Introduction
Changes in the salivary glands and in the composition of saliva can occur as a result of ionising irradiation of salivary gland cells. Salivary function is extremely responsive to radiation. At least 50 per cent of function is lost after only 1000 cGy in one week of radiotherapy. In conventional radiotherapy for head and neck cancer, patients usually receive about 6000 to 7000 cGy over six to seven weeks. One study found that such radiation doses cause 80 per cent salivary dysfunction, which may last for years after radiotherapy.Reference Mossman1
Radiation causes damage to both the acinar and ductal systems of the salivary glands; however, the principal damage is to the acinar parenchyma. Serous acini are more responsive to radiation therapy than mucinous acini. At the beginning of irradiation, there is a prompt, marked elevation of serum amylase levels and a decrease in serum immunoglobulin A, both of which progress with increasing dose.Reference Marks, Davis, Gottsman, Purdy and Lee2, Reference Cheng, Downs, Herbert and Aramany3 The effects of radiotherapy on salivary glands are probably due to effects on the fine vasculature of the glands, with secondary changes in the parenchymal epithelium and interstitial and interlobular fibrosis, leading to degeneration of the functional acini. Later, the damaged secretory acinar cells within the glands are replaced by ductal elements and inflammatory cells. This damage causes an apparent increase in salivary sodium and chloride and a decrease in salivary bicarbonate.Reference Mossman, Shatzman and Chencharick4 These changes account for the thick, tenacious and acidic character of post-irradiation saliva.
When the major salivary glands are exposed within the radiation field, salivary gland dysfunction commonly develops.Reference Valdez5 Basal whole salivary flow may reach a measurable minimum two to three weeks after delivery of 2300 cGy of fractionated radiotherapy.Reference Wescott, Mira, Starcke, Shannon and Thornby6
Dreizen et al. Reference Dreizen, Brown, Handlers and Levy7 showed that salivary flow progressively decreases with radiotherapy, and that after three years the percentage of decrease exceeds 90 per cent and is presumably permanent. The extent of salivary dysfunction is primarily determined by the radiation field and the total dose of radiation.Reference Valdez5
The aim of this study was to objectively assess the salivary gland function in post-irradiated head and neck tumour patients, the majority of whom had received radiotherapy at the radiotherapy and oncology unit of the Hospital of the Science University of Malaysia. The radiotherapy service provided by this hospital began in 1996 and mainly serves the population of the east coast of the Malaysian peninsula. However, in some instances patients were also referred from other parts of the country.
Patients and methods
We conducted a cross-sectional study of head and neck tumour patients who had received radiotherapy treatment between July 2002 and September 2003 at the otorhinolaryngology clinic of the department of otorhinolaryngology and the department of radiotherapy and oncology, School of Medical Sciences, University Sains Malaysia. Patients were seen in the combined oncology–otorhinolaryngology and head and neck clinic during their follow up. Informed consent was obtained prior to inclusion in the study. There were two groups; a study group of patients who had undergone radiotherapy for various head and neck tumours, and a control group comprising patients seen in the University Sains Malaysia otorhinolaryngology clinic for other problems.
The inclusion criteria comprised head and neck tumour patients who had completed radiation therapy over one year previously and were aged between 20 and 70 years. A one-year post-radiotherapy period was necessary because acute radiation effects such as mucositis usually resolve by about three months, and late sequelae (in this case, salivary gland hypofunction) usually present eight months or more after completion of treatment.Reference Kendall, McKenzie, Leonard and Jones8
The following exclusion criteria were applied: any medication which might influence salivary production, secretion or flow (in particular, drugs with anticholinergic properties, including antidepressants, antihistamines and diuretics); systemic medical illness (e.g. diabetes mellitus, renal diseases and thyroid diseases); autoimmune or connective tissue diseases; previous disease or surgical removal of any major salivary gland; pre-existing xerostomia or any related oral morbidity, due to any cause; and a post-radiation interval of less than one year.
Saliva sampling
The appearance of each patient's oral cavity saliva was categorised into one of the following groups; normal, frothy, minimal, or dry (i.e. no saliva).
Salivary flow rate measurement
Salivary gland function was assessed objectively based on two parameters, the unstimulated whole salivary flow rate and the stimulated whole salivary flow rate. The latter required use of 2 per cent citric acid solution. These measurements were performed on both study and control subjects. The mean value for each measurement, for both groups, was compared.
Unstimulated whole salivary flow rate
The unstimulated whole salivary flow rate was measured by the draining method, as follows.Reference Navazesh and Christensen9 Patients refrained from eating or smoking for 90 minutes before the test session. The saliva was collected during a five-minute session under supervision. Subjects were asked to lean forward, keep their eyes open, minimise movement of the tongue and lips, and to allow saliva to drip off the lower lip into a container. At the end of five minutes, the patients were asked to expectorate into the container. The total amount of saliva collected and the flow rate (in ml/min) were measured. An average unstimulated whole salivary flow rate of 0.3–0.4 ml/min has been designated as normal; a rate of less than 0.1 ml/min is considered abnormal.10
Stimulated whole salivary flow rate
The stimulated whole salivary flow rate was measured by the spitting method under supervision, as follows. Patients refrained from eating or smoking for 90 minutes before the test session. Two per cent citric acid solution was used to stimulate saliva production.Reference Valdez, Atkinson, Ship and Fox11 One ml of this solution was placed on the dorsum of the subject's tongue. The subject was then instructed to void the mouth of saliva at one-minute intervals for three minutes. The amount of total saliva collected and the flow rate (as ml/min) were determined. An average stimulated whole salivary flow rate of 1–2 ml/min has been designated as normal; a rate of less than 0.5 ml/min is considered abnormal.10
Radiotherapy
The majority of the patients had received radiotherapy at the radiotherapy and oncology department of the Hospital of the Science University of Malaysia. All data regarding their radiotherapy were traced and recorded for further analysis. For patients who had received radiotherapy at other centres, data were obtained from the respective centres and recorded. Data included tumour type, tumour–node–metastasis (TNM) staging, total radiation dose, duration of radiation, volume of field irradiated, and type of radiation.
Statistical analysis
Study results were recorded and analysed using the Statistical Package for the Social Sciences version 11 software. A p value of less than 0.05 was considered to be significant. Non-parametric correlations were used to analyse the relationship between xerostomia and the variables of interest.
Results
Demographic data
A total of 60 patients were recruited for the study. The study group comprised 30 patients who had undergone radiotherapy for various head and neck tumours and who fulfilled the selection criteria (described above). The control group comprised 30 patients selected randomly on attendance at the otorhinolaryngology clinic of the Hospital of the Science University of Malaysia for other problems. The study group consisted of 21 men (70 per cent) and nine women (30 per cent), aged between 18 and 69 years, with a mean age of 48.3 ± 15.46 years. In the control group, the gender distribution was more equal (16 men (53.3 per cent) and 14 women (46.7 per cent)), aged between 25 and 61 years, with the mean age of 41.7 ± 11.27 years.
Saliva sampling
Figure 1 shows the distribution of patients' saliva into the four categories (normal, frothy, minimal or dry). Only eight patients (26.7 per cent) were found to have normal saliva.
Fig. 1 Patients' oral cavity saliva. Numbers in bars indicate totals.
Whole salivary flow measurements
Salivary gland hyposecretion secondary to radiotherapy was assessed objectively. Unstimulated and stimulated whole salivary flow rates were calculated for both groups. Table I shows the range, median and interquartile range for the unstimulated and stimulated whole salivary flow rate measurements of both groups.
Table I Whole Salivary Flow Rate in Patients and Controls
Unstim = unstimulated; stim = stimulated; min = minimum; max = maximum; IQR = interquartile range
Statistical analysis of whole salivary flow rate measurements was performed using the non-parametric test and the Wilcoxon signed ranks test. There was a significant difference in flow rates between the study and control groups (p = 0.0001). Figure 2 shows the unstimulated and stimulated whole salivary flow rates for both groups. The control group showed a statistically significantly greater increase in whole salivary flow rate after stimulation of the salivary glands with 2 per cent citric acid solution, compared with the study group.
Fig. 2 Unstimulated and stimulated whole salivary flow rates for patients and controls.
In the study group, 24 patients (80.0 per cent) had abnormal unstimulated whole salivary flow rates (i.e. less than 0.1 ml/min), while 20 (66.7 per cent) had no unstimulated salivary flow at all (Figure 3). The median unstimulated whole salivary flow rate for the study group was 0.00 ml/min, with a range of 0–0.30 ml/min.
Fig. 3 Unstimulated whole salivary flow rates in study patients. Numbers in bars indicate totals. The x-axis shows uneven increments as not all numbers have been shown.
Seventeen study group patients (56.7 per cent) had abnormal stimulated whole salivary flow rates (i.e. less than 0.5 ml/min) (Figure 4). In the study group, the median stimulated whole salivary flow rate was 0.38 ml/min, with a range of 0.06–2.73 ml/min.
Fig. 4 Stimulated whole salivary flow rates in study patients. Numbers in bars indicate totals. The x-axis shows uneven increments as not all numbers have been shown.
This difference between unstimulated and stimulated whole salivary flow rates was statistically significant (Spearman's rank correlation; r = 0.515; p < 0.004). In all patients, the stimulated whole salivary flow rate was greater than the unstimulated whole salivary flow rate.
Radiotherapy data collection
All the necessary data regarding patients' radiotherapy treatment was traced through their medical records and analysed for correlation with the study parameters. The diagnoses of the study group patients are shown in Table II. Nasopharyngeal carcinoma constituted 73.3 per cent (22 patients) of the diagnoses. Other types of cancers were laryngeal cancer (three patients, 10 per cent), hard palate cancer (two, 6.7 per cent), oral cancer (one, 3.3 per cent), thyroid cancer (one, 3.3 per cent) and nasopharyngeal angiofibroma (one, 3.3 per cent).
Table II Tumour Types Treated by Radiotherapy
For the 29 patients with head and neck cancer, TNM classification was used to stage the tumours prior to commencement of radiotherapy. Three patients (10.3 per cent) were at stage I, eleven (38.0 per cent) at stage II, five (17.2 per cent) at stage III and ten (34.5 per cent) at stage IV. The patient with nasopharyngeal angiofibroma was at stage IV (Fisch's classification)Reference Fisch12 post-surgery but still had residual intracranial extension.
All of the study patients had received conventional radiotherapy (2 Gy/fraction daily, five times a week). All had received irradiation doses of more than 40 Gy with parallel opposed fields or anterior or posterior fields. All of the study patients had been irradiated in the head and neck region (the parotid, submandibular and sublingual glands were included in the field), except for three patients with laryngeal cancer and one patient with thyroid cancer, who had been irradiated only in the neck region. The median total radiation dose was 70 Gy and the median duration of radiation was 59.5 days.
Discussion
This study was conducted to objectively assess salivary gland function among patients who had undergone radiotherapy for head and neck tumours. The confounding effects in this study were age, gender and tumour type. Roesink and TerhaardReference Roesink and Terhaard13 evaluated stimulated salivary function of the parotid glands in 108 patients with head and neck malignancies prior to radiotherapy. They found that the stimulated parotid salivary output was not correlated with patients' physiological or clinical variables, including sex, age, cigarette and alcohol use, parotid volume, tumour location, T and N stage, and pre-radiotherapy surgery. Although the present study measured whole salivary flow rather than parotid salivary flow, both parameters are comparable since most of the salivary output comes from the parotid glands, and the parotid glands are the most affected by radiotherapy. Therefore, it is reasonable to assume that both parameters would produce a similar outcome.
In general, radiotherapy which includes the salivary glands in the radiation field causes disruption of salivary gland function. Reduction in salivary flow has been documented previously during a course of radiotherapy which included major proportions of salivary tissue within the treatment volume. In humans especially, the parotid glands are markedly sensitive to radiation, with a reduction in salivary flow being detected after just one fraction of a course of treatment.Reference Shannon, Tridahl and Starcke14 Table III lists several studies that have demonstrated early salivary flow changes after radiotherapy.
Table III Studies Reporting Altered Salivary Flow Post-Radiotherapy
* As percentage of pre-RT flow. Reprinted with permission.Reference Leslie and Dische15 Pts = patients; RT = radiotherapy; wk = week
All of the patients in the present study had been irradiated in the head and neck region (the parotid, submandibular and sublingual glands were included in the field), except for three patients with laryngeal cancer and one with thyroid cancer, who had been irradiated only in the neck region, thus sparing most of the salivary glands. The four latter patients were found to have no evidence of abnormally reduced whole salivary flow rates. Since in these cases the salivary glands were not included in the radiation field, this would explain why these patients had no evidence of salivary gland hypofunction.
It is evident from the above observations that inclusion of salivary glands in the radiotherapy treatment field is one important predictor of salivary gland hyposecretion, which may later result in xerostomia. ValdezReference Valdez5 stated that the radiation field is one of several important factors which determine the extent of radiation-induced salivary dysfunction. When all the major salivary glands are included within the radiation field, the mean salivary output can be reduced by as much as 93 per cent.Reference Dreizen, Brown, Handlers and Levy7 Patients requiring bilateral irradiation, with the major salivary glands entirely within the radiation field, usually suffer severe xerostomia and salivary dysfunction. On the contrary, where tumours allow part of the salivary glands to be spared from irradiation (such as laryngeal or thyroid cancers), patients usually suffer less salivary dysfunction and xerostomia.Reference Valdez5
Our study used both unstimulated and stimulated whole saliva collection to measure salivary gland hyposecretion. This method – in particular, unstimulated saliva collection – has been proven to be effective for confirming the degree of xerostomia, compared with measurement of salivary output from individual salivary glands (in particular, the parotid glands).Reference Vitali, Moutosopoulos and Bombaidieri19
In this study, mixed whole saliva was collected by the drainage and spitting methods. These methods involved only minimal effort from the patients, required no difficult technique and used very minimal equipment (only a saliva collecting container). Thus, these methods proved to be simple and cost-effective. Other techniques of collecting whole saliva reported in the literature include the use of suction, a cotton roll and other absorbent devices. These methods are more difficult to employ, as they are more technical and require special and expensive equipment. The suction method requires special instruments to collect the saliva. The method using the cotton roll and absorbent devices can also be used for quantitation purposes. In this process, the cotton roll is weighed before and after it has been soaked with saliva. However, saliva is filtered through the cotton fibre in the process, preventing whole saliva from being collected.Reference Cawson, Gleeson and Gleeson20
Salivary flow studies can confirm the functional status of all the salivary glands. The results of the current study demonstrate that salivary gland hyposecretion is common in post-irradiated head and neck tumour patients. Eighty per cent (n = 24) of the study group patients had abnormally low unstimulated whole salivary flow rates (i.e. less than 0.1 ml/min), whilst 42 per cent (n = 17) had abnormally low stimulated whole salivary flow rates (i.e. less than 0.5 ml/min). Although more than half of the study group patients did not have abnormal stimulated whole salivary flow rates, their stimulated salivary output was probably still inadequate. Both the unstimulated and stimulated whole salivary outputs of the study group patients were significantly lower than those of the control subjects. In addition, the unstimulated whole salivary output is a more important predictor of salivary gland dysfunction, as described above. The results of previous studies also suggest that salivary gland hyposecretion or hypofunction is common in post-irradiated head and neck tumour patients. Valdez et al. Reference Valdez, Atkinson, Ship and Fox11 reported a significant reduction of salivary flow in 50 patients with radiation-induced xerostomia, compared with controls. Out of 24 patients with abnormal unstimulated salivary flow rates, 83.3 per cent (20 patients) had no unstimulated whole saliva flow, implying no resting salivary function. These patients can be considered to have salivary gland dysfunction rather than hypofunction.
• Ionising irradiation of salivary gland cells can result in changes in the glands and in saliva composition
• This study was performed to objectively evaluate salivary gland function in post-irradiated head and neck tumour patients
• Both the unstimulated and stimulated whole salivary flow rates of the patients were significantly reduced, compared with normal subjects, suggesting marked salivary gland hypofunction
Two radiotherapy factors were evaluated in the present study and analysed for correlation with the study parameters: the total radiotherapy dose and the radiotherapy duration. We found no significant correlation between these radiotherapy parameters and the salivary flow rates, both unstimulated and stimulated. However, the study showed a weak relationship between unstimulated whole salivary flow rates and total radiotherapy dose (p = 0.114). As the total radiotherapy dose increased, there was a tendency for the unstimulated salivary flow to decrease more than the stimulated salivary flow rates. This finding is consistent with many reported studies which have shown that higher radiation doses are associated with reduced salivary flow rates, particularly the unstimulated rate.Reference Valdez, Atkinson, Ship and Fox11
This can be further supported by an observation in the present study. In all patients in whom the salivary glands were included within the radiation field, the salivary flow rates were obviously reduced, indicating salivary gland hyposecretion, hypofunction or even dysfunction. However, this was not observed in one patient with hard palate cancer and one with nasopharyngeal angiofibroma. The patient with hard palate cancer had an unstimulated whole salivary flow rate at the cut-off point of 0.1 ml/min. The patient with nasopharyngeal angiofibroma had a normal unstimulated whole salivary flow rate (0.16 ml/min). Interestingly, this latter patient received a total radiotherapy dose of 45 Gy; at this radiation level, the patient may still have had some residual salivary function. Many studies have reported that the extent of salivary dysfunction is determined not only by the radiation field but also by the radiation dose.Reference Valdez5, Reference Valdez, Atkinson, Ship and Fox11, Reference Fox21
Valdez et al. Reference Valdez, Atkinson, Ship and Fox11 stated that complete loss of salivary function was observed following irradiation doses over 60 Gy. Doses below 50 Gy were found to have less influence on salivary gland function, with gradual recovery observed a few years after radiotherapy.Reference Markitziu, Zafiropaulos, Tsalikis and Cohen22
Conclusion
The results of this study indicate that salivary gland hyposecretion is a common problem in patients who have undergone radiotherapy for head and neck tumours. The salivary function of these patients (measured by salivary flow rates) is significantly reduced compared with normal control subjects, suggesting marked salivary gland hyposecretion, which in many cases may have progressed to salivary gland dysfunction.
