Gemcitabine and Irinotecan as First-Line Therapy for Carcinoma of Unknown Primary: Results of a Multicenter Phase II Trial

Metastatic carcinoma of unknown primary (CUP) has a very poor prognosis, and no standard first-line therapy currently exists. Here, we report the results of a phase II study utilizing a combination of gemcitabine and irinotecan as first-line therapy. Treatment was with gemcitabine 1000 mg/m2 and irinotecan 75 mg/m2 weekly times four on a six week cycle (Cohort I). Due to excessive toxicity, the dose and schedule were modified as follows: gemcitabine 750 mg/m2 and irinotecan 75 mg/m2 given weekly times three on a four week cycle (Cohort II). The primary endpoint was the confirmed response rate (CR + PR). Secondary endpoints consisted of adverse events based upon the presence or absence of the UDP glucuronosyltransferase 1 family, polypeptide A1*28 (UGT1A1*28) polymorphism, time to progression, and overall survival. Thirty-one patients were enrolled with a median age of 63 (range: 38–94), and 26 patients were evaluable for efficacy. Significant toxicity was observed in Cohort 1, characterized by 50% (7/14) patients experiencing a grade 4+ adverse event, but not in cohort II. The confirmed response rate including patients from both cohorts was 12% (95% CI: 2–30%), which did not meet the criteria for continued enrollment. Overall median survival was 7.2 months (95% CI: 4.0 to 11.6) for the entire cohort but notably longer in cohort II than in cohort I (9.3 months (95% CI: 4.1 to 12.1) versus 4.0 months (95% CI: 2.2 to 15.6)). Gemcitabine and irinotecan is not an active combination when used as first line therapy in patients with metastatic carcinoma of unknown primary. Efforts into developing novel diagnostic and therapeutic approaches remain important for improving the outlook for this heterogeneous group of patients. Trial Registration ClinicalTrials.gov NCT00066781

Despite an often extensive search, the primary site of disease is not usually discovered. The most common sites found at autopsy series include lung and GI sites (pancreas, colon, stomach, and liver).

1.11
In the work-up of a patient with an unknown primary, it is important to determine if the patient falls into one of the more treatable subcategories. These include: • women with axillary adenopathy, suggesting breast CA primary.
• women with peritoneal carcinomatosis, suggesting ovarian/primary peritoneal CA. • men with prominent blastic bony metastasis, suggesting prostate CA.
• squamous cell CA of the neck nodes/inguinal nodes.
• younger men with poorly differentiated mediastinal or retroperitoneal mass, suggestive of germ cell origin.
• carcinoma with neuroendocrine features Immunostains can be helpful in the diagnosis when H&E is not clear. These should be performed to rule out subsets of malignancies that are treated in a different manner (1). These include: • Keratin or epithelial membrane antigen (positive in carcinoma) • S-100 or HMB45 (melanoma markers) • LCA (CD45) (hematolymphoid) • Chromogranin or synaptophysin (neuroendocrine) 1.12 Patients with well-differentiated carcinomas that do not fit into these groups have a poor prognosis. Median survival is usually <8 months. Patients with poorly differentiated carcinoma (PDC) also have a poor prognosis, but may respond better to CDDP-based chemotherapy (1,2). Single agents have been evaluated in patients with unknown primaries, including 5FU, CDDP, oral etoposide, and mitomycin, with response rates of 10-20%. In general, CDDP and 5FU-based regimens have shown response rates of 20-30%, with median survival of 5-8 months (1,2). Newer, broad-spectrum combinations that include Taxol have yielded even better results (3). Parvadis et al conducted a phase II study using Taxol/CDDP, reporting a response rate of 41% with median survival not yet reached (>8mo) (4).

1.2
Recently, combination chemotherapy using paclitaxel, VP-16 and CDDP has shown promising activity in both PDC and well-differentiated carcinoma of unknown primary. Greco  Two drugs that also have a broad spectrum of antineoplastic activity include gemcitabine and irinotecan. Gemcitabine is a fluoridated pyrimidine that inhibits DNA and RNA synthesis and repair (5). Gemcitabine has been used extensively in the treatment of lung and pancreatic cancer, and also shows activity against stomach, breast and ovarian cancers. Gemcitabine is well tolerated, with mild myelosuppression and minimal nonhematologic toxicity (6)(7)(8)(9)(10)(11). It has been shown to improve tumor-related symptoms in patients with NSCLC and improves QOL in patients with pancreatic cancer (8)(9)(10)(11). Gemcitabine has also been used as a single agent in treating patients with unknown primary carcinomas who had progressed on first line treatment. Response rates were 8%, with 25% of patients having stable disease (12). In a recent report, Greco et al. reported a new combination of gemcitabine, Taxol, and carboplatin for patients with unknown primary cancer. The response rate with the three-drug regimen was 25%, with median survival of 9 months, which is similar to other commonly used regimens. In this study, responding or stable patients continued to receive weekly Taxol, but did not benefit further from this additional treatment. The most common side effects with this regimen were myelosuppression, with grade 3/4 leucopenia in 46%, and grade 3/4 thrombocytopenia in 43% (13).

1.31
Irinotecan is a camptothecan analog. Its active metabolite, SN-38, inhibits topoisomerase I by stabilizing the enzyme-DNA complex. This leads to a DNA stand break and inhibition of replication (14,15). Irinotecan has good activity in both gastrointestinal malignancies as well as lung cancer. Response rates in NSCLC range from 15-34%, with median survivals 6.2-9.8 months (16- 19).
Studies have also shown activity in breast, ovarian, cervical, SCLC, and pancreatic cancer (20). The main toxicities of irinotecan are diarrhea and neutropenia (16-20).

1.32
There is preclinical evidence that the combination of gemcitabine and irinotecan are synergistic (21,22). Bahadori et al. conducted experiments using a small-cell lung cancer cell line, and found the combination to be synergistic. They postulate that if gemcitabine is available at the time of a DNA strand break, G triphosphate can be inserted into DNA, leading to irrepairable damage and cell death. CPT-11, a topoisomerase I inhibitor, causes DNA strand breaks, and could facilitate this incorporation of G triphosphate, leading to synergistic effect of the combination (22). Studies looking at dose and schedule-dependent effects of this combination of drugs are underway. At Mayo Clinic, the BxPC-3 hu pancreatic CA cell line showed synergistic activity with the sequence of gemcitabine followed by SN-38 (personal communication -S. Kaufmann). Given the extremely broad spectrum of antineoplastic activity, different mechanisms of action, and potential for synergism, the combination of these two drugs seems ideal for the empiric treatment of carcinoma of unknown primary.

1.33
Three phase I studies have combined these drugs in different schedules.
1.331 The first study treated 19 patients with solid tumors (refractory to standard therapy, or for which no standard therapy is available) (23). Gemcitabine was given at 1000mg/m 2 over 30 min, d1 and d8 Q3wks. CPT-11 was given in increasing doses (50,75,100,115m /m 2 ) over 90min, after gemcitabine, d1 and d8, Q 3 weeks. Dose-limiting toxicity (DLT) was diarrhea, which was grade 3 or 4 in 3/19 patients at a dose level of N004E 7 Addendum 8 CPT-11 100-115 mg/m 2 . Cholinergic toxicity was noted in 2 patients at 115 mg/m 2 . Hematologic toxicity was not dose dependent (maybe due to day 8 dose reductions) and cumulative toxicity was only observed in two patients at the highest dose level. Grade 2 and 3 neutropenia occurred in 4 patients at doses of 100-115mg/m 2 . Grade 2-3 thrombocytopenia occurred in only 2 patients at 100-115mg/m 2 . Non-hematologic toxicity was mild, and consisted of the following: • Nausea grade 3: 3/19 • Vomiting grade 3: 1/19 • Diarrhea grade 3 or 4: 3/19 1.332 The authors concluded that the maximum tolerated dose (MTD) of irinotecan, when given with fixed dose gemcitabine (1g/m 2 ) was 100 mg/m 2 . At the MTD, one of five patients was unable to receive day 8 therapy as a result of grade 3 thrombocytopenia. Eighteen patients were evaluable for response; three had PR's. Two of these patients had pancreatic CA and one had unknown primary adenocarcinoma. This latter response duration was 15 months.
1.333 O'Reilly also reported a phase I study with fixed dose gemcitabine (1g/m 2 ) and irinotecan (45,60,80, 100 mg/m 2 ) given on d 1, 8 and 15, Q 4 weeks (24). Arm A received gemcitabine followed by CPT-11, and Arm B received CPT-11 followed by gemcitabine. Grade 3 diarrhea and fatigue were DLT. MTD was found to be CPT-11 at 60mg/m 2 for arm A. Arm B is still accruing. So far, there were no differences in pharmacokinetics between either arm of the study, but hematologic toxicity appears greater with Arm B. There was one response in a patient with gastric cancer. Fourteen patients had stable disease.
1.334 A recently completed Mayo Clinic phase I study also used the combination of escalating dose gemcitabine (600, 800, 1000 mg/m 2 ) followed by irinotecan (75, 100, 125 mg/m 2 ) weekly for 4 weeks out of six weeks (25). Twenty-four patients were enrolled. No major responses were seen, but patients were heavily pre-treated. Dose limiting toxicity (DLT) was hematologic, and occurred at doses of gemcitabine 1000 mg/m 2 and irinotecan 125 mg/m 2 . At the maximum tolerated dose (gemcitabine 1000 mg/m 2 , irinotecan 100 mg/m 2 ) 1/6 patients completed 2/4 weeks of treatment, and three other patients completed 3/4 weeks. Of the remaining patients, one required a dose reduction but both received all four weeks of treatment. Most of these missed days were due to hematologic toxicity. At the MTD, two patients had grade 3 neutropenia, and grade 3/4 thrombocytopenia occurred in 17%. Diarrhea was the most common non-hematologic toxicity. At the MTD, three patients had grade 1 diarrhea, and one patient had grade 4 diarrhea. Other mild toxicities included anorexia (12 patients), alopecia (9 patients), fatigue, nausea, vomiting, and weight loss. 1.342 The second phase II study was also in patients with advanced pancreatic cancer (27). Twenty-one chemo-naïve patients were treated with gemcitabine 900 mg/m 2 d1, 8 and irinotecan 300 mg/m 2 on d8, Q 3 weeks. Grade 3 or 4 neutropenia was seen in 36%, with 4 neutropenic fevers (one death). Grade 3 or 4 thrombocytopenia was seen in 8%. Grade 3 diarrhea occurred in 1 patient, and grade 2 or 3 asthenia was observed in 43%. Twenty patients were evaluable for response. PR was seen in 15% (3pts), stable disease in 40% (8 pts). Median TTP was 30 weeks, and median survival not yet reached.

1.35
As technology improves, and more discoveries are made about the genetic and molecular nature of cancer, it is likely that many cases of unknown primary cancer will be defined on the basis of immunohistochemistry and molecular techniques. This may direct treatment with a particular regimen. Currently, the immunohistochemical stains are useful to categorize patients into broad groups, such as hematologic, carcinoma, melanoma, sarcoma, etc. but have yet been refined enough to pinpoint the actual organ of origin. Immunostains such as CK7/20 have been used in the research setting, but are not widely clinically used, and are not specific for diagnosis (28). However, certain patterns of staining are characteristic of a given primary source, and may be useful to direct therapy. The translational component of this study proposes to stain all specimens for CK7/20 in an attempt to ascertain which patterns may be predictive of response to this regimen. This will be an exploratory analysis, as the numbers in this phase II study are too small to make any firm conclusions. We will also request the tissue blocks be retained here so that we may pursue additional translational studies in the future in conjunction with our pathology colleagues.

1.36
Irinotecan and gemcitabine, as shown in the above studies, is potentially active and tolerable when given in combination. Given the broad spectrum antineoplastic activity of these two agents, with well-known activity in both lung and GI malignancies, we propose a phase II trial using this combination in patients with carcinoma of unknown primary.

1.4
Response to treatment, or lack of, is dependent upon a variety of factors. Potentially important factors include sensitivity to the agent, achieving appropriate levels of the agent, and avoiding unacceptable toxicity. The latter two factors are influenced by the uptake, metabolism, and distribution of the agent. Individual variations in drug metabolism may greatly influence the efficacy and tolerability of therapy. However, the N004E 9 Addendum 8 ability to measure or predict individual variations in drug metabolism is quite limited. The evolving field of pharmacogenetics is helping to define factors that may contribute to individual variations by assessing factors such as genetic polymorphisms. As a better understanding of individual variation in response to therapy is obtained it may be possible to being tailoring therapy for individual patients.

1.41
The results of irinotecan clinical trials and in vitro work now suggest that the major dose limiting toxicities of irinotecan, diarrhea and myelosuppression may be genetically determined. Irinotecan, a semisynthetic derivative of camptothecin, is converted in vivo to SN-38, which is 1000 times more potent of an inhibitor of topoisomerase I (29)(30). SN-38 covalently stabilizes the enzyme-DNA complexes, (32) resulting in strand breaks and subsequent cytotoxicity (33-36). SN-38, is inactivated by the polymorphic hepatic uridine diphosphate glucuronosyl-transferase 1A1 (UGT1A1) enzyme (37). A dinucleotide repeat polymorphism in the TATA sequence of the promoter region of UGT1A1, leads to significantly lower SN-38 glucuronidation rates in liver samples heterozygous or homozygous for the (TA) 7 TAA polymorphism (38). In a retrospective study, the presence of the UGT1A1*28 polymorphism was associated with a 7-fold increased risk for severe toxicity (leukopenia/diarrhea) when compared to patients without the *28 polymorphism (39). The first prospective phase I pharmacogenetic study of CPT-11 was recently reported (40). In this study, patients receiving irinotecan were noted to have significantly lower SN-38 glucuronidation rates than those without the *28 polymorphism. Furthermore, severe toxicity (grade 3/4 diarrhea or neutropenia), was seen only in those patients with the *28 polymorphism.
The first large prospective study reporting the association between the UGT1A1 *28 polymorphism and CPT-11 related toxicity and response was recently presented (41). This data is derived from a large prospective NCCTG study (N9741) in which patients with metastatic colorectal cancer were randomized to one of 3 different arms. In 2 of those arms, patients received either CPT-11, 5-FU, and leucovorin, or CPT-11 + oxaliplatin. The findings from this study showed that patients who carried the *28 polymorphism had significantly higher rates of grade 4/5 neutropenia (35% for 7/7 variant, 16% for the 6/7 variant) compared with patients with the 6/6 variant (8%). Perhaps even more importantly, response rates were significantly different, as patients who carried the 6/6 variant had significantly decreased response rates compared with those patients who carried the *28 variant.
Based on the accumulated evidence, it is clear that one dose (based on body surface area) does not fit all for patients receiving CPT-11. Therefore, stratifying by UGT1A1 genotyping is one way in which CPT-11 induced toxicity may be reduced.
Based on the Mayo Clinic phase I study of gemcitabine and CPT-11 (42), the MTD of gemcitabine was 1000 mg/m 2 weekly and CPT-11 (100 mg/m 2 ) weekly. However, at this dose level only 1/6 patients were able to complete all four weeks of treatment. Of the remaining 5 patients, 3 completed 3/4 weeks, 1 completed 2/4 weeks and the last patient required dose reductions in order to complete all 4 weeks of treatment. Toxicities accounting for missed weeks of N004E 10 Addendum 8 treatment or does reductions were predictable and included hematologic toxicity (4 patients) and diarrhea (1 patient).
Based on these results, it is unclear whether those patients who carry the UGT1A1 *28 polymorphism, will be able to tolerate a dose of weekly gemcitabine (1000 mg/m2) and CPT-11 (100 mg/m2). Furthermore, those patients who do not carry this polymorphism (6/6 repeats) may receive subtherapeutic doses of CPT-11 if they are dosed in the same manner as those with the *28 polymorphism.
Therefore, in this study, we will enroll patients to 2 different cohorts, and all patients will be prospectively genotyped for UGT1A1. For Cohort I, we will begin by enrolling the minimum number of patients required such that at least 6 evaluable patients carry the UGT1A1*28 polymorphism, and at least 6 evaluable patients do not carry it. These Cohort I patients will receive gemcitabine at dose level = 1000 mg/m 2 , and CPT-11 at dose level = 75 mg/m 2 (dose level 0). Note that the dose of CPT-11 is 25 mg/m 2 lower than the MTD of CPT-11 reported on the Mayo Clinic phase I study. (42) Following the enrollment of the Cohort I patients (i.e., at least 12), accrual will be suspended and the first 6 evaluable patients from each polymorphism group will be observed for a minimum of 6 weeks for adverse events before any additional patients will be entered. For Cohort II, a maximum of 37 patients (see Section 16.14 for details) will be prospectively genotyped and doses will be determined and subsequently administered based on the results of the feasibility portion (Cohort I portion) of this study.
In addition to the UGT1A1*28 polymorphism, we will also explore the impact of the recently described common variations (-3279G>T and -3156G>A) within the UGT1A1 phenobarbital-responsive enhancer module (43,44). The -3279 variant has been associated with a significant (35%) increase in bilirubinemia in Asian patients with Gilbert's syndrome (43). Furthermore, because both the -3279 and the -3156 variants are in linkage disequilibrium with the (TA) n polymorphism (*28), common haplotypes comprising these variants may impact UGT1A1 glucuronidation (44). The haplotypes comprising -3279T and TA 6 (0.53) and -3279G and TA 7 (0.31) are most common in Caucasians where the presence of the TA 7 allele is known to impact SN-38 glucuronidation (40,45). In addition, the haplotype comprising the -3279G and TA 6 variants (q=0.07 in Caucasians) may also impact SN-38 glucuronidation compared with the common -3279T and TA 6 haplotype. Accordingly, we will develop assays to detect both the -3279 and-3156 variants and explore whether haplotypes which comprise these variants impact toxicity and outcome in this phase II trial.
Finally, common allelic variants in exons 12, 21, and 26 of the MDR1 gene are in linkage disequilibrium and may be associated with MDR1 expression and clinical outcome (46). Because P-glycoprotein plays an important role in the biliary excretion of CPT-11, MDR1 gene polymorphisms may impact CPT-11 pharmacokinetics (47). Indeed, a recent study found that the MDR1 1236 C>T variant was associated with a significant increase in CPT-11 and SN-38 exposure (48). Accordingly, we will explore whether the common polymorphisms in exons 12, 21, and 26 are associated with the clinical outcomes of response and toxicity in this phase II study. The pharmacogenetics of gemcitabine is not well understood. With the current investigational approaches now available to better define individual variations in response to therapy, we are in a unique position to explore these variations in this trial. We plan to obtain appropriate blood samples for pharmacogenetic studies with gemcitabine. For gemcitabine, the laboratories of Drs Weinshilboum and Ames are currently involved in resequencing studies with several nucleoside transporters, cytidine deaminase, dCMP deaminase, deoxycytidine kinase, and deoxycytodilate kinase. Upon characterization of functional polymorphisms in these genes, we will be prepared to genotype for these mutations.

1.6
Background for 92 gene RT-PCR cancer classification assay. The availability of technology allowing the study and comparison of gene expression signatures provides the opportunity to base the classification of human malignancies on a more objective and comprehensive set of data than allowed by the presently used clinical pathology criteria, that even when dealing with a known primary cancer site are hindered by a degree of subjectivity of interpretation. (63) AviaraDx has developed a technology which enables the classification of 39 cancer types (64). This is a molecular-based methodology that quantifies the expression of 92 genes within a tissue cancer biopsy via real time RT-PCR. The clinical utility of this technology is the determination of the primary anatomical origin of a metastatic cancer.
A full description of this is found in the Ma et al (64). In brief terms, the development of the 92-gene assay was completed in five major steps. First, a 22,000 gene microarray was used to build a gene expression dataset of 578 cancer samples representing 39 tumor types. Second, 1001 genes were selected in-silico from this database based on their varying degrees of tumor-specific expression profiles. This was completed by dividing up the dataset into a training and test set and selecting the top 30 genes with the highest ability to discriminate at a specific node within the tumor taxonomy hierarchy of known cancer classification (e.g., a node for epithelial versus non-epithelial cancers). Thirdly, these 1001 genes were further narrowed down in-silico by the use of genetic algorithms to a non-redundant set of 126 genes that represented the overlap of the most optimal set of genes (74 genes) plus those genes most frequently found in the 100-derived genetic algorithm gene sets (90 genes). Fourth, real-time RT-PCR assays were developed for the 126 genes and this was further narrowed to the 87 genes whose individual analytical performance re-capitulate the expression patterns originally observed within the microarray-based dataset. Finally, a set of 5 genes were identified to enable the normalization of data from varying amounts of sample input (i.e., RNA) into the assay. In total, the 92-gene assay contains 87 genes for tumor classification and 5 genes for sample input normalization.
Once the 92-gene assay was developed, it was then used to generate a dataset consisting of the gene expression profiles of 87 genes for 578 cancer samples representing 39 cancer types. To classify a sample of unknown class, we first calculated its distance d to each instance in the training set as the 1-Pearson correlation coefficient. The top k nearest neighbors were then examined for their class labels, and the unknown case was assigned to the class with the largest summed weight (1/d). The overall success rate for classifying 39 cancer types was 88% (64).

2.11
To evaluate the response rate of the combination of irinotecan and gemcitabine in patients with unknown primary cancer.

2.12
For Cohort I patients, to assess the adverse event profile and tolerability of the combination of gemcitabine and irinotecan in patients with unknown primary carcinoma, based upon the presence or absence of the UGT1A1 *28 polymorphism.

2.13
For Cohort II patients, to assess the adverse event profile and tolerability of the combination of gemcitabine and CPT-11, based on the interval dose reductions and schedule change.

2.21
To assess time to progression (TTP) and overall survival (OS).

2.22
To perform immunohistochemical studies using CK7/20 to determine if certain patterns of staining are associated with response to therapy; we will also store tissue blocks at the NCCTG Tissue Repository for future translational studies.

2.23
To assess whether variation in multiple different genes whose protein products are involved in the uptake, metabolism, and distribution of Gemzar and CPT-11 affect clinical outcomes (response, toxicity).

2.24
To determine whether cellular measurements of dFdCTP reflect the aggregate consequences of cellular uptake, anabolism and catabolism of gemcitabine by measuring intracellular triphosphates in the mononuclear cells. This pharmacologic or intervening phenotype will be used for correlation with genotypes determined in this study

2.25
Determine primary origin of cancer of unknown primary (CUP) samples by completing a 92-gene RT-PCR cancer classification assay.

2.26
Determine whether the 92-gene assay results are correlated with clinical response to CPT-11/Gemcitabine If these stains cannot be performed at the treating institution, biopsy blocks should be sent to the NCCTG Pathology Coordinator for testing before patients can enter study (see Section 17.1).

3.13
Patients must have measurable disease as defined in Section 11.0. For patients having only lesions measuring at least 1 cm to less than 2 cm must use spiral CT imaging for both pre-and post-treatment tumor assessments.

3.15
Age ≥18 years. Ca, glucose required only as clinically indicated.

5.
Colonoscopy to be performed in patients with liver metastasis or strong suspicion of a colon primary; ≤30 days prior to registration. 6.
For women of childbearing potential only. Must be done ≤7 days prior to registration. 7.
Biopsy blocks should be sent to NCCTG Pathology Coordinator for testing before patients can enter study (see Section 17.1). If the institution is unable to provide a block, submit 15 regular charged slides (5 microns). 9.
Note: Submission of blood samples is mandatory (gemcitabine genotyping and phenotyping studies). Kits are provided and must be used. See Section 14.0 11.
Note: This test is mandatory: Blood will be collected and sent by overnight express. Once blood is received in the laboratory of Dr. Ames, results will be reported within 3

6.3
At the time of registration/randomization, Randomization Center personnel will verify the following: • IRB approval at the registering institution • Patient eligibility • Existence of a signed consent form • Existence of a sign authorization for use and disclosure of protected health information (USA institutions only) At the time of registration/randomization, the following will be recorded: • Patient has/has not given permission to store and use his/her blood sample(s) for future research of cancer. • Patient has/has not given permission to store and use his/her tissue sample(s) for future research of cancer.
• Patient has/has not given permission to store and use his/her blood sample(s) for future research to learn, prevent, or treat other health problems. • Patient has/has not given permission to store and use his/her tissue sample(s) for future research to learn, prevent, or treat other health problems.
• Patient has/has not given NCCTG permission to give his/her blood sample(s) to outside researchers. • Patient has/has not given NCCTG permission to give his/her tissue sample(s) to outside researchers.
6.4 Randomization Center will automatically register patients separately to the mandatory translational component of this study (see Section 14.0).

6.5
Treatment on this protocol must commence at the accruing membership under the supervision of a NCCTG member physician.

6.6
Treatment cannot begin prior to registration and must begin ≤7 days after registration.

6.7
Pretreatment tests must be completed within the guidelines specified on the test schedule.

6.8
All required baseline symptoms must be documented and graded on the on-study form.

6.9a
Study drug availability checked.

6.9b
Blood draw kit availability checked. Cohort I: All patients accrued during the Cohort I phase will be treated at the starting dose level 0 = GEMZAR 1000 mg/m 2 , CPT-11 75 mg/m 2 . Only the first 6 evaluable patients in each group (i.e., first 6 evaluable patients who carry (are positive for) the UGT1A1*28 polymorphism and first 6 evaluable patients without (are negative for) the UGT1A1*28 polymorphism) will be used in the analysis of Cohort I patients (see 16.14). At the point in which at least 6 evaluable patients in each group have been enrolled, accrual will be suspended and Cohort I patients will be observed for a minimum of 6 weeks for adverse events before any additional patients will be entered. The starting dose for additional patients (referred to as Cohort II) will depend on the adverse events and missed number of doses seen in the first 12 evaluable (6 from each *28 polymorphism group) Cohort I patients. This initial analysis demonstrates that grade 3 or 4 toxicity was not restricted to just the UGT1A1 *28 group, and that toxicity was seen broadly, regardless of genotype. Furthermore, the classic UGT1A1*28 related toxicity (which is grade 4 neutropenia), was seen in only 1/5 patients with the UGT1A1*28 genotype, suggesting that the dose of CPT-11 does not need to be modified for this group. Based on this, we will modify both the schedule of these drugs as well as the dose of Gemcitabine (but not CPT-11). These changes are reflected below.

7.51
Cohort II: Changes to the Cohort II dosing and schedule are as follows: 1) Day 22 (week 4) will be dropped and the cycle will be reduced from 6 weeks to 4 weeks. Treatment will proceed weekly with both drugs for 3 weeks, followed by a 1 week rest period.
Additionally, we will perform an interim toxicity analysis and efficacy analysis once at least 6 patients from each genotype group have been accrued (UGT1A1 *28 versus UGT1A1 6/6). If the study meets the criteria for efficacy at the interim, additional recommendations for dose reductions or genotypeguided dosing will be based on the toxicity analysis at that time. (See section 16.0) 7.6 Efficacy and toxicity analysis for Cohort II An interim efficacy and toxicity analysis will be performed jointly once at least 13 patients have been enrolled to Cohort II and furthermore, at least 6 patients with and 6 patients without the *28 polymorphism (UGT1A1 6/7 and 7/7) have been enrolled (at least 13 and up to 15 patients). If the study does not meet the criteria for efficacy (see section 16), further accrual will cease to this study. If the trial does meet the criteria for efficacy, we will determine the frequency of toxicities in both UGT1A1 genotype groups. Furthermore, the overall cycle 1 toxicity rate for both genotype groups will be reviewed and a decision rendered regarding further dose reductions at that time (see Section 16.0 for details). Retreatment of patients: Decisions regarding retreatment of any patient will be based upon the type, severity, duration, and reversibility of the adverse event reactions. In general, patients who show improvement or stability of their disease will continue to receive treatment. Patients having objective progression of disease or clinical deterioration will not continue to receive treatment on this study. Patients who develop CNS metastasis will not continue to receive treatment on this study.

7.8
During the tolerability assessment portion of the study, if a patient fails to complete the initial course of therapy for reasons other than adverse events, the patient will be regarded as treatment intolerant and will be replaced with an additional patient, to be treated at that same dose level. All adverse event information, however, will be used in the analysis.

AT TIME OF NEXT CYCLE (EVERY 6 WEEKS) BASED ON INTERVAL ADVERSE EVENTS
If dose reductions were used in weeks 1-4 but treatment was not held, use the lowest dose level of drug from the past cycle to start this cycle.
If treatment was held for an adverse event in the last cycle, begin new cycle at one dose level below the last one given. All patients will be instructed to take Loperamide (Imodium or Kaopectate I-D) at the earliest sign of diarrhea (i.e., first poorly formed or loose stool, first episode of an increase from baseline of >2 bowel movements in one day) that occurs more than 12 hours after receiving CPT-11. Loperamide should be taken in the following manner: 4mg at the first onset of diarrhea, then 2 mg every 2 hours around the clock until diarrhea free for at least 12 hours. Patients may take loperamide 4 mg every 4 hours during the night. All patients will be given an instruction sheet (Appendix II) to take home on the first day of their treatment.

9.2
Atropine: Patients should remain in the treatment area for a minimum of one hour following completion of CPT-11 infusion for the first 2 infusions. Thereafter, the patients may leave after the completion of the infusion at their, and the investigator's, discretion. Diarrhea or abdominal cramping that occurs during, or within one hour after receiving CPT-11, should be treated with up to 1 mg of atropine IV (unless there is a relative or absolute contraindication to its use such as, but not limited to, closed angle glaucoma, urinary retention, or previous untoward reactions to the administration of atropine). Additional antidiarrheal measures should be used at the discretion of the treating physician.

9.3
Antiemetics: CPT-11 is moderately emetogenic and pretreatment with dexamethasone and ondansetron (or similar drug) should be used per standard protocols at each treating institution.

9.4
No other chemotherapeutic agents, biologic agents, or radiation may be used while the patient is on study.

9.5
Patients may receive all pre-study concomitant medications. If a patient is on warfarin, a prothrombin time (PT) should be monitored weekly. An increase in the PT has been reported in one patient receiving CPT-11.

9.6
Granulocyte Colony Stimulating Factor (G-CSF): Treatment with G-CSF may not be used to maintain a patient's blood counts for additional cycles of therapy.

Assessment of Attribution
When assessing whether an adverse event is related to a medical treatment or procedure, the following attribution categories are utilized: Definite -The adverse event is clearly related to the agent(s). Probable -The adverse event is likely related to the agent(s).
Possible -The adverse event may be related to the agent(s). Unlikely -The adverse event is doubtfully related to the agent(s). Unrelated -The adverse event is clearly NOT related to the agent(s). All forms noted above are available for downloading from the NCCTG web site. Upon receipt of the above reports, the NCCTG Operations Office will forward reports to all regulatory agencies (including NCI, FDA, and others) as applicable. Specifically, and in accordance with the relevant section of the NCI Guidelines: Expedited Reporting Requirements for NCI Investigational Agents, the Operations Office will forward FDA Form 3500 (MedWatch) reports to the FDA and to NCI within the allotted (total) 10 working days and using one of the cited submission mechanisms.

Additional Instructions or Exceptions to AdEERS Expedited Reporting Requirements for Phase 2 and 3 Trials Utilizing an Agent Under a CTEP IND:
• In the rare event when Internet connectivity is disrupted, a report may be prepared using the If an AdEERS report has been submitted, this form does not need to be submitted.
Enter into the remote data entry system within 5 working days of notification. • All measurements should be recorded in metric notation (i.e., decimal fractions of centimeters) using a ruler or calipers.
• The same method of assessment and the same technique must be used to characterize each identified and reported lesion at baseline and during follow-up. For patients having only lesions measuring at least 1 cm to less than 2 cm must use spiral CT imaging for both pre-and post-treatment tumor assessments.
• Imaging-based evaluation is preferred to evaluation by clinical examination when both methods have been used at the same evaluation to assess the antitumor effect of a treatment.
11.32 Acceptable imaging modalities for measurable disease: CT scan (conventional and spiral), MRI, chest x-ray, and physical examination.
• Conventional CT and MRI must be performed with cuts of 1.0 cm or less in slice thickness contiguously.
• Spiral CT must be performed using a 5 mm contiguous reconstruction algorithm. This specification applies to tumors of the chest, abdomen, and pelvis, while head and neck tumors and those of the extremities require specific procedures.
• Ultrasound (US) is not acceptable to measure tumor lesions that are clinically not easily accessible.
• Color Photography: In the case of skin lesions, documentation by color photography including a ruler to estimate the size of the lesion is recommended.
11.33 Measurement at Follow-up Evaluation: • A subsequent scan must be obtained 4 weeks following initial documentation of an objective status of either complete response (CR) or partial response (PR).
• In the case of stable disease (SD), follow-up measurements must have met the SD criteria at least once after study entry at a minimum interval of 6 weeks (see Section 11.44).
• The cytological confirmation of the neoplastic origin of any effusion that appears or worsens during treatment when the measurable tumor has met criteria for response or stable disease is mandatory to differentiate between response or stable disease (an effusion may be a side effect of the treatment) and progressive disease.
• Cytologic and histologic techniques can be used to differentiate between PR and CR in rare cases (e.g., residual lesions in tumor types such as germ cell tumors, where known residual benign tumors can remain.)

Target Lesions
All measurable lesions (as defined in Section 11.21) up to a maximum of 10 lesions representative of all involved organs should be identified as target lesions and recorded and measured at baseline. If the protocol specified studies are performed, and there are fewer than 10 lesions identified (as there often will be), there is no reason to perform additional studies beyond those specified in the protocol to discover new lesions. For any one organ, no more than 5 lesions need to be measured. Target lesions should be selected on the basis of their size (lesions with the longest diameter) and their suitability for accurate repetitive measurements (either by imaging techniques or clinically).
A sum of the longest diameter (LD) for all target lesions will be calculated and reported as the baseline sum LD. The baseline sum LD will be used as reference to further characterize the objective tumor response of the measurable dimension of the disease.

Non-Target Lesions
All other lesions (or sites of disease) should be identified as non-target lesions and should also be recorded at baseline. Measurements are not required, and these lesions should be followed in accord with 11.433.
11.43 Response Criteria 11.431 All identified sites of disease must be followed on re-evaluation. Specifically, a change in objective status to either a PR or CR cannot be done without rechecking all identified sites (i.e., target and non-target lesions) of pre-existing disease.

Evaluation of Target Lesions
• Complete Response (CR): Disappearance of all target lesions.
• Partial Response (PR): At least a 30% decrease in the sum of the LD of target lesions taking as reference the baseline sum LD.
• Progression (PD): At least a 20% increase in the sum of LD of target lesions taking as reference the smallest sum LD recorded since the treatment started or the appearance of one or more new lesions.
• Stable Disease (SD): Neither sufficient shrinkage to qualify for PR nor sufficient increase to qualify for PD taking as references the smallest sum LD.

Evaluation of Non-Target Lesions
• Complete Response (CR): Disappearance of all non-target lesions.
• Stable Disease (SD): Persistence of one or more non-target lesions.
• Progression (PD): Appearance of one or more new lesions. Unequivocal progression of existing nontarget lesions.
NOTE: Although a clear progression of "non-target" lesions only is exceptional, in such circumstances, the opinion of the treating physician will prevail, and the progression status will be confirmed at a later time by the study chair or a review panel.

Overall Objective Status
The overall objective status for an evaluation is determined by combining the patient's status on target lesions, non-target lesions, and new disease as defined in the following table.

13.1
Patients who are at least stable at the time of their reassessment and who have not experienced intolerable adverse events will continue with the protocol treatment at the same dose level until PD. The exception is for those patients in Cohort I who are negative for (do not carry) the UGT1A1*28 polymorphism (6/6 promoter TA repeats) who complete the first 6-week cycle of treatment at the starting dose (GEMZAR 1000 + CPT-11 75) without any missed or delayed doses and do not require any dose reductions. For these patients, the dose may be escalated to dose level +1 with subsequent cycles (GEMZAR 1000 mg/m 2 , CPT-11 100 mg/m 2 ). Further dose escalation for this Cohort is not permitted.

13.2
Those patients with disease progression, unacceptable adverse events, or refusal of further treatment will go onto the event-monitoring phase and will be followed every 3 months for a maximum of 2 years. (As of Addendum 8, patients no longer need to be followed)

13.3
Patients who develop evidence of CNS disease will be taken off study and go to event monitoring. (As of Addendum 8, patients no longer need to be followed)

13.4
If a patient does not receive treatment and is classified as a cancel, it is not necessary to provide follow-up information. On-study material is to be submitted.

Description of Assays
NOTE: This blood sample is required for UGT1A1 *28 genotyping and for research testing. MDR genotyping, gemcitabine pharmacogenetics and gemcitabine phenotyping are mandatory. Blood sample kits are available and must be used.
The laboratory utilizes allele-specific restriction fragment length polymorphism (RFLP) assays with GeneScan detection for detecting the presence of functional polymorphisms. The laboratory is continuously assessing and updating genotyping assay methodologies as well as adding assays for additional functional polymorphisms using the resources in Dr. Ames' and Dr. Weinshilboum's laboratories, the Mayo Cancer Center MicroArray Shared Resource and the Mayo Genomics Center Developmental Laboratory. An assay for the detection of the UGT1A1*28 promoter polymorphism (utilizing the method of Akaba et al (49) has been developed in the laboratory of Dr. Ames. Assays to detect both the UGT1A1 -3279 and -3156 variants have been developed. In addition, assays for the detections of functional polymorphisms in MDR1 exons 12, 21, and 26 are currently in place.
Additionally, the collected DNA will be used for Gemcitabine pharmacogenetic studies. For gemcitabine, the laboratories of Drs Weinshilboum and Ames are currently involved in resequencing studies with several nucleoside transporters, cytidine deaminase, dCMP deaminase, deoxycytidine kinase, and deoxycytodilate kinase. Functional polymorphisms in these genes will be correlated with the gemcitabine phenotype studies.

Gemcitabine phenotypic studies
After gemcitabine is transported into cells, the parent drug is sequentially phosphorylated to the triphosphate (dFdCTP). dFdCTP is incorporated into DNA and this mechanism is believed to be associated with drug response. Cellular measurements of dFdCTP reflect the aggregate consequences of cellular uptake, anabolism and catabolism. The hypothesis of this phenotypic study is that intracellular triphosphates in the mononuclear cells will reflect variation in the transport, anabolism and catabolism of gemcitabine as a consequence of genetic variation in the genes and protein products associated with those processes. This pharmacologic or intervening phenotype will be used for correlation with genotypes determined in this study. Gemcitabine concentrations will be measured in the plasma and dFdCTP concentrations will be measured in the mononuclear cells. 14.321 Collect one 10 ml of whole blood in an EDTA tube (lavender top) prior to treatment and immediately refrigerate sample.
14.322 Collect whole blood in 2-10 ml tubes (one heparinized tube -green top, one CPT tube -tiger top) each containing 5 micromole/L tetrahydrouridine (THU; see kit instructions) at the following time points: ▪ pre-infusion ▪ 30 minutes (end of infusion) ▪ 90 minutes after start of infusion (60 minutes after completion of infusion).
Process all samples according to kit instructions. Plasma, buffy coat, and red cells will be isolated from the heparin samples and buffy coat will be isolated from the CPT samples. Immediately freeze specimens at -20°C 14.3311 Kits will be sent via FedEx® Ground at no additional cost to the participating institutions. Allow at least two weeks to receive the kits.
14.3312 Kits will not be sent via rush delivery service unless the participating institution provides their own FedEx® account number or alternate billing number for express service. NCCTG will not cover the cost for rush delivery of kits.
14.332 Ship specimens and NCCTG Blood Specimen Submission Form to Mayo Central Laboratory for Clinical Trials (MCLCT) as follows: 14.3321 Ship EDTA tube with a solidly frozen cold pack (see kit instructions for proper packing of blood and cold pack to avoid freezing of specimen.) 14.3322 Ship frozen samples (i.e., three tubes each of plasma, buffy coat, and red cells isolated from the three heparin tubes, and three tubes each of buffy coat isolated from the three CPT tubes) on dry ice.
14.333 All samples should be collected Monday-Thursday ONLY. Do not send samples on weekends or holidays.
14.334 Make sure the specimen tubes are correctly labeled with patient initials, NCCTG patient ID number, protocol number, and time and date drawn.
14.335 ALL sections of the form/specimen collection labels must be completed.
14.336 NOTE: A small, but sufficient, supply of the specimen collection kits should be ordered prior to patient entry.
14.337 NOTE: The kit contains instructions for collecting and processing specimens for shipping.
14.338 Use kit mailing labels for shipment to MCLCT.
14.339 MCLCT will receive the samples and forward specimens within two hours of receipt to the NCCTG Research Base Biospecimens Accessioning and Processing (BAP) Shared Resource, Stabile 13-10A, attention BAP Supervisor.
14.3391 BAP will record receipt of pretreatment EDTA sample. BAP will isolate the DNA, immediately aliquot 10 ul (before quantification) for UGT1A1 genotyping in the Matthew Ames' Laboratory and store the remaining DNA indefinitely. BAP Ames' laboratory for immediate pickup of the 10 ul aliquot and the frozen specimens. UGT1A1 results will be reported in three business days from sample receipt. UGT1A1 genotyping results will be faxed from Dr. Ames' laboratory to the NCCTG member site, NCCTG randomization office, and the principal investigator. As additional relevant functional polymorphisms are identified, future analyses will be conducted depending on the patient consent permission.
14.3392 BAP will record receipt of all specimens for the gemcitabine phenotype studies. BAP will contact Stephanie or Mary (507-284-4303) in Dr. Ames' laboratory for immediate pickup of the frozen specimens (i.e., three tubes each of plasma, buffy coat, and red cells isolated from the three heparin tubes, and three tubes of buffy coat isolated from the three CPT tubes).

14.34
Banking of tumor tissue for future research of tissue correlates of response and/or toxicity: Submission of tissue samples for research testing is mandatory.
14.341 The following materials are required for tissue samples submitted for this protocol: • NCCTG Tissue Specimen Submission Form • A copy of the operative and pathology reports.
• At least one (three, if possible) paraffin-embedded blocks with representative tumor. The formalin fixed tissue sample is preferred; however, if an institution is unable to release tissue blocks, they must be willing to submit 6 unstained charged slides cut at 5 microns.
14.342 The blocks will be available upon specific request to accommodate individual patient management. The institutional pathologist must be notified that the block may be depleted. At the completion of the study, remaining blocks will be retained in the NCCTG Operations Office.
14.343 Additional stains will be performed on all tissue for eligible patients. These are investigational only, and results will be used in a separate analysis. These stains include: CK7 and CK20 profile. Certain patterns of staining are characteristic of primary sites, but not specific enough for diagnosis (52). We will determine if particular staining combinations are associated with response rate in an exploratory manner (there are too few patients to perform a planned subset analysis). Samples will be analyzed in the labs of Marie Christine Aubry, M.D. and Timothy F. Drevyanko, M.D.
14.344 92 gene RT-PCR cancer classification assay. Blinded, de-identified, and de-linked cancer samples in the form of three 5 micron unstained tissue sections plus an adjacent H&E stained section will be received by AviaraDx from the Mayo Clinic. The H&E stained section will be examined by either a designated pathologist at Mayo Clinic or AviaraDx's medical director (Bernie Chang, M.D.) to confirm the presence of cancer as well as outline the cancerous area with a "sharpie pen" to facilitate subsequent macrodissection of the tumor of interest. 14.345 Blocks or slides should be placed in individual plastic bags and each bag labeled with the cooperative group membership name, study patient number (local ID and NCCTG enrollment number), patient's initials, protocol number, surgical accession number, and source (e.g., primary).
14.346 The material will be stored for future research depending on the patient consent permission (see Section 6.3) in order to be suitable for immunohistochemistry (IHC) and to perform DNA extraction. When a protocol is developed, it will be presented for IRB review and approval. The blocks may also be used to construct tissue microarrays (TMAs) for future studies. TMAs will be analyzed to assess predictive biomarkers, changes in expression pattern with therapy, and correlation with response and/or adverse events. 15.126 Other: Headache, shortness of breath, pneumonitis, mild chills and fever, arthralgias, tachycardia, flu-like symptoms, asthenia, malaise, and fatigue, weakness with or without myalgia, cardiac insufficiency, irregular heartbeat, elevated bilirubin, capillary leak syndrome, change in blood chemistry.
15.127 Lilly recommends a two-week delay between the conclusion of radiation and the start of gemcitabine, provided the acute effects of radiation treatment have resolved.

Nursing guidelines:
15.131 Monitor CBC, differential, PLTs prior to each dose. Myelosuppression is the principal dose-limiting factor. Modification may be considered by physician when bone marrow suppression is suspected.
15.132 Evaluate hepatic and renal function prior to initiation of therapy and periodically thereafter. Closely observe those patients with a history of preexisting mild renal impairment or hepatic insufficiency. Encourage hydration.
15.133 GEMZAR clearance is affected by age and gender. Grade 3/4 thrombocytopenia has been more common in elderly women.
15.134 Antiemetics may be required for probable mild to moderate nausea and vomiting.
15.135 Instruct patient in management of possible mild diarrhea and stomatitis.
15.136 GEMZAR may cause fever in the absence of clinical infection. Fever can be accompanied by other flu-like symptoms. Instruct patient to report fever or flu-like symptoms to healthcare team. Treat symptoms as they occur. 15.137 Macular or finely granular maculopapular eruptions were experienced by 30% of patients tested. Instruct patients to report any skin changes.
15.138 Instruct patient to report any cough, shortness of breath, or chest pain immediately. • Hypersensitivity reactions -hypotension and/or bronchospasm or generalized rash/erythema, flushing, rash with or without pruritus, chest tightness, back pain, dyspnea, drug fever, and chills. • Dermatologic -Alopecia. A generalized rash with or without pruritis may occur. Irinotecan may cause local irritation at the infusion site. Extravasation necrosis of the skin has not been reported in US studies.
• Gastrointestinal -nausea, vomiting, diarrhea. Irinotecan can induce both early and late forms of diarrhea that appear to be mediated by different mechanisms. Both forms of diarrhea may be severe. Late diarrhea has been dose limiting in some phase I and II studies. Early diarrhea (occurring during or within 8 hours of administration) may be preceded by complaints of diaphoresis, abdominal cramping, and lacrimation. This syndrome is thought to be cholinergically mediated and may be ameliorated by atropine. Diarrhea or abdominal cramping that occurs during or within one hour after receiving irinotecan should be treated with 0.25 to 1 mg of atropine IV (unless there is a relative or absolute contraindication to its use such as, but not limited to, closed-angle glaucoma, urinary retention/obstructive uropathy, or previous untoward reactions to the administration of atropine). Late diarrhea (occurring more than 8 hours after administration of CPT-11) can be prolonged, may lead to dehydration and electrolyte imbalance, and can be life-threatening. Late diarrhea should be treated promptly with loperamide (suggest: loperamide 4 mg at first onset of diarrhea, then 2 mg q2h around the clock until diarrhea free for at least 12 hours. If diarrhea is not controlled after 3 days, loperamide should be stopped, and hydration and hospitalization should be considered). Patients with severe diarrhea should be carefully monitored and given fluid and electrolyte replacement if they become dehydrated. Infrequent occurrences of mucositis or colitis, sometimes with gastrointestinal bleeding, have been reported. 15.282 Cholinergic symptoms of lacrimation, nasal congestion, diaphoresis, flushing, ABD cramping, and diarrhea can occur at the beginning, during, or immediately after the CPT-11 infusion. It is suggested that the patient remain in the treatment area for a minimum of one hour following the completion of the very first CPT-11 infusion. If diarrhea occurs within one hour of infusion, refer to Section 9.2 for management. 15.283 Patient education is extremely important. Impress on the patient the importance of compliance with treatment of diarrhea management (see Section 9.1 and Appendix III). Stress the need for prompt recognition and early intervention. Motivate the patient to report any complications immediately. The cholera-like syndrome can be unresponsive to conventional antidiarrheals and can result in severe dehydration. 15.284 Ondansetron and diphenhydramine should provide good relief from the nausea/vomiting/cramping. Avoid prochlorperazine on the day of treatment due to its association with akathisia (motor restlessness). Prochlorperazine may be taken between treatments. 15.289a Monitor CBC closely. Leukopenia occurs primarily as neutropenia but can be severe and dose-limiting. The simultaneous occurrence of grade 4 diarrhea and grade 4 neutropenia is rare but may render the patient more susceptible to polymicrobial sepsis and potentially death.
15.289b Advise patients of probable hair loss.

16.1
Overview: As of Addendum 4, this is the new study design to use for this study. The original study design is listed starting in section 16.6. We developed a new study design because (1) the accrual rate was found to be much lower than expected (1/mo vs. 3-4/mo) and (2) due to the level of toxicity observed for cohort I (regardless of genotype), a recommendation was made to reduce the gemcitabine dose and to reduce the cycle from 6 weeks to 4 weeks by dropping the last day of treatment (d22). Due to these major changes to the treatment schedule, we redesigned the study by reducing the overall sample size and to assess (for efficacy), only Cohort II patients. Cohort II consists of all patients accrued on or after the Addendum 4 effective date. This new design is setup as a one-stage phase II study with an interim analysis. This study will assess the toxicity and confirmed response rate associated with the treatment regimen of gemcitabine and irinotecan for patients with unknown primary carcinoma. 16.14 Toxicity Stopping Rule: We'll assess the first 6 evaluable patients in each genotype group at the new Addendum 4 dose level and treatment schedule. We expect about 50% of patients to carry the UGT1A1*28 polymorphism, and the other 50% to be negative for it. With this in mind, we expect to only have to accrue about 15 patients during this phase to get 6 patients in each group (i.e., 6 patients with the UGT1A1*28 polymorphism, and 6 without it). If the trial does meet the criteria for efficacy, we will determine the frequency and nature of the cycle 1 adverse events. If the overall frequency of cycle 1 grade 4+ non-hematologic adverse events (at least possibly related to treatment) is > 33% across all patients, consideration will be given to stopping the trial. If however, the frequency of grade 4+ non-hematologic adverse events is acceptable (< 33%), we will go on to determine the incidence of the following cycle 1 hematologic adverse events (regardless of attribution): Grade 4+ neutropenia, grade 3+ febrile neutropenia and grade 3+ infection in the presence of grade 3 or 4 neutrophils. The following scenarios apply for assessment of these hematologic adverse events: 1) If ≥33% of UGT1A1 6/7 or 7/7 patients experience at least one of the above cycle 1 hematologic adverse events while < 33 % of UGT1A1 6/6 patients experience these cycle 1 adverse events, an addendum will be written to establish genotype-based dosing with a reduced dose for patients who carry the UGT1A1*polymorphism. 2) If in the overall group (all patients) the incidence of these cycle 1 hematologic adverse events is ≥33% and there does not exist a clear relationship with UGT1A1 genotype, consideration will be given to closing the trial at that time. 3) If in the overall group (all patients) the incidence of these cycle 1 hematologic adverse events <33%, and there does not exist a clear relationship between these adverse events and UGT1A1 genotype, the trial will continue at the same dose, regardless of genotype. As the trial proceeds beyond the initial assessment of adverse events described above, we will continue to use the following adverse event stopping rules in all patients. If either (1) or (2) occurs below, we will suspend the trial and do a full review of the data. 16.23 Decision Rule: The largest success proportion where the proposed treatment regimen would be considered ineffective in this population is 20%, and the smallest success proportion that would warrant subsequent studies with the proposed regimen in this patient population is 40%. The following Three-Outcome Phase II study design (51) with an interim analysis uses a maximum of 25 patients to test the null hypothesis that the true success proportion in a given patient population is at most 20%. An interim analysis will be performed at the time the 13 th patient becomes evaluable for the Addendum 4 dose level. If at least 3 successes are observed in these initial 13 evaluable patients (regardless of genotype for those patients receiving the Addendum 4 dose level), we will continue enrollment to a maximum of 25 evaluable patients (regardless of genotype or dose level after the dose was changed for Addendum 4). Otherwise, we will discontinue enrollment and conclude that the regimen is not sufficiently active in this patient population.
16.231 "Not promising": This regimen will be classified as not promising with respect to the success rate in this patient population if at most 6 successes are observed in a total of 25 evaluable patients.
16.232 "Inconclusive": The results of this study will be classified as inconclusive with respect to this regimen demonstrating an improved success rate if 7 successes are observed in 25 evaluable patients. In this case, toxicity, quality of life, and duration of response observed in this have had tumor progression at the time of their death unless there is sufficient documented evidence to conclude no progression occurred prior to death. If the patient is declared to be a major treatment violation, the patient will be censored on the date the treatment violation was declared to have occurred. In the case of a patient starting treatment and then never returning for any evaluations, the patient will be censored for progression on day 1 post-registration. The distribution of time to progression will be estimated using the method of . Time to disease progression will be calculated for all evaluable patients combined and by group (ie. for patients with or without the UGT1A1*28 polymorphism).
16.33 Toxicity: As per NCI CTCAE v3.0, the term toxicity is defined as adverse events that are classified as either possibly, probably, or definitely related to study treatment. The maximum grade for each type of adverse event will be recorded for each patient, and frequency tables will be reviewed to determine adverse event patterns. In addition, we will review all adverse event data that is graded as 3, 4, or 5 and classified as either "unrelated or unlikely to be related" to study treatment in the event of an actual relationship developing. Toxicity will be calculated for all evaluable patients combined and by group (ie. for patients with or without the UGT1A1*28 polymorphism).

Over Accrual:
If more than the target number of patients are accrued, the additional patients will not be used to evaluate the stopping rule or used in any decision-making processes; however, they will be included in final point estimates and confidence intervals.
16.35 Routine Monitoring: Efficacy, toxicity, and administrative information for this trial will be reviewed by the study team twice per year in conjunction with production of the semiannual NCCTG Group Meeting reports. They will monitor the trial for evidence of severe adverse events and feasibility problems.

16.4
Translational Component: A blood sample will be collected ≤14 days prior to registration to assess whether variation in multiple different genes whose protein products are involved in the uptake, metabolism, and distribution of Gemzar and CPT-11 affect clinical outcomes (i.e., response, toxicity, survival, time to progression, etc.). All analyses with respect to the translational component of this study are intended to be hypothesis-generating and descriptive in manner. The correlations between gene type (obtained at baseline) and response (and toxicity) will be explored via frequency tables (i.e., Chi-square analyses). The correlation of gene type with other patient outcomes (i.e., survival, time to progression, etc.) will be explored via Cox proportional hazards regression (52). The distributions of survival, and time to progression defined by the different genes will be estimated via Kaplan-Meier methodology (50).
The translational component of this study also proposes to stain all specimens for CK7/20 in an attempt to ascertain which patterns may be predictive of response to this regimen. This will be an exploratory analysis, as the numbers in this phase II study are too small to make any firm conclusions. We will also request the tissue blocks be retained here so In addition, we plan to assess the relationship between the site of origin (by gene expression profiling) and response/time-to-progression.

Inclusion of Women and Minorities
16.51 This study will be available to all eligible patients, regardless of race, gender, or ethnic origin.
16.52 There is no information currently available regarding differential effects of this regimen in subsets defined by race, gender, or ethnicity, and there is no reason to expect such differences to exist. Therefore, although the planned analysis will, as always, look for differences in treatment effect based on racial and gender groupings, the sample size is not increased in order to provide additional power for subset analyses.
16.53 Based on prior studies involving similar disease sites, we expect about 5% of patients will be classified as minorities by race and about 33% of patients will be women. Estimates of racial by gender subsets are shown in the following American Indian or Alaskan Nativea person having origins in any of the original peoples of North, Central, or South America, and who maintains tribal affiliations or community attachment. Asiana person having origins in any of the original peoples of the Far East, Southeast Asia, or the Indian subcontinent including, for example, Cambodia, China, India, Japan, Korea, Malaysia, Pakistan, the Philippine Islands, Thailand, and Vietnam. (Note: Individuals from the Philippine Islands have been recorded as Pacific Islanders in previous data collection strategies.) Black or African Americana person having origins in any of the black racial groups of Africa. Terms such as "Haitian" or "Negro" can be used in addition to "Black or African American." Native Hawaiian or other Pacific Islandera person having origins in any of the original peoples of Hawaii, Guam, Samoa, or other Pacific Islands. Whitea person having origins in any of the original peoples of Europe, the Middle East, or North Africa.

16.6
Original Statistical Design (Prior to Addendum 4) Design Overview: This is a one-stage phase II study with an interim analysis and a lead-in phase to assess tolerability. This study will assess the tolerability and confirmed response rate associated with the treatment regimen of gemcitabine and irinotecan for patients with unknown primary carcinoma.
16.61 Primary Endpoint: The primary endpoint is confirmed response rate. If measurable disease is present, a confirmed tumor response is defined to be either a CR or PR noted as the objective status on 2 consecutive evaluations at least 6 weeks apart (see Section 11.0). All registered patients meeting the eligibility criteria that have signed a consent form and have begun treatment will be evaluable for response. With this in mind, we expect to only have to accrue about 15 patients during the Cohort I phase to get 6 patients in each group (i.e., 6 patients in the UGT1A1*28 polymorphism group, and 6 patients in the group without the 28 polymorphism). If more than 15 patients are needed during Cohort I, an addendum will be developed to revise the projected accrual for the Cohort I phase. We anticipate accruing an additional 2 patients to account for ineligibility, cancellation, major treatment violation, or other reasons during Cohort I. Therefore maximum accrual during Cohort I is expected to be 17 patients. All patients enrolled during the Cohort I phase will be treated at dose level 0 (see Section 7.2). After the first 6 evaluable patients have been enrolled into each group (i.e., 6 patients with the UGT1A1*28 polymorphism, and 6 without it), accrual will be suspended, and adverse events will be evaluated before any additional patients will be entered. Only the first 6 evaluable patients entered into each group (i.e., with or without the *28 polymorphism) will be used to determine the appropriate Cohort II dose levels, no matter how many patients end up being enrolled during Cohort I.
• If the Cohort I starting dose level gets adjusted for both groups (i.e., group 1: patients with UGT1A1*28 polymorphism, group 2: patients without the *28 polymorphism) for the Cohort II phase, a maximum of 37 additional patients (regardless of whether they have the UGT1A1*28 polymorphism or not) will be treated (per study design) at the new dose levels determined during the Cohort I phase (see Section 7.4). Only the patients treated at the appropriate Cohort II dose levels will be included in the decision rule described below in Section 16.23.
• If the Cohort I starting dose level is adjusted for one group only (i.e., either group 1 or 2), a maximum of 31 additional patients (regardless of whether they have the UGT1A1*28 polymorphism or not) will be treated (per study design) at the dose levels determined during the Cohort I phase (see Section 7.4). Only the patients treated at the appropriate Cohort II dose levels will be included in the decision rule described below in Section 16.23.
• If the Cohort I starting dose level is not adjusted for either group, a maximum of 25 additional patients will continue to be treated (per study design) at dose level 0 (see Section 7.2), and all patients will be included in the decision rule described below in Section 16.23.  (50) we would expect at most 20% of patients to have a confirmed response. For an agent to provide significant clinical improvement from standard therapy, the confirmed response rate should be at least 40%.
16.72 Definition of a Success: A success is defined as a patient who has a confirmed response as defined in Sections 11.0 and 16.11.
16.73 Decision Rule: The largest success proportion where the proposed treatment regimen would be considered ineffective in this population is 20%, and the smallest success proportion that would warrant subsequent studies with the proposed regimen in this patient population is 40%. The following Three-Outcome Phase II study design (51) with an interim analysis uses a maximum of 37 patients to test the null hypothesis that the true success proportion in a given patient population is at most 20%. An interim analysis will be performed at the time the 17 th patient becomes evaluable for the appropriate Cohort II dose levels.
If at least 4 successes are observed in these initial 17 evaluable patients, we will continue enrollment to a maximum of 37 evaluable patients treated at the appropriate Cohort II dose levels. Otherwise, we will discontinue enrollment and conclude that the regimen is not sufficiently active in this patient population.
16.731 "Not promising": This regimen will be classified as not promising with respect to the success rate in this patient population if at most 10 successes are observed in a total of 37 evaluable patients.
16.732 "Inconclusive": The results of this study will be classified as inconclusive with respect to this regimen demonstrating an improved success rate if 11 successes are observed in 37 evaluable patients. In this case, toxicity, quality of life, and duration of response observed in this study will be used in addition to confirmed response rate to make the final determination as to whether or not this treatment is considered promising and worthy of further study in this patient population.
16.733 "Promising": This regimen will be classified as promising with respect to increasing the success rate in this patient population if at least 12 successes are observed in 37 evaluable patients. Subsequent larger confirmatory studies may be recommended. In order to assure prompt handling, please call Christine Maszk at (507) 266-8919 to alert of the time/date sent and courier contracted.
Marie Christine Aubry, M.D., will review the slides for confirmation of disease.
17.13 The NCCTG Pathology Coordinator will take the blocks/slides to NCCTG Tissue Acquisition and Cellular/Molecular Analysis Shared Resource laboratory. Once the blocks/slides are prepared Dr. Aubry will receive them for review. The NCCTG Pathology Coordinator will call the institution notifying them of the results. The NCCTG Pathology Coordinator will fax a copy of the completed Marie-Christine Aubry, M.D., will review the slides for confirmation of disease.
17.23 If pathology review finds the patient ineligible based on the criteria, this patient will be considered a protocol violation but may continue on study. All slides, remaining paraffin block, and forms will be returned to the submitting institution.