Cryotherapy: one trial was terminated early because the outcomes were worse than expected

A 2018 report in J Plast Reconstr Aesthet Surg from Erasmus MC and others in the Netherlands. It reports two multicentre, randomised, controlled, open trials.

The two trials asked different questions.

• primary keloids — intralesional cryotherapy versus excision + corticosteroid injections
• therapy-resistant keloids — intralesional cryotherapy versus excision + brachytherapy

The primary outcome was scar appearance assessed with the Patient and Observer Scar Assessment Scale (POSAS). Secondary outcomes were patient-reported quality of life (Skindex-29, SF-36, EQ-5D-5L), recurrence rates and scar volume reduction. Power analysis indicated 33 patients in each group were needed.

And the first sentence of the results is the content of this entry.

The trial was prematurely terminated after inclusion of 26 patients due to unexpectedly inferior outcomes after intralesional cryotherapy.

The rest of the results, as written.

• for primary keloids no convincing difference between treatments was found. But surgery improved scar appearance while cryotherapy did not.
• for resistant keloids, excision followed by brachytherapy significantly improved scar appearance (POSAS) and scar symptoms (itch and pain) (p < 0.001, p < 0.001 and p = 0.006 respectively). Cryotherapy did not.
• neither treatment caused indisputable improvements in quality of life.

Conclusion: intralesional cryotherapy is inferior to keloid excision followed by brachytherapy for resistant keloids. In primary keloids, intralesional cryotherapy reduced keloid volume and therefore may be used in these patients and specific cases. Primary keloid group size was too small to draw valid conclusions; further research on the efficacy of intralesional cryotherapy for primary keloids is warranted.

Now how to read it.

First, read the early termination precisely. The trial stopped because the investigators judged one arm's outcomes to be poor. That is strong information. It is also weak information — stopping means the planned numbers were not reached, which lowers the precision of the conclusion.

Second, the authors wrote both sides of that. For resistant keloids they wrote plainly inferior; for primary keloids they wrote that the group size was too small to draw valid conclusions. Within one paper they drew a conclusion on one arm and declined to on the other. This site carries that distinction across as it is.

Third, volume did fall. For primary keloids the authors wrote that cryotherapy reduced keloid volume. Yet it did not improve scar appearance (POSAS). The distinction this site keeps returning to appears here too — measuring and looking can come apart. Volume can fall without it looking better.

Fourth, these were open trials. Surgery and cryotherapy cannot be masked. And the primary outcome includes self-assessment (the patient part of POSAS).

Fifth, the abstract does not state the follow-up period. So the duration field on the badge above is empty.

Sixth, what this trial compared was not cryotherapy against doing nothing. It compared it against surgery. So whether cryotherapy beats doing nothing is not something this trial says.

So it sits at 'not established'. But here that badge means not 'it has not been measured' but 'it was being measured and stopped, and the reason was the result'.

A 2016 randomised controlled trial in Dermatol Surg from JIPMER, India.

60 patients with 101 keloids were randomised into two groups. On day 1 the keloids were ablated with either CO2 laser or cryotherapy, and then both groups had intralesional triamcinolone injected at baseline and at four-week intervals for three months. They were followed for 12 months.

The results. Rates of successful therapeutic response (more than 50% improvement), CO2 versus cryotherapy.

• reduction in thickness — 55.55% vs 70.37%
• reduction in volume — 61.1% vs 77.8%
• patient's self-assessment — 75% vs 77.78%
• observer's assessment — 61.12% vs 85.18%
• Vancouver Scar Scale — 52.78% vs 62.96%
• and at 12 months the difference in therapeutic response between the two groups was not statistically significant

Conclusion: both CO2 laser and cryotherapy in combination with intralesional triamcinolone were found to be equally effective in the treatment of keloids.

Now how to read it.

First, all five figures look higher on the cryotherapy side. Observer assessment especially: 61% against 85%. And yet it was not significant. Numbers leaning one way without separating statistically is common, and this site does not turn a visible lean into a conclusion. These are figures from 60 people.

Second, and this matters more — both arms had steroid injections. Intralesional triamcinolone at four-week intervals for three months. Intralesional corticosteroid is a method this site already covers for keloids.

So what this trial shows is 'there was no difference between putting a CO2 laser in front and putting cryotherapy in front', not 'cryotherapy reduces keloids'. A substantial part of the improvement may belong to the steroid.

This site drew the same distinction on the pycnogenol, N-acetyl glucosamine and laser-assisted drug delivery pages. 'What was it added on top of' is a recurring question here.

Third, 12 months of follow-up. The longest period on this page, and that matters for keloids — recurrence is a common problem.

Fourth, 60 people with 101 lesions. There are more lesions than people. Counting several keloids in one person separately can inflate the statistics. How the abstract handled that is not stated.

Fifth, it was randomised but the abstract says nothing about masking. Whether the observer assessment was blinded is unknown. We do not invent it.

So it sits at 'moderate'. Good conditions — 60 people, 12 months — but both arms had steroid, and the masking is unclear.

Put the two entries side by side and you can see where this procedure sits. Against surgery and radiotherapy it came out inferior; alongside CO2 laser it tied — but in the trial where it tied, both sides were having steroid injections.