Retrospective cohort study

B - Click here for explanation of classification scheme.

POSITIVE: See Quality Criteria Checklist below.

The purpose of the research was:

• To identify differences in parenteral nutrition [PN] requirements associated with different standard protocols for bone marrow transplantation
• To determine the difference between the number of times PN was indicated and the number of times PN was actually administered
• To identify the characteristics of patients who were able to meet their nutritional needs without the administration of PN

Data was analyzed retrospectively for all consecutive patients who:

• Received intensive cytotoxic therapy in the study facility between 1993 and 1998
  • High-dose chemotherapy with or without total body irradiation
  • Eventually followed by BMT
  • Resulted in a neutropenic period (< 0.1 x 10⁹/1 neutrophils) for at least 7 days
• Participated in a nutritional monitoring program while on the BMT unit

• (1st patient cohort) Patients requiring a second remission induction course due to refractory acute myelogenous leukemia [AML]
• (2nd patient cohort) Only patients receiving their first BMT were included
• (3rd patient cohort) Patients who received a protocol which did not include total body irradiation, or which included an HLA-mismatched protocol

Recruitment:

• Data was analyzed retrospectively for all consecutive patients meeting inclusion criteria

Design:

• The researchers carried out three analyses of the data; each analyses had its own patient cohort:
  • (Analysis #1) Comparative longitudinal analysis of three clinical phases of treatment for acute myelogenous leukemia [AML]
    • Patient cohort for analyses #1 consisted of patients treated for primary AML according to the EORTC LAM-10 protocol
      • Remission induction phase
        • Cytarabine 100 mg/m² by continuous IV infusion/10 days, for total dose of 1000 mg/m²
        • Etoposide 100 mg/m² once daily, days 1 -5 for total dose 500 mg/m²
        • Either idarubicin 10 mg/m² (total dose 30 mg/m^2), daunorubicin 50 mg/m² (total dose 150 mg/m²), or mitoxantrone 12 mg/m² (total dose 36 mg/m²), on days 1, 3, and 5
      • Consolidation phase
        • Cytarabine 500 mg/m² twice daily from days 1 to 6 (total dose 6000 mg/m²)
        • Either idarubicin 10 mg/m² (total dose 30 mg/m^2), daunorubicin 50 mg/m² (total dose 150 mg/m²), or mitoxantrone 12 mg/m² (total dose 36 mg/m²), on days 4, 5, and 6
      • Conditioning phase
        • Cyclophosphamide 60 mg/kg once daily on days -5 and -4 (total dose 120 mg/kg)
        • Single dose total body irradiation 9 Gy on day 0
  • (Analysis #2) Comparative cross-sectional analysis of four conditioning regimens for BMT
    • Patient cohort for analyses #2 consisted of patients undergoing BMT for hematological malignancies
      • Four conditioning regimens were compared:
        • BEAC regimen
          • 300 mg/m² BCNU one daily on day -6
          • 200 mg/m² cytarabine and 200 mg/m² etoposide on days -5, -4, -3, -2, and 140 mg/m² on day -1
          • Autologous bone marrow transplant
        • Cycolphosphamide 60 mg/kg once daily on days -5 and -4, followed by total body irradiation [TBI] 9 Gy on day 0, followed by autologous BMT
        • Cycolphosphamide 60 mg/kg once daily on days -5 and -4, followed by TBI 9 Gy on day 0, followed by allogeneic BMT from an HLA-matched donor
        • Cycolphosphamide 60 mg/kg once daily on days -5 and -4, followed by TBI 9 Gy on day 0, and 5 consecutive days  5 mg Campath-1G (days -8 to -3), followed by allogeneic BMT from an HLA-mismatched matched donor
  • (Analysis #3) Comparative cross-sectional analysis of personal and medical characteristics of patients requiring or not requiring PN in a homogenous group of BMT patients
    • Patient cohort for analyses #3 consisted of patients who underwent a conditioning regimen consisting of Cycolphosphamide 60 mg/kg once daily on days -5 and -4, followed by total body irradiation 9 Gy on day 0, followed by either autologous or allogeneic BMT from an HLA-matched donor
      • Regimens which did not include total body irradiation, or which included an HLA-mismatched protocol, were excluded from this analyses in order to provide a heterogenous patient group
• Nutritional monitoring
  • Energy and protein requirements were estimated
    • Energy: TEE was estimated as 140% Harris-Benedict [HB]
    • Protein: Protein requirement was estimated at 1.5 g/kg actual body weight
  • Goal intake was 70% estimated energy requirement, and 1.0 g/kg protein
  • Oral food intake was recorded three times per week
    • Energy and protein content of oral intake was calculated using a computerized Dutch nutrient table
  • Patients received nutrition counseling aimed at improving oral food/beverage intake
    • Snacks and supplements were recommended on an individual basis
  • Criteria for initiating PN was 2 consecutive oral energy intakes < 50%, with no improvement expected within 1 week
• GI care
  • All patients received prophylactic GI care, including:
    • Stomatitis prevention
      • Local antibiotics 3 times daily
      • Careful tooth brushing
    • Nausea prevention
      • Routine seratonin antagonist
      • Other anti-emitics as indicated
    • Protective isolation
    • Low microbial diet
    • Decontamination of the gut with oral non-absorbable antibiotics
• Criteria for administration of PN (NOTE: These criteria were applied retrospectively to the patients in the data base)
  • Severe malnutrition at start of therapy
    • Presence of at least one of the following: baseline serum albumin <30 g/l; body mass index [BMI] < 18.5 kg/m²
  • OR a prolonged period of inadequate oral intake
    • Three or more measurements of oral intake (over a period of at least 7 days)  <10% estimated energy requirement; or five or more measurements of oral intake (over a period of at least 11 days) with oral energy intake < 50% the individual's estimated energy requirement
  • OR severe weight loss during admission for BMT, defined as weight loss > 10% admission weight between admission and discharge

Blinding used: N/A

Statistical Analysis:

In Analysis #1, non-parametric variables were tested by the Friedman-rank test; dichotomous variables were tested by the Cochran Q test

In Analysis #2, comparison of the four sub-groups was carried out by the Kruskal-Wallis test for non-parametric data, and by the Fisher exact test for dichotomous variables

In Analysis #3, comparison of the two groups was carried out by the Student's t-test for parametic variables, and by the Mann-Whitney test for non-parametric variables

• Throughout the study:
  • Results of parametric data were expressed as means and standard deviation
  • Non-parametric data were expressed as median and inter-quartile range
  • Dichotomous variables were expressed as frequency and percentage

 

 

 

Timing of Measurements:

• Data of all consecutive patients who met inclusion criteria between 1993 and 1998 were analyzed retrospectively

Dependent Variables

• PN requirements associated with different conditioning regimens
• PN indicated vs PN administered
• Characteristics of patients who met nutritional goals without PN

Independent Variables

• Intensive cytotoxic therapy
• Type of conditioning regimen

Control Variables

• Hematological malignancy

 

Initial N:

• Analysis #1:  N = 20 (13 males, 7 females)
• Analysis #2:  N = 93
  • Auto-BMT without TBI: N = 16 (75% male)
  • Auto-BMT with TBI: N = 28 (61% male)
  • Allo-BMT with TBI, HLA-matched: N = 30 (60% male)
  • Allo-BMT with TBI, HLA-mismatched: N = 12 (50% male)
• Analysis #3: N = 58
  • No PN support: N = 16 (50% male)
  • PN support: N = 42 (64% male)

Attrition (final N):

• Analysis #1: Initial N was 50 patients admitted for remission induction; only 20 of these patients proceeded to consolidation treatment and BMT
• Analysis #2:  Initial N was 93 patients; 7 patients died within 1 week following BMT, and their data was not included in analysis due to incompleteness
• Analysis #3: N/A

Age:

• Analysis #1:  Age 44 + 10 years
• Analysis #2:
  • Auto-BMT without TBI: Age 47 (35 - 58 years)
  • Auto-BMT with TBI: Age 46 (34-54 years)
  • Allo-BMT with TBI, HLA-matched: Age 39 (30-47 years)
  • Allo-BMT with TBI, HLA-mismatched: Age 31 (21-38 years)
• Analysis #3:
  • No PN support: Age 42 + 10 years
  • PN support: Age 40 + 12 years

Ethnicity: Not mentioned

Location: Center for Rehabilitational and Nutritional Sciences, University Hospital Utrecht, The Netherlands

 

 Analysis #1: Clinical characteristics of patient cohort (N = 20)*

Clinical Characteristics Days since diagnosis	Remission induction phase 0	Consolidation phase 48 (40-57)	BMT 139 (120-180)	Statistical Significance of Group Difference
Initial weight (kg)	72.5 (61-85)	70.5 (57-82)	72.3 (59-85)	p = 0.009
BMI (kg/m2)	23.5 (21.1-26.1)	22.5 (20.9-24.8)	23.1 921.8-26.7)	p = 0.002
Serum albumin (g/l)	40 (38-46)	41 (39-44)	42 (40-44)	NS
Duration of hospitalization (days)	33 (29-39)	29 (26-31)	36 (30-43)	p = 0.004)
Mean oral intake (%REE)	64 (46-81)	106 (82-139)	93 (71-113)	p = 0.007
Duration PN (days)	18 (12-25)	0 (0-11)	9 (7-26)	p = 0.002
Weight loss (% start weight)	4.0 (1.7-7.4)	2.0 (0.8-4.2)	5.6 (3.8-9.6)	p = 0.001

 *Weight and BMI were lowest at beginning of the consolidation phase.  The consolidation phase reported the shortest length of stay, highest mean oral intake, least number of days on PN, and lowest loss of weight.

 Analysis #1: Number of patients meeting criteria for PN in each phase of treatment**

PN criteria	Remission induction phase (%)	Consolidation phase (%)	BMT	Statistical Significance of Group Difference
Serum albumin < 30 g/l	2 (10)	0	0	NS
BMI < 18.5 kg/m2	3 (15)	2 (10)	2 (10)	NS
>7 days with no intake per os	11 (55)	1 (5)	6 (30)	p = 0.002
>11 days with <50% required energy intake per os	13 (65)	4 (20)	8 (40)	p = 0.01
Clinical weight loss >10%	2 (10)	0	5 (25)	p = 0.04
PN indicated	16 (80)	7 (35)	12 (60)	p = 0.004
PN given	19 (95)	9 (45)	16 (80)	p = 0.002

*The consolidation phase had the lowest requirement for PN.  In all phases, PN was administered more than it was indicated.

Analysis #2: Clinical characteristics of patient cohort (N = 93)*

Clinical Characteristics	Auto-BMT without TBI N = 16	Auto-BMT + TBI N = 28	Allo-BMT +TBI, HLA-matched N = 30	Allo-BMT + TBI, HLA-mismatched N = 12	Statistical Significance of Group Difference
Initial weight (kg)	78 (64 - 89)	77 (67-88)	77 (67-85)	74 (67-86)	NS
BMI (kg/m2)	23.8 (21.9-28.7)	24.7 (22.2-27.1)	24.7 (22.2-26.6)	25.4 (21.2-26.7)	NS
Serum albumin (g/l)	44 (41-47)	42 (40-44)	43 (40-46)	44 (43-47)	NS
Duration of hospitalization (days)	28 (25-34)	36 (29-41)	37 (33-43)	48 (42-82)	p = 0.001
Mean oral intake (%REE)	84 (69-122)	105 (76-140)	94 (74-130)	76 (47-89)	NS
Duration PN (days)	8 (0-14)	7 (0-19)	9 (2-22)	20 (5-55)	NS
Weight loss (% start weight)	4.0 (1.7-7.4)	2.0 (0.8-4.2)	5.2 (3.0-8.2)	5.6 (3.8-9.6)	NS

*Duration of hospitalization was significantly longer for patients undergoing allogeneic transplant.

Analysis #2: Number of patients meeting one criteria for PN in four BMT groups (N=93)**

Clinical Characteristics Days since diagnosis	Auto-BMT without TBI N = 16	Auto-BMT + TBI N = 28	Allo-BMT +TBI, HLA-matched N = 30	Allo-BMT + TBI, HLA-mismatched N = 12	Statistical Significance of Group Difference
Serum albumin < 30 g/l	0	0	0	0	NS
BMI < 18.5 kg/m2	0	1 (4%)	4 (13%)	0	NS
>7 days with no intake per os	5 (30%)	7 (25%)	9 (30%)	6 (50%)	NS
>11 days with <50% required energy intake per os	6 (37%)	11 (39%)	12 (40%)	11 (92%)	p = 0.01
Clinical weight loss >10%	1 (6%)	5 (18%)	5 (17%)	2 (17%)	NS
PN indicated	6 (37%)	14 (50%)	17 (58%)	11 (92%)	p = 0.03
PN given	12 (75%)	19 (70%)	23 (77%)	10 (83%)	NS

**The allo-BMT+TBI, HLA-mismatched group demonstrated the longest period of reduced oral intake, as well as the greatest percentage of patients requiring PN.  In all groups, PN given exceeded PN indicated. 

Analysis #3: Differences between patients requiring vs not requiring PN in patients conditioned for BMT with cyclophosphamide and TBI (N=58)*

Clinical Characteristics	No PN N = 16 (28%)	PN N = 42 (72%)	Statistical Significance of Group Difference
Gender (% male)	50%	64%	NS
Age (years)	42 + 10	40 + 12	NS
Diagnosis: AML Other	56% 44%	67% 33%	NS NS
Transplant type: Auto Allo	56% 44%	45% 55%	NS NS
Serum albumin at baseline (g/l)	41.8 +  4.3	42.4 + 3.0	NS
BMI at baseline (kg/m2)	27.0 + 4.5	23.9 + 3.6	p = 0.01

*The no-PN group maintained oral intake of at least 70% and 1.0 g/kg protein.  BMI at admission was significantly different in the two groups.  There were no significant differences in weight loss during admission between the two groups (-5% in no-PN group vs -6% in PN group).

 

The authors found that PN is not indicated or needed for all patients undergoing intensive cytotoxic therapy.  They concluded that routine screening for severe malnutrition at outset of therapy, coupled with careful monitoring of oral intake during therapy, are practical clinical tools that promote appropriate and individualized decisions regarding the need for PN in this patient population.

Industry:

NUMICO Research Food Company:

• Well-designed retrospective cohort study, with weaknesses inherent in the retrospective design
  • Study results dependent on the accuracy of medical records compiled without intent for use in research
• Clinical measurements such as body weight and oral food intake are prone to error
• The researchers mentioned that the occurence of vomiting/diarrhea may have led to overestimation of energy intake
• Changes in hydration status related to administration of PN may cause errors in measurement of actual body weight; however, the researchers attempted to compensate for this by measuring weight at admission and discharge only
• Large number of eligible patients admitted for remission induction were excluded from analysis #1 (only 20 of 50 patient records were analyzed) due to high drop-out from treatment, including patients who did not reach remission, patients who relapsed early, and patients who died before transplant could occur
  • The researchers compared PN indicated/PN given for included and excluded patients, and identified no significant differences
• Patient population may not have been representative