So I am seeing a young guy who had a solitary cerebellar met from colorectal cancer. Was subtotally resected initially and had a 3x 3 cm type cavity. I treated the cavity + 3mm to 30Gy in 5 fx. No progression for about 1.5 years, then had slow radiographic progression over serial scans with stable clinical picture. All the typical systemic stuff was tried, but it continued to progress slowly. I had the surgeon look at him and he did a repeat crani with perhaps a more aggressive surgery. He's had several post op complications requiring a graft revision and VP shunt etc... But that has all healed and now he comes back in walking and talking normally, no symptoms, and only slight CN XI and XII asymmetry on exam. Cavity is large on MRI, about 5 cm with peripheral enhancement 8 weeks out from surgery.
Overall, he is >2 years out from previous XRT. Still only site of disease.
Any suggestions? I'm thinking a larger, partial brain volume covering the cerebellar hemisphere and dura in maybe 15 fractions. 2.5Gy fractions to cover edema and dose paint the enhancement a bit. Does that sound reasonable?
1) Colorectal cancer in the young: not on anyone's radar (
great video)
2) You know I like to take clinical situations and reverse engineer them to math...
a) In Hall in chapter 3 there's this cell surv curve graph, and then DNA electrophoresis of irradiated cell lines. The "laddering" of the DNA electrophoresis indicates radiosensitivity. There were prostate, ovarian, lymphoma, and colon cancer cell lines. The lymphoma cell line showed extreme laddering (lots of radiosensitivity). Ovarian and prostate showed some laddering. The colon cancer lines showed ZERO laddering, indicating, surprisingly, radioresistance.
b) One mark of radioresistance in a tumor is a high D0.* The "average" tumor D0 is ~3 Gy. A radioresistant tumor's D0 might be 5 Gy.
c) The amount of cell kill of 30/5 for D0 = 5 Gy is (e^-(6/5))^5 = 0.002. This is not much cell kill. Postop tumor cell number might be as high as 100,000.
0.002 * 100,000 = ~200
Thus 30/5 postop would leave behind about ~200 cancer cells after surgery.
d) If you decided to do 60/30 postop in this situation, and D0 = 5 Gy, and assume 100,000 cells left behind postop, the cell kill would be:
(e^-(2/5))^30 = 0.000006
And 0.000006 x 100,000 = 0.6, or ~60% chance for LC.
I show #2 as an example that 1) in certain situations, hypofractionation is not the panacea we tend to rely on, and 2) depending on a tumor's radioresistance, pure dose escalation trumps hypofractionation. Hypothetical but provocative.
(However, 60/30 would have more late effects than 30/5. Yet there seems something appealing about using standard fractionation in this re-irradiation situation... if you re-irradiate. My take.)
* D0 equals dose that achieves 1/e cell kill
