The predictive value of clinical evaluation of response to neoadjuvant chemoradiation therapy for rectal cancer

Multimodality therapy has become the standard treatment for patients with locally advanced (T3 and T4) rectal carcinoma. Accurate preoperative staging of the patients with rectal cancer has increased in importance because the selection of patients with transmural rectal cancer (T3 or T4) or node-positive disease leads to a previous nonsurgical neoadjuvant treatment. The purpose of this study was to evaluate the predictive value of the clinical response to neoadjuvant therapy on the basis of pathological results obtained on rectal cancer patients treated by chemoradiotherapy and surgery.

From 1994 to 2003, 58 patients with a primary diagnosis of rectal cancer were studied at our department and enrolled in a neoadjuvant protocol of chemoradiotherapy followed by surgery. All patients were treated by 30 days of chemoradiotherapy. At the end of the chemoradiotherapy, each patient underwent clinical examination, including digital rectal examination, proctoscopy and abdominal-pelvic computerized tomography to define the clinical response to the chemoradiotherapy. Surgical resection was performed in all patients three weeks after the end of chemoradiotherapy, and histological analysis was performed on all resected specimens.

The clinical complete response rate corresponded to the pathological complete response rate, whereas the clinical evaluation overestimated partial response and stable disease. The pathologic examination revealed that 3.5% of clinical partial responses and 3.4% of clinical stable disease were really pathological progressive disease. Clinical partial response and clinical stable disease positive predictive values were 92.8% and 90.9%, respectively, whereas the clinical progressive disease negative predictive value was 20%. Then, 6.9% of patients believed to have responded to the therapy, or not to have responded or worsened, actually had worsened by the end of the chemoradiotherapy.

Positive and negative predictive values, in particular for partial response and stable disease, of clinical evaluation of the response to chemoradiotherapy were not high enough to consider clinical evaluation accurate enough to make treatment decisions.