• Treated in Departments of Surgery and Medical Oncology, Aberdeen Royal infirmary
• T2>cm,T3,T4 or N2 node status associated with a primary tumor of any size (UICC criteria)
• pre and post menopausal women
• Dx confirmed by mammography, breast ultrasonography, fine needle aspiration cytology and/or core biopsy
• Staged with serum biochemistry (urea, electrolytes, liver function tests),full blood counts and differentials and imaging modalities,
• All had echocardiography prior to study entry, were ambulatory, Estern Co-operative Oncology Group 0-1 and Karnofsky>80%.
• All received cyclophosphamide, 1g/m²  (maximum dose per cycle: 1.8 g); vincristine 1.5 mg/m² (maximum dose per cycle: 2 mg); doxorubicin 50 mg/m² (maximum dose per cycle; 90 mg), all given as i.v. bolus injections, followed by prednisolone 40mg orally for 5 days.
• Surgery performed 3 weeks after completion of chemotherapy; type determined by protocol.  Received tamoxifen, 20mg/day,following surgery 
• After surgery, radiotherapy given as 20 daily fractions (4500 cGy to breast and 4500 cGy to lymph draining areas).

• If patient had distant metastases

Recruitment Method not noted although all patients were treated in Departments of Surgery and Medical Oncology, Aberdeen Royal Infirmary.

Design 96 female patients (30-73 years of age, median age 50) with large or locally advanced breast cancer were randomized to 1 of 2 groups before being treated with chemotherapy (doxorubicin, cyclophosphamide, vincristine, prednisolone). 

•  Group 1(n=48) received 30g L-arginine per day for 3 days immediately prior to each round of chemotherapy, 6 pulses at 21-day intervals.
• Group 2 (n=48) received placebo per day for 3 days immediately prior to each round of chemotherapy, 6 pulses at 21 day intervals
• Protocol used to adjust or delay chemotherapy per blood white cell count (WCC) and platelet count(Plts)

Blinding used (if applicable)yes according to a stratified ( for menopausal status), permuted blocks design

Intervention (if applicable) L-arginine orally in dose of 30 g/day x 3 days immediately prior to each cycle of chemotherapy, 6 pulses.

Statistical Analysis

• Power calculations: 80% power of detecting 20% difference in clinical respone rates at .05 level of significance 
• Analyses carried out on an intention to treat basis
• Tumor response rates between L-arginine and placebo group compared using Chi-square test, with p<0.05 accepted as statistically significant.  All tests were two-tailed and data were analyzed using SPSS MS for Windows.

Timing of Measurements

• clinically measured tumor size prior to each cycle of chemotherapy (6 pulses)
• performed mammography and breast ultrasound at baseline and after 4th and 6th course of chemotheropy
• measured histological responses of excised breast tissue for tumor stage and clinical nodes after surgery (3 weeks after completion of chemotherapy cycles) 
• Dependent Variables
• Variable 1: overall clinical response after 6 cycles of chemotherapy using standard UICC criteria (combined complete response  and partial response together) using following measures  to assess response:
  •  product of two maximal perpendicular diameters  of tumor evaluated with calibrated skin callipers (four diameters, at 45 degree intervals) 
  • Product of two maximal perpendicular diameters of mammography
  • reductions in tumor volume on ultrasound
• Variable 2: overall clinical response according to initial tumor size(<6 cm, 6-7 cm, > 7 cm) useing same measures as Variable 1 above.
• Variable 3: overall pathological response after 6 cycles of chemotherapy histologically examining excised tumors using previously described protocol (J. Brittenden et al. Br J Surg 69:918-921,1994),
  • Type I: changes in tumor cells but tumor nests not destroyed; type II: tumor structure destroyed to minor degree; type III: tumor structure destroyed to  moderate degree; type IV: tumor structure destroyed to severe degree; type V: no tumor cells in any specimens
• Variable 4: overall pathological response after 6 cycles of chemotherapy according to initial tumor size(<6 cm, 6-7 cm, > 7 cm)
  • median split performed with
    • a) type I and type II (minimal or minor degree of tumor cell destruction) grouped together 
    •  b) types II, IV and V (moderate to complete destruction of tumor cells) grouped together. 

Independent Variables  dietary supplementation with L-arginine

Control Variables randomized into 2 groups, blinded ( although protocol not described), same protocol for all with exception of L-arginine

Initial N:  96 women  split into 2 randomized groups: L-arginine =48 and placebo group=48

Attrition (final N): 93 ( 3 patients died of MI's so were not included in pathological responses after surgery) 3% mortality.

Age:  Mean of L-arginine group 49 (31-71) and Mean of Placebo group 50(30-72) 

Ethnicity:  unknown although treated in Scotland

Other relevant demographics:  93 participants assessed for pathology: tumor types: 74 invasive ductal carcinomas, 7 lobular and 2 tubular carcinomas, 10 not able to be assessed because of complete pathological response, 51 participants were node negative and 42 had positive nodes.

Anthropometrics  No significant differences noted.  See p 223, Table I in article for details on tumor stage, clinical nodal status, menopausal status, body weight, and body mass index if needed.

Location: Aberdeen Royal Infirmary in Scotland

Other Findings: Pathological responses by type for all tumors ( sum of both E and C groups):

Type I n=20, typeII n=25, type III n=25, type IV n =13, and type V  n=10

• Well-designed and carefully implemented study
• Study was able to address the hypothesis; however no significant effect ws seen on clinical response or on pathological response in tumors  of 6 or >cm.
• However, L-arginine supplementation significantly impacted pathological response in tumors < 6 cm (p=0.04) with better histopathological responses versus placebo group (88% vs 52%).  This may have important implications for clinical practice since pathological response to treatment are best predictors of likelihood of overall survival.
• Better pathological response in tumors < 6 cm may have resulted, in part, because of an increased drug delivery as a result of a nitric oxide-induced increase in tumor blood flow. L-arginine is the precursor for nitric oxide synthesis.  Changes may be dependent on tumor volume and explain lack of overt benefit in larger breast cancers.
• Due to short length of follow-up, can't determine if L-arginine supplementation will result in increased survival but it is first human RCT study to show that administration of selected nutrients can modulate the response to chemotherapy.
• L-arginine is readily available, cheap and well tolerated when ingested in large doses by patients. ( diarrhea is only side effect directly attributed to it versus toxicities of other protocols).