diff --git a/vllm/lora/layers/fused_moe.py b/vllm/lora/layers/fused_moe.py
index 7c6d86b3602f..9a4da0619728 100644
--- a/vllm/lora/layers/fused_moe.py
+++ b/vllm/lora/layers/fused_moe.py
@@ -130,7 +130,6 @@ def _inject_lora_into_fused_moe(self):
                 prepare_finalize, self.base_layer
             ),
             self.base_layer.shared_experts,
-            getattr(self.base_layer, "shared_experts_stream", None),
         )
         if quant_config.use_mxfp4_w4a16:
             assert isinstance(