The Effect of Octreotide on Wound Drainage After Mastectomy

Study design and location: This was a randomised controlled trial carried out at the University College Hospital Ibadan, a premier teaching hospital in South West Nigeria.

Study population: Study participants were recruited from female breast cancer patients who presented through the Surgery Out-Patient (SOP) clinic of the University College Hospital, Ibadan to these general surgical divisions over the study period.

Sample size determination: Assuming a standard normal deviate (Zα/2) at 5% type I error of 1.96, standard normal variate for power (Zβ, at 80% power) of 0.84, standard deviation of 1.9 to detect a difference in mean duration of drain of 1.7 days, and an attrition rate of 5%, a sample size formula for comparing 2 independent means was used to obtain a total sample size of 41 study participants.

Sampling and randomisation technique: Convenience sampling of breast cancer patients presenting during the study period was employed. Study participants were randomized into 2 groups, a treatment and a control group, using blocked sequence randomization. Consecutive female breast cancer patients were prospectively enrolled in turns into the two study groups via the surgical out-patient department of the University College Hospital, Ibadan over the study period till the required sample size was attained. A computer-generated blocked randomisation sequence was used to insert the instruction "1" or "2" (representing treatment or control groups respectively) consecutively to 30 serially numbered envelopes, all of which were sealed afterwards.

Data collection: Data was obtained through a detailed history and physical examination and includes sociodemographic variables like age, occupation, educational status, religion and tribe. Other pre-operative data obtained included BMI, diagnosis, clinical (TNM) stage, histology, tumor grade, comorbidities and ASA grade. Intra-operative data obtained included cadre of surgeon, length of skin incision, intra-operative blood loss, total number of lymph nodes removed, weight of mastectomy specimen and duration of surgery. Post-operatively, the daily volume of wound drainage, post-operative day of drain and suture removal, total drain volume, time to drain removal, occurrence of seroma formation, frequency and duration of seroma aspirations and total number of positive lymph nodes on histology were recorded. All the above data were entered into a questionnaire proforma.

Procedure: All patients were admitted a day prior to surgery. Informed consent was obtained from all patients that met the inclusion criteria. Surgery was done under general anaesthesia and was performed by a consultant general surgeon or a senior registrar supervised by a consultant general surgeon.

Each patient was positioned supine and in slight reverse Trendelenburg position and the ipsilateral arm abducted at 90 degrees with a sandbag under the ipsilateral axilla. The skin was cleaned twice with savlon (chlorhexidine + cetrimide) followed by a single cleaning with 70% alcohol, each time from the level of the suprasternal notch to the umbilicus. The operation field was isolated using sterile drapes. An appropriate breast skin incision incorporating the nipple-areolar complex (with underlying breast mass and previous biopsy scar if any) was made and deepened to the subcutaneous tissue plane using monopolar diathermy. In all patients, electrocautery was used for skin flap dissection up to the clavicle superiorly, upper border of the rectus abdominis muscle inferiorly, sternum medially and the anterior axillary line laterally. Lane's tissue forceps was applied on the nipple-areolar areafor traction while Allis tissue forceps was applied on the flap edges for retraction. Removal of all breast tissue and pectoralis fascia began from the sternal border using a diathermy till the breast tissue was shaved off the pectoralis major muscle. A cuff of adjoining pectoralis major muscle was resected wherever the tumor was attached to the muscle or its fascia. The dissection was carried towards the axilla, stripping the fascia over the upper border of the rectus sheath and serratus anterior muscle, till the clavipectoral fascia was opened to gain access to the axillary fat and lymph nodes. Using blunt dissection, axillary clearance up to level III nodal area was done. During axillary clearance, the pectoralis minor was retracted to enable dissection of levels II and III nodes/fat. Haemostasis was secured using both electrocautery and ligatures. An improvised passive wound drain (urine bag; Meheco®, China) was inserted under the flaps and axilla to run from the lateral side of the wound up to the axilla and down to the medial side of the wound. The drain was brought out through a separate stab wound on the dependent side of the wound in the inferior flap and anchored using Silk 1 suture (Mersilk®, Ethicon). Wound closure was done in one layer using a disposable skin stapler (Advan®, China). All patients had an external wound dressing with meshed gauze over the incision site, followed by Gamgee dressing, which was held in place by a circumferential elastic bandage on the chest wall.

The sealed envelopes bearing the type of intervention was pulled serially with each consecutive patient on the morning of the first post-operative day. Patients in the treatment group were given 100µg of octreotide (Actide, Samarth®, India) intravenously(slowly over 3minutes) 8 hourly for 5days while those in the control group received 1ml of sterile water intravenously 8 hourly for 5days from the first post-operative day. The octreotide and sterile water for injection was administered by a designated surgical resident. Ipsilateral upper limb elevation above the shoulders was commenced with active but gentle upper limb exercises on the first post-operative day. Once daily emptying of the drain and recording of drain output was done by the ward nurses while the patients were on admission. In the first 4days after surgery, patients were shown how to empty the drainage bag and record the drainage volumes at home daily. Measurement of drainage fluid was done by first emptying the fluid in a container, then aspirating it in aliquots using a 10ml syringe till all fluid was aspirated and discarded and the total volume noted and recorded.

In compliance with the current unit protocol in my hospital, the mastectomy dressing was opened for initial wound review on the morning of the 5th postoperative day, following which another firm compression dressing was applied over the chest and axilla, in all patients. In the absence of an indication for continued hospital stay, patients were discharged home on the 5th post-operative day with their drains and were instructed to chart the daily drain output. The unit closed user group (CUG) number and my phone number were given to them in case of an emergency or need for a clarification/complaint. Further wound review was done on the 8th-12th post-operative day after removing the compression dressing. A drain output of \<40ml in the preceding 24hours was a criterion for drain removal.

Subsequent reviews were weekly till 28 days post-operatively at the surgical out-patient (SOP) clinic. During the clinic visits, they were assessed for seroma formation by a senior registrar blinded to the patients' randomization groups. Seroma was defined as clinically evident serous fluid collection at the operation site. Seroma was aspirated under aseptic technique with the volume, number and duration of aspirations recorded for each patient.

Data analysis: Descriptive statistics used for nominal variables (like histology, occupation) and ordinal data (like degree of BMI, ASA grade, disease stage, cadre of surgeon, level of axillary clearance) were proportions, rates and ratio while numerical variables (like age, volume of drain output, time to drain removal, frequency and volume of seroma aspirations) were analysed using means and standard deviations. Tables and graphs were used where applicable. Groups were compared primarily for the time to drain removal and total drain volume using the Mann Whitney-U test. Categorical variables (like incidence of seroma) were compared between the 2 groups using the Chi-square (χ2) or Fishers exact test as appropriate. The correlation between numerical perioperative factors with the time to drain removal and total drain output was analysed using Pearson or Spearman rank correlation as appropriate while the correlation between categorical perioperative factors with the time to drain removal and total drain output was analysed using the Kruskal-Wallis or Mann-Whitney U test as appropriate. Statistical significance was set at a p-value of \<0.05. Version 23 of the Statistical Package for Social Sciences for Windows was used to analyze all data obtained from the study.