Gastroresistant gelatin films prepared by addition of cellulose acetate phthalate.

Gastroresistant capsules are obtained mostly by using modified-release fill in hard capsules, or by coating the gelatin shell with acid-resistant polymers. Modification of the material used at the stage when the capsule shell is produced would reduce the complexity and cost of introducing new products to the market. Gastroresistant gelatin films were obtained by using commercial cellulose acetate phthalate (aqueous dispersion Aquacoat® CPD). Only films casted from non-alkalized mixtures showed no visible disintegration at pH from 1.2 (simulated gastric fluid) to 4.5 (phosphate buffer). Elasticity of the dry films was comparable with the one determined for non-modified gelatin films, however tear resistance was 2-fold smaller, but still acceptable for practical application.

Split-course accelerated hyperfractionated irradiation (CHA-CHA) as a sole treatment for advanced head and neck cancer patients-final results of a randomized clinical trial.

OBJECTIVE: Evaluation of the efficacy and toxicity of split-course accelerated hyperfractionated irradiation (CHA-CHA) as a sole treatment for advanced head and neck (H&N) cancer patients. METHODS: We enrolled 101 patients (39 in CHA-CHA and 37 in conventional (Conv.) arm completed the treatment). The CHA-CHA arm patients were irradiated twice a day, 7 days a week, using a fraction dose (fd) of 1.6 Gy up to 64 Gy with an 8-day gap in midterm. Patients in the control (Conv.) arm group were irradiated with a fd of 2 Gy, five times a week to a total dose of 72-74 Gy in the overall treatment time of 50-53 days. Quality of life (QOL) and acute mucosal reaction were evaluated during radiotherapy (RT). After RT, we followed the effect of treatment, QOL, performance status and adverse effects of radiation. For statistical analysis mainly a hierarchical multilevel modelling was used. RESULTS: QOL was most deteriorated in the CHA-CHA arm; the CHA-CHA scheme also caused a relatively stronger acute injury. There were no significant differences in late adverse effects. In the CHA-CHA arm in 35% and in Conv. arm in 30% of patients, disease was controlled during follow-up. Tumour regression 1 year after the treatment was significantly better in the CHA-CHA arm. However, the overall survival rate analysis did not show significant difference between both arms. CONCLUSION: Despite differences in treatment results, we cannot conclude that split-course accelerated hyperfractionated irradiation is superior to conventionally fractionated RT as a sole treatment for advanced H&N cancer patients. ADVANCES IN KNOWLEDGE: Obtained results in the context of published data support the statement that altered fractionations alone do not give an advantage for advanced H&N cancer patients.

Prognostic role of tumor volume for radiotherapy outcome in patient with T2 laryngeal cancer.

BACKGROUND AND PURPOSE: Tumor volume (TV) is recognized as a prognostic factor of treatment outcome for head and neck tumors but is not routinely included in the treatment decision-making process. The purpose of the study was to define its prognostic role for patients with T2 laryngeal cancer. MATERIAL AND METHODS: TV of 160 patients who underwent RT between 2002 and 2006 for T2 laryngeal squamous cell carcinoma were reviewed. The tumor was located in the glottis and epiglottis in 82 (51%) and 78 (49%) patients, respectively. TV was manually contoured on pretreatment, planning, contrast-enhanced CT scans and the volumetric measurement (cm3) was calculated by the volume algorithm. RESULTS: The median TV value was 2.01 cm3 (range 0.15-21.68 cm3). The median TV was significantly lower in patients with glottic tumors (p<0.0001), N0 (p<0.001), or well histopatologically differentiated tumors (p=0.01). A significant correlation between TV, hemoglobin concentration (p<0.01), and total dose (TD; p<0.001) was observed. On univariate analyses, TV influenced local control (LC; p=0.02) and overall survival (OS, p<0.001). On multivariate analysis, both age (HR 1.038, p=0.03) and TV (HR=1.075, p=0.01) remained significantly related to LC and OS (age: HR 1.038, p=0.005; TV: HR 1.097, p=0.0001). CONCLUSION: Large TV worsen prognosis of patients with T2 laryngeal cancer. A large TV is more common for supraglottic, poorly differentiated tumors and may suggest higher risk of nodal spread. The routine estimation of TV prior to therapy may be essential in order to select the best treatment option for patients with T2 laryngeal cancer.

Predictors of radiotherapy outcome in patients with T2 supraglottic carcinoma.

The decision regarding treatment of early supraglottic carcinoma remains controversial. Single institution clinical data of patients with T2 supraglottic carcinoma treated exclusively with radiotherapy in terms of prognostic factors and treatment results were analyzed. Patient-related factors that would potentially by useful for optimal therapeutic decision to be undertaken were especially investigated. Between 1994 and 2004, 78 patients with T2 supraglottic carcinoma underwent radiotherapy (RT) in Maria Sklodowska-Curie Memorial Cancer Centre and Institute of Oncology, Gliwice Branch, Poland. There were 54 (69%) males and 24 (31%) females in the median age of 61 years. There were 17 (22%) patients with N+. Median body mass of patients before (BM0) and after RT (BM1) was 74 kg (range 45.2-130 kg) and 72.9 kg (range 49.9-122.5 kg), respectively. Median hemoglobin concentrations before (Hb0) and after (Hb1) RT were 14.3 and 13.4 g/dl, respectively. Median change of Hb concentration during RT (dHb) was -0.8 g/dl. All were treated up to total doses (TD) ranged from 62.5 to 72 Gy. The overall treatment time (OTT) ranged from 30 to 70 days. Estimates of local control (LC), ultimate local control (uLC), and overall survival (OS) were calculated using the Kaplan-Meier method. Log rank statistics, Cox proportional hazard model and step-wise Cox regression hazard model were employed to identify prognostic factors for LC, uLC, and OS in univariate and multivariate analyses. The 5-year LC, RC, uLC and OS rates were 85, 92, 88, and 56%, respectively. In multivariate analysis N+ (p = 0.01) and prolonged OTT (p = 0.03) significantly decreased LC. Females (p = 0.02), higher BM0 (p = 0.03), and HB0 (p = 0.006) significantly prolonged OS. Patient-related factors like gender, hemoglobin concentration, and body mass may predict treatment outcome. Radiotherapy is effective for T2 supraglottic carcinoma of the larynx unless higher dose intensity is provided. Involved regional lymph nodes significantly deteriorate locoregional cure.

Split-course accelerated hyperfractionation (CHA-CHA) for advanced head and neck cancers -- preliminary results.

The purpose of the study was to evaluate toxicity and effectiveness of the split course, accelerated hyperfractionation with a planned gap (CHA-CHA) on a base of the pilot group treatment results. The material consisted of 27 patients with advanced (T2N2c-3, T3-4N2), oral cavity, oro- hypopharynx and larynx cancers, irradiated with a df 1.6 Gy twice a day, with 6-hour gap, up to TD 64 Gy in 28 days, with 8-day gap in a middle. Mean tumor dimensions were 4.2 and 3.9 cm in X and Y axes, respectively and mean nodes dimensions 3.9 cm in both axes. The course and intensity of the mucositis were evaluated in the Dische scale together with the degree of a tumor and nodes regression. Survival analysis and dependencies evaluation between physical/biological factors and treatment results were performed. The mucositis intensity was high from 2nd to 5th week (the highest mean - 15 Dische points in 4th week), but on acceptable level. 54.5% of CR of the tumor and 57.1% of nodes were noted. Survival analysis showed 44% of a 2-year and 35% of a 4-year OS and 44% of a 2-year and 38% of a 4-year survival excluding deaths from distant metastases. Correlation between the hemoglobin level and the nodal regression two months after the radiotherapy (p=0.01), showed the tumor diameter and TCP (p=0.047), the OS and the nodes diameter (p=0.004), the PS (p=0.02) and the hemoglobin level (p=0.04). CHA-CHA could be a valid method in advanced head and neck cancer treatment but its efficacy should be confirmed in randomized phase of trial.

[Unconventional accelerated irradiation of advanced pharyngolaryngeal cancer as an alternative treatment compare to surgery. The interim results of randomized clinical trial]. / Przyspieszona niekonwencjonalna radioterapia zaawansowanego raka krtani i gardla dolnego jako alternatywa dla leczenia operacyjnego. Wczesne wyniki kontrolowanego doswiadczenia klinicznego.

The paper presents preliminary results (3 year tumour control and survival) of clinical randomized trial comparing accelerated fractionation (7 fx a week) with conventional (5 fx a week) radiotherapy (RT) for advanced pharyngo-laryngeal cancer. Significant treatment gain (p < 0.001) and good tolerance associated with accelerated RT was noted (82% of disease-free and 79% of overall survival). These results are similar with combined (total laryngectomy + postoperative RT) policy and, additionally, accelerated RT offers a high chance of larynx preservation.

Randomized clinical trial on 7-day-continuous accelerated irradiation (CAIR) of head and neck cancer - report on 3-year tumour control and normal tissue toxicity.

PURPOSE: To evaluate tumour and normal tissues 3-year response to 7-day-a-week continuous accelerated irradiation (CAIR) compared to a conventional treatment (5 days per week) in a randomized trial. MATERIALS AND METHODS: One hundred patients with squamous cell carcinoma of the head and neck in stage T(2-4)N(0-1)M(0) were entered into the trial between December 1, 1993 and June 30, 1996. Dose per fraction of 2.0 Gy (to the end of 1994), and 1.8 Gy (since January 1, 1995) was the same in both arms and delivered once a day at regular 24-h intervals to total dose in the range of 66-72 Gy (depending on tumour stage). The only difference was overall treatment time being 5 weeks in the CAIR and 7 weeks in control arm. RESULTS: Actuarial 3-year local tumour control was 82% in the CAIR and 37% in the control group (P<0.0001) with reduction in local recurrence rate of 83%. Actuarial 3-year overall survival was 78 and 32% (P<0.0001), respectively. Confluent mucositis was significantly more severe and lasted longer in the CAIR than in control arm. After 2.0 Gy fractions five of 23 patients (22%) in the CAIR developed early necroses over a period of 2-4 months of follow-up which can be considered as a consequential to severe protracted acute mucosal reactions (CLE). For this reason dose per fraction was lowered to 1. 8 Gy and the CLE was not observed again until now. Thus the overall rate of CLE decreased to 10%. CONCLUSIONS: The gain in tumour control is likely the effect of shortening of overall treatment time by 14 days and regular continuous dose delivery during the whole course of radiation therapy including weekends. A 7-day schedule produces more severe acute mucosal reactions lasting longer than in conventional fractionation, however tolerable by patients. Relatively high rate (22%) of CLE in the 7-day arm observed during the first year of the study was eliminated by decreasing dose per fraction from 2.0 Gy to 1.8 Gy.

Role of short TE 1H-MR spectroscopy in monitoring of post-operation irradiated patients.

Post-surgical radiation therapy is a routine procedure in the treatment of primary malignant brain tumors. Along with modest therapeutic effects conventional fractionated radiotherapy, in spite of any modifications, produces damage to non-malignant brain tissues lying within the treatment volume, the extent of which depends on radiation dose. Serial 1H-MRS allows non-invasive investigation of tissue metabolic profiles. In the present study the ratios of resonance signals assigned to the major 1H-MRS-visible metabolites (N-acetylaspartate, choline, creatine, inositol, lactate and lipid methylene group) were evaluated before, during and after post-surgical fractionated radiotherapy in brain regions close to and more distant from the tumor bed, receiving different radiation exposures (60 and < 40 Gy, respectively). The study group consisted of ten patients (aged 28-51). A MRI/MRS system (Elscint 2T Prestige) operating at the field strength of 2 T and the proton resonance frequency of 81.3 MHz has been used and the 1H-MR spectra were acquired using single voxel double-spin-echo PRESS sequence with a short TE. The spectra were post-processed with automatic fitting in the frequency domain. It was found that although the metabolite profiles depend on the dose obtained, but other stress factors (like surgery) seem to contribute to the overall picture of the metabolic status of the brain as well. In studies of early irradiation injuries, an increase of choline related ratios may serve rather as cell proliferation indictors than as cell injury ones, whereas the mI/Cr ratio appears as one of the first indicators of local irradiation injury. In order to establish the prognostic marker for early radiation damage, however, it seems necessary to analyze all visible metabolites as well. None of the metabolites separately may serve as such an indicator due to the complexity of tissue metabolism. Interestingly, MRI reveals no changes during the therapy process, whereas the metabolite ratios are being affected in the course of time, thus supporting the presumption that the 1H-MRS is a valuable method of radiation therapy monitoring.

Comparative estimation of cure rates for supraglottic and glottic cancer in radiotherapy.

There are some clinical evidences, that the same types of tumors originated from neighboring anatomical structures can significantly differ in their response to radiation therapy. Squamous cell cancer of supraglottis and glottis could be good examples of this phenomenon. The purpose of the study was to compare the radiocurability of cancers localized in the upper and medium level of the larynx. From 1985 to the end of 1989, 544 patients with squamous cell cancer of the larynx were treated by radiotherapy alone. There were 388 patients with supraglottic cancer and 156 patients with glottic cancer. The total dose was in the range of 59-74 Gy. The end-point criteria were overall (OS) and disease-free survival (DFS). Generally, 5-year overall and disease-free survival rates were significantly more favorable for glottic cancer patients than for supraglottic cancer (67 and 63% vs. 40 and 36%, respectively). Significant differences in both disease-free and overall survival between supraglottic and glottic cancer in aspect of several analyzed clinical prognostic factors were found for: male sex, age, pattern of tumor growth, clinical performance status, radiation total dose lower than 70 Gy, fraction doses and overall treatment time. In all these prognostic categories 5-year survival rates were lower for supraglottic cancer patients. This tendency disappeared when the treatment results were compared in aspect of tumor stage (T). Tumor cure doses for 50% probability of local control (TCD50) in supraglottic cancer were estimated as: 61 Gy (T(1+2)) and 66 Gy (T3). In glottic cancer the lower TCD50 values of 54.5 Gy (T(1+2)) and 61 Gy (T3) were found in comparable treatment time. The comparative estimation of cure rates (i.e. OS and DFS) of laryngeal cancer treated by radiation alone showed that in aspect of almost all analyzed prognostic factors the greater risk of treatment failure was significantly associated with supraglottic origin.

Clinical radiobiology of glottic T1 squamous cell carcinoma.

PURPOSE: Radiation therapy is the treatment of choice for early glottic squamous cell cancer in many institutions over the world. Despite a relatively homogenous clinical model of T1 glottic tumors for the fractionation studies, the relationships between dose-time parameters remain unclear. To analyze the influence of fractionation parameters and hemoglobin level on tumor cure, this study has been performed. MATERIALS AND METHODS: This is a retrospective review of 235 patients with T1N0M0 glottic cancer treated by radiation therapy alone given in a conventional schedule with 5 fractions each week. The individual total dose, dose per fraction, and overall treatment time (OTT) ranged from 51-70 Gy, 1.5-3.0 Gy, and 24-79 days, respectively. The median follow-up was 48 months. Patient data--total dose, dose per fraction, OTT, and hemoglobin level (Hb) measured before the radiation treatment--were fitted by the mixed LQ/log-logistic model. RESULTS: The 5-year local relapse-free survival rate was 84%. All parameters included in the mixed LQ/log-logistic model improved the fit significantly. The dose-response curve for 235 patients with T1 glottic cancer was well defined and steep, and showed significant decrease in tumor control probability (TCP) when total doses were below 61 Gy. The 10-day prolongation of OTT, from 45 to 55 days, decreased the TCP by 13%. The dose of 0.35 Gy/day, compensated repopulation during the 1 day of prolongation, which indicates a potential doubling time (Tpot) for glottic T1 tumor clonogens of 5.5 days. The drop of Hb level of 1 g/dl (from 13.8 g/dl to 12.8 g/dl) gave a 6% decrease of TCP, provided that OTT was 45 days. CONCLUSION: The significant correlation between the total dose, overall treatment time, hemoglobin concentration, and tumor control probability has been found for T1 glottic cancer.

Dose-response relationship for elective neck irradiation of head and neck cancer--facts and controversies.

The aim of this study is to assign dose-response relationship for subclinical neck metastases of squamous cell head and neck cancer based on extensive survey of 24 data sets collected from the literature. Neck relapse rates (NRR) without and after elective (ENI) or preoperative irradiation were estimated for each site and stage of primary tumor and the reduction in neck relapse rate was calculated. An average NRR without ENI was 22% (12-35%) and only 2.5% (0-10%) after the ENI with total dose of 46-50 Gy which gives high reduction rate in the risk of neck recurrences being on the average 89% and 42% (0-46%) after preoperative irradiation using 22-30 Gy. Dose response curve for elective and preoperative irradiation have shown that 50 Gy in 2 Gy fraction reduces the incidence of neck relapses in the N0 patients by more than 90% and only by less than 50% after total doses lower than 30 Gy. No correlation between the risk of neck metastases without ENI and the reduction in neck relapses after ENI was found.

Prognostic value of hemoglobin concentration in radiotherapy for cancer of supraglottic larynx.

PURPOSE: The aim of this work is the estimation of correlations between hemoglobin concentration either before or after radiotherapy and local tumor control probability for laryngeal cancer. METHODS AND MATERIALS: Retrospective analysis of 847 cases of laryngeal supraglottic squamous cell carcinoma treated with radiation alone was performed using maximum likelihood estimations, and step-wise logistic regression. All patients were in good initial performance status (Karnofsky index >70). The minimum follow-up time was 3 years. RESULTS: Logistic regression showed that the hemoglobin concentration after radiotherapy is an important prognostic factor. There was a very strong correlation between hemoglobin concentration and tumor local control probability. Hemoglobin concentration at the beginning of radiotherapy does not correlate with treatment outcome, but any decrease of hemoglobin during therapy is a strong prognostic factor for treatment failure. CONCLUSIONS: Although regression models with many variables may be instable, the present results suggest that hemoglobin concentration after treatment is at least as important as overall treatment time. It was not possible to find out whether the low concentration of hemoglobin is an independent cause of low TCP or whether it reflects other mechanisms that may influence both hemoglobin level and the TCP.

Incidence and kinetics of distant metastases in patients with operable breast cancer.

The purpose of this paper is to evaluate the incidence and kinetics of distant metastases in operable breast cancer and to relate these estimates to various tumor and patient characteristics. The records of 309 consecutive patients with operable breast cancer in stage T1-4N0-1M0 were reviewed, and the incidence of distant metastases (DM) and death due to DM were evaluated. 195 patients had positive axillary nodes with the following distribution of the number of nodes: 45% had 1-2 node, 16% had 3-4 nodes, 14% and 25% had 5-7 and more nodes, respectively. All patients were treated with radical mastectomy with axillary nodes dissection (the only treatment in 39% of cases). In 198 cases radical mastectomy was combined with radiotherapy and/or chemotherapy given pre- or postoperatively. Hormonal treatment was given in 27% of cases. Minimum follow-up was 10 years. Distant metastases were found in 150 cases (49%) and in 78 cases (25%) they developed early, during the first 18 months follow-up. Average rate of DM in N0 cases was 25%. Number of involved nodes and extracapsular invasion were found significant and independent prognostic factors. High risk (50%) of DM and death due to DM correlate with age < 40 y, premenopausal status, tumor stage > or = T3, more than 2 axillary nodes and/or extracapsular invasion. The linearity of the curves for freedom from DM and for freedom from death due to the DM suggest uniform distribution of progression rates with a median value for halving time for freedom from early DM of about 8 months, and of about 40 months for freedom from the DM occurring later than 18 months, being for whole group an average of 20 months. High incidence of DM is a significant cause of poor long-term survival. Early appearance (< 18 month follow-up) of about half of the DM suggests that they are already present as subclinical micrometastases at the time of initial loco-regional treatment. The time of appearance of distant metastases is consistent with a wide range of metastatic cell burdens among patients. Systemic therapy, at least for selected group of patients, might decrease the incidence of DM and improve long-term results.

Randomized clinical trial on accelerated 7 days per week fractionation in radiotherapy for head and neck cancer. Preliminary report on acute toxicity.

PURPOSE: Toxicity of an accelerated 7 days per week fractionation schedule (arm A) was evaluated and compared with a conventional 5 days per week treatment (arm B) in a randomized trial. MATERIALS AND METHODS: Forty-four patients with squamous cell carcinoma of the head and neck in stage T2-4Nzero-1Mzero were included in the study. Total dose and dose per fraction of 2.0 Gy given once-a-day at 24 h intervals were the same in both arms of the trial. The only difference was the overall treatment time being 5 weeks in arm A and 7 weeks in arm B. RESULTS: Analysis of severe mucosal reactions shows significant difference between arm A and B, with regard to both maximum score and duration of severe mucositis. Confluent mucositis (score > 15 according to the Dische system) lasting longer than 3 weeks developed in 48% of patients in arm A and only in 5% in arm B. In group A seven (30%) late effects (osteo- and soft tissue necrosis) occurred during 7-12 month follow-up with two reactions (10%) in group B being suspected as late effects. There was significant association between acute reactions and late effects in arm A, suggesting that the late effects are consequential. CONCLUSION: The high incidence of severe acute reactions and consequential late effects suggests that the accelerated treatment in arm A (using daily fractions of 2.0 Gy, 7 days per week) gives unacceptable toxicity.

Sequential combination of radio-chemotherapy (5-fluorouracil, cis-platinum and irradiation) in the management of locally advanced head and neck cancers.

Thirty-seven previously untreated patients with advanced, inoperable head and neck were treated with a sequential courses combining hypofractionated irradiation with chemotherapy (5-fluorouracil and cis-platinum). Each course was repeated every 4 weeks. Tumor response was evaluated and for 15 patients (41%) with a partial or complete regression after 3 radio-chemotherapy courses conventional radiotherapy was added. Eleven percent of all patients were in complete remission at the end of a treatment. This tumor response rate and the 50% rate of pain subside after first course for symptomatic patients contributed for a good palliative effect in the present study. However, the median survival of 7.2 months was considered unsatisfactory.

[Prognostic factors in radiotherapy of supraglottic cancer]. / Czynniki prognostyczne w radioterapii raka górnego pietra krtani.

In 971 patients with squamous cell carcinoma of supraglottic larynx several clinical and physical prognostic factors were evaluated. There were 55% of patients with advanced primary tumours and 71% of patients with no regional neck metastases. All patients were irradiated radically using external megavoltage beam. The total dose was in range 60-70 Gy for 95% of patients. The 3-year local control rate and 3-year overall in whole group were 41% and 50% respectively. Clinical stage, haemoglobin level at the end of treatment, overall treatment time, sex and total dose were estimated as an independent and important prognostic factors for the outcome in radiotherapy of supraglottic larynx cancer.

Normobaric oxygen as a sensitizer in radiotherapy for advanced head and neck cancer.

The aim of this study was to evaluate radiosensitizing effect of normobaric oxygen breathing in radiotherapy for advanced head and neck cancers. Forty seven patients with advanced squamous cell carcinomas of the head and neck (7% in Stage III and 93% in Stage IV) were entered in the study. Breathing the pure, normobaric oxygen was given for 15-20 min in the treatment room. Irradiation started immediately after oxygen breathing. Conventional, megavoltage radiotherapy to the total doses in the range of 46-67.5 Gy was used. The control group was 46 patients with the same diagnosis and stage treated by radiotherapy alone. Locoregional tumor control was 36% in the study group compared to 15% in the control (p < 0.05). Mean survival time was 15.8 and 11.8 months, and 3-year survival was 19% and 2%, respectively (p < 0.05). Survival depended on total tumor dose and total nodal dose. No significant influence of the tumor location on local control and importance of the size of dose per fraction and overall treatment time were found. The most common failure in both groups was persistent tumor. Mean recurrence time was 5 months in the study group and 8 months in the control. Present results suggest that the use of normobaric oxygen breathing prior to irradiation could increase effectiveness of conventional radiotherapy for advanced squamous cell carcinomas of head and neck.

Planned and unplanned gaps in radiotherapy: the importance of gap position and gap duration.

Local tumour control in 971 patients with squamous carcinoma of the supraglottic larynx has been examined in relation to the occurrence of gaps in radiation therapy. The minimum follow-up time was 3 years. The reasons for a gap in radiotherapy fell into four categories: independent of the patient (national holidays, machine break-down, etc.), planned gaps (split-course therapy), severe normal-tissue reactions, and intercurrent disease. Only 11.7% of patients had no gap at all, 75.5% a single gap, and 2.1% had more than four gaps. The probability of tumour control increased with dose in all patient sub-groups; the average percentage increase for a 1% increase in dose was 4.3. The data were subjected to multivariate analysis, leading to the following conclusions. Patients in whom there was a single gap showed a remarkable trend of local control: if the gap began before day 19 after the start of therapy, the local tumour control was considerably below that in patients who did not suffer a gap in treatment. The local tumour control in patients whose gap began at day 20-29 was indistinguishable from that in patients who had no gap in treatment. A gap further towards the end of treatment was again associated with a severe drop in local control. This trend was independent of the recorded cause of the gap. The mechanism of this phenomenon is not clear. The effect of the timing of a treatment gap appears in this data set to have had a considerable impact on outcome and our observations should stimulate further study of this phenomenon in other clinical settings.