Description and prediction of clinical radiosensitivity : emphasis on normal tissue reactions

Abstract

There is a range in the severity of normal tissue reactions observed when cancer patients receive radiotherapy (RT). The limits of dosage in RT have empirically been set by doses that have caused severe reactions in the most sensitive patients. The understanding of factors governing variations of radiosensitivity is therefore very important in RT. This thesis describes specific normal tissue effects after irradiation and explores factors behind this. In this work special emphasis has been put on sensitivity of the thyroid and the lung to RT.

RT of head and neck cancer frequently involves the thyroid gland and can cause thyroid dysfunction. The thyroid function was prospectively studied in a cohort of 391 patients with head and neck cancer admitted for external RT. The median follow-up was 19 months. Seventeen of 264 evaluable patients (6%) developed clinical hypothyroidism (increased TSH/decreased FT4) whereas chemical hypothyroidism (increased TSH/normal FT4) developed in 57 patients (22%). At 3 years after treatment, the actuarial risk of hypothyroidism was 15% (clinical) and 40% (chemical), respectively. The thyroid fraction included in the treatment volume was a significant risk factor (p = 0.041).

Using the cohort described above, we further studied the long-term incidence of clinical hypothyroidism in 308 evaluable patients. The median follow-up in this study was 4.2 years. The incidence of clinical hypothyroidism was 17%, and the median time to development of hypothyroidism was 1.8 years. The 5 and 10 years Kaplan-Meier actuarial risk of hypothyroidism was 20% and 27%, respectively. Multivariate analysis of data showed that bilateral neck irradiation (relative hazard [RH] 0.37, p = 0.02), pre-TSH value (RH 1.58, p < 0.001), addition of surgery (p < 0.001), and if surgery involved the thyroid gland (RH 4.74. p < 0.001) were significant risk factors. If surgery did not involve the thyroid gland, there was, however, no increased risk compared with no surgery (RH 1.44, p = 0.305).

Individual sensitivity of the thyroid to radiation may also be investigated in cohorts that have received radioiodine (131I). Several studies have attempted to determine factors that may predict response to 131I, but none have been adopted widely. We have examined if clinical features, thyroid 24-h 131I uptake test before and 12 weeks after 131I therapy and treatment factors may predict the response of 1290 patients with thyrotoxicosis. Multivariate logistic regression of pre-treatment factors showed that a low 131I uptake was a statistically significant covariate for remission after one 131I treatment (odds ratio 0.61,p < 0.001). Clinical type was also a significant variable in the multivariate assessment (overall, p = 0.001). The post-treatment 131I uptake test data suggest that an uptake of 20% or less is related to a favourable outcome after 131I therapy for thyrotoxicosis.

A well-known effect of exposure to RT is a cell cycle arrest in proliferating cells with a very characteristic accumulation of cells in G2 phase. In a study on 49 head and neck cancer patients admitted for RT, we evaluated radiation-induced cell cycle delay in lymphocytes and related it to treatment outcome. Cell cycle changes were measured in mitogen-stimulated peripheral blood lymphocytes before and after irradiation. Non-responding patients had a high level of S-phase cells compared with partial (p < 0.001) and complete responders (p = 0.016). An inverse relationship was found when analysing the fraction of cells in G2 (p = 0.009 and p = 0.034, respectively). Responding patients thus had an increased G2 phase accumulation in comparison with non-responders.

From a cohort of 177 breast cancer patients, 16 patients (9%) developed severe early lung reactions after RT. In these patients we measured radiation-induced cell cycle delays of lymphocytes and analysed them together with matched control patients. Cells from the control patients responded to a higher extent to mitogen stimulation than cells from the cases (p < 0.05). Analysis of S and G2 phase accumulation showed a significant dose-response effect (p < 0.001) but no significant difference between the two groups. We also studied whether there is a correlation with serum factors known to be involved in the pathogenesis of lung injuries. There was no difference between patients who did and did not develop pulmonary injury after RT with respect to plasma levels of TGF-01 or GSH as has been suggested by other authors.

When using the thyroid and the lungs for the analysis of side-effects of RT, we found factors like initial-TSH and 24-h 1311 thyroid uptake tests to be of significant importance for the outcome. It is likely that the background of the variations in sensitivity should be explored in the genetic polymorphism of individuals probably involving several genes.