Sixty-seven studies were included in the review (over 9,000 patients treated with foam sclerotherapy): nine RCTs; eight non-randomised comparative studies; 43 case series; six case reports; and one registry.
Of these, 26 were full-papers published in English (seven RCTs, one non-randomised comparative study, 13 case series, four case reports and one registry). Adequate allocation concealment was reported in one RCT. The one non-randomised study reported clear inclusion/exclusion criteria and clearly described the foam sclerotherapy techniques used. Twelve case series reported clear inclusion/exclusion criteria. All 13 case series clearly described the foam sclerotherapy techniques used. Length of follow-up was considered adequate in most of the studies. No studies reported methods of follow-up or details of how outcomes were ascertained. Completeness of follow-up varied across studies.
Safety
Adverse events such as visual disturbance, transient confusion and other systemic symptoms were reported in rates of less than six per cent (range 0 to 5.9 per cent). Occurrence of headache ranged from 0 to 23.05. Local adverse events were generally more common: minor vein thrombosis (0 to 17.6 per cent); thrombophlebitis (0 to 45.8 per cent); skin matting/staining/pigmentation (0 to 66.7per cent); and pain at site of injection (0.6 to 41.0 per cent) and local neurological injury (0 to 0.7 per cent). Other adverse events included haematoma, local allergic reaction and lower back pain (range 0 to 11.2 per cent). In the comparative studies, the relative risks of most adverse events did not significantly differ between foam sclerotherapy, liquid sclerotherapy and surgery. In the French registry a statistically significant difference in favour of liquid sclerotherapy compared with foam sclerotherapy was found for the risk of visual disturbance. In one RCT a statistically significant difference in favour of surgery compared with foam sclerotherapy was found for risk of skin matting/staining/pigmentation. Overall, the comparative studies were too small to reliably detect differences in rare adverse events at the level of the reported rates.
Serious adverse events associated with foam sclerotherapy occurred in 0 to 5.7 per cent of treatments. No anaphylaxis, intra-arterial injections or deaths were reported. In two conference abstracts (n=253) the median rate of arterial events was 2.1 per cent (range 1.4 per cent to 2.8 per cent). Five English language case series (n=1,316) reported that one patient suffered from pulmonary embolism. Deep vein thrombosis occurred at a median rate <1, ranging from 0 to 5.7 per cent across all studies. In four English language case series (n=781) cutaneous necrosis occurred at a median rate of 1.3 per cent (range 0.3 to 2.6 per cent) and at 0 (range 0 to 0.2 per cent) in five conference abstracts/non-English language studies (n=766). One non-English language study (n=28) reported that one patient suffered from cutaneous ulceration. One case report reported one case of myocardial infarction (30 minutes after injection) and another case report reported one case of an epileptic fit (grand mal) 40 minutes after injection.
Efficacy
The median rate of venous occlusion across all studies was 60 per cent or more. In the English language RCTs the median rate was 84.4 per cent (range 67.4 per cent to 93.8 per cent) and 84.4 per cent (range 60.0 per cent to 98.2 per cent) in the English language case series. No significant between group difference was found for complete venous occlusion when foam sclerotherapy was compared with liquid sclerotherapy (RR 1.39, 95% CI: 0.91, 2.11), based on three small studies. No significant between group difference was found when foam sclerotherapy was compared with surgery involving stripping (RR 0.86, 95% CI: 0.67, 1.10), based on three studies. Evidence of substantial heterogeneity was found in both analyses.
Overall range of recurrence or development of new veins ranged from 0.5 per cent to 51.2 per cent. Overall, the risk of recurrence or development of new veins did not significantly differ between foam sclerotherapy and comparator treatments in the RCTs and non-randomised comparator studies. However, one RCT, with a 10-year follow-up, found that the risk of developing new veins was significantly higher after foam sclerotherapy than surgery (ligation: RR 1.4, 95% CI: 1.02, 1.8; ligation combined with liquid sclerotherapy: RR 1.4, 95% CI: 1.1, 1.9).
One study (n=272) found that patients' return to normal activity was significantly less after treatment with sclerotherapy compared with surgery (median two days versus 13 days). No statistically significant difference in patient satisfaction, disappearance of varicosities or change in disease status was found between foam sclerotherapy and liquid sclerotherapy.