Utility of Antiphosphorylated H2AX Antibody (y-H2AX) in Diagnosing Metastatic Renal Cell Carcinoma
Matthew J. Wasco, MD and Robert T. Pu, MD, PhD
Abstract: The differential diagnosis of metastatic renal cell carcinoma (RCC) includes, although is not limited to, hepato- cellular carcinoma (HCC) and adrenocortical carcinoma (ACC) due to overlapping morphology. Immunohistochemical mar- kers, including RCC marker (RCC-Ma) have been employed with varying success in the differential diagnosis of RCC. Our preliminary tissue microarray study demonstrated that y-H2AX, an antibody that specifically reacts with phosphorylated histone H2AX, stained many primary RCC strongly and did not stain HCC or ACC, prompting us to evaluate its utility in these tumors and to compare it with RCC-Ma. Seventy-one cases of metastatic RCC, 18 HCC, and 21 ACC were stained with y-H2AX and RCC-Ma and the sensitivity and specificity of each marker was compared. RCC-Ma demonstrated a membranous pattern of staining in 70% of RCC cases (50/71), and none of the ACC or HCC (100% specificity for RCC). Nuclear staining by y-H2AX had a similar sensitivity of 70% for RCC but a lower specificity of 77%, as it was seen in 1 of 18 HCC (5%) and 8 of 21 (38%)ACC. In metastatic RCC, 83% (39/47) of tumors with a higher nuclear grade stained with y-H2AX, compared with 46% (11/24) of low nuclear grade (equivalent of Fuhrman 2 and lower) tumors. RCC-Ma had a similar rate of staining in low and high-grade tumors, 75% (18/24) and 68% (32/47), respectively. More importantly, of RCCs that were negative for RCC-Ma, 14 of 21 (67%) were positive for y-H2AX. The results suggest y-H2AX is a useful adjunct in diagnosis of metastatic RCC when RCC-Ma is negative and in higher grade RCC, which are often a diagnostic challenge. A nuclear pattern of staining of y-H2AX has a comparable sensitivity with RCC-Ma, and the interpretation is easier and more reliable. RCC-Ma is 100% specific for RCC, but only when a membranous pattern of staining is interpreted as positive.
Key Words: metastatic RCC, differential diagnosis, RCC-Ma, γ-H2AX
(Appl Immunohistochem Mol Morphol 2008;16:349-356)
From the Department of Pathology, The University of Michigan Medical School, Ann Arbor, MI.
Disclosure/Conflict of interest: The authors have no relevant financial disclosures or conflict of interest in the products and methods used in this paper.
Reprints: Robert T. Pu, MD, PhD, Department of Pathology, The University of Michigan Medical School, 1500 E. Medical Center
Drive, Room 2G332, Ann Arbor, MI 48109 (e-mail: robertpu@ umich.edu).
Copyright @ 2008 by Lippincott Williams & Wilkins
I n certain cases, diagnosing metastatic renal cell carci- noma (RCC) can be a challenge. Although many RCCs have a classic morphology, higher grade tumors and metastases often have features that make the tumor difficult to distinguish morphologically from other possible diagnoses. In particular, adrenocortical carci- noma (ACC) and hepatocellular carcinoma (HCC) are often in the differential diagnosis of a metastatic tumor when the primary tumor is not known or confirmed, or when the tumor is located within the liver or adrenal gland.1 Both ACC and HCC can also have clear cell variants that histologically mimic RCC.
Immunohistochemistry is often not required in classic examples of these 3 diagnoses or when pertinent clinical information is available, but the spectrum of morphology is such that there are many overlapping features, particularly when the tumors take on a clear cell morphology. In addition, the diagnosis of metastatic disease is frequently attempted with relatively small samples, such as needle core biopsy or fine needle aspiration, which can make accurate diagnosis all the more difficult.2
Many immunohistochemical stains have been stu- died to attempt to find markers that will reliably identify or exclude these entities, and the studies have had varying success rates.3 To date, however, there are no reliable stains or panels to clearly distinguish these tumors. For example, hepatocyte paraffin-1 is a relatively specific marker for HCC that is generally negative for RCC,4 and inhibin A is a marker relatively specific for ACC.5,6 In general, however, the diagnosis usually needs to be confirmed with a combination of stains and the frequency of false positive and negative staining still remains high.
The antibody known as RCC marker (RCC-Ma) was developed as a monoclonal antibody to normal human kidney proximal tubule,7 and then found to be highly specific for RCC, both primary and metastatic, in many studies.2,8 Other studies, however, and our own anecdotal working experience have demonstrated that the interpretation of this stain is often difficult and sensitivity is not high, particularly in small biopsies.9,10 In particular, the stain can be focal, and many normal tissues, including liver and adrenal, often have a significant amount of background cytoplasmic staining, which complicates the interpretation. Staining by RCC-Ma is also most clearly demonstrated in classic cases of RCC when immunohistochemical adjuncts to diagnosis are not always necessary.
Y-H2AX is an antibody that specifically reacts with phosphorylated histone H2AX (at position Ser-139) and which has been shown to be a marker of activated DNA damage in tumor cells.11 It has been used to detect cancers and precursor lesions in organs, including the bladder, colon, breast, and lung.11 During a preliminary tissue microarray study at this institution, the antibody was found to react strongly with many examples of primary RCC while being negative in examples of HCC and normal kidney, liver, and adrenal (Fig. 1).
The purpose of this study therefore was to study the use of y-H2AX in a cohort of RCC, HCC, and ACC to determine its diagnostic utility in distinguishing these lesions. A comparison with RCC-Ma was also done to completely evaluate this marker and to compare the 2 antibodies directly.
MATERIALS AND METHODS
A search of the University of Michigan pathology database was performed with IRB approval. A total of 71 examples of metastatic RCC, 21 examples of ACC, and 18 examples of HCC were included in our study. The specimens included classic versions of the tumors as well as undifferentiated tumors and those with overlapping features. Patient characteristics of the 3 groups are shown in Table 1. The majority of the tumors were clear cell
RCC, but several papillary and mixed papillary/clear cell tumors were also included.
Formalin-fixed, paraffin-embedded tissue sections were stained with y-H2AX and RCC-Ma. Phospho- histone (Ser139) H2AX (y-H2AX) was obtained from Cell Signaling Technology (Cat. No 2577, Danvers, MA). Antigen retrieval pretreatment consisted of 10-minute citrate buffer wash (pH 6.0, microwaved), followed by 10-minute cooling and 10-minute wash in running water. After this pretreatment, immunohistochemical staining was performed on the Dako AutoStainer incubated with the primary antibody at a dilution of 1:50 for 60 minutes at room temperature. After buffer rinse, slides were incubated with rabbit EnVision + labeled polymer for 30 minutes. After an additional buffer rinse, incubation with DAB chromagen was performed for 10 minutes. RCC- Ma (clone 66.4.C2) was obtained from NovoCastra Laboratories Ltd (Newcastle upon Tyne, UK). Antigen retrieval was performed with 15 minutes citrate buffer (pH 6.0, microwaved) followed by 15-minute cooling and 15-minute running water wash. After antigen retrieval and 5 minutes peroxidase block, incubation with primary RCC-Ma antibody was performed at a dilution of 1:100 for 30 minutes. Secondary antibody incubation was performed for 30 minutes, and then tertiary reagent (Streptavidin label) for 30 minutes. Incubation with DAG chromagen was performed for 5 minutes.
A
B
C
D
| Metastatic RCC | HCC | ACC | ||
|---|---|---|---|---|
| Age, mean (y) | 61.6 | 60.4 | 50.8 | |
| Male/female | 71.8%/28.2% | 66.7%/33.3% | 57.1%/42.9% | |
| Tumor sites | No. Cases | |||
| Adrenal | 8 | 2 | 11 (primary) | |
| Bone | 15 | — | — | |
| Liver | 9 | 10 (primary) | 7 | |
| Lung | 9 | 3 | 2 | |
| Lymph node | 5 | — | — | |
| Central nervous system | 6 | — | — | |
| Skin or soft tissue | 6 | 3 | 1 | |
| Gastrointestinal tract | 2 | — | — | |
| Pancreas | 3 | — | — | |
| Parotid | 2 | — | — | |
| Spleen | 1 | — | — | |
| Testes | 1 | — | — | |
| Thyroid | 1 | — | — | |
| Vagina | 1 | — | — | |
| Vulva | 1 | — | — | |
| Total | 71 | 18 | 21 | |
Stains were reviewed by both pathologists (M.J.W. and R.T.P.) and the patterns (membranous, cytoplasmic, etc) of staining and the amount (in percentage) of tumor cells staining were recorded. These characteristics were identified by visual estimation. Any background staining was also recorded. Percentages of cells staining were independently recorded by both pathologists and the majority of the results were in agreement. Rare cases with minor discrepancy were resolved and the consensus percentage was used for such cases. For cases of RCC,
| RCC-Ma | H2AX | |||
|---|---|---|---|---|
| (+) | (-) | (+) | (-) | |
| RCC | 50 (70.4%) | 21 (29.6%) | 50 (70.4%) | 21 (29.6%) |
| HCC | 0 | 18 | 1 (5.6%) | 17 (94.4%) |
| ACC | 0 | 21 | 8 (38.1%) | 13 (61.9%) |
each case was given a Fuhrman nuclear grade as per criteria enumerated in the WHO description.12 As a control, a tissue microarray of primary RCC was also stained with y-H2AX. McNemar test was used for statistical analysis (matched-pairs analysis).
RESULTS
A total of 71 metastatic RCC cases were identified and used for the study. Specimen types included resec- tions and biopsies, consisting of multiple cases from the liver and adrenal as well as lung, bone, brain, soft tissue, pancreas, and thyroid (Table 1). Twenty-four (24) cases were Fuhrman grade 2, 38 Fuhrman grade 3, and 9 Fuhrman grade 4. The majority of the cases were clear cell RCC or had been originally diagnosed as such before metastasis. Three of the cases were papillary RCC and 1 was unclassified. Eighteen cases of HCC and 21 cases of ACC were identified and used for the study (Table 1). The HCC cases included primary tumors (9 cases) and those metastatic (9 cases) to adrenal, lung, and skin. Clear cell variants as well as sarcomatoid and anaplastic variants
A
B
C
D
A
D
B
E
C
F
were included. The ACC cases included primary tumors (11 cases) and those metastatic (10 cases) to liver, lung, and chest wall. Low-grade tumors and high-grade, undifferentiated tumors were included.
Our preliminary tissue microarray study showed that y-H2AX antibody reacted strongly with RCC while being negative in examples of normal kidney, liver, and adrenal (Fig. 1). In our cohort of metastatic RCC, 50 of 71 (70%) cases were positive for RCC-Ma with a membranous pattern of staining in at least 20% of tumor cells (Table 2, Figs. 2A, B). There was significant background staining, predominantly cytoplasmic, in the liver and adrenal, which can be misinterpreted as positive (Figs. 2B-D). However, if only a membranous pattern of staining was considered as positive (with or without cytoplasmic staining), no cases of ACC or HCC were positive (Table 2).
For the y-H2AX antibody, 50 of 71 RCC were positive (70%, Table 2), including the papillary and mixed cases, defined as staining in more than 25% of
tumor nuclei. Most had strong staining (Figs. 3C, F). Background staining was largely limited to cytoplasmic debris, skeletal muscle, some skin nuclei, and actively dividing cells. For HCC, 17 of 18 cases were negative (Fig. 4). Only 1 case of clear cell HCC was positive, a case in which greater than 50% of the nuclei were strongly positive, whereas the other cases of clear cell HCC were negative (less than 25% staining, the majority completely negative). For ACC, 8 of 21 (38%) cases were positive in greater than 25% of tumor cells, including both low- grade and high-grade tumors (Fig. 5, representative ACC, 1 positive and 1 negative for this antibody). A similar percentage of primary RCC in a tissue microarray also stained with y-H2AX (data not shown).
Interestingly, RCC-Ma stained a lower percentage of RCC cases as the nuclear grade of the tumor increased, whereas y-H2AX staining increased (Table 3). Subcate- gorized, 46% (11/24) of Fuhrman 2 tumors stained for Y-H2AX and 75% (18/24) for RCC-Ma. For Fuhrman 3 to 4 tumors, 83% (39/47) stained for y-H2AX and
A
B
C
68% (32/47) for RCC-Ma. However, the difference did not reach statistical significance. In all, 37.5% (9/24) of Fuhrman 2 tumors stained only with RCC-Ma, as compared with 10.6% (5/47) of Fuhrman 3 to 4 tumors; 8.3% (2/24) of Fuhrman 2 tumors and 23.4% (12/47) of
Fuhrman 3 to 4 tumors stained for y-H2AX only. For 9 cases of RCC interpreted as Fuhrman nuclear grade 4, Y-H2AX stained all of them whereas RCC-Ma only stained 4 of 9 cases. Of the 14 RCC that stained only for Y-H2AX, 12 were grade 3 to 4 tumors.
Both RCC-Ma and y-H2AX stained a similar portion of tumor cells when the tumor was positive for the antibody. RCC-Ma stained an average of 69% (median 75%, range 20% to 95%) of tumor cells when positive, and y-H2AX stained an average of 68% (median 75%, range 30% to 90%). When positive, the antibodies stained > 75% of tumor cells in a majority of cases. As noted previously, however, a positive y-H2AX stain was easy to interpret, as it is a nuclear stain, whereas RCC-Ma can have false positive staining and its membranous pattern of staining can be more difficult to detect.
The overall specificity and sensitivity of y-H2AX for RCC was 77% and 70%, respectively (Table 4). For the 21 RCC that were negative for RCC-Ma, 14 of 21 were positive for y-H2AX (67%, Figs. 3D-F). Similarly, for the 21 RCC that were negative for y-H2AX, 14 of 21 were positive for RCC-Ma (67%). Only 7 (10%) RCC were negative for both antibodies, and 36 (51%) were positive for both. If either antibody was positive, sensitivity was 88%, specificity increased to 82%. A summary and comparison of the 2 antibodies alone and in combination can be seen in Table 4.
DISCUSSION
The diagnosis of metastatic RCC can be difficult, particularly with the characteristic of RCC to metastasize to many different sites, and its capacity for varied histologic appearances.12 Common difficult differential diagnoses faced in surgical pathology include the differ- entiation of metastatic RCC to liver from HCC, or metastatic RCC to adrenal gland from ACC.13
The results of this study confirm that RCC-Ma is a highly specific (100%), moderately sensitive (70%) anti- body for RCC, but suggest difficulties with interpretation of the antibody stain, which have been noted previously.3 In particular, there was significant cytoplasmic staining of normal hepatocytes and adrenal cortical cells, as well as tumors. Therefore, one must be careful when interpreting RCC-Ma to ensure that the correct pattern of staining is identified. RCC-Ma is easier to interpret when used on classic RCC, but these are not the cases which so often defy diagnostic ability. The more poorly differentiated RCC cases are more likely to be confused with another neoplasm histologically, and in these cases unfortunately RCC-Ma is less useful. One criticism of the use of RCC- Ma in prior studies has been that it often has a focal staining pattern. We found that, when positive, RCC-Ma stained an average of 68% of tumor cells, with the majority of cases staining 75% or greater. However, the membranous staining pattern can at times be difficult to distinguish from background staining.
We report the use of a new antibody, y-H2AX, as potentially useful marker in diagnosing RCC. This
A
D
B
E
C
F
antibody had a similar sensitivity (70%) for the detection of RCC as did RCC-Ma, but was more useful in higher grade tumors (Table 4) where it stained a higher proportion of Fuhrman 3 and 4 tumors than did RCC Ma (83% of tumors vs. 68%). In fact, y-H2AX stained 9/9 tumors interpreted as Fuhrman nuclear grade 4, whereas RCC-Ma only stained 4/9 cases. Five cases of grade 4 RCC were not stained by RCC-Ma but all of them were positive for y-H2AX (Table 3). As these higher grade tumors are the ones that more often defy conventional diagnostic methods, utilization of y-H2AX may be a useful adjunct to diagnosis in these circum- stances.
y-H2AX has a nuclear pattern of staining (generally strongly) and has, in our experience, less of the back- ground staining that makes RCC-Ma a difficult antibody for result interpretation. RCC-Ma, while in a membra- nous pattern of staining (with or without cytoplasmic staining) is highly specific for RCC, often has background cytoplasmic staining of other tissues that can cause
diagnostic dilemmas, particularly in small biopsies when diagnostic material is limited and artifacts can be hard to ascertain. Although RCC-Ma stains a majority of the tumor cells in our study, it has been reported that its stain can be focal and therefore absent on small biopsies.9 As seen in an image of normal adrenal gland, the pattern of staining can mimic the membranous pattern of staining which is specific for RCC (Fig. 2). And in cases where the tumor is in the liver, significant cytoplasmic staining of hepatocytes can also complicate diagnosis. y-H2AX does not have as much background staining. Our results show a significant proportion of ACC cases do stain for y-H2AX, which may make the stain less useful in cases where the differential diagnosis is between RCC and ACC. Only 1 of 18 cases of HCC stained, however, making this antibody highly valuable if the differential diagnosis is between HCC and RCC.
Although our study was limited to consideration of 3 different tumors, they are the ones that often come into the differential diagnosis because of overlapping
| Grade (No.) | Both (+) | RCC-Ma Only | H2AX Only | Both (-) | H2AX (+) | RCC-Ma (+) |
|---|---|---|---|---|---|---|
| 2 (24) | 9 (37.5%) | 9 (37.5%) | 2 (8.3%) | 4 (16.7%) | 11 (45.8%) | 18 (75%) |
| 3 (38) | 22 (57.9%) | 5 (13.2%) | 8 (21.1%) | 3 (7.9%) | 30 (78.9%) | 27 (71.1%) |
| 4 (9) | 5 (55.6%) | 0 | 4 (44.4%) | 0 | 9 (100%) | 5 (55.6%) |
| Total (71) | 36 (51%) | 14 (20%) | 14 (20%) | 7 (10%) | 50 (70%) | 50 (70%) |
morphologic features. Other studies have reported that tumors of different sites (such as breast) can be positive for RCC-Ma.2 Similarly, y-H2AX antibody was also found positive in lung and ovarian carcinoma (our observation) and breast, prostate, and urinary bladder cancers.11 Thus while limited to a comparison of RCC, ACC, and HCC, our results suggest relatively high specificity and sensitivity for those 2 antibodies in the most common differential diagnostic setting. As stated above, the use of y-H2AX in a clinical setting may be most helpful in higher grade tumors, where RCC-Ma is less often helpful, and will likely be useful as part of a panel of immunohistochemical stains to provide further evidence and support for diagnosis. In RCC-Ma negative tumors, for example, y-H2AX has a negative predictive value of 81% and a positive predictive value of 61%. Further clinical utility for y-H2AX in additional clinical situations will depend on additional studies and compar- ison with other antibodies, although the strength of stain, ease of interpretation, and moderately high sensitivity suggest that y-H2AX is likely to be of some use whenever RCC is in the differential diagnosis. To this end, we have started testing this antibody in a panel with PAX-2 and RCC-Ma in fine needle aspiration material (Wasco and Pu, in preparation).
H2AX and other linker histones have been shown to be crucial for nucleosome formation and consequently gene expression controls. Posttranslational modification such as phosphorylation of a protein has functional consequence. Phosphorylation of H2AX at serine 139, a marker for double-stranded DNA damage, which is important for tumor development, has been demon- strated in several cancers and their precursor lesions.11 We showed in this study that most RCC were marked
| For RCC | Sensitivity (%) | Specificity (%) | PPV | NPV (%) |
|---|---|---|---|---|
| (%) | ||||
| RCC Ma (+) | 70.4 | 100 | 100 | 65.0 |
| H2AX (+) | 70.4 | 76.9 | 84.8 | 58.8 |
| RCC Ma and H2AX both (+) | 50.7 | 100 | 100 | 52.7 |
| H2AX (+) when RCC Ma (-) | 65.0 | 76.9 | 59.1 | 81.1 |
| RCC Ma (+) when H2AX (-) | 66.7 | 100 | 100 | 81.1 |
| Either marker positive | 87.8 | 81.8 | 90.1 | 76.9 |
NPV indicates negative predictive value; PPV, positive predictive value.
with y-H2AX, whereas HCC were not. The different staining pattern by y-H2AX suggests that the double- stranded DNA damage control pathway might be involved differently during the tumorigenesis of these 2 tumors. Similarly, epigenetic change such as DNA methylation also involves different tumor suppressor genes in RCC and HCC.14,15 As H2AX is a known substrate of ATM kinase, drugs targeting ATM kinase pathway might have different effects on RCC as compared with HCC.16 There may be additional clinical implications in selecting potential treatment agents such as histone deacetylase inhibitor (sodium butyrate), which can sensitize tumor cells for radiotherapy by suppressing DNA repair activity.17,18
In conclusion, y-H2AX is an antibody with comparable sensitivity to RCC-Ma for the diagnosis of RCC, but has the benefit of improved performance on identifying higher grade, more poorly differentiated tumors. There is significant staining of ACC (38% of tumors), which limits its diagnostic utility somewhat if ACC is in the differential diagnosis. However, a combination of RCC-Ma and y-H2AX carries a diag- nostic sensitivity of 88% and a specificity of 82% if either marker is positive.
ACKNOWLEDGMENTS
The authors thank Nancy McAnsh, Alan Burgess, and Tina Fields for assistance with the preparation of the slides and immunohistochemistry. They also thank Kent Griffith for his help with statistical analysis.
REFERENCES
1. Ghorab Z, Jorda M, Ganjei P, et al. Melan A (A103) is expressed in adrenocortical neoplasms but not in renal cell and hepatocellular carcinomas. Appl Immunohistochem Mol Morphol. 2003;11:330-333.
2. McGregor DK, Khurana KK, Cao C, et al. Diagnosing primary and metastatic renal cell carcinoma. Am J Surg Pathol. 2001;25: 1485-1492.
3. Simsir A, Chhieng D, Wei XJ, et al. Utility of CD10 and RCCma in the diagnosis of metastatic conventional renal-cell adenocarcinoma by fine-needle aspiration biopsy. Diagn Cytopathol. 2005;33:3-7.
4. Zimmerman RL, Burke MA, Young NA, et al. Diagnostic value of hepatocyte paraffin 1 antibody to discriminate hepatocellular carcinoma from metastatic carcinoma in fine-needle aspiration biopsies of the liver. Cancer. 2001;93:288-291.
5. Fetsch PA, Powers CN, Zakowski MF, et al. Anti-alpha-inhibin: marker of choice for the consistent distinction between adrenocor- tical carcinoma and renal cell carcinoma in fine-needle aspiration. Cancer. 1999;87:168-172.
6. Renshaw AA, Granter SR. A comparison of A103 and inhibin reactivity in adrenal cortical tumors: distinction from hepatocellular carcinoma and renal tumors. Mod Pathol. 1998;11:1160-1164.
7. Yoshida SM, Imam A. Monoclonal antibody to a proximal nephrogenic renal antigen: immunohistochemical analysis of for- malin fixed, paraffin-embedded human renal cell carcinomas. Cancer Res. 1989;49:1802-1809.
8. Avery AK, Beckstead J, Renshaw AA, et al. Use of antibodies to RCC and CD10 in the differential diagnosis of renal neoplasms. Am J Surg Pathol. 2000;24:203-210.
9. Gokden N, Mukunyadzi P, James JD, et al. Diagnostic utility of renal cell carcinoma marker in cytopathology. Appl Immunohisto- chem Mol Morphol. 2003;11:116-119.
10. Abrahams NA, MacLennan GT, Khoury JD, et al. Chromophobe renal cell carcinoma: a comparative study of histological, immuno- histochemical and ultrastructural features using high throughput tissue microarray. Histopathology. 2004;45:593-602.
11. Bartkova J, Horejsi Z, Koed K, et al. DNA damage response as a candidate anti-cancer barrier in early human tumorigenesis. Nature. 2005;434:864-870.
12. Grignon DJ, Eble JN, Bonsib SM, et al. Tumors of the Kidney. In: Eble JN, Sauter G, Epstein JI, et al, eds. Tumours of the Urinary System and Male Genital Organs. Lyon: IARC Press; 2004:23-25.
13. Pan CC, Chen PC, Tsay SH, et al. Differential immunoprofiles of hepatocellular carcinoma, renal cell carcinoma, and adrenocortical carcinoma: a systemic immunohistochemical survey using tissue array technique. Appl Immunohistochem Mol Morphol. 2005;13: 347-352.
14. Yang B, Guo M, Herman JG, et al. Aberrant promoter methylation profiles of tumor suppressor genes in hepatocellular carcinoma. Am J Pathol. 2003;163:1101-1107.
15. Morris MR, Hesson LB, Wagner KJ, et al. Multigene methylation analysis of Wilms’ tumour and adult renal cell carcinoma. Oncogene. 2003;22:6794-6801.
16. Bartkova J, Bakkenist CJ, Rajpert-De Meyts E, et al. ATM activation in normal human tissues and testicular cancer. Cell Cycle. 2005;4:838-845.
17. Munshi A, Kurland JF, Nishikawa T, et al. Histone deacetylase inhibitors radiosensitize human melanoma cells by suppressing DNA repair activity. Clin Cancer Res. 2005;11:4912-4922.
18. Munshi A, Tanaka T, Hobbs ML, et al. Vorinostat, a histone deacetylase inhibitor, enhances the response of human tumor cells to ionizing radiation through prolongation of gamma-H2AX foci. Mol Cancer Ther. 2006;5:1967-1974.