Lawrence W. Davis

F-18 FDG PET-CT fusion in radiotherapy treatment planning for head and neck cancer.

The fusion of fluoro‐2‐deoxy‐d‐glucose–positron emission tomography (FDG‐PET) with CT scans has been shown to improve diagnostic accuracy and staging in non‐small cell lung cancer. We report on the influence of PET‐CT fusion on the management of patients with head and neck cancer.

TREATMENT OF PANCREATIC CANCER TUMORS WITH INTENSITY- MODULATED RADIATION THERAPY (IMRT) USING THE VOLUME AT RISK APPROACH (VARA): EMPLOYING DOSE-VOLUME HISTOGRAM (DVH) AND NORMAL TISSUE COMPLICATION PROBABILITY (NTCP) TO EVALUATE SMALL BOWEL TOXICITY

The emergent use of a combined modality approach (chemotherapy and radiation) in pancreatic cancer is associated with increased gastrointestinal toxicity. Intensity-modulated radiation therapy (IMRT) has the potential to deliver adequate dose to the tumor volume while decreasing the dose to critical structures such as the small bowel. We evaluated the influence of IMRT with inverse treatment planning on the dose-volume histograms (DVHs) of normal tissue compared to standard 3-dimensional conformal radiation treatment (3D-CRT) in patients with pancreatic cancer. Between July 1999 and May 2001, 10 randomly selected patients with adenocarcinoma of the pancreatic head were planned simultaneously with 3D-CRT and inverse-planned IMRT using the volume at risk approach (VaRA) and compared for various dosimetric parameters. DVH and normal tissue complication probability (NTCP) were calculated using IMRT and 3D-CRT plans. The aim of the treatment plan was to deliver 61.2 Gy to the gross tumor volume (GTV) and 45 Gy to the clinical treatment volume (CTV) while maintaining critical normal tissues to below specified tolerances. IMRT plans were more conformal than 3D-CRT plans. The average dose delivered to one third of the small bowel was lower with the IMRT plan compared to 3D-CRT. The IMRT plan resulted in one third of the small bowel receiving 30.2+/-12.9 Gy vs. 38.5+/-14.2 Gy with 3D-CRT (p = 0.006). The median volume of small bowel that received greater than either 50 or 60 Gy was reduced with IMRT. The median volume of small bowel exceeding 50 Gy was 19.2+/-11.2% (range 3% to 45%) compared to 31.4+/-21.3 (range 7% to 70%) for 3D-CRT (p = 0.048). The median volume of small bowel that received greater than 60 Gy was 12.5+/-4.8% for IMRT compared to 19.8+/-18.6% for 3D-CRT (p = 0.034). The VaRA approach employing IMRT techniques resulted in a lower dose per volume of small bowel that exceeded 60 Gy. We used the Lyman-Kutcher models to compare the probability of small bowel injury employing IMRT compared to 3D-CRT. The BIOPLAN model predicted a small bowel complication probability of 9.3+/-6% with IMRT compared to 24.4+/-18.9% with 3D-CRT delivery of dose (p = 0.021). IMRT with an inverse treatment plan has the potential to significantly improve radiation therapy of pancreatic cancers by reducing normal tissue dose, and simultaneously allow escalation of dose to further enhance locoregional control.

Long-term outcomes of IMRT for breast cancer: a single-institution cohort analysis.

PURPOSE To evaluate long-term outcomes of adjuvant breast intensity-modulated radiation therapy (IMRT), with a comparison cohort receiving conventional radiation (cRT) during the same period. METHODS AND MATERIALS Retrospective review identified patients with Stages 0-III breast cancer who underwent irradiation after conservative surgery from January 1999 to December 2003. Computed tomography simulation was used to design standard tangential breast fields with enhanced dynamic wedges for cRT and both enhanced dynamic wedges and dynamic multileaf collimators for IMRT. Patients received 1.8-2-Gy fractions to 44-50.4 Gy to the whole breast, followed by an electron boost of 10-20 Gy. RESULTS A total of 245 breasts were treated in 240 patients: 121 with IMRT and 124 with cRT. Median breast dose was 50 Gy, and median total dose was 60 Gy in both groups. Patient characteristics were well balanced between groups. Median follow-ups were 6.3 years (range, 3.7-104 months) for patients treated with IMRT and 7.5 years (range, 4.9-112 months) for those treated with cRT. Treatment with IMRT decreased acute skin toxicity of Radiation Therapy Oncology Group Grade 2 or 3 compared with cRT (39% vs. 52%; p = 0.047). For patients with Stages I-III (n = 199), 7-year Kaplan-Meier freedom from ipsilateral breast tumor recurrence (IBTR) rates were 95% for IMRT and 90% for cRT (p = 0.36). For patients with Stage 0 (ductal carcinoma in situ, n = 46), 7-year freedom from IBTR rates were 92% for IMRT and 81% for cRT (p = 0.29). Comparing IMRT with cRT, there were no statistically significant differences in overall survival, disease-specific survival, or freedom from IBTR, contralateral breast tumor recurrence, distant metastasis, late toxicity, or second malignancies. CONCLUSIONS Patients treated with breast IMRT had decreased acute skin toxicity, and long-term follow-up shows excellent local control similar to a contemporaneous cohort treated with cRT.

Prognostic Accuracy of Computed Tomography Findings for Patients With Laryngeal Cancer Undergoing Laryngectomy

PURPOSE The indications for upfront laryngectomy in the management of laryngeal cancer are a functionless larynx and extralaryngeal extension. Practically, clinicians rely on imaging to predict which patients will have T4 disease. Our goal was to review the accuracy of preoperative computed tomography (CT) scanning in determining the necessity for initial laryngectomy for advanced laryngeal cancer. PATIENTS AND METHODS In total, 107 consecutive untreated laryngectomy specimens with high-quality, preoperative CT imaging interpreted by our neuroradiologists were reviewed. Radiographic findings, including sclerosis, invasion, penetration, extralaryngeal spread, and subglottic extension were correlated with pathologic findings. CT images were not reinterpreted, since our purpose was to assess the original interpretations. RESULTS CT imaging reported 23 cases of thyroid cartilage penetration and 27 cases of extralaryngeal spread. Pathology reported 12 cases of thyroid cartilage invasion, 29 cases of penetration, and 45 cases of extralaryngeal disease. CT imaging identified 17 (59%) of 29 cases of pathologically documented thyroid cartilage penetration and 22 (49%) of 45 cases of pathologically documented extralaryngeal spread. Pathologically proven extralaryngeal spread without thyroid cartilage penetration occurred in 18 (40%) of 45 cases. The positive predictive values for thyroid cartilage penetration and extralaryngeal spread were 74% and 81%. Sclerosis was of limited value in predicting thyroid cartilage invasion or penetration. Cricoid or arytenoid destruction predicted for thyroid cartilage penetration at rates of 57% and 63%. CONCLUSION CT imaging has clear limitations when deciding whether there is thyroid cartilage penetration or extralaryngeal spread of advanced laryngeal cancer. Extralaryngeal spread without thyroid cartilage penetration was more common than expected. Alternate methods of pretreatment assessment are needed.

Three-Year Outcomes of Breast Intensity-Modulated Radiation Therapy With Simultaneous Integrated Boost

PURPOSE To report our clinical experience using breast intensity-modulated radiation therapy with simultaneous integrated boost (SIB-IMRT). METHODS AND MATERIALS Retrospective review identified 354 Stage 0 to III breast cancer patients treated with SIB-IMRT after conservative surgery between 2003 and 2006. The most common fractionation (89%) simultaneously delivered 1.8 Gy to the ipsilateral breast tissue and 2.14 Gy to the resection cavity, yielding a breast dose of 45 Gy (25 fractions) and cavity dose 59.92 Gy (28 fractions), biologically equivalent for tumor control to 45 Gy to the breast with sequential 16-Gy boost (33 fractions). RESULTS A total of 356 breasts in 354 patients were treated: 282 with invasive breast cancer, and 74 with ductal carcinoma in situ (DCIS). For left breast radiation, median cardiac V(15) was 2.9% and left ventricular V(15) 1.7%. Median follow-up was 33 months (range, 4-73 months). Acute toxicity was Grade 1 in 57% of cases, Grade 2 in 43%, and Grade 3 in <1%. For invasive breast cancer, the 3-year overall survival was 97.6% and risk of any locoregional recurrence was 2.8%. For ductal carcinoma in situ, 3-year overall survival was 98% and risk of locoregional recurrence 1.4%. In 142 cases at a minimum of 3 years follow-up, global breast cosmesis was judged by physicians as good or excellent in 96.5% and fair in 3.5%. CONCLUSIONS Breast SIB-IMRT reduced treatment duration by five fractions with a favorable acute toxicity profile and low cardiac dose for left breast treatment. At 3 years, locoregional control was excellent, and initial assessment suggested good or excellent cosmesis in a high percentage of evaluable patients.

Concurrent platinum‐based chemotherapy and simultaneous modulated accelerated radiation therapy for locally advanced squamous cell carcinoma of the tongue base

Randomized data support use of chemotherapy concurrently with radiation in treatment of advanced squamous cell carcinoma (SCC) of the oropharynx. Intensity modulated radiation therapy (IMRT) is increasingly being used to deliver such radiotherapy; no published reports specifically describe results of chemotherapy with IMRT for SCC of the base of tongue (BOT). We present outcomes data using simultaneous modulated accelerated radiation therapy (SMART) combined with platinum‐based chemotherapy in treatment of locally advanced SCC of the BOT

Dose verification for patients undergoing IMRT

At Emory Clinic intensity-modulated radiation therapy (IMRT) was started by using dynamic multileaf collimators (dMLC) as electronic tissue compensators in August 1998. Our IMRT program evolved with the inclusion of a commercially available inverse treatment planning system in September 1999. While the introduction of electronic tissue compensators into clinical use did not affect the customary radiation oncology practice, inverse treatment planning does alter our basic routines. Basic concepts of radiation therapy port designs for inverse treatment planning are different from conventional or 3D conformal treatments. With inverse treatment planning, clinicians are required to outline a gross tumor volume (GTV), a clinical target volume (CTV), critical normal structures, and to design a planning target volume (PTV). Clinicians do not designate the volume to be shielded. Because each IMRT radiation portal is composed of many beamlets with varying intensities, methods and practice used to verify delivered dose from IMRT portals are also different from conventional treatment portals. Often, the validity of measured data is in doubt. Therefore, checking treatment planning computer output with measurements are confusing and fruitless, at times. Commissioning an IMRT program and routine patient dose verification of IMRT require films and ionization chamber measurements in phantom. Additional specialized physics instrumentation is not required other than those available in a typical radiation oncology facility. At this time, we consider that routine quality assurance prior to patient treatments is necessary.

Percutaneous feeding tubes in patients with head and neck cancer: rethinking prophylactic placement for patients undergoing chemoradiation☆

OBJECTIVES Although intensified therapy has contributed to improved outcomes for patients with head and neck cancer, acute toxicity has increased as well. To lessen the severity of nutritional compromise in these patients, our institutional protocol has been to routinely place feeding tubes before the initiation of therapy. This investigation details the toxicities associated with feeding tube placement and predictors for duration of tube dependence. MATERIALS AND METHODS The records of the Radiation Oncology Department at Emory Clinic were reviewed for patients receiving definitive radiotherapy between 6/1/2003 and 6/1/2006. The records of the subset of patients with feeding tube placement before the initiation of therapy were then reviewed for toxicities as well as length of time of tube dependence. RESULTS There were 102 eligible patients. Radiotherapy was delivered with concomitant chemotherapy in all. Median time with feeding tube in place for all patients was 4.4 months (range, 0.2-28.9 months). For 82 patients with eventual tube removal, the median time of tube dependence was 3.8 months (range, 1.4-28.9 months). Risk factors for prolonged tube dependence are analyzed; on multivariate analysis, patient age, T stage, and nodal status remained significant. The most common complication was tube replacement, with 11.8% of all tubes requiring replacement. Infection and pain occurred in 8.8% and 5.9% of patients, respectively. CONCLUSION Feeding tubes are required for more than 2 months after combined modality treatment of head and neck cancer. They are generally well tolerated, but toxicities are not trivial: more than 10% require replacement and more than 8% of patients develop infection at the insertion site. We are assessing their routine placement in light of these data.

A totally integrated simulation technique for three-field breast treatment using a CT simulator.

A method was devised to simulate patients with breast cancer in the actual treatment position utilizing a diagnostic CT spiral scanner, 3-D Image Workstation for virtual simulation, and a laser coordinate system to transfer planning parameters to the patients skin. It was desired to produce non-divergent tangential beams through the lung as well as a matched line for tangential and supraclavicular fields. The patients were immobilized in an Alpha CradleTM cast. Radio-opaque markers were placed on the superior, inferior, medial, and lateral margins of the field so as to afford appropriate initial field set-up approximations. The patient was scanned. The data set was then transferred to the workstation where an isocenter was chosen. The patient was marked. Virtual simulation was then performed. This method employed a half beam technique for the posterior edge of the tangential fields. Table rotation and blocking of the superior margin of the tangential fields were used to produce a vertical edge to match a supraclavicular field. Using a beams eye view the lateral tangent was matched to the medial exit. A digitally reconstructed radiograph was created to define the tangent fields and place the supraclavicular block. Our initial experience with 50 patients verifies that this is a reproducible and accurate technique. Time required for immobilization and tangential field simulation is approximately 30 minutes. Data is available for 3-D treatment planning or 2-D treatment planning on a reconstructed transverse slice angled to match the collimator angle through the patient. Using a CT simulator for simulation of breast cancer affords accuracy of at least equal magnitude to conventional simulators as determined by beam films and ease of set-up. This technique also affords greater ease in changing treatment parameters without having to resimulate the patient.

EARLY CLINICAL EXPERIENCE WITH KILOVOLTAGE IMAGE-GUIDED RADIATION THERAPY FOR INTERFRACTION MOTION MANAGEMENT

Interest in image-guided radiation therapy (IGRT) reflects the desire to minimize interfraction positioning variability. Using a kilovoltage (kV) imaging unit mounted to a traditional LINAC allows daily matching of kV images to planning digitally reconstructed radiographs (DRRs). We quantify and evaluate the significance of calculated deviation from the intended isocenter. Since September 2004, 117 patients with various malignancies were treated using the On-Board Imaging (OBI) system, with 2088 treatment sessions. Patients were positioned by the treating therapist; orthogonal images were then obtained with the OBI unit. Couch shifts were made, aligning bony anatomy to the initial simulation image. Routine port films were performed weekly (after that days OBI session). Ninety percent of all lateral, longitudinal, and vertical shifts were less than 0.8 cm, 0.6 cm, and 0.7 cm, respectively. The median vector shift for each anatomic site was: 0.42 cm for head and neck, 0.40 cm for CNS, 0.59 cm for GU/prostate, and 0.73 cm for breast; shift magnitude did not change with successive OBI sessions. The use of OBI effectively corrects setup variability. These shifts are typically small and random. The use of OBI likely can replace weekly port films for isocenter verification; however, OBI does not provide field shape verification.