.
- Proportional Hazards Assumption: The authors note that the proportional hazards assumption was violated, which is a common challenge in survival analysis and was addressed by using weighted Kaplan-Meier curves for direct comparison.
- Future Work: The authors emphasize the need for results from ongoing randomized trials (like the SERC trial) and meta-analyses to confirm these findings and establish a stronger consensus for omitting cALND. They also highlight the importance of avoiding axillary dedicated treatment and its associated morbidity, especially for patients who might benefit from systemic therapies.
Summary updated