Richard W. Tsang

Radiation therapy for pituitary adenoma: Treatment outcome and prognostic factors

PURPOSE Radiation therapy is often an integral part of postoperative treatment in patients with nonfunctional pituitary adenomas. The Princess Margaret Hospital (PMH) experience was reviewed and analyzed to establish the role of radiation therapy in local control relative to its complications, and to see if subgroups of patients with a greater or lesser risk of recurrence postsurgery can be defined. METHODS AND MATERIALS Records of 160 patients with nonfunctional pituitary adenoma treated between 1972 and 1986 were reviewed retrospectively. The review focused on 128 patients treated with surgery and postoperative radiation as initial therapy. The median total dose was 45 Gy. Local tumor control was defined as lack of progression or recurrence of adenoma as assessed clinically and by imaging studies. The following factors were analyzed for prognostic significance in local tumor control: age, sex, direction of tumor extension, radiation dose, and preoperative tumor size as reflected by the radiation field size. Complications including hypopituitarism and second tumors were analyzed. Hypopituitarism was defined as requirement for permanent hormone replacement therapy. RESULTS With a median follow-up duration of 8.3 years, the 10-year actuarial local control rate was 87% for the entire 160 patients and 91% for the 128 patients given postoperative radiation as initial treatment. For the 29 patients referred for treatment of recurrent tumor, the 10-year local control rate was 78%. Prognostic factors for local control identified in univariate analysis included age (p = 0.005) and radiation field size (p = 0.0001). Older patients and those with larger tumors requiring large radiation portals were less likely to achieve durable local control. These two factors remained significant in a multivariate analysis (p < 0.005). The major complication, hypopituitarism requiring hormonal replacement with thyroxine, glucocorticoid, and sex hormone was observed to date in 65% (100 out of 155), 68% (105 out of 154), and 67% (85 out of 127) of evaluable patients, respectively. Radiation was the contributing cause of the hypopituitarism in only 23%, 16%, and 13%, respectively. There were no cases of brain necrosis or radiation damage to the optic pathways. Two patients developed a fatal in-field glioma of the brain stem at 10 and 15 years following radiation. CONCLUSION Postoperative external beam radiation therapy is highly effective in preventing recurrence of hormonally inactive pituitary adenomas. Hypopituitarism is commonly observed, but radiation can only be incriminated as the contributing cause in approximately one-fifth of the cases. Treatment of patients at the time of recurrence gave comparable local control rates to those irradiated initially. Favorable patients (age < or = 50, with small tumors removed totally) probably can be safely observed postoperatively with radiation reserved for recurrence.

ROLE OF RADIATION THERAPY IN CLINICAL HORMONALLY-ACTIVE PITUITARY ADENOMAS

BACKGROUND AND PURPOSE The outcome following radiation therapy (RT) of hormonally-active pituitary adenomas was assessed. The purpose of this analysis was to determine the control rate after radiation, identify any prognostic factors and evaluate the late toxicity. MATERIALS AND METHODS From 1972 to 1986, 145 patients received RT for hormonally-active pituitary adenomas. The median age was 39 years (range 15-76), with 81 males and 64 females. There were 52 patients with acromegaly, 64 with prolactinoma, and 29 with Cushings disease. The median follow-up was 7.3 years. RT was given as primary treatment in 17 patients, after initial surgery in 65 patients, and as part of salvage therapy in 63 patients. The median total dose was 50 Gy (daily fraction: 2 Gy). Tumor control was defined as normalization of basal hormonal level and lack of progression of adenoma assessed by imaging studies. The following factors were analyzed for prognostic significance in tumor control: age, sex, tumor type, direction of tumor extension, radiation dose, and radiation field size. RESULTS The 10-year actuarial proportion of patients with persistent elevated hormone level were 61% following RT alone, and 44% with the addition of medical management. The progression-free rate was 96% at 10 years. Of the 20 deaths, three patients died with uncontrolled pituitary adenoma and three died of treatment complications. The actuarial 10-year overall and cause-specific survival rates were 86% and 97%. The actuarial rates of radiation-induced hypopituitarism were 35%, 22% and 22% at 10 years for thyroid, glucocorticoid and gonadal functions, respectively. None of the factors examined were found to be significant predictors of tumor control. CONCLUSIONS Post-operative external beam RT is highly effective in preventing recurrence of space-occupying effects of hormonally-active pituitary adenomas. However, long-term biochemical remission is observed only in approximately 40% of patients (at 10 years), with an additional 20% requiring medical therapy. Malignancies of the CNS can develop as an infrequent late event.

External-beam radiation therapy in the treatment of differentiated thyroid cancer

The role of external-beam radiation therapy (EBRT) in differentiated thyroid cancer is reviewed. In the presence of gross residual disease after attempted surgical excision, retrospective series have reported local control is possible with EBRT. If, in addition to Iodine-131 (I131), there is a role for adjuvant EBRT in differentiated thyroid cancer, it would be only in patients in whom there is a high risk of relapse in the thyroid bed. Evidence is presented that suggests that EBRT can improve the local relapse-free rate in selected patients (over the age of 45, with microscopic residual disease or extensive extrathyroid invasion). For patients with recurrence in the thyroid bed, EBRT can be given in addition to surgery and I131. In bone metastases that are demonstrable radiographically, I131 therapy is often unsuccessful and EBRT also should be given. The technique of thyroid bed radiation is described. EBRT has acceptable acute toxicity and rarely produces serious long-term complications.

Squamous cell carcinoma of the lip: analysis of the Princess Margaret Hospital experience

We reviewed 117 patients with squamous cell carcinoma of the lip who were treated at the Princess Margaret Hospital between 1976 and 1985. Ninety-eight cancers arose from the lower lip, 18 from the upper lip and 1 from the commissure. Two patients had lymph node metastases at presentation. Sixty-one patients were treated with radiation therapy following a biopsy, 28 underwent surgery followed by post-operative radiation, and 28 had surgery alone. With a median follow-up time of 5.4 years, the 5-year actuarial overall and cause-specific survival rates were 81% and 99%, respectively. Local failure developed in 4 patients after radiation treatment, 3 of whom were salvaged by surgery. Six patients developed regional metastases after initial treatment, 4 of whom were salvaged with surgery and/or radiotherapy. Two patients died from lip cancer. After a univariate analysis, the only factor which predicted for nodal failure was T stage of the primary lesion, with a 4% risk of nodal failure for T1 lesions vs. 20% for T2/3 lesions (p = 0.03). No other patient, tumour or treatment variables influenced loco-regional control or survival in a statistically significant manner. Cosmetic and functional outcome were evaluated in 8 patients whose radiation treatments were administered 13 years ago. No patients had compromised lip function, and the majority had minimal cosmetic sequelae from their radiation therapy. Based on the excellent results of this review, we would continue to recommend radiation therapy as an effective treatment modality for patients with lip cancer because of the ease by which the entire tumour can be encompassed whilst maintaining excellent cosmetic and functional outcome.

Tumour proliferation and apoptosis in human uterine cervix carcinoma II: correlations with clinical outcome

PURPOSE The prognostic value of tumour proliferation and apoptosis measurements were studied prospectively in patients with carcinoma of the uterine cervix, relative to other established clinical factors. MATERIALS AND METHODS The labelling index (LI) for bromodeoxyuridine was determined by flow cytometry (fc) and also by immunohistochemistry. Apoptosis was assessed histologically using morphological criteria. Patients were treated with radical radiation therapy (RT). RESULTS The median/mean LI-fc were 6.7%/7.9% (range 1.52-3.9%). The median/mean apoptosis index (AI) were 1.0%/1.6% (range 0-6.8%). To date, 27 patients have died of disease, and the median follow-up for alive patients is 3.2 years (range 0.4-6.0 years). Among 64 patients who completely responded to treatment, 25 patients have relapsed (six pelvic, 17 distant and two pelvic and distant). In univariate analysis, the most significant factors for disease-free survival (DFS) were large tumour size (P=0.0001), low haemoglobin (P=0.01 ), LI-fc (DFS 67% for LI < 7%, 33% for LI > or = 7%, P=0.03), and T(pot) (DFS 66% for T(pot) > 5 days, 35% for T(pot) < or = 5 days, P=0.04) Stage, overall treatment time (OTT), S-phase fraction, ploidy, T(s), LI by histology, mitotic index, and AI were not significant. Multivariate analysis (Coxs model) showed that the only significant prognostic factors for DFS were tumour size and OTT. However, for small tumours (diameter < 6 cm), either a high LI-fc ( > or = 7%) or a high AI ( > 1%) was associated with poorer DFS, whereas patients with larger tumours (diameter > or = 6 cm) fared poorly regardless of LI-fc and AI. CONCLUSIONS Tumour size was the most important prognostic factor in cervix carcinoma. Although none of the biologic parameters have independent prognostic significance when the effect of initial tumour size was taken into account, our data suggests that LI and AI may be useful in discriminating outcome for patients with smaller tumours when managed by radical RT. These findings support the hypothesis that rapidly proliferating tumours are less likely to be controlled with a conventional course of RT.

Interrelationship of proliferation and hypoxia in carcinoma of the cervix.

PURPOSE In human cervix cancer treated with radiotherapy, we have previously shown from separate groups of patients that tumor hypoxia and proliferation rate as measured by bromodeoxyuridine (BrdU) labeling index (LI) are important determinants of clinical outcome. We now examine the relationship of these two pre-treatment predictive assays in 43 patients studied prospectively from 1994-98 where both tests were performed for each patient. MATERIAL AND METHODS Newly diagnosed patients with carcinoma of the cervix were examined under anesthesia for staging purposes. Patients were given BrdU (200 mg) by intravenous route prior to the procedure. Tumor oxygenation was measured with the Eppendorf pO2 histograph. Biopsy of tumor was then performed and the BrdU LI was obtained by flow cytometry. The degree of tumor hypoxia for each tumor was expressed as median pO2 values, and as the percentage of pO2 readings <5 mm Hg (HP5). RESULTS The median age was 53 years (range 23-79 years). There were 32 squamous, and 11 non-squamous carcinomas. FIGO stages were: IB and IIA, 8; IIB, 17; IIIB, 18; with a median tumor size of 6 cm (range 2-10 cm). The patients received uniform treatment with radical radiation therapy. There were 22 diploid and 21 aneuploid tumors. The median LI, pO2, and HP5 were 8.0%, 5.4 mm Hg, and 46.8%, respectively. Tests for linear associations showed no significant correlation between median pO2 vs. LI (r = 0.078, p = 0.62), and HP5 vs. LI (r = -0.14, p = 0.38). CONCLUSIONS The clinical outcome in this group of patients is immature, but these results suggest that tumor hypoxia and proliferation measurements are independent and potentially complementary predictive assays in cervix carcinoma. Further investigations are required to examine the distribution of proliferating tumor cells and its relationship with hypoxic tumor cells in tissue sections with the use of immunohistological techniques and image analysis systems.

Tumor proliferation and apoptosis in human uterine cervix carcinoma I: correlations between variables

PURPOSE Parameters for tumor proliferation and apoptosis were studied prospectively in 84 previously untreated patients with a diagnosis of carcinoma of the uterine cervix. MATERIALS AND METHODS Tumor proliferation was assessed by in vivo labeling with bromodeoxyuridine (BrdU), followed by a biopsy of the tumor 4-10 h thereafter during an examination under anesthesia. The potential doubling time (Tpot) was obtained by deriving the BrdU labeling index (LI) and S-phase duration (Ts) using flow cytometry. The LI for BrdU and its staining pattern were also determined immunohistochemically. Apoptosis was assessed histologically using morphological criteria. RESULTS Seven patients were excluded and the FIGO stages of the remaining 77 patients were as follows: IB and IIA, 20 patients; IIB, 29 patients; IIIB and IV, 28 patients. The median tumor diameter was 6 cm. There were 61 squamous cell, 11 adeno- and five adenosquamous carcinomas. Of the 63 patients in whom the tumor grade could be determined, 37 were well or moderately well differentiated and the remaining 26 were poorly differentiated. The median mitotic index (MI) was 0.7%. There were 43 diploid and 34 aneuploid tumors. Median values for Ts and S-phase fraction (SPF) were 9.9 h and 16%, respectively. The median BrdU LI by flow cytometry (LI-fc) was 6.7%. There was a significant correlation between LI-fc and LI by histology, although values for the latter (median 11.1%) were consistently higher than those determined by flow cytometry by a factor of 1.5. The median Tpot value was 5.0 days. The median apoptotic index (AI) was 1.0% and AI correlated positively with LI-fc. Median values for LI-fc increased with increasing tumor size and were 5.1%, 6.4%, 7.5% and 11.0% for tumors measuring < or = 4 cm, 4-6 cm, 6-8 cm and > 8 cm, respectively. The remaining proliferation parameters, however, showed no correlation with tumor size, stage, grade or histologic type. CONCLUSIONS In carcinomas of the cervix, tumor proliferation is positively associated with apoptosis and tumor size. These findings suggest that parameters for tumor proliferation and apoptosis are associated with tumor progression and may thus be predictive of clinical outcome.