insert into cloud sql sql server column names with spaces in them

I am trying to load a CSV file in GCS into Cloud SQL SQL Server database using sqlalchemy and dataframe.to_sql

The problem I am running into is the columns in the database have spaces in them. I replaced the spaces with "_" character and out example looks like this.

Works (@From_Date NVARCHAR(MAX)) INSERT INTO dbo.calendar ([From_Date]) VALUES (@From_Date)

Fails (@From Date NVARCHAR(MAX)) INSERT INTO dbo.calendar ([From Date]) VALUES (@From Date)

Is there anyway to make it behaves like this (@FromDate NVARCHAR(MAX)) INSERT INTO dbo.calendar ([From Date]) VALUES (@FromDate) or this? (@From_Date NVARCHAR(MAX)) INSERT INTO dbo.calendar ([From Date]) VALUES (@From_Date)

Python Script

import os
import pandas as pd
import pytds
import sqlalchemy

from fast_to_sql import fast_to_sql as fts
from google.cloud import storage
from google.cloud.sql.connector import connector
from io import StringIO

column_names = [..., "From Date",...]

def init_connection_engine() -> sqlalchemy.engine.Engine:
    def getconn() -> pytds.Connection:
        conn = connector.connect(
            "instance_name",
            "pytds",
            user="sqlserver",
            password="",
            db="my_db"
        )
        return conn

    engine = sqlalchemy.create_engine(
        "mssql+pytds://localhost",
        creator=getconn,
    )
    engine.dialect.description_encoding = None
    return engine
# GCS client
storage_client = storage.Client(project=project_name)

# storage bucket connection
bucket = storage_client.get_bucket(bucket_name)
blob_file = bucket.get_blob(file_name)
gcs_file_byte = blob_file.download_as_string()
gcs_file_string = gcs_file_byte.decode()
file_data_string = StringIO(gcs_file_string)
calendar_dataframe = pd.read_csv(file_data_string, sep=',', header=None, usecols=[*range(0, 15)],
                                 names=column_names)

pool = init_connection_engine()

with pool.connect() as db_conn:
    calendar_dataframe.to_sql(table_id, db_conn, schema='dbo', if_exists='append', chunksize=None, index=False)