The Incidence of Stroke in Patients with Head and Neck Cancer with or without Radiotherapy

Reviewer: Christine Hill, MD
Abramson Cancer Center of the University of Pennsylvania
Last Modified: September 24, 2008

Presenter: D.J. Huang
Presenter's Affiliation: Mount Sinai School of Medicine, New York, NY
Type of Session: Scientific


  • Radiotherapy is an essential component of treatment for many cancers of the head and neck, and may be delivered in combination with surgery and/ or chemotherapy.
  • As the number of cancer survivors living in the United States and around the world continues to increase, attention to late effects resulting from cancer and its treatments is particularly important. 
    • When head and neck cancer patients undergo radiotherapy, one or both sides of the neck are often included in the radiation field.
    • High doses of radiation are often delivered to one or both carotid arteries.
    • The presence of carotid arterial disease in patients treated with radiation to the neck has been found to be clinically significant (Steele, Am J Surg, 2004), and this process may be accelerated when compared to patients not having had radiation (Cheng, J Vasc Surg, 2004).
  • Patients who develop cancers of the head and neck often have confounding risk factors for increased atherosclerosis, including tobacco history.
  • While several groups have examined the risk of carotid disease in head and neck cancer patients who have received radiotherapy, these patients have not been compared to control patients with head and neck cancer who have not received radiation. Furthermore, risk of clinically significant stroke in this population has not been examined.
  • The study presented here is an investigation of the impact of radiotherapy to the neck on long-term risk of clinically significant stroke.

Materials and Methods

  • The Surveillance, Epidemiology, and End Results (SEER)/ Medicare database was utilized to identify the patient population investigated in this study.
  • Queries were performed regarding inpatient hospitalization for stroke and stroke-specific mortality in patients with squamous cell carcinomas of the head and neck, stage I-IVB, diagnosed between 1988 and 2000.
  • All patients included in this study were required to have at least five years of follow-up. This restriction was imposed because the authors felt that strokes occurring within five years of diagnosis were unlikely to be radiation-induced.
  • Patients were required to be at least 21 years of age at diagnosis, and were excluded if they had evidence of distant metastasis at the time of diagnosis, if they were noted to have in situ carcinoma, or if treatment modalities were unknown.
  • Data abstracted included age at diagnosis, gender, race/ ethnicity, surgical procedures undergone, radiation treatments received, lymph node status, tumor size, primary tumor site, and grade of disease.
  • The Kaplan-Meier method was utilized to generate time-to-event curves, and univariate analysis was performed using chi-square and log-rank tests. Multivariate analysis was performed using the Cox model.


  • A total of 9738 patients meeting the selection criteria were identified.
  • Of these, 6164 had received radiotherapy as part of their treatment.
  • Median follow-up was 8.5 years (range 5 – 17.4 years)
  • Median patient age was 65 years.
  • When patients having received radiotherapy were compared to those not treated with radiotherapy, patients treated with radiotherapy were older on average, with more advanced primary and nodal disease (p < 0.01). 
  • Of all patients analyzed, 741 were documented to have experienced stroke. Of these, 490 had received radiotherapy.
  • The 10-year freedom from stroke rate was 90% for patients having received radiotherapy and 92.5% for those not having received radiotherapy (p = 0.01).
  • The incidence of stroke was 8.3/ 1000 person-years in the radiotherapy group versus 7.6/ 1000 person-years in the non-radiotherapy group.
  • No difference in the 10-year stroke specific mortality was observed for patients having received radiotherapy versus those not (1.9% versus 2.0%, respectively, p = 0.57).
  • On Cox multivariate analysis, factors identified as increasing stroke risk included treatment with radiotherapy [hazard ratio (HR) 1.38, p < 0.01], year of diagnosis (HR 1.13, p < 0.01), age (HR 1.06, p < 0.01), and race (HR 1.50, p < 0.01).
  • Increased risk of stroke in patients having received radiotherapy was greatest for those 60 years or younger (HR 2.18, p < 0.01), and with negative lymph nodes (HR 1.44, p < 0.01).

Author's Conclusions

  • The authors conclude that this study demonstrates a 38% increase in the relative risk of stroke in head and neck cancer patients having received radiotherapy versus those not receiving radiotherapy.
  • They note that this risk translates to a 1.6% increase in absolute risk, and that no difference in stroke-related mortality was observed.
  • They describe the study presented here as hypothesis generating, and representing the first attempt to study this important issue with appropriate control group analysis.

Clinical/Scientific Implications

  • As the cancer survivor population grows, issues regarding late effects of cancer treatment continue to gain importance.
  • Clinical counseling for patients with head and neck cancer for whom radiotherapy is recommended generally includes discussion of some increased risk of carotid artery sclerosis and/ or damage following radiotherapy. This risk is difficult to quantify, and its clinical relevance poorly understood.
  • This study represents a well-designed attempt to determine the clinical impact of radiotherapy to the neck in head and neck cancer patients with regard to risk of stroke, and its findings are a valuable contribution to patient care.
  • This study is limited by its retrospective nature, as well as certain elements of data that are not available from the SEER/ Medicare database.
    • These include data regarding radiotherapy dose to specific regions of the neck, namely the carotid arteries. Certainly, correlation of dose-volume histograms with stroke risk would be of great interest.
    • Having further information regarding specifics of treatment might also assist us with interpretation of the authors’ results indicating that patients with negative lymph nodes were at higher risk of stroke following radiotherapy than were those with positive nodes. This finding is somewhat difficult to interpret as one would expect patients with negative nodes to have received lower doses of radiation to the neck when compared to those with positive lymph nodes. Further information on treatment plans for these patients would be helpful.
    • Along these same lines, further information on surgical procedures undergone, as well as chemotherapies received, would be of assistance in data interpretation. Each of these treatments would be expected to potentially impact stroke risk, and could confound the findings of this study.
    • Finally, addition of information regarding co-morbidities such as diabetes, obesity, and smoking status, all of which are recognized to increase stroke risk, would be of interest. The authors note that patients having received radiotherapy in this study tended to be older and with more advanced cancer than those not receiving radiotherapy. These differences in patient population may contribute to increased risk of stroke in the radiotherapy group, and co-morbidities could likely also contribute.
  • Despite these limitations, this study is interesting and well-designed. Its findings have important implications for patient care, and will be particularly useful in discussions with patients regarding stroke risk following radiotherapy.
  • Although this study demonstrated increased relative risk of stroke in patients undergoing radiotherapy, absolute risk was increased only modestly, and stroke-related mortality was not affected. These findings may reassure patients and clinicians that increased risk of stroke likely does not offset the benefits of radiotherapy in treatment of head and neck cancer; however, patients and clinicians alike should be aware of increased risk of stroke, and incorporate this into follow-up care and behaviors during and following treatment.