investigated the potential of Gallium-67 (Ga )-Citrate as a bone scanning agent and noted its concentration in the cervical lymph nodes of a patient with Hodgkin's
6 7
disease. Ga was subsequently described as "tumour- seeking", not only for head and neck malignancy
(Kashima et al, 1974) but for tumours in general (Andrews and Edwards, 1975). It has been extensively evaluated
as a tumour imaging agent in h e a d and neck squamous carcinoma and its uptake has been r eported and confirmed in a variety of primary sites and sub-sites including metastatic
lymphadenopathy within the neck (Kashima et al, 1974? Kornblut et al, 1974? S i l b erstein et al, 1974? Smith
et al, 1975? Higashi et al, 1977a-b? Teates et al, 1980).
6 7
However, Ga has d istinct disadvantages. It has a normal biodistribution w i t h i n the head and neck which includes the nasopharynx, the lacrimal glands and the salivary glands (Kashima et al, 1974), the uptake in the
167
latter two being permanently enhanced by radiotherapy (Kashima et al, 1974; Bekerman and Hoffer, 1976). Uptake also occurs, not only in a variety of malignant tumours including lymphoma, but also in benign tumours and inflammatory tissue (Kashima et al, 1974?
Kornblut et al, 1974; Ohta et al, 1984a; Ohta et al, 1988). Lesions less than 2 cm in size are not usually detected
with planar scintigraphy (particularly cervical lymph n o d e s ) , by which time they are usually clinically palpable
(Cummings et al, 1981). Sensitivity for all lesions is significantly increased for masses greater than 3 cm in size (Teates et al, 1980) and is also affected by site
(Bland and Rose, 1981) , being highest for lesions not situated in or near either the salivary glands or the nasopharynx. Such phenomenon all contribute to a low sensitivity and specificity w h i c h varies from 53% and 89%
(Silberstein et al, 1974), 56% and 64% (Teates et al, 1980) and 85% and 51% (Endo et al, 1985). In Silberstein*s
paper all the patients (52) had proven squamous carcinoma. However, in Teates* paper, only 51% had proven head and neck squamous carcinoma (the rest w e r e unspecified, except
8% had lymphoepithelioma (lymphomas were excluded)),and in Endo's paper 75% of patients had squamous carcinoma
(the rest were unspecified? lymphomas were e x c l u d e d ) . 6 7
clearance which delays the scanning time up to 48-72 hours. Consequently, its current clinical use in head and neck tumour imaging is largely confined to the
evaluation of lymphoma (Watkinson, 19 90)/ although uptake in squamous carcinoma continues to be evaluated
(Solfanelli et al, 1987).
Many other radiopharmaceutical agents have been used to evaluate squamous carcinoma of the head and neck.
57
Cobalt-57 (Co )-Bleomycin (Poulouse et al, 1975? Sawas-Dimopoulou et al, 1978; Woolfenden et al, 1979? Cummings et al, 1981); Indium-III (In II1)-Bleomycin
(Goodwin et al, 1981; Hofer et al, 1987); In ^ ^ - T r a n s f e r r i n 197
(Goode et al, 1973); Mercury-197 (Hg )-Dichloride (Clj) (Johnson et al, 1965? Aversa et al, 1978);
Q Q m Q Q m
Tc -Bleomycin (Lin et al, 1974) and Tc - T c O “
(Grant and Smith, 1974) have all been evaluated but, like 6 7
Ga , have distinct disadvantages. They all exhibit a low sensitivity and specificity, considerable cost and a
99m
prolonged blood clearance (except Tc - T c O J ) • Of these 57
radiopharmaceuticals, Co -Bleomycin is the one most extensively evaluated and the one which has been shown to
6 7
be more sensitive than both Ga -Citrate and
In III-Bleomycin, with a sensitivity and specificity of 84% and 50% respectively (Cummings et al, 1981). Although
169
it has no normal biodistribution in either the nasopharynx or the salivary glands, it is accumulated by normal thyroid cartilage (Cummings et al, 1981)• In addition, it has a h a lf-life of 270 days and since 80% of the radioactivity is excreted via the urine within 24 hours, the urine must be collected during this period to reduce environmental c o n t a mination (Woolfenden, 1979)• It is not possible to demonstrate lesions, particularly lymph nodes, which measure
57
less than 2 cm m size with Co -Bleomycin
(Cummings et al, 1981). In view of their limitations, 57
neither Co -Bleomycin nor any of the other
r a d i o pharmaceutical agents evaluated in head and neck 6 7
squamous c a rcinoma (except Ga ) have been adopted into routine clinical practice and, therefore, play no role in staging or the d e t e c t i o n of residual or recurrent disease.
6 7
One of the major criticisms of both Ga -Citrate
57 .
and Co - B l eomycin has been their unreliability in the d e t e c t i o n of lesions less than 2 cm in size. One of the reasons for this is that planar scintigraphy provides a t wo-dimensional image of a three-dimensional object and
contr a s t is comp r o m i s e d not only by superposition of activity above, and below, the area of interest, but also tissue
s uperimposition and attenuation, scatter, motion artefacts, image noise or statistics, and spatial resolution
Recent advances in scanning techniques have allowed tomographic nuclear medicine studies using single photon
9 9 m
emitters such as Tc (Sousaline, 1982). This
technique is called Single Photon Emission Computerised Tomography (SPECT) and uses a rotating gamma camera w h i c h acquires 64 projections over 360°, each view for a 5.6° rotation of the camera. This information can be combined with data already available from conventional planar views and using a data processor, mathematically reconstructed tomographic images can be produced in the coronal, sagittal and transaxial planes by utilising a filtered back-
projection technique (Piez and Holman, 1985)• This is based on algorithms developed in radio-astronomy to
reconstruct a two-dimensional brightness distribution over a source from fan-beam scans taken in various position angles (Bracewell and Riddle, 1967). SPECT improves depth interpretation and reduces tissue superimposition artefacts and leads to an increase in sensitivity, image quality and spatial resolution (Heller and Goodwin, 1987).
99m •
Tc Dimercaptosuccinic Acid (DMSA) is a new
radiopharmaceutical which has been used to evaluate head and neck squamous carcinoma and, for which, an increased sensitivity has been reported using SPECT (Ohta et al, 1985a
171
Aw et al, 1986? Ohta et al, 1988), Although its uptake has been reported at a number of primary head and neck sites and sub-sites and early reports are encouraging no studies have been made comparing tumour stage and size w i t h
sensitivity. Further studies are necessary. The hist o r y 99m
of Tc (v) DMSA is discussed m Section 1.6.
Techniques such as lymphoscintigraphy have b e e n used to demonstrate the cervical lymphatics in health and disease and are discussed in Section 1.4.
The concept of using antibodies to identify tumours was conceived by Paul Erlich who invented the term
"magic bullet" (Himmelweit, 1960). The magic bullets used today are monoclonal antibodies, first isolated by Kohler and Milstein at C a mbridge in 1975 using hybridoma
131
technology. Tranter et al (1984) used I monoclonal anti-CEA antibodies to evaluate five patients w i t h primary and metastatic head and neck squamous carcinoma. A lthough positive uptake was demonstrated at the primary sites, at the sites of metastatic cervical l y m p hadenopathy, and at the sites of pulmonary metastases, no lesions were demonstrated which measured less than 2 cm in size and residual and recurrent disease was not detected.
in the development and regulation of human cancer and there are increased levels of EGF receptors, both in tumour
biopsies, and in a number of established cell lines derived from head and neck squamous carcinoma
(Gusterson, 1984; Cowley, 1986). Soo et al (1987) used In 111 labelled E G F R I , (a monoclonal antibody against the EGF receptor (Waterfield, 1982)) to evaluate 11 patients w ith primary and cervical metastatic disease. Although he obtained a- sensitivity of 73%, no lesions less than 3 c m were identified.
Monoclonal antibodies have made a huge impact on the in-vitro diagnosis of head and neck squamous carcinoma
(Price, 1987). In theory, in -vivo radio-immunoscintigraphy provides the ideal answer to both tumour imaging and
target directed therapy, but in practice, it suffers from d istinct problems such as high background activity, low target to non-target ratios and non-specific accumulation
in organs such as the liver and spleen. There is, at present, a variety of techniques attempting to solve some of these
problems (Davies, 1985) , but images are far from ideal and lesion uptake is well below that required for
effective radioimmunotherapy (Vaughan et al, 1987).
173
in the evaluation and staging of head and neck squamous carcinoma, nor in the d e tection of residual and recurrent disease following surgery and irradiation.
1.3.3. PRIMARY DISEASE
The majority of the anatomical boundaries and
pathological staging criteria using CAT, MRI and ultrasound have been based on AJC c riteria (AJCC, 1978; AJCC, 1983).
It is not the intention of this section to discuss the indications for sophisticated imaging techniques (usually CAT or MRI) at each head and neck site. However, in * general terms, their use is only indicated if the result will significantly alter management, or where an accurate knowledge of tumour volume is mandatory for radiotherapeutic purposes. The normal head and neck anatomy, together
w i t h patterns of spread of malignant disease, as seen on CAT