| All comers | Which squamous cell carcinomas are the dangerous ones | Most of these are cured by removing them, and the British Association of Dermatologists says exactly that: most squamous cell carcinomas are low-risk skin cancers and can be cured, and a small number recur locally or spread to the lymph nodes or elsewhere. The useful thing is to know which is which, and two cohorts did the work. In a prospective study of 615 patients followed for a median of 43 months, no tumour 2.0 mm thick or less metastasised at all; metastases occurred in 12 of 318 tumours between 2.1 and 6.0 mm thick (4 percent) and in 14 of 90 thicker than 6.0 mm (16 percent). On multivariate analysis the factors that mattered were increasing thickness (hazard ratio 4.79), immunosuppression (4.32), a location on the ear (3.61) and increasing width (2.22); local recurrence was driven by thickness and by desmoplastic growth (16.11). In a ten-year cohort of 985 patients with 1,832 tumours, local recurrence occurred in 4.6 percent, nodal metastasis in 3.7 percent and death from the cancer in 2.1 percent, and the independent predictors of nodal spread and of death were a diameter of at least 2 cm, poor differentiation, invasion beyond the fat, and an ear or temple location, with perineural invasion also associated with death from the cancer. Those are the words to look for on your pathology report, and they are the ones that decide whether radiotherapy after surgery, imaging or a specialist team referral are discussed. Perineural invasion comes in two forms and the distinction changes the outlook: found only on the slide with no symptoms, or clinically evident through pain, altered sensation or weakness, which is the worse kind and the reason those symptoms are on the red cards. Treatment is surgical: the BAD says removal with a margin of normal skin under local anaesthetic, closed with stitches or sometimes a graft, with curettage and cautery for some lesions and Mohs in some circumstances, and radiotherapy as an alternative. It is the BAD's transplant leaflet rather than its squamous cell one that says when radiotherapy is chosen: for skin cancers that are difficult to remove with surgery or have a high risk of returning after it. A pooled analysis of observational studies put local recurrence at 3.0 percent after Mohs, 5.4 percent after standard excision and 6.4 percent after radiotherapy, while warning that the tumours sent to each were not comparable and that photodynamic therapy, at 26.4 percent, is not a treatment for an invasive squamous cell carcinoma. | | Brantsch et al., prospective study of 615 patients (Lancet Oncology 2008) and Schmults et al., cohort of 985 patients (JAMA Dermatology 2013), with British Association of Dermatologists patient information on squamous cell carcinoma (updated April 2022) | 93 |
| All comers | Radiotherapy after surgery, and what not to add to it | For high-risk disease, surgery followed by postoperative radiotherapy of 60 to 66 Gy. The habit of adding chemotherapy, carried over from head and neck oncology, was tested in TROG 05.01: 321 patients randomised to postoperative radiotherapy with or without weekly carboplatin, with freedom from locoregional relapse at five years of 87 against 83 per cent, hazard ratio 0.84 (0.46 to 1.55, p=0.58), and no difference in disease-free or overall survival. Carboplatin should not be added. The control arm is the number to carry forward: surgery and radiotherapy alone controlled 83 per cent of these tumours locoregionally at five years, and only 7 per cent failed first at a distant site. | | TROG 05.01 (Journal of Clinical Oncology 2018) | 82 |
| All comers | Before surgery, when the operation would be disfiguring | Up to four doses of cemiplimab before an operation planned with curative intent. In 79 patients with resectable stage II to IV disease, a pathological complete response, meaning no viable tumour cells anywhere in the specimen on central review, was found in 40 (51 per cent, 39 to 62) and a pathological major response in a further 10 (13 per cent). Imaging understated it, at 68 per cent objective response. Grade 3 or higher adverse events occurred in 18 per cent. What the trial does not answer is whether the surgery can then be reduced or omitted, because everyone was resected; that is the next question and it is the same one being asked in rectal and bladder cancer. | | NCCN · Category 2A | 89 |
| All comers | The transplant recipient and the immunosuppressed patient | A group with 65 to 100 times the ordinary risk, tumours that are multiple and far more likely to spread, and almost no evidence, because every registration trial excluded them. Three levers. Switching immunosuppression: TUMORAPA randomised 120 kidney transplant recipients who had already had one of these cancers and found new squamous cell carcinomas in 22 per cent after a switch to sirolimus against 39 per cent continuing a calcineurin inhibitor (relative risk 0.56, 0.32 to 0.98), at the price of 60 serious adverse events against 14 and 23 per cent stopping the drug. Chemoprevention: acitretin and nicotinamide, with the nicotinamide effect not reproduced in transplant recipients. Immunotherapy, long refused because of the risk to the graft: a twelve-patient phase 1 study that cross-tapered to a mammalian target of rapamycin inhibitor and pulsed prednisone around each cycle reported no rejection and no graft loss, with responses in 5 of 11 evaluable patients. That is the whole prospective evidence base. | | TUMORAPA (New England Journal of Medicine 2012) and the Dana-Farber phase 1 (Journal of Clinical Oncology 2024) | 89 |
| All comers | What is available in England | NICE technology appraisal TA802 recommendation 1.1 recommends cemiplimab for metastatic or locally advanced cutaneous squamous cell carcinoma in adults when curative surgery or curative radiotherapy is not suitable, only if it is stopped at 24 months or earlier on progression and the company supplies it under the commercial arrangement. It replaced TA592, which had recommended it within the Cancer Drugs Fund. That single appraisal is the extent of NICE's appraisal guidance for this disease, though not of its guidance: pembrolizumab and cosibelimab have not been appraised for it, and the adjuvant indication that C-POST supports has no appraisal either, while HTG333 on electrochemotherapy, HTG99 on photodynamic therapy, NG12 recommendation 1.7.4 and QS130 all cover it. | | NICE TA802 recommendation 1.1 (cemiplimab recommended, stopped at 24 months) | 98 |