OncoLink

Recurrences after Conformal Parotid-Sparing Radiotherapy

Reviewed by: Thomas J. Dilling, MD
The Abramson Cancer Center of the University of Pennsylvania

Source: Radiotherapy and Oncology 72 (2004) 119-127
Authors: Barbara Bussels, Annelies Maes, Robert Hermans, Sandra Nuyts, Caroline Weltens, Walter Van den Bogaert

Background

  • One of the most difficult long-term complications of radiation treatment of head and neck cancer is xerostomia.
  • Radiation-induced xerostomia leads to difficulties chewing, swallowing, talking, as well as increased risk of dental caries.
  • By sparing one of the two parotid glands, the hope is to preserve enough salivary flow to prevent subjective complaint of xerostomia.
  • The concern, however, is that sparing the contralateral parotid would also spare contralateral junctional (high level II) lymph nodes, potentially leading to an increased risk of tumor recurrence.
  • This paper reports 19 month follow-up on patients selected to undergo radiotherapeutic treatment with sparing of the contralateral parotid gland.

Methods

  • 72 patients with histologically-confirmed head/neck cancer were treated between June, 1999 and July, 2002.
  • All patients had a negative metastatic work-up.
  • Patients with tumors originating from midline or crossing midline were excluded, as were all patients with nasopharyngeal cancers or contralateral nodal metastasis.
  • The majority were oropharyngeal tumors (57%), with 93% being squamous cell carcinomas.
  • 72% had stage III or IV cancer.
  • 49/72 had radiotherapy alone, with 23/72 receiving post-operative radiation.
  • All surgical patients underwent modified radical neck dissection (ipsilateral in 19/23 and bilateral in 4/23)
  • Three patients received chemotherapy, either before radiation (1/3) or concurrent with radiation (2/3), as per various study protocols in their department.
  • All patients underwent 5mm CT scans, with delineation of tumor volumes on the scan.
    • GTV1 was CT-visible primary tumor volume.
    • GTV2 included lymph nodes which were clinically or radiologically suspect for invasion.
    • These volumes were extended by 1 cm in all directions, except through air or bone, to generate CTV1 and CTV2.
    • CTV3 included the elective nodal volume, which included all nodal regions that would have been irradiated using standard 3-field technique, excluding contralateral junctional nodes adjacent to the contralateral parotid gland. This region was defined radiologically as extending to the base of skull cranially, caudally to the crossing of the internal jugular vein with the posterior belly of the digastric muscle, laterally to the pterygoid muscle and parotid gland, and medially to the pharyngeal wall and lateral retropharyngeal nodes.
    • In post-operative patients, the CTV of the tumor resection bed was determined, based upon pre-surgical CT scans, presence of surgical clips, and including potential sites of local tumor extension, with 1.5 cm margin.
    • 5 mm were added uniformly to the CTVs to generate PTVs.
    • Organs outlined included both parotid glands, the spinal cord, and the brainstem.
  • Maximal dose of 48-50 Gy was tolerated to the spinal cord. Maximal dose to the brain stem was likewise limited to 50 Gy. The goal of parotid-sparing was to administer a mean dose ≤26 Gy to the contralateral parotid, with no dose constraints placed upon the ipsilateral parotid.
  • 68/72 patients were treated with a single-isocenter 3D-conformal technique, utilizing two anterior obliques and a lateral field superiorly, matched to a low-neck anterior field with half-beam blocking technique, as described in another paper.
  • Total doses ranged from 50-66 Gy for post-operative patients, and 70 Gy for definitive patients, though 17 patients were treated with an accelerated radiation schedule.
  • Loco-regional control and overall survival were estimated according to the Kaplan-Meier product-limit method, with local control differences between treatment groups assessed using the log-rank test.

Results

  • Minimal follow-up was 6 months, with a median of 19 months.
  • Of the 72 patients, 18 had locoregional recurrence and 2 had simultaneous locoregional and distant recurrence. These 20 patients were analyzed for patterns of failure. This group included 13/49 patients treated with primary radiotherapy and 7/23 treated with post-operative irradiation.
  • 12/13 patients treated with primary radiotherapy had failures which occurred in-field, with one recurrence categorized as extending out-of-field.
  • In the post-operative group, 3/7 presented with in-field failures, while 4/7 included failures with originated in the high-dose region but extended outside the borders of the treated volume.
  • There were no out-of-field recurrences were found in any patients. In particular, there were no documented recurrences in the contralateral high level II region which had been omitted from the treatment field.

Discussion

  • The authors note that there were no documented recurrences in the region of the contralateral parotid gland.
  • While this paper included no specific analysis of salivary flow or review of patients’ subjective complaints of xerostomia, the hope would be that by sparing the contralateral parotid gland, one could minimize the xerostomia and thereby improve the patients’ quality of life. Other researchers have performed parotid-sparing head/neck irradiation, with demonstrable long-term improvement in salivary flow, with diminished subjective complaints of xerostomia by the patients.
  • The authors make mention of these two previously-published papers, and point out that, of all the patients in these three studies, only one had a recurrence in the contralateral lymph nodes which had been spared from high-dose irradiation.
  • The assumption remains, therefore, that sparing of the contralateral parotid gland (mostly typically in the US via an intensity-modulated radiation technique), should diminish somewhat the long-term risk of xerostomia, without adversely affecting locoregional recurrence.
  • A flaw of this study is that it fails to adequately describe the means by which patients were selected for treatment with this technique, except to say that they were patients with lateralized tumors.