Surgery for buffalo hump shows variable results depending on method

Simon Collins, HIV i-Base

It has been suggested that there may be a higher incidence of buffalo hump in the USA compared to Europe, but it has still been reported as a side effect in the UK and referrals for surgical removal on the NHS is appropriate.

Three posters at the conference presented results from various surgical operations to remove cervical fat pads across the shoulders, buffalo hump (BH), dorsal fat pads (DFP) and lipohytrophy of face and neck (submandular fat, often referred to as ‘moon face’).

The case examples shown in these studies all had severe symptoms, involving a significant change in appearance, serious restriction of normal movement, prevention of normal sleep and sometimes pain.

Gervasoni and colleagues from L Sacco Hospital, Milan reported on 14 men and four women who had BH, and two patients who additionally had DFP enlargement. At time of first diagnosis of BH, 16 patients were on PI-containing regimens (indinavir, saquinavir or nelfinavir), with two previously naïve patients on a first combination that was NNRTI-based. No improvement was reported in the 11 patients who switched the PI to an NNRTI for a median 14 months prior to surgery (range 10-24).

Fifteen patients underwent liposuction (preceded by local infiltration of saline solution, cold adrenaline 4°C and lidocaine using the wet technique) and three had classical surgical removal. No surgical or local complications were reported. ARV therapy was continued in all patients after surgery, five including PIs. After a follow-up of a median 12 months (range 8-30) BH returned in only one patient (on an efavirenz/3TC/d4T combination), occurring three months after surgery.

DeWesse and colleagues from St Francis Memorial and Kaiser Permanente Hospitals in San Francisco reported less successful results from 28 patients (median age 47; 23 men, five women) with dorso-cervical fat accumulation (DC), 16 of whom also had submandibular (SM) fat accumulation who were treated with ultrasound-assisted liposuction (UAL) of dorso-cervical, submandibular, trapezio-occipital and mastoid fat deposits.

Liposuction was performed under general anaesthetic and tumescent technique was used. This involves injecting a large volume (two to three litres) of dilute anaesthetic into the area to be treated. The liquid causes the compartments of fat to become swollen and firm or ‘tumesced’ and the expanded fat compartments allow the liposuction cannula to travel more smoothly beneath the skin as the fat is removed. Standard ultrasonic equipment is used to liquefy the fat by cellular fragmentation at a frequency that targets fat cells leaving other tissue and nerve structures intact.

Results were measured by change in patient symptom score, surgeon assessment of pre- and post-operation photographs and patient post-op satisfaction.

Mean symptoms scores (0= no pain 10=greatest severity) improved from 5.3 pre-op to 3.1 at 2-3 months and 2.8 at > 6months. Mean patient satisfaction score (0=dissatisfied, 5=most satisfied) post op (time not specified) was 3.8. Surgeon assessment DC reduction as >75% reduction in 75% cases, and 25-75% reduction in a further 18% of operations, but 7/25 (28%) recurrences were reported. Reduction of SM was less successful with no patients achieving >75% reduction, 29% patients achieving 25-75% reductions and 71% patients achieving <25% reductions.

Of concern was the rate of major complications in 5/28 operations and minor complications in 10/28 operations. Patient photographs on the poster which can often show the most encouraging results for lipodystrophy treatment included three men with severe DS, two of whom showed partial recurrence at one and two years, one of whom showed very successful and encouraging result out to one year.

Piliero and colleagues reported retrospectively evaluated results from 12 ultrasound-assisted liposuctions, similar to that described above, performed on 10 patients (six men, four women; mean age 46, range 37-60) treated at Albany Medical College, New York. Fewer details were provided for the evaluation of results in this study, and although no patient had full resolution of buffalo hump all had initial partial reduction in size. Buffalo hump returned to pre-treatment levels in five patients. Two subjects developed pneumococcal bacteremia at one and three months post-UAL, both despite having received vaccinations twice in their lifetime.


Liposuction to remove buffalo hump is the preferred method for removal as the scarring is minimal, and this is always important, particularly so around the neck and face. However, a common complication of any liposuction is contour deformity – the degree of “bumpiness” – which is difficult to avoid completely and which could also be seen in some of the examples presented. This is almost impossible to correct.

Problems regarding using “dry-wet-suprawet-tumescent” techniques have generally been solved and there is wide acceptance that tumescent techniques generally have fewer complications. Also, more local anaesthetic can be used if local anaesthetic is diluted (wet technique) or very diluted in tumescent technique. Superdilution of local anaesthetic keeps lidocaine longer in the tissues (a lot of it is also “sucked out” together with fat). Adding adrenalin not only causes vasoconstriction helping thus to prevent local haematomas, but also delays absorption of local anaesthetic. It is known that it is possible to inject more than maximum dose of local anaesthetic with the tumescent technique because of this effect. However, it remains unclear exactly how much of the injected local anaesthetic gets into circulation and how much is removed together with the fat. Lidocaine is metabolised by the hepatic cytochrome P450 enzyme, and although all these patients remained on antiretroviral treatment over the period of surgery, the risk from excessive lidocaine dosing was not addressed.

In general lidocaine can cause hypersensitivity reaction as the most common side effect. Cardiac arrhythmia is a potential complication of high doses of lidocaine. In addition fat embolism may be a complication of liposuction itself, particularly if large amount of fat are removed.

A limitation in injecting large volumes to the nape of the neck involves avoiding any circumferential or semi-circumferential injecting to avoid pressure on the neck structures. For this reason it would be expected for these patients to have the posterior and anterior part of the neck done at separate sessions, which again was not the case in these studies. Fat around the neck in Madelung’s disease, which more surgeons are familiar with, is very well vascularised, so bleeding, haematomas and subsequent delayed healing is not that uncommon, and it is unusual that surgical or local complications were not reported here.

Ultrasonic liposuction is more efficient in fibro-fatty tissue (which is the case for buffalo hump), than standard liposuction. But risk of burning or other damaging of skin is higher. Equipment is expensive and because the difference is not significantly noticeable, many plastic surgeons returned to the classical liposuction method.

Liposuction is a specialised area of medicine and even the poster presentations for these studies provided few details of the actual procedures used. Different techniques may or may not be appropriate for HIV-related lipodystrophy and use of more recent techniques involving greater fluid volumes may also be related to the greater complications seen in the UAL studies. The sooner specialised HIV-associated lipodystrophy clinics are established, the sooner effective treatment can be determined.

Comment prepared with the assistance of Lada Lysakova, Charing Cross Hospital.


Unless stated otherwise, all references are to the Programme and Abstracts of the 10th Conference on Retroviruses and Opportunisitc Infections (CROI), 10–14 February 2003, Boston.

  1. Gervasoni C, Vaccarezza M, Galli M et al. Long-term efficacy of buffalo hump surgical treatment in patients continuing antiretroviral therapy. Abstract 723.
  2. DeWesse J, DeLaney A et al. Surgical treatment of HIV lipohytrophy of head and neck. Abstract 721.
  3. Piliero P, Hubbard M, King J et al. Ultrasound-assisted liposuction of HIV-related buffalo humps. Abstract 724.


A brief summary of different liposuction techniques is available at:

Links to other websites are current at date of posting but not maintained.