Nasal mucociliary transport in laryngectomees

European [~]1 It',ltz~r Journal of I ~l[.A~.Jl~r_.~l

Eur J Nucl Med (1986) 11:470-473

Medicine © Springer-Verlag 1986

Nasal mucociliary transport in laryngectomees J. Pavia 1, A. Gareia t, R. Abell61, M. Franch 2, R. de Espafia 2, and R. Herranz 1 1 Department of Nuclear Medicine and 2 Department of Otorhinolaryngology, Hospital Clinic i Provincial, Casanova 143, 08036 Barcelona, Spain

Abstract. A study has been carried out to assess the mucociliary function of nasal mucosa in 30 patients: 16 had undergone surgery due to larynx malignancy and the remaining 14 were divided in two groups. The first group was composed of seven patients suffering from larynx cancer but not operated, and the second group of seven had pathology not related to the airways (control group). The method used a 99mtechnetium sulphur colloid drop deposited on the nasal mucosa of one nostril. To calculate the rate of transport, a new method of quantification has been introduced, based on the formation of a parametric image. The mean velocity in laryngectomees was 3.6 mm/min (range 1.1-6.4). It was compared with the group of 14 patients without tracheostomy (mean 5.3, range 3.3-8.2). An impaired Student's-t-test gave a significant difference between both groups ( P < 0.005). The comparison between patients with nonoperated larynx cancer and normals gave a nonsignificant difference. The 16 operated patients were arbitrarily divided into two groups as per the time elapsed from laryngectomy; seven were studied within the first 3 years of operation and seven other patients after this term. The comparison between them was not significant. Key words: Parametric images - Mucociliary transport Laryngectomees

The purpose of our study is to find a new method for calculating the mucociliary transport rate by introducing a parametric image, and also to assess mucociliary damage in the laryngectomees nasal mucosa.

Materials and methods Of the 38 patients studied, 8 had to be excluded from the study mainly because of movements of the head during detection, and also due to bad deposition of the radioactive drop on the nasal mucosa. The 30 patients whose detection was good enough to be included in the study were males (mean age 45 years, range 18-78), 16 of whom had undergone surgery over a term ranging from 1 week to 30 years. The remaining 14 were patients without tracheostomy, seven of whom also had larynx cancer but had not been operated and seven had different pathology not related to the airways; the latter group was therefore considered as the control group (two had lip cancer, one tympanic perforation, one tongue malignancy, one otitis, one external auditive duct cancer, and one otosclerosis). The method consisted of placing a drop of 99roTe-sulp h u r colloid on the lower turbinate of one nostril, the patient in supine decubitus position and with his head lateral under the gamma-camera. In this way, we intended to avoid any influence of gravity on the drop motion over the mucosa.

Different authors have carried out studies quantifying the transport rate of mucociliary nasal mucosa. These studies included the utilization of metal particles, saccharine, and radioactive-tagged particles, most of which were detected by external probes (Quinlan et al. 1969; Proctor and Andersen 1976; Puchelle et al. 1981 ; Traserra et al. 1982). The average transport rate published for normal subjects showed a very wide range (0.5-23.6 mm/min, Proctor and Andersen 1976). Up to now, no satisfactory method for calculation of the transport rate has been reported. Patients suffering from larynx cancer who have been laryngectomized in a variable term develop damage of the mucociliary nasal mucosa, undergoing atrophy of the epithelium, deprivation of mucous secretion by the caliceal cells, and lesion of the mucociliary transport function. Offprint requests to: Javier Pavia, Hospital Clinic i Provincial, Casanova 143, 08036 Barcelona, Spain

The drop contained 25 lal measured with a micropipette, and radioactivity was approximately 37 MBq (1 mCi). Each study lasted 30 min, with analog images taken every 10 min. At the same time the studies were recorded in a computer, one image per minute in a 64 x 64 word matrix type. We used a Picker Dyna 4/15 gamma-camera fitted with a low-energy, high-resolution collimator, and connected to a Digital PDP 11/34 computer. The images were smoothed in time and space to minimize the error in further calculations. Afterwards, a time-ofarrival parametric image (Pavia et al. 1984) was obtained by means of a F O R T R A N IV program. Each pixel of this parametric image contained the time spent by the tracer in arriving, starting from the initial point where the tracer was placed at the beginning of the study. If the pixel considered was outside the pathway of the drop, it was given a zero value. The method used for calculation is schematically described in Fig. 1. For evey elemental time-activity curve

471 Ac

DIXEL i,j

•..":"if'':'"".. i

oI

~+2sD

; •*.,'.*,*,**,'**" TA

i

i

i L)

Fig. 1. Elemental time-space curve corresponding to the pixel in position i, j through the time of the study. For each one of these curves, the time of arrival is calculated, and the value obtained placed on the pixel in the same position, thus generating the parametric image TA (i, j) [a ÷ [ b [a-

Fig. 2. Time-of-arrival parametric image. Each pixel contains the time spent by the tracer in arriving to the position of the pixel, as measured from the beginning of the study. An increase of the T A values is observed in the direction of the movement

the time of arrival was calculated as the time taken to surpass a threshold ( m e a n + S D ) previously calculated in a background area outside the pathway. If throughout the study no activity reached a determinate pixel, its contents was zero. This showed us the track left by the tracer while passing through the nostril. At the beginning, that track had lower values which increased as the drop advanced (Fig. 2). The pathway of the drop on the parametric image was divided in a polygonal line formed by a maximum of 11 segments. To do this, up to 12 one-pixel ROIs were marked which limited each one of the 11 segments (Fig. 3 A). The first pixel was considered as the beginning of the pathway. To determine the space covered by the tracer, it was necessary to calculate the length of the polygonal. This was done using geometric methods, using the co-ordinates of each one of the pixels under consideration. Therefore, to obtain length units, we had to first calculate the size of 1 pixel, which was performed by placing two spot radioactive sources on the collimator surface, one 10 cm apart from the other. The factor obtained was 2.56 mm/pixel. If we chose several pixels, each one contained a value representing the time spent by the tracer in arriving from the initial point, i.e., the point at which the tracer had been deposited. The space covered in that time was equal to the length of the polygonal between those pixels. In this way, we obtained a maximum of 11 points in the space-time curve (Fig. 3 B). F r o m the time-space curve, we calculated the transport

ta tb te Time Fig. 3A, B. A The polygonal line drawn after selection of the ROIs from the origin 0 to other points a, b .... e, along the trajectory. This line corresponds to the pathway followed by the drop front. The length of each segment is denoted as la, lb ... le. Each of these pixels represents the time taken by the drop to move from the origin 0 to the place it is occupying. B In the time interval t,, the drop moves from the origin 0 to the point a and travels distance 1~. For the time tb, the space covered to b is la+lb. In this way the curve space-time is plotted. From these data the mean velocity of the drop can be calculated

rate in any fragment of the pathway by fitting a linear least-square regression to the points forming that fragment. We validated our method by giving a slow and even movement to a small radioactive source and recording its motion with the computer. The velocity obtained using our method coincided with the velocity given to the source. The patients were always studied in the same room, where the temperature (22°_+ 2 ° C) and relative humidity (45% + 5%) were controlled and remained constant. All the results obtained with the different groups of patients were compared by an impaired Student's-t-test.

Results In several studies we observed that the trajectory followed by the drop (Fig. 4) formed an angle; the horizontal part due to the passage of radioactivity through the nostril and the vertical part probably produced by its passage into the rhinopharynx. We assumed that either the deglutition of saliva or the anatomical configuration of that area could influence the vertical part of the trajectory. Perhaps gravity also affected it, though it is quite improbable. We decided not to include this second part and, therefore, the transport rate calculation refers only to the horizontal portion. The rates of transport of all patients are given in Table 1. For group 1, the mean rate of transport was 3.6 + 1.2 and for group 2, 5.3 + 1.4. The difference was statistically significant (P < 0.005). We arbitrarily divided the first group of patients in two subgroups, depending on the time elapsed since laryngectomy: F o r subgroup a the time elapsed was less than 3 years, and for subgroup b it was 3 years or more. The

472 Table 2. Comparison of experimental subgroups, as determined by time since surgery

Velocity (mm/min) Patient

Fig. 4. An example of the most frequent pattern of pathway. It is composed of a horizontal portion, due to the passage of the drop through the nostril, and a vertical portion, which is not taken into account in this study

Subgroup a Surgery < 3 years

Patient

1

2.2

2

5 7 16 19 20 22 23

4.3 1.1 2.5 4.6 3.6 3.1 6.4

3 4 6 9 10 17 18

n=8 mean 3.5 ram/rain SD 1.6 mm/min

Subgroup b Surgery > 3 years 2.2 4.2 3.6 3.6 4.3 4.3 4.4 3.5 n=8 mean 3.8 mm/min SD 0.7 ram/rain

Not significant Table 1. Rates of transport velocity for all patients

Patient

Velocity (mm/min) Group l Operated

Group 2 Non-operated + control

Table 3. Comparison of controls and unoperated larynx cancer patients

Velocity (mm/min) 1

2 3 4 5 6 7 8 9 10

2.2 2.2 4.2 3.6 4.3 3.6 1.1 4.9 4.3 4.3 4.7 5.8 6.7 6.1 5.7

It

12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 3O

(P