A 75 year old female has a Squamous Cell Carcinoma on the dorsal side her right index finger over the proximal phalanx. The size of lesion is 2cm with depth of 3mm and it is mobile on the bone.
There are some things to consider -
- you want to cover the whole tumour area, even if it goes around the finger for some distance (more than 30 degrees of sloping skin is going to be a problem for electrons)
- the area you want to cover is going to be quite small (this is going to be a problem for electrons because of the lack of electronic equilibrium with small electron fields
- the critical structures in the area are the skin on the other side, the bone and the tendon (the tendon itself has a very high radiotolerance because it is predominantly collagen and very few cells) and peritendon/synovial tissues.
The problem/solution also includes orientation. Direct ortho or MV photons will treat through and through. Parallel opposed beams with MV photons will be hard to set up accurately and will need to have a substantial amount of bolus to get over the skin sparing. Bolus might be blocked by the other fingers (proximal or distal phalanx is therefore relevant). Ortho fields will be very hard to set up in a parallel opposed fashion as there is no gantry or stabilisation to assist, although you could use a finger casing to stabilise the finger and have POP borders drawn on.
There is a problem with skin sparing also. Ortho has none when used in apposition, but as a POP the field definition is not anywhere near as sharp. MV photons have good field definition as a POP setup but the penumbral width will be 7mm. The main problem is that the block width for the POP field will be a minimum of 5cm wide (a problem with other fingers in the area).
With an appositional electron beam, you can construct a much better setup. The realities of e- beams for this circumstance are :
- that for 6MeV e- there is a ~0.7cm skin sparing area before reaching the more superficial d90%. The 'plateau' (depth between superficial and deep d90%) ends at ~1.7cm, so the useful treatment depth is only 1cm. So if the required depth is <1cm, a 6MeV e- beam with 7+mm of bolus can provide a beam with great cut-off at depth and very good homogeneity.
- the lesion of 2cm is too small for the electron beam, when a 6MeV e- beam gets below 4cm there is a constriction of the high dose region so that instead of being 0.4-0.6cm from the edge of the field, it moves in towards 1.0-1.2cm from the rim (so if you use a 3cm field with 1cm of lateral high dose constriction at either side, the useful dose will only be across the central 1cm. There are two ways around this problem:
- increase the field size to a minimum of ~4-4.5cm (it is very reasonable to ask the physicists here for detailed advice, it is unlikely that your bosses will actually have these details floating around in their heads)
- use a little known modification (that your physicists can test to show that it actually works!); use the large electron cut-out field but under the bolus place over the area to be treated a 2mm thick piece of moulded lead (1-2mm wax layer on the skin side). It is important that the electron field coming out of the machine be at least 5x5cm (that is larger than the cut out for the lesion on the skin). The dosimetry of the electrons between the Pb cut out will be within 3-5mm of the field edge. (they won't believe you, so insist they do the dosimetry!)
The other option not included is an iridium wire mould (a variation of a single plane implant with wires that are about 0.5cm apart and about 15% longer than the target area), the catheters of which can be inlaid into a wax block that will later sit on the finger surface.
Personally I would use the 6MeV e- and the second approach above (which I am familiar with). I would built a lower wax block to hold the finger (like a bivalve clamping arrangement). The top portion would have a large 2mm Pb shield moulded to the top of the finger and surrounding hand, overlaying this would be wax that filled the troughs to make a flat surface, but over the open area this top portion would be at least 0.7mm thick and as big as the surface PTV + 0.5cm.