Appropriate Chemotherapy Dosing for Obese Adult Patients with Cancer: American Society of Clinical Oncology Clinical Practice Guideline

Home , Body surface area

Table of Contents

Data Supplement 1 - Evidence Table

Data Supplement 2- Evidence Table: Pharmacokinetic Studies

Data Supplement 3: Search strategy

Data Supplement 4: Quorom Diagram

Data Supplement 5: Glossary and BSA Formulas

Data Supplement 6: Interpretation of BMI for adults

Data Supplement 1 - Evidence Table

Author, Year Objectives Methods Efficacy Toxicities Conclusions

Lyman, 2003 To assess practice patterns in Review BMI as a measure of obesity was associated with increasing reductions in RDI. Patients with high BMI or BSA of 2 m2 or greater also experienced Multivariate analysis identified several independent predictors for reduced adjuvant chemotherapy for greater reductions in dose-intensity. RDI, including increased age; chemotherapy with CMF, AC, CAF; a 28-day ESBC and to define the Chemotherapy The association between BMI based on WHO criteria and RDI was related to a reduction in schedule; BSA > than 2 m2; and no primary CSF prophylaxis. incidence and predictive AC, CMF, CAF, planned dose-intensity. Most of the difference in RDI observed with increasing obesity was in factors of RDI planned dose, with virtually no difference in the need for unplanned N= 1,243 community oncology practices surveyed nationwide; data extracted Dose reductions ≥ 15% occurred in 36.5% of patients, and there were treatment delays ≥7 days dose modifications owing to toxicity. Therefore, there seems to be a from records of 20,799 ESBC patients treated with adjuvant chemotherapy in 24.9% of patients, resulting in 55.5% of patients receiving RDI less than 85%. Nearly two consistent tendency to underdose obese or larger patients. thirds of patients received RDI less than 85% when adjusted for differences in regimen dose- Assessments included demographic and clinical characteristics, chemotherapy intensity. dose modifications, incidence of febrile neutropenia, and patterns of use of colony stimulating factor (CSF). Dose-intensity was compared with published reference standard regimens.

Lyman, 2004 To survey practices nationwide Review MVA for FN - BSA >2 m2 Adj OR=0.85 (95% CI: 0.7-1.04) Dose reduction ≥15% The primary outcome of the survey was the average reduced relative dose- treating patients with NHL and Average RDI for each regimen equaled chemotherapy dose delays ≥7 intensity ( RDI) for each treatment regimen. Also assesses the incidence of establish RDI and incidence of Chemotherapy Risk of FN in patients w/ BSA <=2 m2 (22.6%) days. FN, patterns of CSF use, and average RDI in high-risk sub groups. FN and CSF use CHOP, CHOP-R, CNOP No data on overweight or obese RDI Patients with aggressive and potentially curable NHL treated with CHOP, N= 567 oncology practices and 4,522 patients CHOP-R, or CNOP frequently receive reduced RDI. Predictive models based Patients with aggressive-histology NHL treated with CHOP and CHOP-like chemotherapy in on the risk factors identified for reduced RDI should enable the targeted use clinical practices throughout the United States found substantial reductions in chemotherapy of appropriate supportive care, facilitating the delivery of full chemotherapy dose-intensity in half of the patients. This large practice-based study has shown that both doses on schedule. planned and unplanned chemotherapy dose modifications occur, resulting in reduced dose intensityI. A number of factors were found to be significantly associated with reduced dose- The multivariate analysis reported here suggests that, after adjustment for intensity in both univariate and multivariate analysis, including older age, poor performance age and the other risk factors considered, the prophylactic use of CSF is status, advanced disease stage, year of treatment, and practice site. The age distribution of associated with less reduction in RDI. Therefore, the significant impact of patients in this study reflects that of the broad population of patients in the United States with primary prophylaxis on FN risk and reduced dose-intensity after adjustment NHL: More than half are age 60 or older. Given that advanced age is associated with a higher for other known risk factors is reassuring. incidence of neutropenic complications, the delivery of standard dose-intensity may be particularly problematic in the elderly.

Author, Year Objectives Methods Efficacy Toxicities Conclusions

Modesitt, To evaluate the association Retrospective Review There was no significant difference in OS across BMI groups for etiher stage III ( P=0.02) or Toxicities decreased with increasing BMI, perhaps secondary to BMI was not predictive of PFS, although morbidly obese patients had 2007 between BMI and outcomes in stage IV P=0.87) or recurrent patients (P=0.77). but after adjusting for age , status, stage capped dosing. decreased OS in primary stage III/IV patients. Toxicities decreased with women with advanced or Chemotherapy: histologic type, grade and protocol, there was significant evidence that BMI was assosciated increasing BMI, perhaps secondary to capped dosing. recurrent endometrial cancer cisplatin with OS P=0.02 among patients with primary stage III or or IV disease. The morbidly obese Signigicant decrease in grade 3 to 4 leukopenia (37% vs 60.3%; p=.018 treated with doxorubicin (Adriamycin) patients were at increased risk of death compared to the normal weight patients ( HR 1.86, adjusted for age and prior RT) and neutropenia (46.3% vs 77.8% doxorubicin/cisplatin. A 95% CI 1.16-2.99). The association between BMI and OS was not evident for recurrent patients adjusted p=.002) was also observed in those patients with capped secondary objective was to Total N=945/949 (P=0.63). No significant relationship between BMI and PFS nor treatment response )P=0.92)in dosing. report the proportion of (N=131 from GOG107, N=193 from GOG122, N=169 + N=171 from GOG139, either group. women whose chemotherapy N=157 from GOG163, N=128 from GOG177); 945 in analyses (4 excluded due Grade 3-4 leukopenia or neutropenia were less frequently observed dosing was capped (<95%) and to missing height or weight data). Of 117 patients with BSA > 2.0, dosing was capped in 43.6% for doxorubicin, 44.4% for amont patients with higher BMI (p<.001; test for trend adjusted for to determine if that was cisplatin, and 46.2% for either one. Compared with patients with uncapped dosing, capped age, prior RT and protocol); 90.8% of normal weight patients (BMI associated with outcomes. Endometroid 475/949 (50.1%), BMI 30.4+-8.2; Clear cell 38/949 (4.0%), BMI dosing was not statistically significantly associated with a worse prognosis (HR 1.46, CI 0.89 to <25) suffered Grade 3 or 4 toxicity which decreased to 72.2% among 27.1+-6.1; Serous 180/949 (19.0%), BMI 29.1+-6.5; Other 256/949 (27.0%), BMI 2.41; P=.133) adjusted for age, pre-treatment performance status, disease stage, histologic morbidly obese patients (BMI >=40)/ Women above normal weight 29.5+-7.0; p=.024; tumor grade p=.014 type, tumor grade, and protocol). had more GU AEs (any grade) than normal-weight patients (18.3% vs 9.6%, adjusted p=.014) stage III 189/949 (19.9%); stage IV 227/949 (23.9%); Recurrent 533/949 (56.2%); p=.047

Lindman, To investigate (Phase II Study) Prospective - Phase II Study Overall response rate for 21 patients was 81%; CI, 66% to 96%), including six complete Dominating grade III/IV toxicities were nausea (12% of patients) and Delivery of standard doses of chemotherapy based on patients' BSA will 2007 a tailored and dose-escalated responses (23%; CI 7 %-39%). For the 13 patients receiving the the highest dose intensity, 11 febrile neutropenia (31% of patients). G-CFF used. result in marked inter-individual variations in toxicity, which may be regimen that is active and Chemotherapy: responded compared with 10 patients among the 13 patients with the lowest dose intensity.No explained by the large differences in drug clearance and sensitivity between feasible in metastatic breast 5 FU with or w/o leucovorin (F) significant correlation between Epirubicin and Cyclophosphamide intensity and OS or TTP. patients. Retrospective studies in early breast cancer have indicated that lack cancer and to provide a Epirubicin (E) of toxicity could be a sign of undertreatment which could compromise pragmatic way of overcoming Cyclophosphamide (C) Median TTP 14 months; median OS 36 months, with a median f/up of 113 months. survival. shortcomings of BSA-based dosing. N= 26 patients Delivered dose intensity was high but varied substantially between patients (ranges F 126-202, E 14.4-36.0, C 160-510 mg/m2/w). The given average dose intensity was F 185 mg/m2/2, E 26.4 mg/m2/2 and C 336 mg.m2/2 (representing 93%, 132% and 168% of intended doses of standard regimen).

Adbah- To evaluate the chemotherpy Prospective - Single Arm No data SCRT- dose limiting (e.g. neutropenic fever, severe The conclusion was that calcuation of standard chemotherapy dose Bortnyak, induced toxicity in obese thrombocytopenia associated with significant bleeding and/or according to ABW in obese patients is relatively safe. 2003 cancer patients with dosing Chemotherapy: requiring platelet transfusion) was noted in 16/147 patients (11% in based on ABW cyclophosphamide, methotrexate, 5 FU 21%; cyclophosphamide, doxorubicin, first cycle, 7% in second cycle and 4% in third cycle). The incidence of 5FU 12%; Other doxorubician regimens 16%; 5 FU with leucovorin 14%; Cisplatine regimens 17%; Carboplatin regimens 5%; other 18% Grade III-IV nonhematologic toxicity (e.g alopcia, nausea, vomiting) were not dose limiting in obese patients. N=178/606 (29%) patients were obese (BMI ≥ 27.3 kg/m2 for females and 27.8 kg/m2 for males)

Arm A: N=147/178 obese cancer patients 24% received full doses and were evaluable

Author, Year Objectives Methods Efficacy Toxicities Conclusions

Madarnas, To determine the distribution Retrospective % of group with 2 or 3 drug reducations - Overall 373/532 70%, <1.5 BSA 26.44 (59%) has Overall 373/532 70%, <1.5 BSA 26.44 (59%) has reduction, BSA 1.5- No differences in dosing patterns of epirubicin or doxorubicin were noted on 2001 of body size and prevalence of reduction, BSA 1.5-1.9 311/449 had 69% reduction, and >2 BSA 36.39 had 93% reduction 1.9 311/449 had 69% reduction, and ≥ 2 BSA 36.39 had 93% reduction the basis of BSA above or below 2m2, nor on the basis of BMI above or obesity in the breast cancer Chemotherapy: below 25 kg/m2. Significantly greater reductions in the dose of methotrexate population and to determine 5 FU with or w/o leucovorin (F) Results of subgroup: 7% received a level II/III* reduction at the start of their adjuvant chemo. among women deemed large on the basis of BSA and BMI. Methods for clinician chemotherapy dosing cyclophosphamide (C) While the majority of women in both BSA categories received ≥ 85% of ideal dose at cycle 1. reducing doses were rarely stated in records. When identified the rationale patterns. methotrexate(M) The mean reduction in dose at cycle one was significantly higher in the high ≥ 2m2 BSA group was a desire to "scale back" the dose for a large individual. (5.3± 11.3% vs 9.9 ± 11/3% for < 2m group. P=0.02. AND 4.3 ± 8.2% versus 6.7 ± 13.1 % in the N=10,128 - Initial population was 3,048 patients with invasive breast cancer BMI <25 and > 25 , respectively P=0.008. Only 24% of chemotherapy reductions of ≥ 15% were had received systemic therapy (neoadjuvant, adjuvant, or palliative); 63% = in the BSA ≥ 2m2, 76% were in the BMI ≥ 25 kg/m2 group. 1922/3048 were evaluable (mean BMI 26.4.± 5.3 kg/m2, 54% were overweight, 2% severely obese, 18% moderately obese. Mean BSA was 1.7 ± 0.2m2 and *Level 1 ≥ 85% of ideal dose; Level II 65-84% of ideal dose, Level III < 65% of ideal dose only 5% had a BSA ≥2m2 Some dose reductions may have been overestimated because of rounding Cyclophosphamide Subgroup treated for curative intent N= 532 ;27 ± 5.4 kg/m2; 58% were tables, but when this was factored out, again the mean dose reduction was hight in the high overweight, 2% were severely obese; 23% were moderately obese. Mean BSA BSA ≥ 2 m2 (2.9 ± 7.6% vs 8.3 ± 10.5% for lower BSA. P =0.003 1.7 ± 0.2m2 and 8% had BSA ≥2m2 Median follow up is 74.6 months and median survival has not yet been reached but an initial 2 52% CMF; 3% MF; 18% AC; 3% CAF.FAC; 24% CEF/FEC; 45% had survival analysis was done on the basis of body size with no apparent delterious effect of anthracylcline based chemo with epirubicin ( CEF) in 55% . increased BMI on OS.

Wright, 2008 This study assessed the Retrospective No stastically significant difference in PFS or OS across 3 groups Significantly less thrombocytopenia, dose delays, and dose Obese ovarian cancer patients treated with carboplatin experience association between BMI and modifications in obese women. substantially less toxicity than normal weight women. The lower loxicity outcome for ovarian cancer Chemotherapy: "Trend" toward increased risk for disease progression in obese patients suggests that obese patient may be receiving a substandard dose. patients treated with carboplatin : AUC 7.5 carboplatin-based paclitaxel (Taxol) : 175 mg/m2 RR: 1.25, 95% CI: 0.93-1.69, P=0.14 in obese patient compared to normal chemotherarpy based on after adjustment for age, PS, histology, and residual disease Jelliffe formula, so obese N= 387/392 ovarian cancer patients women often receive Arm A: grade 3/4 thrombocytopenic 49.5%, leukopenia 70.1%; subtherapeutic dose.) Arm A (N= 194) : BMi<25.0 neutropenia 93.3% Arm B (N=122) : BMI 25.0-29.9 Arm B : grade 3/4 thrombocytopenic 32%; leukopenia 51.6%; Arm C (N=71) : BMI>=30.0 neutropenia 91.8% Arm C : grade 3/4 thrombocytopenic 26.8%; leukopenia 43.7%; neutropenia 81.7% Significance: thrombocytopenic P<0.0004 ;leukopenia P <0.0001;neutropenia P< 0.01

Author, Year Objectives Methods Efficacy Toxicities Conclusions

Meyerhardt, To study the influence of BMI Retospective The administration of 95% of the expected first 5-FU dose was not predictive of overall Obese patients had significantly lower rates of Grade 3–4 leukopenia Among women with Stage II–III colon carcinoma, obesity was associated with 2003 on recurrence, OS, and toxicity mortality or disease recurrence in the multivariate model. There were no differences between and lower rates of any Grade 3 toxicity compared with patients of a significant increase in overall mortality as well as a borderline significant in a large adjuvant Chemotherapy: men and women with regard to the completion of therapy (78.1% men vs. 76.5% women; normal weight increase in disease recurrence. Nonetheless, obesity was not associated with chemotherapy trial of patients 5 FU with or w/o leucovorin : LDLV: low-dose leucovorin; HDLV: high-dose P=0.30 and underdosing of chemotherapy (3.6% men vs. 2.6% women); similar results were any increase in chemotherapy-related toxicity. The authors hypothesized with high-risk, stage II and leucovorin; LEV: levamisole; 5-FU: 5-fluorouracil; observed when this comparison was restricted to obese patients that this gender interaction may be related to the effects of estrogen. stage III colon carcinoma studied. Immunotherapy: Despite protocol specifications to dose patients based on their actual weight, adjuvants (BCG) there were modest differences in the percentages of women who were underdosed (2.9% of women with BMI 21.0 kg/m2, 2.1% of women with N=3438/3579 evaluable patients BMI 21.0 – 24.9 kg/m2, 1.6% of women with BMI 25.0 –27.49 kg/m2, 2.9% of women with BMI 27.5–29.9 kg/m2, and 4.9% of women with BMI 30.0 Arm A: Underweight - BMI < 21.0 N=488 kg/m2; = 0.18). Arm B; Normal weight BMI 21.0-24.9 N=1166 Arm C: Overweight BMI 25.0-27.49 N=712; BMI 27.5-29.9 N=472 Arm D: Obese BMI >= 30.0 N=600

Meyerhardt, To study the relationship Retrospective Obese men with rectal cancer were also more likely than normal-weight men to have a local Among all study participants, obese patients had a significantly lower Increasing BMI in male patients with rectal cancer is associated with a 2004 between body mass index recurrence (hazard ratio [HR], 1.61; 95% CI, 1.00 to 2.59). In contrast, obesity was not rate of grade 3 to 4 leukopenia, neutropenia, and stomatitis and a decreased likelihood of sphincter preservation and a higher chance of local (BMI) and rates of sphincter- Chemotherapy: predictive of cancer recurrence in women, nor was BMI predictive of overall mortality in either lower rate of any grade 3 or worse toxicity when compared with recurrence. For both men and women, overweight and obese patients preserving operations, 5 FU with or w/o leucovorin men or women. Underweight patients had an increased risk of death (HR, 1.43; 95% CI, 1.08 to normal-weight individuals. experience less toxicity associated with adjuvant chemoradiotherapy, overall survival, cancer Other : leukovorin 1.89) compared with normal-weight patients but no increase in cancer recurrences. Given the suggesting that ABW of fluorouracil for obese patients is justified. recurrence, and treatment- Other : levamisole small percentage of patients receiving less than 95% of the expected first 5-FU dose, there was related toxicities in patients not sufficient statistical power to stratify patients by underdosing or not Not statistically powered to determine effct of other under dosing on with rectal cancer. N=1688/1792 outcomes. Among patients who were normal-weight or heavier, there were no appreciable differences in Arm A: 5-FU (500 mg/m2/day) days 1-5 & 29-33 followed by RT (5040-5400Gy) the rates of chemotherapy underdosing (0.0% of patients with BMI 20 kg/m2, 2.0% of patients + 5-FU (500 mg/m2/day) days 57-59, 85-87 followed by 5 FU (450 mg/m2.day) with BMI of 21 to 24.9 kg/m2, 1.9% of patients with BMI of 25 to 26.9 kg/m2, 2.0% of patients days 1-5 & 29-33 with BMI of 27 to 29.9 kg/m2, and 2.3% of patients withBMI 30 kg/m2: P = .66). Arm B: 5-FU (425 mg/m2/day) and leucovorin ( 20mg/m2/day) days 1-5 & 29- 33 followed by RT (5040-5400Gy) + 5-FU (400 mg/m2/day) days 57-60, 85-88 + . leucovorin (20 mg/m2/day) days 57-60 & 85-88 followed by 5-FU (380 mg.m2/day) and leucovorin (20 mg/m2/day) days 1-5 & 29-33 Arm C: 5-FU (450 mg/m2/day) days 1-5 & 29-33 + levamisole (150 mg/day x 3 days, every 14 days x 4) followed by RT (5040-5400Gy) + 5-FU (500 mg/m2/day) days 57-59 & 85-87 followed by 5-FU (450 mg/m2/day) days 1-5 & 29-33 + levamisole ( 150 mg/day x 3 days, every 14 x 4) Arm D: 5-FU (425 mg/m2/day) and leucovorin )20 mg/m2/day) days 1-5 & 29- 33 + levamisole (150 mg/day x 3 days, every 14 days x 4) followed by RT (5040-5400Gy) + 5-FU (400 mg/m2/day) days 57-60 & 85-88 + leucovorin ( 20 mg/m2/day) days 57-6- & 85-88 followed by 5-FU (380 mg/m2/day) and leucovorin ( 20 mg.m2/day) days 1-5 & 29-33 + levamisole (150 mg/day x 3 days, every 14 x 4)

Author, Year Objectives Methods Efficacy Toxicities Conclusions

Barrett, 2008 To analyze the relationship Retrospective OS:There was no statistically significant association between BMI category and OS (log-rank no data Obese patients with epithelial ovarian cancer do not have a poorer SCOTROC I between BMI and OS was test statistic 6.33, n = 1067). P = 0.10. PFS: There was no statistically significant association prognosis, provided that they receive optimal doses of chemotherapy based Trial assessed in patients with Chemotherapy: between BMI category and PFS (log-rank test statistic 2.31, n = 1067). P= 0.51 on measured GFR and actual body weight. ovarian cancer. carboplatin : all patients N=1067 docetaxel (Taxotere) : DC- N=538 paclitaxel (Taxol) : PC N=538

Arm A: <18.5 BMI, N=59/1075; PC=36%%; DC=64% Result Arm A: Median OS was 32.9 months (95% CI 23.5-42.4, n = 59). Median PFS was 14.7 months [95% confidence interval (CI) 11.6-17.9, n = 59 Arm B: 18.5-24.9 BMI, N=582/1075, PC=51%; DC=49% Result Arm B: Median OS was median not attained in the ideal weight group (n = 576). Median PFS was 14.7 months (95% CI 13.3-16.1, n = 576) Arm C: 25-29.9 BMI, N=305/1075, %PC=50%; DC=50% Result Arm C: Median OS was 30.1 months (95% CI 25.4=34.7, n = 303). Median OS was Median PFS was14.7 months (95% CI 12.6=16.8, n = 303) Arm D: >30 BMI, N=129/1075, PC=53%; DC=47% Result Arm D: Median OS was 34.3 months (95% CI 26.9-41.7, n = 129). Median PFS was16.6 months (95% CI 11.8-21.4, n = 129)

Griggs, 2005 To review retrospective data Retrospective - cohort study Severe obesity was independently associated with a lower likelihood of admission for febrile Occurrence of Febrile Neutropenia: 462/9672 (5%) were hospitalized Overweight and obese women with breast submitted by nurses (multi- neutropenia, even among those subjects given full weight-based doses (odds ratio, 0.61; 95% for FN. Women hospitalized were more likely to be treated with G- cancer often receive intentionally reduced doses of adjuvant chemotherapy. site) on consecutive patients Chemotherapy: confidence interval, 0.38-0.97) CSF at each cycle of their chemotherapy compared with women not Administration of initial and overall full weight-based doses of adjuvant with ESBC recently treated cyclophosphamide needing hospitalization (P<.001). The use of G-CSF increased chemotherapy in overweight and obese women is likely to improve with adjuvant chemotherapy doxorubicin (Adriamycin) Among the 513 practices that had at least 5 patients in the database who were overweight, throughout the first 3 cycles of chemotherapy from 16% in the first outcomes in this group of patients including patients with obese, or severely obese (57% of the practices), 33% of practices did not administer a first-cycle cycle to 25% in the third cycle. While there were no significant localized or locoregional N=9672 dose reduction in any of these patients. An additional 10% of practices reduced the dose in 10% differences in the mean dose proportion for the entire course of disease (stages I, II, and III), or fewer of their patients. In 9% of practices more than 50% of overweight and obese patients treatment between women who were and were not hospitalized for dose, and toxicities. Actual Underweight < 18.5 = 129/9672 (1%); Healthy Weight (>/=18.5 yo <25) = were administered first-cycle dose reductions. Among the 113 practices for which there were FN, the women requiring hospitalization had a statistically height and weight were used 3588/9672 (37%) Overweight (>/= 25 to <30) = 2984/9671 (31%); Obese (>/= at least 5 overweight subjects and 5 obese subjects included in the study, the practice level significantly (although not clinically significant) higher dose to calculate BMI and BSA and 30 to <35) = 1666/9672 (17%); Severely ovese (>/= 35) = 1305/9672 (14%) correlation between the use of first-cycle dose reductions in overweight and obese patients proportion for their first cycle of chemotherapy (0.970 vs 0.953, WHO critieria - Mean 27.9 ± was 0.44 (P<.001).The % of obese and overweight subjects having a first cycle dose reduction P<.001). 6.5 First-cycle dose reductions (defined as a dose proportion of 0.9 compared with overall was not correlated with the # of obese and overweight women subjects from the standard published doses) were administered to 9% of the healthy weight,11% practice. There was, however, a significant negative correlation between the % of obese Overweight, obese, and severely obese women were no more likely of the overweight, 20% of the obese, and 37% of the severely obese women subjects administered first cycle dose reductions and the number of obese women subjects to require admission for FN compared with healthy weight and (P= .001). First-cycle reduction was independently associated with being from the practice (r=0.19; P<.05). underweight women, regardless of whether a dose reduction was overweight (P=.03), obese (P .001), severely obese (P .001), older than 60 years administered. Severe obesity was associated with lower likelihood of (P .001) admission for FN (odds ratio, 0.61; 95% confidence interval, 0.38- 0.97).

Author, Year Objectives Methods Efficacy Toxicities Conclusions

Geogiadis, To evaluate the clinical course Retrospective All patients were treated with doses of chemotherapy based on BSA Obesity at the start of treatment was not associated with increased toxicity 1995 of cohorts of patients treated calculated from actual body weight. No consistent differences from treatment or shortened survival. There is no support for empiric for SCLC to determine if obese Chemotherapy: associations of significance between obesity and toxicity from chemo. chemotherapy dose reduction based on ideal body weight. patients had an increase in cisplatin : after 1986 N= 54 with etoposide plus 2 x daily chest radiotherapy There wer 14 association between BMI or BMI level and toxicity that toxicity cyclophosphamide : before 1986 N= 47, with or without radiotherapy reached significance. P≤ .05 etoposide (VP-16) : after 1986 N= 54 with cisplatin chest radiotherapy

N=262 evaluable patients [146/262 (56%) had limited stage SCLC and 11/262 (44%) had extensive disease]

Arm A: Normal BMI Arm B: Obese BMI Arm C: Severely obese Result Arm A: 19.5% (95% CI 11%-28%) alive at 3 years 22.1% were obese, 10.7% were severely obese. The percentage of the Result Arm B: 31.8% ( 95% CI 12.4% -51.2%) alive at 3 years intended dose of chemotherapy actually given was more that 99% for both Result Arm C: 6.5% (95% CI 0-19%) alive at 3 years cycles of chemo within all cohorts. Notes : The survival of the very thin patients did not differ significantly from the patients with normal weight P=0.3

Rosner, 1996 To examine the data from a Retrospective The overall risk ratio (obese v nonobese) of treatment failure among women who received Among women who received weight-based doses of the most dose- Obese patients initially dosed (within 5%) by actual weight did not large clinical trial to etermine if weigh tbased doses during cycle 1 was 1.02 (95% confidence interval, 0.83 to 1.26), stratified intensive CAF regimen, 47% of obese and 51 % of nonobese women experience excess cycle 1 toxicity or worse outcome compared with chemotherapy dosing Chemotherapy by treatment and adjusted for number of positive nodes, menopausal status, hormone experienced severe nonobese women dosed similarly. The data suggest that obese women who according to actual body cyclophosphamide (C) receptor status, and tumor size. The overall adjusted failure risk ratio (weight-based v reduced hematologic toxicity (grade 2 3) during cycle 1 (P = .51). received reduced doses in cycle 1 experienced worse failure-free survival. weight places obese stage 2 doxorubin (A) initial doses) was 0.73 (95% confidence interval, 0.53 to 1.00) among obese women. We recommend that initial doses of CAF be computed according to actual breast patients at greater risk 5-FU (F) The risks of grade 3 or higher nonhematologic toxicity were similar body weight. of toxicity. among obese N= 1,471/1,572 evaluable Of 397 women randomized to the most dose-intensive arm who received weight-based doses and nonobese women who received weight-based doses. No (within 5%), 119 (30%) were obese, compared with 51 (75%) of 68 women randomized to this apparent relationship between obesity and the risk of grade 3 or 4 3 levels of dose intensity treatment arm but receiving less than 95% of weight-based doses for each cycle 1 drug (P < toxicity (either hematologic or nonhematologic), even though the .001). Among patients randomized to the least dose-intensive arm, obese women composed median total amount of drug received by obese women randomized Arm A: CAF 600/60/600; # of obese patients n = 194 33% of the group that received doses within 5% of weight-based doses for all cycle I drugs, but to the 600/ 60/600 treatment arm and dosed according to actual 84% of women randomized to this treatment arm whose cycle I doses were less than 95% of weight (within 5%) was 120% (range, 111% to 159%) of what they Arm B: CAF 400/40/400; # of obese patients n = 208 weight-based doses were obese (P < .001). Thirty-seven (28%) of 134 obese women who would have received if dosed according to IBW. received less than 95% of the weight-based doses during cycle 1 had been randomized to the Arm C: CAF 300/30/300; # of obese patients n = 190 300/30/ 300 arm, whereas 46 (34%) and 51 (38%) of the obese women who received dose reductions were randomized to the 400/40/400 and 600/60/600 arms, respectively (P= .15).

Author, Year Objectives Methods Efficacy Toxicities Conclusions

Colleoni, To assess dose response in 4 Prospective Obese patients initially treated with expected doses of chemotherapy A higher proportion of obese patients (39% [97 of 249]) received less than 2005 RCTs from IBCSG that (>/=85%) did not have more grade 3/4 toxicity than patients who 85% of protocol specified dose during the first course of CMF compared with according to BMI and ER Chemotherapy: received reduced (< 85%) doses (14% [22 of 152] vs 12% [12 of 97] patients with normal and intermediate BMI (16% [298 of 1891]; P<0.0001. expression of the primary 5 FU with or w/o leucovorin cyclophosphamide respectively; P = 0.62). tumor in premenopausal methotrexate Considering all patients together in analyses stratified by BMI group, reduced patients with node + breast dose in the first cycle compared with protocol specified dose was associated cancer treated with CMF. Arm A: N=368; ER- obese, intermediate, normal≥ 85% of dose Result Arm A : DFS - HR 95% CI 0.68 (0.54-0.86) p= 0.0013; OS- HR 95% CI 0.72 (0.56-0.94) p= with a significantly worse outcome for the ER negative or low ER tumors. 0.0142 Reduction in chemotherapy should be avoided. Arm B: N= 62; ER- obese, intermediate, normal < 85% of dose Result Arm B: DFS - HR 95% CI 0.68 (0.54-0.86) p= 0.0013; OS- HR 95% CI 0.72 (0.56-0.94) p= For obese patients reducing the dose of chemotherapy was associated with a Arm C: N=1149; ER+ obese, intermediate, normal≥ 85% of dose 0.0142 significantly worse outcome for the ER-negative cohort (total population hazards ratio ≥ 85% vs < 85% 0.68 [95% CI 0.54–0.86] for disease free Arm D: N= 252; ER+ obese, intermediate, normal ≤ 85% of dose Result Arm C: DFS- HR 95% CI 1.16 (0.97-1.40) p=0.1028; OS- HR 95% CI 1.16 (0.94-1.44) survival; 0.72 [95% CI 0.56–0.94] for overall survival) but not for the ER- p=0.1707 positive cohort (95% CI 1.16 [0.97–1.40] 1.16 [95% CI 0.94–1.44] for overall survival) [interaction p values=0.0001 for disease free survival and =·0019 for Result Arm D: DFS- HR 95% CI 1.16 (0.97-1.40) p=0.1028; OS- HR 95% CI 1.16 (0.94-1.44) overall survival]. p=0.1707

Sculier, 1999 To compare a combination of Retrospective There was an excellent linear correlation between btween the different ways of measuring Hematological toxicity appeared to be significantly associated with They concluded that for a moderate carboplatin dose in non-small cell lung carboplatin and cisplatin with response. But with the Chatelut method the calcuatled administrered AUC was lower the carboplatin AUC dose delivered. CC Leuopenia Grade 1-2 33%, cancer, the therapeutic index could be improved if dosage is calculated or without a isosfamide for Chemotherapy: Notes : Whichever method was used, carboplatin AUC was not significantly associated with Grade 3-4 7%, CCI Grade 1 -2 43%, gread 3-4 28% according to the AUC. NSCLC and Cervical, Ovarian, carboplatin 200 mg/m2 and cisplatin 50 mg/m2 (CC) antitumour response rate nor patient survival. Whichever method was used to calculate the Endometrial cancer patients. ifosfamide AUC and whether assessed continuously or categorised into three classes, there was no Thrombocytopenia appeared to be highly related to the carboplatin With the Calvert-Cockcroft method, the odds ratio for increased toxicity was significant association between the carboplatin AUC dosage administered and the objective AUC dose. CC Grade 1-1 21%, Grade 3-4 12% , CCI grade 1-2 46%, OR 2.90 (P =0.003) in favour of the CC arm and OR 2.04 (P= 0.001) in favour Accessed according to BSA or N=505 patients response rate. The results in terms of impact of the AUC on response rate, survival and toxicity Grade 3-4 30% of low carboplatin AUC. AUC (Calvert formula using the were very similar. Cockroft to evaluate GFR rate Arm A: N=248 CC regimen combination of cisplatin (30 mg/m2 on days 2 and This effect was observed whatever AUC variable was considered, i.e. With the Chatelut method, the OR was 2.90 (P =0.003) for the treatment arm and the Chatelut equation)) 3) plus carboplatin (200 mg/m2 on day 1) total dosage at course one, total dosage during the first 3 and 2.19 (P< 0.001) for the AUC dose. chemotherapy courses or dose intensity during the first three Arm B: m=257 CCI regimen consisting of the CC combination plus ifosfamide courses. The effect remained highly significant after adjustment for With the Calvert-Cockcroft method, the OR were, respectively, for treatment (1.5 g/m2 on days 1, 2 and 3). treatment arm. arm and AUC dose, 4.01 (P= 0.0001) and 1.41 (P < 0.0001). With the Chatelut formula, they were, respectively, 3.98 (P =0.001) and 1.45 (P< 0.001) For analysis - 3 groups: Calvert AUC <3, N=137 ; AUC 3-4, N=260; AUC>4 N=99. Chatelut: AUC <2.5, N=203, AUC 2.5-3.5 N=151, AUC > 3.5 N=142

Data Supplement 2- Evidence Table: Pharmacokinetic Studies

Author, Year Objectives Methods Efficacy Toxicities Conclusions

Poikonen, To examine the effect (DFS and PK Study - Single Arm No association found between body weight, height, BMI, or BSA and DFS or OS (P> 0.1 for all no data When the blood leukocyte nadir is used as a surrogate marker for the drug 2001 OS) of parameters reflecting comparisons) effect, obese patients receiving intravenous CMF have higher leukocyte the body size, body weight and Chemotherapy: nadirs than the lean ones. Therefore, the drug doses should not be reduced height, BMI, and BSA on the 5 FU with or w/o leucovorin Patients with high BMI had higher leukcyte nadirs than lean patients (r= 0.3, p < .001). High because of obesity, and even when obese patients are treated according to depth of the blood leukocyte cyclophosphamide leukocyte nadirs also associated with high body weight and large BSA, but not height (refer to the scheduled doses they may remain slightly underdosed. nadir in breast cancer patients methotrexate Table 2 for more details). Variation in deprth of leukocyte nadir substantial even between receiving adjuvant patients with similar BMI. Despite adjusting for age and CYC dose, association between BMI chemotherapy, when drug N=340/368 patients [34% (N=115) patients mildly obese (25-30 kg/m2); 8.2% and leukocyte nadir was still highly significant (P < .001). dosing was based on the BSA. (N=27) markedly obese (BMI > 30 kg/m2)]. High BMI associated with high leukocyte nadir in subgroup of patients who had all received Adjuvant chemotherapy consisted of 6 cycles of CYC (600 mg/m2), METHO (40 similar mg dose of CYC (1000 to 1100 mg, n = 121); r = 0.25; p < .001. mg/m2), and FLUO (600 mg/m2) administered IV on day 1 of cycle at 3-week intervals Patients within the highest BMI had the highest leukocyte nadir values (P < 0.001). A high body weight and a large BSA were also associated with high leukocyte nadirs.

Dobbs, 1998 To address questions in PK Study - Single Arm There was no difference in the CV of adjusted and unadjusted clearance (39.4% and 37.7% In 11/32 patients an analysis of actual and predicted neutropenia Unadjusted clearance did not correlate with surface area, height, weight, per patients with advanced respectively) confirmed that unadjusted dosing would have had no significant cent ideal body weight or body mass index. There was no difference in the primary breast cancer: (1) do Chemotherapy: effect on the pattern of myelosuppression. CV of adjusted and unadjusted clearance (39.4% and 37.7% respectively). physical epirubicin : 12.5-120 mg/m2 There was wide variability in both unadjusted epirubicin clearance (mean 49.5 1 h-1. range 17.7- This predicted AUC was accurate, unbiased and had the same CV as the characteristics influence the 91.7) and adjusted clearance (mean 30.5 1 h-1 m-2. range 11.1-58.0). There was a linear There was a strong correlation between epirubicin AUC and the actual AUC. Normalization of epirubicin dosage according to surface area variability in epirubicin N=32 relationship between total epirubicin dose and AUC (r = 0.80. P < 0.001). Over the dose range observed neutrophil nadir (r = 0.85, P < 0.001). This was reflected in appears not to reduce either pharmacokinetic or pharmacodynamic pharmacokinetics : (2) does 12.5-12h mg/m2 there was no difference in the variation of the adjusted and unadjusted the significantly higher AUC of the five patients with grade 3/4 variability. adjustment of dose according Patients were treated with epirubicin 12.5-120 mg/m2 given as a slow bolus epirubicin CV = 39.4% and 37.7%, respectively, P>0.05. NO relationship between unadjusted neutropenia compared with the six with less severe myelosupression to surface area appear to intravenous injection. Treatment dose was selected by the clinician. To reduce epirubicing CL and height (r=-0.01, body weight ( r=-0.15, IBW ( r=-0.17 , BMI )r=-0.16 or age ( (4363 and 1611 ng mll h respectively, P = 0.01 Mann-Whitney test) There is no relationship between total plasma clearance of epirubicin and reduce this variability: (3) what the possible bias of the older and less fit women being treated at lower doses, r=-0.12). any clinical or biochemical parameter or physical characteristic, including would be the effect on some of the younger patients were treated using divided doses. In these Epirubicin AUC had the strongest relationship with both the absolute body surface area. epirubicin pharmacokinetics women the 'study' dose of epirubicin was given on day 1 and the remainder neutrophil nadir (r2 = 0.72) and total WBC nadir (r2 = 0.63). This and pharmacodynamics of administered 48 h later. After completing pharmacokinetic sampling. These relationship was stronger than that of epirubicin dose with neutrophil abandoning surface area dose patients were excluded from the pharmacodynamic analysis of nadir or WBC nadir (r2 = 0.51 and 0.47 respectively). Similarly, normalization? myelosuppression. Treatment continued on a 3-weekly schedule. epirubicin AUC was the only parameter strongly associated with the surviving fraction of both neutrophils and total WBC (r2 = 0.62 and 0.57 respectively).

Author, Year Objectives Methods Efficacy Toxicities Conclusions

Gibbs, 1999 To measure oral clearance PK study Data suggest that the apparent reduction in busulfan CL/F in NHL Routine dosing of oralbusulfan on the basis of BSA or AIBW in adults and (CL/F, Ml/min) of busulfan of patients could result in enhanced toxicity after a fixed 1 mg/kg dose. adolescents does not require a specific accommodation for the obese. cancer patients - breast Chemotherapy: The ratio of BW:IBW for each respective BMI category was: However, the small number (n = 10) of NHL patients included in this However, dosing based on BSA may be improved by considering CL/F (ER/HER2 status and pre or busulfan analysis suggests that this issue differences in certain diseases. Adjusting dose for body size or disease does post menopausal)(n = 55); should be investigated further. not diminish interpatient variability sufficiently to obviate plasma level ovarian (N=7); N=279 adolesent and adults monitoring in many indications.In conclusion, absolute busulfan CL/F is Leukemia: AML (n = 60), CML Busulfan CSS was related to regimen-related toxicity and graft elevated in obesity. There appears to be a potentially important difference (n = 73); MDS (N=49); NHL Arm A (N=7): Underweight Results Arm A: 78.4% ± 7.4% (range, 65.6% to 88.1%); no significant diffference in CL/F; CL/F rejection in 42 patients with a variety of diseases undergoing between NHL patients and those with CML in busulfan CL/F expressed (N=10; ) relative to BW (mL/min/kg) was 32% higher hematopoietic stem cell transplantation. Severe-grade 3-4 RRT was relative to BW, BSA, or AIBW. Even when expressed relative to BSA or AIBW, Multiple Myeloma (N=25) Arm B (N=173): Normal Results Arm B: 106% ± 12% (range, 76.4% to 138%); no significant diffference in CL/F only observed in patients with CSS > 900 ng/mL. interpatient variability in busulfan CL/F expressed relative to any measure of Arm C (N=89): Obese Results Arm C: 136% ± 13% (range, 117% to 166%; Absolute CL/F was elevated in obese was body size is large relative to the therapeutic window in certain indications 17% higher; CL/F relative to BW (mL/min/kg) was 12 higher% The need for adjusting busulfan dose based on AUC or CSS measured in the Results Arm D: 177% ± 23% (range, 151% to 215%); Absolute CL/F was elevated in severely individual patient remains in certain settings regardless of body size Arm D (N=10): Severely obese obese was 32% higher; CL/F relative to BW (mL/min/kg) was 21% higher measure.

There was not a statistically significant difference in CL/F relative to BSA (mL/min/m2) or AIBW (mL/min/kg AIBW) among underweight, normal, obese, and severely obese patients.

There was a statistical difference when comparing the mean CL/F expressed relative to BW in patients with NHL to those with AML (2.15 ± 0.22 v 2.82 ± 0.66 mL/min/kg, P < .045, respectively), and NHL v CML (2.15± 0.22 v 2.92± 0.70 mL/min/kg, P< .007, respectively). Significant differences in mean busulfan CL/F expressed relative to BSA were found in patients with NHL in comparison to those with CML (87.9 ± 12.3 v 116 ± 25 mL/min/m2, P< .006, respectively). The mean CL/F expressed relative to AIBW was statistically significantly different in patients with NHL compared with those with BrCa (2.41 ± 0.42 v 3.15 ± 0.62 mL/min/kg, P<.017, respectively), NHL v CML (2.41 ± 0.42 v 3.20 ± 0.70 mL/min/kg, P <.006, respectively), and NHL v MM (2.41 ± 0.42 v 3.24 ± 0.61 mL/min/kg, P <.012, respectively).

Mathijssen, To evaluate relationships PK Study -Phase II The mean irinotecan clearance was 33.6 ± 10.8 L/h, with an interindividual variability of 32.1%. no data The nonscientifically based BSA-based dosing strategy should be replaced by 2002 between various body-size When clearance was adjusted for BSA, the interindividual variability was similar at 34.0%. alternative strategies. Despite the lack of basic fundamentals, BSA-based measures and irinotecan Chemotherapy: dosing still seems “untouchable” in clearance and metabolism in irinotecan @90-minute IV infusion (dose range, 175 to 350 mg/m2 clinical oncology. cancer patients, and to provide future dosing N=82 patients recommendations for this agent.

Author, Year Objectives Methods Efficacy Toxicities Conclusions

Rudek, 2004 To explore the influence of PK Study - Single Arm The mean clearance was 63.6 ± 22.7 L/h for doxorubicin and 42.8 ± 14.9 L/h for docetaxel. no data It is difficult to make specific recommendations for dosing changes of age, body size, concomitant Normalisation BSA reduced the interindividual variability by only <1.7%. Doxorubicin clearance doxorubicin- or docetaxel-containing chemotherapeutic regimens. Although drugs, dose, infusion duration, Chemotherapy: was significantlyreduced when administered at doses >50 mg/m2 or in combination with monitoring of plasma concentrations and dosage adjustment may be and sex on the clearance for docetaxel (Taxotere) cyclophosphamide. Upper extremes of body size were associated with increased clearance for necessary to optimise anticancer efficacy in patients, therapeutic drug doxorubicin and docetaxel in doxorubicin (Adriamycin) both drugs, whereas no consistent effect of age on clearance was discerned. monitoring is not routinely available for these agents. patients malignant solid tumors. N=243 (Dox N=110; Doc N=152; 19 patients received both doxorubicin and docetaxel in combination therapy Arm A: A positive association was observed between BSA and doxorubicin clearance (r=0.34; Arm A: Dox N=110 P=0.0015); a separate analysis in males and females revealed a stronger correlation in males (r=0.64; P=0.0002) than females (r=0.01; P=0.95). Clearance was 22% higher in patients with BMI > 30 kg/m2 (63.6 ± 19.9 L/h versus 65.7 17.1 L/h versus 78.9 ± 27.3 L/h; P=0.045). No sex differences found.

Arm B: The mean plasma clearance of docetaxel was 42.8 ± 14.9 L/h (range 13.8–84.4 L/h) with Arm B: Doc N= 152 a CV of 34.8%, and was similar among the different drug-administration schedules (P=0.34) . A Dox and Doc N=19 (excluded from analysis) positive correlation was noted between BSA and docetaxel clearance (r=0.30; P=0.0002); a separate analysis revealed a stronger correlation in males (r=0.35; P=0.0032) than in females (r=0.18; P=0.11). Clearance was not higher in patients with BMI > 30 kg/m2 (40.9 ± 15.0 L/h versus 46.1 ± 15.2 L/h versus 42.3 ± 13.3 L/h; P=0.15). Docetaxel clearance was reduced on average by 11% in females compared to males (40.6 14.7 L/h versus 45.6 14.8 L/h; P=0.040). No sex differences were found.

Miller, To study a fixed dose (360 mg) PK Study - Single Arm The median BSA for both studies was approximately the same. The actual doses given to Main toxicity was neutropenia of grade III and IV severity in 21% and These results suggest that fixed dosing of Rosner, 2004 of paclitaxel patients in CALGB 9342 varied widely around the median of 370 mg (Table 2). The median 25% of patients, respectively, in cycle 1. The worst grade of any paclitaxel is feasible in women, which would simplify the given I.V. over 3 hours to Chemotherapy: calculated dose of 203 mg/m2 on CALGB 9763 was only slightly less than the dose of 210 toxicity, nadir WBC and absolute neutrophil count, and survival administration of this drug. female patients, and to paclitaxel (Taxol) mg/m2 used on CALGB 9342. The toxicity profiles for the two studies were similar. fractions were assessed; no significant relationship was found evaluate prospectively the between BSA and any measure of toxicity. Grade III and relationships between the N=32 BSA was inversely correlated with area under the curve (r = 0.67; P < 0.001) and correlated with IVneutropenia occurred in 21% and 25% of patients, respectively, in following: BSA a and toxicity; total body clearance (r = 0.69; P < 0.001), but BSA was not correlated with total clearance > cycle 1. Grade III and IV4 lymphopenia was observed in 8% and 8%, BSA and pharmacokinetics; The dose normalized for body surface area ranged from 162 to 300 mg/m2. 0.05 umol/L (r = 0.038; P = 0.85). respectively, in cycle 1. Grade III (no grade IV) sensory neuropathy and pharmacokinetics and Estimation of area under the concentration versus time curve (area under the and hyperglycemia were noted in 7% and 8%, respectively, in cycle 1. toxicity. curve) used a limited sampling strategy by which AUC (umol/L X hour) was No other kind of toxicity exceeded 5% in incidence. There were no calculated from the following equation: area under the curve = 4.7 significant relationships between BSA and paclitaxel induced (concentration at 1 hour) + 10 (concentration at 6 hours) +0.63. The limited leukopenia or neutropenia. Neither AUC nor total body clearance was sampling strategy used for area under the curve had a mean error (reflecting significantly correlated with absolute neutrophil count surviving bias) of 2.9% and root mean squared error (representing precision) of 9.8%. fraction or nadir absolute neutrophil count. T >0.05 umol/L was The limited sampling strategy used for T > 0.05 umol/L had a mean error of correlated with log(nadir absolute neutrophil count; r = -0.41; P = 1.6% and root mean squared error of 5.6%. 0.04; Fig. 4) but was not correlated with log(absolute neutrophil count surviving fraction) [r = 0.29; P = 0.16]. Paclitaxel pharmacokinetics were assessed in the first cycle of treatment only. Four blood samples were obtained: before the paclitaxel infusion; and at 1, 6, and 24 hours from the start of the 3-hour infusion. The total (not free) paclitaxel concentrations were measured by hig hperformance liquid chromatography.

Total body clearance was calculated from the relationship as follows: clearance = dose / area under the curve.

Author, Year Objectives Methods Efficacy Toxicities Conclusions

Smorenburg, To evaluate the role of BSA in PK Study - Single Arm The (AUC) of unbound paclitaxel were similar in both dosing groups, with mean values ± SD (A None of the patients developed grade II or greater nonhematologic This study indicates that paclitaxel disposition 2003 paclitaxel disposition. vs B) of 1.34 ± 0.158 vs 1.30 ± 0.329 u/M/h, indicating that BSA based dosing reduced the toxicity, and there were no episodes of neutropenic fever or is significantly related to BSA. This provides a pharmacokinetic rationale for Chemotherapy: coefficient of variation by 53.3%. treatment-related deaths. Overall, hematologic toxicity in the first BSA based dosing of this drug. paclitaxel (Taxol) cycle was mild. Unbound and total paclitaxel clearance was also significantly related to various bodysize Randomized cross-over design with paclitaxel (3-hour infusion at a 3-week measures, including BSA (R ≥ 0.617; P ≤ .033, weight (R ≥ 0.621; {≤ .031) and lean body mass (r The absolute clearances of unbound and total paclitaxel were significantly interval) at 175 mg/m2 in cycle 1 (A) and a flat-fixed dose of 300 mg in cycle 2 ≥ 0.630; P ≤ 0.028). related to various measures of body size, except for height. The observed R (B), or vice versa. values indicate that BSA explains around 38% of the total variation in The exposure to unbound paclitaxel, total paclitaxel, and Cremophor EL was similar in both clearance of unbound paclitaxel and around 45% of the total variation in N=12 of patients with fixed dose (7); # with BSA based dose (5) dose groups (BSA-based dose v flat-fixed dose), with overall mean AUC values of 1.34 ± 0.16 vs clearance of total paclitaxel. These values are similar to the percentage of 1.30 ± 0.33 u/M/h ;P=.67, 17.7 ± 3.0 vs 17,3 ± 5.2 u/M/h ;P=-.71 and 57 ± 13.9 vs 55.5 ± 17.7 total variation explained by weight, so weight is probably the more Arm A: one cycle of paclitaxel at a flat-fixed dose of 300 mg followed by a u/M/h ;P = .53, respectively. important factor in explaining the variation in paclitaxel clearance. second cycle of paclitaxel at a BSA-based dose of 175 mg/m2 BSA based dosing reduced the CV by 53.3%. The CV in AUC was substantially lower in the BSA- based dose group (unbound paclitaxel, 11.8% v 25.3%; total paclitaxel, 16.7% v 29.9%; Arm B: a second cycle of paclitaxel at a BSA-based dose of 175 mg/m2 followed Cremophor EL, 24.2% v 32.0%), indicating that BSA might reduce interindividual variability in by one cycle of paclitaxel at a flat-fixed dose of 300 mg exposure to unbound paclitaxel, total paclitaxel, and Cremophor EL by 53.4%, 44.2%, and 24.4%, respectively.

Sparreboom, To assess the PK Study The absolute clearance of cisplatin, paclitaxel, and troxacitabine was significantly increased in The results suggest that a number of widely used empiric strategies for dose 2007 pharmacokinetics the obese (P < .023), but this was not observed for carboplatin, docetaxel, irinotecan, or adjustments in of eight anticancer drugs in Chemotherapy: topotecan (P < .17). For doxorubicin, the systemic clearance was statistically significantly obese patients, including a priori dose reduction or dose capping, should be adults and evaluated the cisplatin reduced in obese women (P = .013), but not in obese men (P = .52). Evaluation of alternate discouraged. potential utility of alternative paclitaxel weight descriptors for dose calculation in the obese, including predicted normal weight, lean weight troxacitabine body mass, (adjusted) ideal body weight, and the mean of ideal and actual body weight, descriptors in dose calculation carboplatin indicated that, for most of the evaluated drugs, weight scalars used to calculate body-surface for the obese. docetaxel area should consider actual body weight regardless of size. irinotecan topotecan

N= 1,206

Schmitt, To evaluate prospectively PK Study - Single Arm The best covariate equation was : carbopaltin clearance ( mL/min)= 117.8 . (Scr/75)-0.450. no data Individual dosing of carboplatin is a current pracice to control plasma drug 2009 (multi-center study) the (cycC/1.00) -03.85. (ABW/65)+0/504 (age/56) -0.366. 0.847 (sex). Using an alternative weight descriptor exposure of this drug. The main limitation of the exsisting equations is that benefit of using cystatin C Chemotherapy: ( ideal body weight or lean body mass) did not improve the prediction. The final covariate they are all based on serum creatinine level as the unique biological levels (as a marker for GFR) for Carboplatin as single agent (N=70) model was validated by bootstrap analysis. The bias (mean percentage error) and imprecision covariate; it has been chown that these formula overstimate carboplatin carboplatin individual dosing Carboplatin with other agents N=278 (mean absolute percentage error) were +1% and +15% respectively, on the total population, clearance in obese patients. for patients with primary ovary 5 FU (N=7) and were of similiar magnitude in each of the three subgroups of patients defined according to (N=156), uterus (N=44), head docetaxel (N=8) their body mass index. For the first time, a unique formula (Modified Thomas) is proposed for and neck (N=15), lung (N=36), doxorubicin (N=9) carboplatin individual dosing to patients, which is shown to be equally valid and others (N=106). etoposide (N=31) Results validate equation based on five parameters: (Scr, cystatin C, actualy body weight, age, for underweight, normal weight, and obese patients. gemcitabine (N=17) and sex), which is show to be equally valid for underweight, normal weight, and obese patients. paclitaxel (N=194) Modified Thomas formula is CL (mL.min)-01.(Scr.75)02 * (cysC/1.0) vinorelbine (N=10) 03*(ABW/65)04 (age/56)05*06 sex other (N=11)

N= 357 patients included in this study were receiving carboplatin as a part of established protocols.

Seven covarites studied we as follows Scr, cysC, age, sex, ABW, ideal body weight, and lean body mass.

Author, Year Objectives Methods Efficacy Toxicities Conclusions

Okamoto, To evaluate the performance PK Study - Single Arm no data Observed standard-dose carboplatin AUCs after aiming for target AUCs of 5 1998 of the three formula (Calvert, and 6 mg x min/ml using the Calvert formula based upon the 24-h Ccr were Cockcroft-Gault, Chatelut) in Chemotherapy: 5.3 ± 0.8 and 5.9 ± 0.8, respectively, indicating a small and acceptable bias predicting standard- and low- carboplatin : standard dose (N=27) and low-dose (N=25) compared with that predicted from the dosing formula. dose carboplatin pharmacokinetics for patients N=52 The pharmacokinetics of standard-dose carboplatin were accurately with locally advanced NSCLC Result Arm A: MAPEs for the prediction of carboplatin CL from the 24-h Calvert, CG-Calvert and predicted by the Calvert formula based upon either 24-h or CG-calculated (N=25); SCLC (N=20); Arm A (N=37) : standard-dose of carboplatin Chatelut formulae were 13%, 12% and 23%, respectively Ccr, but not by the Chatelut formula. Either CG-calculated or 24-h Ccr can be Metastatic NSCLC or recurrent substituted for the GFR in the Calvert formula for the determination of SCLC (N=7) individual doses. The poor predictability of the Chatelut formula found in this study might be the result of a differences in either the Cr assay or the patient Arm B (N=25) : low-dose carboplatin (25mg/m2, i.v.) Result Arm B: MAPEs were 27%, 18% and 44%, respectively. population. Therefore, formulae which attempt to estimate GFR are not necessarily valid if either the Cr assay or the patient population is changed. Observed standard-dose carboplatin AUCs after aiming for target AUCs of 5 and 6 mg x min/ml using the Calvert formula based upon the 24-h Ccr were 5.3 ± 0.8 and 5.9 ± 0.8, respectively, indicating a small and acceptable bias compared with that predicted from the dosing formula.

Ekhart, 2009 To determine the utility of Retrospective cohort In overweight and obese patients with normal renal function a flat carboplatin dose should be no data In overweight and obese patients with a normal renal function, a flat alternative weight descriptors administered based on the population carboplatin clearance (8.38l/h=140L,min. In the case an carboplatin dose should be administered based on the population in the Cockcroft-Gault Clearance values were obtained from individual fits using NONMEM and were AUC of 5 mg min/Ml is desired, the appropriate dose for carboplatin would be 5 x 140-700 mg. carboplatin clearance (8.38l/h = 140 mL/min). In the case of an AUC of 5 mg equation to more accurately compared to predicited carboplatin clearances calcuated with the min/mL is desired, the appropriate dose for carboplatin would be 5 c 140- predict carboplatin clearance Cockcroft-Gault equation using diverse weight descriptors Results suggest that a flat dose carboplatin = target AUC x carboplatin clearance) based on the 170 mg. in patients with many cancers. population carboplatin clearance (8.38l/h = 140 mL/min) will result in less bias in overweight Chemotherapy: and obese patients with adequate renal function. Incorporation of allometric coefficient (range 18.5-18.8%%) did not -PC- Paciltaxel and carboplatin significantly improve the fit of the model (diffferences in OFV were less than -(t)CTC - Carboplatin and high dose cyclophosphamide (1 h infusion) and 6.63) compared to the basic model (19.4%)in which no relation between thiotepa (2 x 0.5h infusion every day for 4 days weight and carboplatin clearance was assumed. -miniCTC - day 1 Carboplatin and cyclophosphamide (1 h infusion), day 2 thiotepa 1 h infusion of carboplatin and thiotepa

Notes : NSCLC PC dose carbo 300-400 mg/m2 per min; NSCLC 2-6 PC (dose carbo AUC 6 mg min/mL administered in 30 min); High risk primary breast (CTC (dose carbo 400 mg/m2 per day or AUC 20 mg min/mL administered in 1 h for 4 days); germ cell CTC (dose carbo 400 mg/m2/day or AUC 20 mg min/mL administered in 1 h for 4 days); Metastatic ovarian CTC (dose carbo 267 mg.m2 per day or AUC 13.3 mg min/mL adminstered in 1 h for 4 days); Epithelial breast- mini CTC (dose carbo 400 mg/m2 per day or AUC 10 mg min/mL administered in 1 h for 2 days)

N= 240 patients with 380 chemotherapy courses and 4478 samples

Author, Year Objectives Methods Efficacy Toxicities Conclusions

Key 5 FU- 5-Fluorouracil; ABW– actual body weigh; AC-doxorubicin and cyclophosphamide; BMI – body mass index; CAF- cyclophosphamide,doxorubicin, and fluorouracil; CCI -Charlson comorbity index; CHOP-cyclophosphamide, doxorubicin, vincristine, and prednisone; CHOP-R-CHOP-rituximab; CNOP-cyclophosphamide, mitoxantrone, vincristine, and prednisone; (CL/F, Ml/min) – oral clearance; CSF – colony stimulating factor; (cysC– cystatin C level; CV- coefficient of variation; ESBC – early stage breast cancer; FN– febrile neutropenia ; GFR– glomerular filtration rate; GOG – Gynecologic Oncology Group; IBW- Ideal body weight; I.V- intravenously; MAPE – median absolute percent error; NHL – Non-Hodgkin’s Lymphoma; Nadir values– white blood cells, neutrophils, hemoglobin, and platelets; NSCLC- non-small cell lung cancer; OS – overall survival; pCR-pathological complete response; PFS – progression-free survival; PK – Pharmacokinetic; PS – performance status; RT- radiation therapy; RDI – reduced relative dose- intensity; RR- relative risk; SCLC- small cell lung cancer; SCRT - Severe chemotherapy related toxicity ; Scr – serum creatinine level; Thal - Thalidomide ; TTP – Time to Progression; WHO – World Health Organization; Vss – volume of distribution

Data Supplement 3: Search strategy

Original search

("neoplasms"[MeSH Terms] OR "neoplasms"[All Fields] OR "cancer"[All Fields] OR tumor [All Fields] OR tumour[All Fields] OR oncolog*[All Fields] OR myeloma*[All Fields] OR lymphoma*[All Fields] OR (Hodgkin*[All Fields] OR lymphoma[All Fields] OR "NHL"[All Fields] OR carcinom*[All Fields] OR adenocarcinom*[All Fields])) NOT (leukemia [All Fields]) AND (dose [All Fields] OR dosing [All Fields]) AND ("drug therapy"[Subheading] OR "drug therapy"[All Fields] OR "chemotherapy"[All Fields] OR "drug therapy"[MeSH Terms] OR ("drug"[All Fields] AND "therapy"[All Fields])) AND ("obesity"[MeSH Terms] OR "obesity"[All Fields] OR "obese"[All Fields] OR “body weight” [All Fields] OR “body weight” [MeSH Terms] OR "overweight"[MeSH Terms] OR "overweight"[All Fields] OR "overweight"[All Fields] OR (“over”[All Fields] and “weight”[All Fields])) AND (((randomized controlled trial[pt] OR controlled clinical trial[pt] OR randomized controlled trials[mh] OR clinical trial[pt] OR "clinical trial"[tiab] OR "clinical trials"[tiab] OR clinical trials as topic[mh] OR controlled clinical trials as topic[mh] OR randomized controlled trials as topic[mh] OR clinical trials, phase II as topic[mh] OR clinical trials, phase III as topic[mh] OR clinical trials, phase IV as topic[mh] OR clinical trial, phase II[pt] OR clinical trial, phase III[pt] OR clinical trial, phase IV[pt] OR random allocation[mh] OR “random allocation”[tiab] OR “randomly allocated”[tiab] OR double-blind method[mh] OR single-blind method[mh]) OR ((random[tiab] OR randomly[tiab] OR randomized[tiab] OR randomised[tiab] OR randomization[tiab] OR randomisation[tiab]) AND (clinical[tiab] OR control[tiab] OR controlled[tiab] or control groups[mh])) OR ((single[tiab] OR single-[tiab] OR double[tiab] OR double-[tiab] OR triple[tiab] OR triple-[tiab] OR multi[tiab] OR multi-[tiab] OR evaluator[tiab] OR assessor[tiab] OR interviewer[tiab]) AND (mask[tiab] OR masked[tiab] OR masking[tiab] OR blind[tiab] OR blinded[tiab] OR blinding[tiab])) OR ((placebos[mh] OR placebo[tiab] OR placebos[tiab] OR random[tiab] OR randomly[tiab] OR randomized[tiab] OR randomised[tiab] OR randomization[tiab] OR randomization[tiab]) AND (research design[mh] OR “comparative study”[tiab] OR comparative study[pt] OR evaluation studies as topic[mh:noexp] OR evaluation studies[pt] OR “evaluation study”[tiab] OR “evaluation studies”[tiab] OR validation studies as topic[mh] OR follow-up studies[mh] OR “follow-up study”[tiab] OR “follow up study”[tiab] OR “follow-up studies”[tiab] OR “follow up studies”[tiab] OR prospective studies[mh] OR prospective[tiab] OR epidemiologic research design[mh] OR epidemiologic methods[mh] OR epidemiologic study characteristics as topic[mh] OR epidemiologic studies[mh] OR intervention studies[mh] OR cross-over studies[mh]))) NOT (clinical trial, phase I[pt] OR clinical trials, phase I as topic[mh]) OR ((meta-analysis[pt] OR meta-analysis as topic[mh] OR “meta- analysis”[tiab] OR “meta analysis”[tiab] OR “meta-analyses”[tiab] OR “meta analyses”[tiab] OR “meta-analyzed”[tiab] OR “meta-analysed”[tiab] OR “meta analyzed”[tiab] OR “meta analysed”[tiab] OR (meta[tiab] AND analysis[tiab]) OR (meta-[tiab] AND analysis[tiab])) OR ((critical[tiab] OR critically[tiab] OR systematic[tiab] OR systematical[tiab] OR systematically[tiab] OR evidence-based[tiab] OR “evidence based”[tiab]) AND (review[pt] OR review[tiab] OR reviews[tiab] OR reviewed[tiab] OR appraise[tiab] OR appraises[tiab] OR appraised[tiab] OR appraisal[tiab] OR assess[tiab] OR assesses[tiab] OR assessed[tiab] OR assessment[tiab] OR evaluate[tiab] OR evaluates[tiab] OR evaluated[tiab] OR evaluation[tiab] OR critique[tiab] OR critiques[tiab] OR analysis[tiab] OR analyses[tiab] OR analyzed[tiab] OR analysed[tiab])) OR ((evidence-based[tiab] OR “evidence based”[tiab]) AND (guideline[tiab] OR guidelines[tiab] OR recommendation[tiab] OR recommendations[tiab] OR consensus[mh] 15

OR consensus[tiab] OR consensuses[tiab])) OR (review[pt] OR review literature as topic[mh] OR consensus development conference[pt] OR consensus development conference, NIH[pt] OR consensus development conferences as topic[mh] OR consensus development conferences, NIH as topic[mh] OR “consensus development”[tiab] OR health planning guidelines[mh] OR guideline[pt] OR practice guideline[pt] OR practice guidelines as topic[mh] OR “practice guideline”[tiab] OR “practice guidelines”[tiab] OR health planning guidelines[mh] OR "standard of care"[tiab] OR “evidence-based medicine”[tiab] OR “evidence based medicine”[tiab] OR “evidence-based care” OR “evidence-based practice”[tiab] OR cochrane database syst rev[ta] OR acp j club[ta] OR health technol assess[ta] OR clin evid[ta]) NOT (case reports[pt] OR case report[tiab] OR editorial[pt] OR editorial[tiab] OR letter[pt] OR newspaper article[pt])))

Supplemental search

(("body surface area"[MeSH Terms] OR ("body"[All Fields] AND "surface"[All Fields] AND "area"[All Fields]) OR "body surface area"[All Fields]) AND dosing[All Fields]) AND ("neoplasms"[MeSH Terms] OR "neoplasms"[All Fields] OR "cancer"[All Fields])

Data Supplement 4: Quorom Diagram: Exclusions and Inclusions of Publications Identified for Systematic Review

Potentially relevant abstracts identified by electronic searching and screened for retrieval

(n = 913)

Articles retrieved for full text Articles excluded after full review text review

(n=148) (n=767)

Articles that met selection Articles excluded after full criteria for data extraction text review

(n=56) (n=92)

Additional articles recommended by Panel members or identified by hand-searching

(n=8 on morbidly obese, excluded)

(n= 4 other papers, excluded)

Data Supplement 5: Glossary and BSA Formulas

Pharmacokinetics is the study of the action of drugs within the body, which can, in many respects, be envisioned more accurately as the actions of the body on an administered drug. It includes studies of the mechanisms of drug absorption, distribution, metabolism, and excretion; onset of action; duration of effect; biotransformation; and effects and routes of excretion of the metabolites of the drug. Pharmacokinetics can be referred to as "PK".

Pharmacodynamics is the study of the physiological effects of drugs on the body or on microorganisms or parasites within or on the body and the mechanisms of drug action and the relationship between drug concentration and effect.[1] One dominant example is drug-receptor interactions as modeled by

where L=ligand (drug), R=receptor (attachment site), reaction dynamics that can be studied mathematically. Pharmacodynamics can be referred to as “PD”.

PKPD In conjunction with pharmacokinetics , pharmacodynamics can be referred to as "PKPD". Toxicity Grades - http://evs.nci.nih.gov/ftp1/CTCAE/CTCAE_4.03_2010-06- 14_QuickReference_5x7.pdf

General definitions that apply across all adverse events (AE), with the caveat that “Not all Grades are appropriate for all AEs. Therefore, some AEs are listed with fewer than five options for grade selection.

Grade 1 Mild; asymptomatic or mild symptoms; clinical or diagnostic observations only; intervention not indicated. Grade 2 Moderate; minimal, local or noninvasive intervention indicated; limiting age-appropriate instrumental Activities of Daily Living (ADL). Grade 3 Severe or medically significant but not immediately life-threatening; hospitalization or prolongation of hospitalization indicated; disabling; limiting self-care ADL. Grade 4 Life-threatening consequences; urgent intervention indicated. Grade 5 Death related to AE.

Calculation Tools for Body Surface Area (BSA) and Body Mass Index (BMI)

Chemotherapy is usually dosed by square meter of body surface area (BSA). BSA has been chosen rather than body weight as the basis for calculation for two reasons. First, BSA has been demonstrated to provide a more accurate comparison of activity and toxicity for certain drugs. Second, BSA can be more closely correlated with cardiac output, which determines the blood flow to the liver and kidneys, thus influencing drug elimination.

Common Formulae – most did incorporate sex and none were developed in a typical 21st century population. ASCO does not endorse any particular calculator.

Carboplatin Calculation Resource

Calvert Formula - https://hccapps.musc.edu/hemonc/carboplatin_dose_calculator.htm

Total Dose (mg) = (target AUC) X (GFR + 25)

BSA Calculation Resources

Boyd Formula - http://www.medcalc.com/body.html; http://www.halls.md/body-surface- area/bsa.htm; http://endmemo.com/medical/bsa.php

BSA (m2) = 0.0003207 x Ht (cm) 0.3 x weight (g) (0.7285 - ( 0.0188 x LOG(grams) ) Based on 197 patients

DuBois and DuBois Formula - http://www.medcalc.com/body.html; http://www.esmo.org/fileadmin/practice/bodymass.php ; http://www.globalrph.com/bsa2.htm; http://www.halls.md/body-surface-area/bsa.htm; http://endmemo.com/medical/bsa.php

BSA(m2) = Wt(kg)0.425 x Ht(cm)0.725 x 0.007184

This is the classic formula, published in 1916, on which many monograms are based. However, it was based on measurements of 9 individuals, one of whom was a child.

DuBois D, DuBois DF. A formula to estimate the approximate surface area if height and weight be known. Arch Int Med 1916;17:863-71.

Gehan and George Formula - http://www.medcalc.com/body.html; http://www.esmo.org/fileadmin/practice/bodymass.php ; http://www.globalrph.com/bsa2.htm; http://www.halls.md/body-surface-area/bsa.htm; http://endmemo.com/medical/bsa.php

BSA(m2) = Wt(kg)0.51456 x Ht(cm)0.42246 x 0.0235

This formula is based on direct measurements of 401 individuals.

Gehan EA, George SL. Estimation of human body surface area from height and weight. Cancer Chemother Rep 1970;54:225-35.

Haycock, et al. Formula - http://www.medcalc.com/body.html; http://www.esmo.org/fileadmin/practice/bodymass.php ; http://www.globalrph.com/bsa2.htm; http://www.halls.md/body-surface-area/bsa.htm; http://endmemo.com/medical/bsa.php

BSA(m2) = Wt(kg)0.5378 x Ht(cm)0.3964 x 0.024265

Haycock et al. reported that the formula of DuBois and DuBois increasingly underestimated BSA as values fell below 0.7 m2. Their formula was based on measurements of 81 individuals ranging from premature infants to adults.

Haycock GB, Schwartz GJ, Wisotsky DH. Geometric method for measuring body surface area: A height-weight formula validated in infants, children and adults. J Pediatr 1978;93:62-6

Mosteller (Adults and Children) Formula -http://www.medcalc.com/body.html; http://www.esmo.org/fileadmin/practice/bodymass.php; http://www.globalrph.com/bsa2.htm; http://www.halls.md/body-surface-area/bsa.htm; http://endmemo.com/medical/bsa.php

This formula is a simple modification of an equation by Gehan and George, and requires the use of a calculator with a square root key. The formula has been confirmed as being applicable to children by Lam and Leung. (SQR RT = Square Root of…)

Mosteller RD. Simplified calculation of body-surface area. N Engl J Med 1987;317:1098.

Data Supplement 6: Interpretation of BMI for adults

For adults 20 years old and older, BMI is interpreted using standard weight status categories that are the same for all ages and for both men and women. For children and teens, on the other hand, the interpretation of BMI is both age- and sex-specific. For more information about interpretation for children and teens, visit Child and Teen BMI Calculator.

Calculation of BMI For adults 20 years old and older, BMI is interpreted using standard weight status categories that are the same for all ages and for both men and women. For children and teens, on the other hand, the interpretation of BMI is both age- and sex-specific. For more information about interpretation for children and teens, visit Child and Teen BMI Calculator.

BMI is calculated the same way for both adults and children. The calculation is based on the following formulas:

Measurement Units Formula and Calculation

Kilograms and Formula: weight (kg) / [height (m)]2 meters (or centimeters) With the metric system, the formula for BMI is weight in kilograms divided by height in meters squared. Since height is commonly measured in centimeters, divide height in centimeters by 100 to obtain height in meters.

Example: Weight = 68 kg, Height = 165 cm (1.65 m) Calculation: 68 ÷ (1.65)2 = 24.98

Pounds and inches Formula: weight (lb) / [height (in)]2 x 703

Calculate BMI by dividing weight in pounds (lbs) by height in inches (in) squared and multiplying by a conversion factor of 703.

Example: Weight = 150 lbs, Height = 5'5" (65") Calculation: [150 ÷ (65)2] x 703 = 24.96