Locally recurrent rectal cancer : aspects on management, surgery and outcome

Abstract

Locally recurrent rectal cancer (LRRC) is a severe condition associated with a high morbidity and mortality, affecting about 150-200 people/year in Sweden. After improvements in management and treatment of primary rectal cancer (RC), including an amelioration and standardisation of surgical technique, improved preoperative staging, addition of preoperative radiotherapy (RT), implementation of multidisciplinary team conferences and postoperative surveillance programmes, the rate of LRRC has decreased dramatically from 30-40% to 5-15%. A curative surgical resection is necessary for a favourable prognosis and five-year survival rates of more than 60% have been reported after curative surgery. However, when looking at all patients with LRRC, the survival rate drops below 10%. The aim of this thesis was to learn more about the characteristics, management and outcome of patients with LRRC in Sweden.

Paper I aimed at investigating whether the time interval from primary surgery of RC to diagnosis of LRRC had any impact on survival after LRRC diagnosis. Population-based data on patients operated for primary RC during the period 1995-2002 and with a diagnosis of LRRC as first event reported to the Swedish Colorectal Cancer Registry (SCRCR) were accessed from the registry. 386 patients were included, of which 25% had an early LRRC (ELR) and 75% had a late LRRC (LLR). Patients with a stage III primary tumour and non-irradiated patients were significantly more common in the ELR group. Factors that influenced survival were age, stage of primary tumour and surgical resection of the LRRC. Time to diagnosis of LRRC did not influence survival and should not be taken into account in treatment decisions of LRRC patients.

The aim of paper II was to assess what factors affect treatment intention in patients with LRRC. Medical records of patients defined by the same criteria as in paper I were collected and analysed. 426 patients were included, of which 35% had been treated with curative intent and 65% with palliative intent. Factors significantly associated with palliative treatment intention were age ≥80 years, presence of symptoms and a non-central location of the LRRC. The same factors also increased the risk of death among patients treated with curative intent. Five-year survival was 23% for patients treated with curative intent and 9% for all LRRC patients. The results present a congruence between factors predictive for treatment and prognostic factors, indicating an adequate selection for treatment.

In paper III, an analysis of time trends, regional variations and prognosis of the same patients as in paper II was performed. 28% of the patients were treated with tumour resection with curative intent. No significant time trends or regional variations were seen regarding treatment intention or resection margins after surgery for LRRC. The proportion of patients with non-central recurrences increased over time. Patients with a centrally located tumour were more likely to have a radical tumour resection. Fiveyear survival rates were 43% for patients treated with R0 resection and 14% after R1 resection. The results confirm that negative resection margins are crucial for a favourable prognosis in patients with LRRC.

In paper IV, details on surgical treatment of the LRRC in the patient cohort included in paper II and III were evaluated. 35% were treated with tumour resection, 19% had surgery without tumour resection and 20% were treated with best supportive care. Abdominoperienal resection was the most commonly performed resection procedure, performed in 49% of the abdominally resected patients. In total, 49% of the patients had a multi-organ resection and another 10% had a total pelvic exenteration. Complications were more common after tumour resection than after surgery without tumour resection, but the postoperative mortality was significantly higher among non-resected patients. Complications should be avoided, possibly by a better selection for surgery.