l' I\-, r ,-r-.
`1 .t ;_Ir t. .
`
`/
`
`NOVARTIS EXHIBIT 2007
`Breckenridge v. Novartis, IPR 2017-01592
`Page 1 of 37
`
`

`

`Cambridge Medical Reviews
`
`Haematological Oncology Volume 2
`
`NOVARTIS EXHIBIT 2007
`Breckenridge v. Novartis, IPR 2017-01592
`Page 2 of 37
`
`

`

`Published by the Press Syndicate of the University of Cambridge
`The Pitt Building, Trumpington Street, Cambridge CB2 !RP
`40 West 20th Street, New York, NY 10011-4211, USA
`IO Stamford Road, Oakleigh, Victoria 3166, Australia ·
`
`© Cambridge University Press 1992
`
`First published 1992
`
`Printed in Great Britain by Redwood Press Limited, Melksham, Wiltshire
`
`A catalogue record of this book is available from the British Library
`
`Library of Congress cataloguing in publication data available
`
`ISBN O 521431905 hardback
`
`WV
`
`NOVARTIS EXHIBIT 2007
`Breckenridge v. Novartis, IPR 2017-01592
`Page 3 of 37
`
`

`

`Clinical features and management of
`localized extranodal lymphomas
`
`S B SUTCLIFFE and MK GOSPODAROWICZ
`
`Introduction
`Malignant lymphomas account for about 5% of human cancers with an age(cid:173)
`standardized incidence of approximately 17 per 100 000 of the population.
`Non-Hodgkin's lymphomas are about four times more common than Hodg(cid:173)
`kin's disease and have an age-peak in the over-SO years age group. Both
`diseases share similar symptomatology in terms of characteristic symptoms
`(fever, night sweats, and weight loss) and with respect to nodal enlargement
`as the most common form of presentation. Both share a common staging
`classification which is based upon anatomical distribution of disease. A major
`difference occurs, however, in the frequency with which non-Hodgkin's
`lymphomas present with apparently localized disease in extranodal sites -
`primary extranodal lymphoma - a circumstance of extreme rarity in Hodg(cid:173)
`kin's disease.
`
`Primary extranodal non-Hodgkin's lymphoma
`The evolution of histological classifications of non-Hodgkin's lymphoma has
`followed from largely morphological observation stressing architecture and
`cytology to functional and biological measurements recognizing lineage and
`differentiation of lymphoma cells. Much of the controversy of lymphoma
`classification is illustrated in the histology of primary extranodal non-Hodg(cid:173)
`kin's lymphoma:
`
`• extranodal sites rarely align themselves with lymph node structure,
`thus the majority of primary extranodal lymphomas manifest 'dif(cid:173)
`fuse' effacement of tissue architecture.
`• lymphoid aggregates within extranodal tissue may lack characteristic
`features of lymphoma in terms of invasion and mitotic activity. Such
`aggregates have been considered pseudotumors and have charac-
`
`AII correspondence to: Dr SB Sutcliffe, Department of Radiation Oncology, Princess Margaret
`Hospital, 500 Sherbourne Street, Toronto, Ontario M4X 1K9, Canada.
`
`Cambridge Medical Reviews: Haematological Oncology Volume 2
`© Cambridge University Press 1992
`
`189
`
`NOVARTIS EXHIBIT 2007
`Breckenridge v. Novartis, IPR 2017-01592
`Page 4 of 37
`
`

`

`S B Sutcliffe and .M K Gospodarowicz
`
`teristically been recorded in orbital, pulmonary and gastrointestinal
`sites. Current functional characterization of the cells within such
`aggregates to define clonality should resolve the issue of 'benign'
`versus 'malignant' lymphoid infiltrates.
`• certain extranodal sites clearly have a spectrum of lymphoma
`histology from low to high grade within the International Working
`Formulation e.g. orbit, head and neck, gastrointestinal tract, lung
`and skin. Other sites have a highly skewed representation of inter(cid:173)
`mediate and high grade tumors e.g. extradural, brain, testis and
`bone lymphomas. The basis for this variable distribution of histology
`by geographic or anatomic site is unknown.
`•/certain extranodal sites have characteristic spectra of B- or T-cell
`disease. Cutaneous lymphoma clearly comprises a range of lym(cid:173)
`phomas of T-cell origin which demonstrate a preferential localization
`for the skin for long periods of the natural history of the disease.
`Similarly, the gastrointestinal tract favours a subset of lymphoma(cid:173)
`mucosa associated lymphoid tissue lymphoma (.MALT lymphoma)
`demonstrating preferential traffic patterns influencing localization
`and recurrence within the gastrointestinal tract. Similar analogies
`apply to thyroid and lung lymphoma.
`• whereas the existing histological classifications apply satisfactorily to
`nodal and B-cell lymphoma, the histological classification of T-cell
`disease within nodes and in extranodal sites is less satisfactory. The
`diversity of T-cell disease is demonstrated in primary cutaneous
`lymphoma with a disease spectrum ranging from chronic, indolent
`lichenoid eruptions
`through
`to
`lymphoblastic disease with a
`fulminant natural history. Even within a single disease entity - lym(cid:173)
`phomatoid granulomatosis - a very wide pattern of clinical behaviour
`is contained within a seemingly uniform histology.
`
`Thus, in terms of prognostic importance, histology is not only dis(cid:173)
`tinguished in its own right, but also in the context of histology in relation to
`primary extranodal site as a component of therapeutic management.
`Despite the obvious clinical diversity of primary extranodal lymphoma,
`there are many common attributes to define prognosis and management
`strategy. Stage, in the context of localization, defines those who are potential
`candidates for radiation as opposed to those who require systemic therapy for
`more advanced disease. However within localized stage, additional factors -
`presence or absence of symptoms ('A' or 'B'), tumor bulk, and extent of local
`invasion - influence the success of local therapies (radiation and/or surgery)
`with respect to both local tumor control and distant relapse rate. Histology
`influences management of those with localized or advanced lymphoma. In
`
`190
`
`NOVARTIS EXHIBIT 2007
`Breckenridge v. Novartis, IPR 2017-01592
`Page 5 of 37
`
`

`

`Localized extranodal lymphomas
`
`certain sites where low grade tumors occur, localized or more conservative
`treatment options may be applicable. In other sites where virtually all tumors
`are of diffuse, aggressive type, initial systemic therapy is a mandatory com(cid:173)
`ponent of potentially curative treatment. Finally, advancing age is an
`independent determinant characterizing poorer survival and disease-free
`rates.
`Whilst the approach to management of localized nodal lymphoma is
`generally well recognized, there is less familiarity with the management of
`localized extranodal lymphoma. This, in part, reflects the infrequency of
`extranodal presentations in certain anatomic sites and the requirement for a
`significant referral practice to acquire familiarity with management. This
`chapter will deal specifically with localized extranodal non-Hodgkin's lym(cid:173)
`phoma, the patterns of disease, and will address three principal questions:
`
`1. To what extent is the natural history and outcome of nodal and extra(cid:173)
`nodal disease similar, given similar management?
`2. • Are all extranodal sites similar in natural history and outcome, given
`similar management?
`3. Can the same principles of management be applied to extranodal lym(cid:173)
`phoma, as are applied to nodal lymphoma?
`
`In addressing these questions, data will be reviewed from the world literature
`and also from the experience at the Princess Margaret Hospital.
`
`General features of extranodal non-Hodgkin's lymphoma
`During the period 1967-1988, the referral of patients for radical treatment of
`non-Hodgkin's lymphoma at Princess Margaret Hospital was approximately
`62% Stage 1-11 and 38% Stage III-IV (Fig. 1). The composition of the stage
`1-11 cohort is shown in Table 1 and Fig. 2.
`The representation of stage I-II extranodal lymphomas by site is shown in
`Table 2 and Fig. 3. In descending order of frequency, the most common
`extranodal presentations were gastrointestinal tract, Waldeyer's ring, brain,
`head and neck, thyroid, soft tissue and testis. Symptomatic presentations (B
`symptoms) were uncommon. Certain presentations were frequently associ(cid:173)
`ated with clinical or surgically defined regional adenopathy, e.g. Waldeyer's
`ring, thyroid, gastrointestinal tract. Others were infrequently associated with
`regional adenopathy, e.g. eye and orbit, head and neck, testis, brain and
`other rare sites. This difference may, in part, reflect the higher probability of
`recording nodal disease in sites presenting primarily through surgical
`management.
`The cause-specific survival rates (i.e. death from lymphoma as the end
`point of analysis) from the time of diagnosis for extranodal and nodal presen-
`
`191
`
`NOVARTIS EXHIBIT 2007
`Breckenridge v. Novartis, IPR 2017-01592
`Page 6 of 37
`
`

`

`S B Sutcliffe and M K Gospodarowicz
`
`St III-IV
`38.3%
`
`*Does not include 79 pts with I'' brain lymphoma
`
`Fig. 1. Non-Hodgkin's lymphoma: PMII 1967-1988. Distribution by clinical stage.
`
`Table 1. Stage I and II 11011-Hodghin's lymphoma: PMH 1967-1988
`
`Number of patients
`.Male : female
`Low grade
`Intermediate and high grade
`
`1391*
`1.17 : 1.0
`27.5%
`69.5%
`
`Nodal
`Extranodal
`B symptoms
`
`49.1%
`50.9%
`6.8%
`
`* Excluding patients with primary brain lymphoma.
`
`192
`
`NOVARTIS EXHIBIT 2007
`Breckenridge v. Novartis, IPR 2017-01592
`Page 7 of 37
`
`

`

`Localized extranodal lymphomas
`
`Extranodal
`50.9%
`
`Nodal
`49.1%
`
`*Does not include 79 pts with 1° brain lymphoma
`
`fig. 2. Stage I and II non-Hodgkin's lymphoma: PMH 1967-1988. Distribution by
`nodal or extranodal site of presentation.
`
`Table 2. Stage I and II head and neck lymphoma presenting extranodal sites:
`PMH 1967-1988
`
`Waldeyer's ring
`
`Tonsil
`Naso pharynx
`Base of tongue
`Larynx and hypopharynx
`
`35.0
`17.5
`5.2
`1.3
`
`Non-Waldeyer's ring
`
`Oral cavity
`Salivary glands
`Paranasal sinuses
`Nasal cavity
`
`16.0
`9.6
`8.0
`2.2
`
`193
`
`NOVARTIS EXHIBIT 2007
`Breckenridge v. Novartis, IPR 2017-01592
`Page 8 of 37
`
`

`

`S B Sutcliffe and M K Gospodarowicz
`
`lung and pleura (1.1%)
`breast (2.0%)
`extradural (2.5%)
`bone (3.7%)
`
`eye and orbit (4.1%)
`
`testis and G.U. (4.4%)
`
`soft tissues (5.2%) _.-
`
`thyroid (6.1%)
`
`"-waldeyer's ring (19.4°1°)
`
`Fig. 3. Stage I and II non-Hodgkin's lymphoma: PMH 1967-1988. Distribution by
`extranodal site of presentation.
`
`brain (10%)
`
`tations demonstrate no significant difference (Fig. 4). These survival analyses
`do not take into account imbalances in the groups with respect to important
`prognostic factors including therapy. Table 3 shows the composition of the
`two groups with the respect to major prognostic variables. Patients with
`primary brain lymphoma arc included in this analysis. There is a slightly
`higher proportion of patients~ 60 years of age in the cxtranodal group. Other
`variables are fairly balanced other than for histology, with diffuse large cell
`(histiocytic) and higher grade histologies accounting for 39% of localized
`nodal presentations compared with 67% oflocalizcd cxtranodal presentations.
`Treatment allocations were similar during this time period with two-thirds or
`more of patients being treated primarily with radiation alone. More patients
`in the extranodal group received combined modality therapy (25% vs 17%),
`and surgery alone was the only form of treatment for 4.2% of extranodal
`presentations compared with 1.6% for the nodal group. Despite these
`imbalances, no significant difference in death from lymphoma was apparent
`within the overall analysis of nodal versus extranodal presentation of
`lymphoma.
`If, indeed, there is n.o overall difference in outcome, between nodal and
`extranodal disease, are all sites of extranodal lymphoma comparable in terms
`
`194
`
`NOVARTIS EXHIBIT 2007
`Breckenridge v. Novartis, IPR 2017-01592
`Page 9 of 37
`
`

`

`Lo.calized extranodal lymphomas
`
`'• .. ,. ··-, ... . , .....
`····-· ....
`.... , ...... . ............. , ··········., __
`
`100
`
`>,
`~
`15
`co
`.0
`0 ,._
`a.
`ro
`>
`-~
`
`::J
`Cf)
`
`80
`
`60
`
`40
`
`20
`
`0
`
`2
`
`4
`
`6
`
`12
`10
`8
`Time in years
`
`14
`
`16
`
`18
`
`20
`
`946W
`
`Fig. 4. Cause specific survival rates from the date of diagnosis for extranodal and
`nodal presentations of Stage I and II non-Hodgkin's lymphoma.
`
`Table 3. Prognostic factor distribution nodal and extranodal lymphoma - stage
`I and II: PMH 1967-1988
`
`Nodal
`
`Extranodal
`
`Age >60 years
`Stage IA & IIA
`Histology ==>DLC
`Bulk >5 cm
`Treatment XRT alone
`
`57
`95
`39
`30
`72
`
`49
`92
`67
`27
`63
`
`of outcome? The cause specific survival rates for various sites of localized
`extranodal lymphoma is shown in Table 4. Outcomes vary from a 5-year
`actuarial cause-specific survival of> 75% for skin, eye and orbit, thyroid and
`gastrointestinal lymphoma to <50% for testis, bone and brain lymphoma.
`Although these results are cause-specific actuarial survivals, given the dis(cid:173)
`tribution of prognostic factors including treatment allocation, it is unlikely
`
`195
`
`NOVARTIS EXHIBIT 2007
`Breckenridge v. Novartis, IPR 2017-01592
`Page 10 of 37
`
`

`

`S B Sutcliffe and M K Gospodarowicz
`
`Table 4. Localized1 extranodal lymphoma cause-specific 5-year survival: PMH
`1967-1988
`
`Skin
`Thyroid
`Gastrointestinal
`Soft tissue
`Orbit
`Gynaecological
`Extradural
`
`92%
`76%
`77%
`77%
`76%
`75%
`73%
`
`1 Clinical stage I & II.
`2 Head and neck lymphoma.
`3 Mostly including testis lymphoma.
`
`Waldeyer's2
`Non-Waldeyer's2
`Breast
`Genitourinary3
`Bone
`Lung and pleura
`Brain
`
`72%
`67%
`63%
`52%
`47%
`44%
`24%
`
`that the variable outcomes reflect changing treatment policy or imbalance of
`prognostic variables.
`Given this heterogeneity of outcomes, the various primary extranodal
`presentations will now be examined, particularly with respect to factors in the
`natural history that should influence management strategy.
`
`Primary gastrointestinal lymphoma
`The gastrointestinal tract is the most frequent site of presentation of localized
`extranodal lymphoma in adults. In the Western world, the most common
`locations are stomach (approximately 50-60%), small intestine (approx(cid:173)
`imately 30%), large intestine (approximately 10%). Localized oesophageal
`lymphoma is extremely rare. Within the small intestine, ileal (ileocecal)
`presentations are most common (approximately 60%), followed by jejunal
`(approximately 30%) and duodenal sites (approximately 10%). These propor(cid:173)
`tions differ by ethnic group and geography with small intestinal lymphomas
`being more common than gastric lymphoma in the Middle East. In such
`Mediterranean lymphomas, not only is there a background of malabsorption
`related to immunoproliferative small intestinal disease, but also a similar
`proportion of presentations in jejunum and ileum and an equal proportion
`(approximately 30%) of presentations at multiple sites within the small bowel.
`Colorectal lymphomas occur most commonly in the cecum, although this
`may be somewhat artificial due to the higher frequency of ileocecal presen(cid:173)
`tation. The rectum is the site of presentation in approximately 20% of large
`bowel presentations with equal (5-10%) incidence in the ascending,
`transverse and descending colon. Presentation frequencies of large bowel
`lymphomas in association with ulcerative colitis show a higher proportion of
`recto-sigmoid and transverse colon lesions.'
`Presenting symptoms are most frequently due to the local lesion (pain,
`
`196
`
`NOVARTIS EXHIBIT 2007
`Breckenridge v. Novartis, IPR 2017-01592
`Page 11 of 37
`
`

`

`Localized extranodal lymphomas
`
`obstruction, haemorrhage). Symptoms of lymphoma (fever and night sweats)
`are uncommon, however, weight loss associated with anorexia, pain or
`obstructive symptomatology are frequent and are more likely related to the
`primary presentation rather than a symptom of lymphoma per se.
`Radiological procedures commonly identify the site and nature of disease.
`Within the stomach, lymphoma is most common at the pyloric antrum,
`followed by the body, then the cardia. Lesions may appear polypoid, ulcer(cid:173)
`ative or diffusely thickening the stomach wall giving rise to rigidity and
`alteration of mucosa! texture. Whilst computed tomography is not well suited
`to imaging of the intraluminal component of lymphoma, visualization of wall
`thickening is possible as well as commentary on regional node size. Imaging
`of small bowel lymphoma more commonly illustrates luminal narrowing or
`obstruction with the more gross impact of mass size and mesenteric node
`status being defined by computerized tomography. In both large and small
`bowel, lesions may be annular, ulcerative, proliferative or polypoid.
`Diagnosis is made by examination of an adequate tissue sample most
`commonly obtained by endoscopy. Given access to fresh, frozen and fixed
`material with availability of immunophenotyping or genotypic studies, a pre(cid:173)
`operative diagnosis should be achievable on most patients with gastroin(cid:173)
`testinal lymphoma. Optimal histological classification remains controversial.
`The vast majority of tumors have a diffuse architectural pattern and are of
`predominantly large cell type, i.e. intermediate grade tumors within the
`Working formulation. Both low grade (follicular or small cleaved cell tumors)
`and high grade lymphomas (Burkitt, undifferentiated non-Burkitt) occur less
`frequently. The overwhelming majority of tumors are of B-cell origin
`although there is a distinct, small subset of T-cell tumors of pleomorphic type
`ranging from low to high grade. 1-4 True histiocytic lymphoma is not recorded
`in circumstances where accurate lineage definition is available. A further
`caveat to the histological classification is provided by the concept of mucosa(cid:173)
`associated lymphoid tissue. 5 Based upon clinical features suggesting preferen(cid:173)
`tial patterns of response and failure (clinical inference of homing), the
`presence of evolutionarily conserved endothelial antigens involved in prefer(cid:173)
`ential lymphocytic traffic between gastrointestinal and nodal sites,6 and the
`distinction of T. lymphocytic recognition systems that differ between gut and
`nodal sites,7•8 it has been suggested that within the histogenesis of gastroin(cid:173)
`testinal lymphoma, MALT lesions be recognized within existing histological
`classification schemes. The validity and utility of such modifications remain
`9
`to be seen. 1
`•
`Prognostic factors for survival include stage, nodal status and extent of
`nodal disease for localized presentations, ESR, and histological subtype. The
`role of tumor bulk, a highly significant prognostic factor in lymphoma, has to
`be evaluated in GI lymphoma in light of the fact that the majority of lesions
`are surgically excised prior to any additional therapy. There is little indication
`
`197
`
`NOVARTIS EXHIBIT 2007
`Breckenridge v. Novartis, IPR 2017-01592
`Page 12 of 37
`
`

`

`S B Sutcliffe and M K Gospodarowicz
`
`that the size of the primary lesion is of prognostic significance, as long as
`complete surgical resection has been achieved. Other factors are also con(cid:173)
`troversial: site within GI tract, depth of tumor penetration through bowel
`wall, and presentation with perforation. 10•11 In part, the role of factors charac(cid:173)
`terized by the primary tumor is influenced by the fact that virtually all
`patients receive therapy in addition to initial surgery.
`Historically, surgery has been the initial procedure of choice for manage(cid:173)
`ment of primary bowel lymphoma. This has been predicated on an overall
`surgical cure rate of the order of 25-40%; 12 an operative mortality of 5-20%; a
`procedure which can define accurate staging information; an effective debulk(cid:173)
`ing mechanism; and a means of restoring bowel integrity with avoidance of
`spontaneous or treatment-induced perforation. Whilst the indications for
`surgery remain unchanged, it is rarely the sole form of therapy. Postoperative
`radiation therapy has been commonly applied to fields involving the tumor
`bed or the whole abdomen with an expectation of cause-specific survival rates
`in excess of 70% at 10 years of follow-up. In the most favourable circum(cid:173)
`stances of totally resected disease or microscopic residual disease (Stage IA,
`IIA stomach and IA small bowel) surgery and radiation therapy resulted in
`cause-specific survival of 80% and a relapse rate of less than 15%. The
`prognosis was less favorable for Stage IIA small bowel lymphoma with a
`relapse rate of 45%. 11 More recently, surgery and chemotherapy have been
`employed given the relapse rate of approximately 30-35% for unselected
`patients with Stage 1-11 gastrointestinal lymphoma and the predominantly
`extraabdominal pattern of relapse, in those with complete resection of the
`primary tumor. With this approach, cause-specific survival rates in excess of
`75% at 5 years have been achieved. 13,14
`Whilst it is now clear that combination chemotherapy is required for
`patients with advanced local (residual local disease, Stage II with multiple
`involved nodes) and disseminated disease, the effectiveness of this treatment
`has engendered a reconsideration of the role of primary surgery. In principle,
`chemotherapy is effective for both debulking localized disease and for control
`of occult or overt disseminated disease. It might be argued that enhanced
`local control could be achieved with radiation to the tumor bed as part of a
`combined modality approach with the only principal additional morbidity
`from radiation. This approach has been pursued by the MD Anderson group
`with apparent success. 15 Such a management plan is based upon effective
`endoscopic diagnosis and removes detailed pathologic information. It does
`not deal with the issue of perforation, however, this risk may have been
`overestimated. In addition, it removes the place of radiation therapy from an
`· adjuvant to surgery to an adjuvant to a combined modality role. The optimal
`approach remains to be established. For patients with surgically resected
`primary bowel lymphoma (Stage I and IIA), a cause-specific survival rate and
`relapse-free rate of approximately 80% should be achieved for selected
`
`198
`
`NOVARTIS EXHIBIT 2007
`Breckenridge v. Novartis, IPR 2017-01592
`Page 13 of 37
`
`

`

`Localized extranodal lymphomas
`
`patients receiving postoperative chemotherapy. Preliminary data suggest a
`similar prognosis for patients receiving combined chemo-radiation in the
`absence of primary surgery.
`
`Upper aerodigestive tract
`Non-Hodgkin's lymphoma occurring in the head and neck region is the
`second most frequent site of localized extranodal presentation. In practice, it
`is preferable to restrict the term head and neck lymphoma to those presen(cid:173)
`tations occurring in the upper aerodigestive tract and to consider thyroid,
`brain, orbit and ocular, skin and cervical nodal lymphoma as separate sites.
`The localization of tumor presentation by site reveals tonsil to be the most
`common, followed by nasopharynx, oral cavity, salivary glands (probably
`incorporating nodal lymphoma presenting within salivary tissue), paranasal
`sinuses and base of tongue (Table 2). Thus, excluding multiple sites of
`presentation, lymphomas involving Waldeyer's ring constitute the majority
`(almost 60%) of upper aerodigestive tract localized lymphomas as in other
`17 In the Princess .Margaret series, localized disease with involvement
`series. 16
`•
`confined to the first echelon draining lymph nodes accounted for 82% of cases
`and intermediate grade histologies were present in 75% of patients. Localized
`presentation with B symptoms are unusual. The precise definition of disease
`bulk poses some difficulty with upper aerodigestive tract presentations.
`Advances in imaging, particularly computerized tomography (CT) and
`magnetic resonance imaging (.MRI) have greatly assisted the definition of
`disease extent. 18 Even so, documentation of margins and establishing the
`correlation of all abnormalities to malignancy, e.g. paranasal sinus opacifica(cid:173)
`tion, constitute problems with description of tumor bulk.
`Traditionally, radical radiation employing involved or extended field tech(cid:173)
`niques to tumor doses of 35 Gy or more has been the treatment method of
`choice. Retrospective analysis of such an approach reveals: 16•19- 25
`
`• head and neck;
`upper aero-digestive
`tract
`• Waldeyer's ring
`
`- actuarial survival
`- relapse-free
`rate/survival
`- actuarial survival
`- relapse-free
`rate/survival
`
`25-60%
`
`38-50%
`25-70%
`
`25-65%
`
`The wide range in survival and relapse-free rate largely reflects the effect of
`stage of disease with most series observing that nodal involvement effectively
`halves the results obtained for Stage I disease. In addition, the use of TN.M
`staging has indicated a substantial effect of primary tumor size and extent.
`Prognostic factor analysis commonly identifies stage as the most significant
`factor. In the P.MH series, lymphomas at sites other than Waldeyer's ring had
`a significantly higher relapse rate compared with Waldeyer's ring, although
`
`199
`
`NOVARTIS EXHIBIT 2007
`Breckenridge v. Novartis, IPR 2017-01592
`Page 14 of 37
`
`

`

`S B Sutcliffe and M K Gospodarowicz
`
`this is not a uniform experience. The prognosis for paranasal sinus tumors
`also appears to be within the expected range for other aero-digestive sites with
`more extensive local tumors (T3 and T4) clearly having an inferior survival
`compared with T 1_2 lesions. 25 ,26 Following irradiation, the isolated local
`failure rate is low (13% at 10 years in the PMH series), and the vast majority
`of failures are distant.
`More recently, chemotherapy usually in combination with radiation, has
`been employed as primary management. In retrospective analysis this
`approach has generally resulted in superior overall survival and relapse-free
`22
`21 The
`rates: 60-80% and 56-100% respectively (PMH series 1967-86). 19
`,
`,
`benefits of combined modality are most evident in patients with Stage II
`disease. In addition, the advantage of initial control of disease by combined
`modality therapy almost certainly outweighs a strategy of chemotherapy for
`salvage of radiation failure given a relatively unfavourable survival rate from
`relapse after radiation.
`An important aspect of the natural history of primary upper aerodigestive
`tract lymphoma is its relationship to gastrointestinal (GI) tract involvement,
`either concurrently at presentation or at subsequent relapse. The association
`at presentation was noted in 11 of 292 cases reported by Banfi et al. (1972).
`Although it is common practice to perform investigation of the GI tract in
`patients with apparently localized upper aerodigestive tract presentations, the
`yield is usually low. 21 The GI tract as a site of first relapse is clearly
`acknowledged and, indeed, is commonly the most frequent extranodal site of
`disease progression beyond the primary site. 16,20,22- 2·1,28 There is no apparent
`predilection for any other extranodal site of relapse - particularly, central
`nervous system progression is not an identified pattern of failure.
`
`Primary thyroid lymphoma
`Primary thyroid lymphoma commonly presents as a rapidly enlarging neck
`mass producing local obstructive and infiltrative symptomatology. The
`median age at presentation is in the mid-60 years and women are more
`commonly affected than men (M:F :: 4: 1). Approximately 90% of tumors are
`of diffuse architecture and of intermediate or high grade classification.
`Tumors are often bulky and complete or total thyroidectomy is rarely accom(cid:173)
`plished (approximately 10-20% of patients). Clinical Stage I and II disease
`accounts for approximately 80% of thyroid lymphomas. Involvement of the
`adjacent neck nodes is common. The predictive association of pre-existing
`chronic lymphocytic thyroiditis and subsequent lymphoma of the thyroid
`gland is well documented. 29
`Overall 5-year survival rates for patients with localized thyroid lymphoma
`range from 40-72%, 30-33 The cause-specific survival rate at 10 years (survival
`rate based on mortality directly due to lymphoma) is 64%, indicating the
`
`200
`
`NOVARTIS EXHIBIT 2007
`Breckenridge v. Novartis, IPR 2017-01592
`Page 15 of 37
`
`

`

`Localized extranodal lymphomas
`
`significant age-related mortality in a patient population with a median age of
`65 years. Relapse-free rates vary from 38-64%. The majority of patients in
`these reports were treated with radical radiation therapy alone. Local control
`of neck disease was achieved in >70% of patients. Overall survival and
`relapse-free rates of 93% and 78% are quoted for the MD Anderson series. 33
`More recently, chemotherapy has been incorporated into the management
`plan. In the MD Anderson series, overall survival and relapse-free rates were
`77% respectively for combined modality and 53% and 30% for chemotherapy
`alone. 33 The role of primary chemotherapy alone for unselected patients with
`thyroid lymphoma remains to be defined 34 given that most experience to-date
`probably reflects outcome for patients with adverse prognostic factors.
`Prognostic factor analysis commonly reveals that age and histology are not
`significant - a reflection of the advanced age of patients with thyroid lym(cid:173)
`phoma and the dominance of diffuse large cell histology. Most reports clearly
`define the role of stage and tumor bulk either as size, unresected neck disease,
`extrathyroid invasion or tumor fixation. In addition, retrosternal nodes or
`mediastinal mass are defined as adverse, bearing in mind the predominant
`role of radiation therapy in most of the reported results. 3i-33
`Patterns of failure analysis record recurrence in the abdomen with gastroin(cid:173)
`testinal, liver and splenic disease. Tonsillar recurrence is recorded and also
`lung and bone sites. The patterns of recurrence linking Waldeyer's ring and
`gastrointestinal sites accord with the view that primary thyroid lymphoma is a
`tumor of mucosa-associated lymphoid tissue. 35 As such, a preferential lym(cid:173)
`phocyte traffic may exist for sites of common embryologic origin. Disease
`progression in the central nervous system is noted very rarely.
`In current practice, the role of surgery is largely diagnostic and modern
`histopathologic techniques applied to fine-needle aspiration specimens or
`tissue biopsies should render thyroidectomy unnecessary. Radiation is highly
`effective for local tumor control and is a curative treatment for patients with
`limited, small bulk neck disease. In all other circumstances, combined
`modality therapy would appear to be optimal therapy with an anticipated cure
`rate of 80% for localized disease. The role of adjuvant radiotherapy in a
`combined modality regimen remains to be defined by prospective study,
`however, the high local control rate with modest dose radiation (35-40 Gy) is
`largely without significant acute or chronic morbidity.
`
`Primary central nervous system (CNS) lymphoma
`Central nervous system involvement by non-Hodgkin's lymphoma is not
`uncommon, however, secondary involvement subsequent to, or concurrent
`with, presentation at other nonneurological sites is the usual circumstance.
`The probability of CNS involvement may be predicted based upon certain
`presentation parameters - diffuse, high grade histology, extensive bone mar-
`
`201
`
`NOVARTIS EXHIBIT 2007
`Breckenridge v. Novartis, IPR 2017-01592
`Page 16 of 37
`
`

`

`S B Sutcliffe and M K Gospodarowicz
`
`row involvement particularly with circulating lymphoma cells, certain extra(cid:173)
`nodal presentations e.g. testis, intraocular or epidural lymphoma and
`widespread, particularly progressive, advanced disease. 3&--3s
`Lymphoma involving the central nervous system in the absence of overt
`systemic disease comprises two major types of presentation:
`
`• primary CNS lymphoma
`• primary leptomeningeal lymphoma
`
`Primary CNS lymphoma
`Primary parenchymal lymphoma of the brain is rare, usually comprising
`about 1-2% of all lymphomas and approximately l % of primary brain
`tumors. It commonly presents in the 50-70 year age group and gives rise to
`symptomatology based upon raised intracranial pressure, cranial nerve pal(cid:173)
`sies, neurologic deficit and, commonly, a significant impairment of mental
`function. Primary CNS lymphoma occurs almost exclusively in the brain
`with spinal cord presentation being extremely rare. 39 The incidence of pri(cid:173)
`mary parenchymal brain lymphoma has increased in the last decade as a
`result of HIV/AIDS, thereby establishing clearly the association previously
`observed in the organ transplantation setting, that primary CNS lymphoma is
`associated with immunodeficiency and immunoincompetenc